journal of tics joe 2010 issue august

Clinical Research

Biofilms and Apical Periodontitis: Study of Prevalenceand Association with Clinical and Histopathologic FindingsDomenico Ricucci, MD, DDS,* and Jose F. Siqueira, Jr., DDS, MSc, PhD†

Abstract
Introduction: This study evaluated the prevalence ofbacterial biofilms in untreated and treated root canalsof teeth evincing apical periodontitis. The associationsof biofilms with clinical conditions, radiographic size,and the histopathologic type of apical periodontitiswere also investigated. Methods: The materialcomprised biopsy specimens from 106 (64 untreatedand 42 treated) roots of teeth with apical periodontitis.Specimens were obtained by apical surgery or extractionand were processed for histopathologic and histobacter-iologic techniques. Results: Bacteria were found in allbut one specimen. Overall, intraradicular biofilmarrangements were observed in the apical segment of77% of the root canals (untreated canals: 80%; treatedcanals: 74%). Biofilms were also seen covering the wallsof ramifications and isthmuses. Bacterial biofilms werevisualized in 62% and 82% of the root canals of teethwith small and large radiographic lesions, respectively.All canals with very large lesions harbored intraradicularbiofilms. Biofilms were significantly associated with epi-thelialized lesions (cysts and epithelialized granulomasor abscesses) (p < 0.001). The overall prevalence of bio-films in cysts, abscesses, and granulomas was 95%,83%, and 69.5%, respectively. No correlation was foundbetween biofilms and clinical symptoms or sinus tractpresence (p > 0.05). Extraradicular biofilms wereobserved in only 6% of the cases. Conclusions: Theoverall findings are consistent with acceptable criteriato include apical periodontitis in the set of biofilm-induced diseases. Biofilm morphologic structure variedfrom case to case and no unique pattern for endodonticinfections was identified. Biofilms are more likely to bepresent in association with longstanding pathologicprocesses, including large lesions and cysts. (J Endod2010;36:1277–1288)
Key WordsApical periodontitis, bacterial biofilm, endodontic infec-tion, endodontic treatment

From *Private Practice, Rome, Italy; and †Department of EndodAddress requests for reprints to Dr Domenico Ricucci, Piazza Ca

0099-2399/$0 - see front matterCopyright ª 2010 American Association of Endodontists.

doi:10.1016/j.joen.2010.04.007

JOE — Volume 36, Number 8, August 2010

In their natural habitats, microorganisms almost invariably live as members of meta-bolically integrated communities usually attached to surfaces to form biofilms (1).

The biofilm community lifestyle provides microorganisms with a series of advantagesand skills that are not observed for individual cells living in a free-floating (planktonic)state including establishment of a broader habitat range for growth; increased meta-bolic diversity and efficiency; protection against competing microorganisms, hostdefenses, antimicrobial agents, and environmental stress; and enhanced pathogenicity(2). The study of microbial biofilms assumes a great importance in different sectors ofindustrial, environmental, and medical microbiology. In medical microbiology, bio-films have been increasingly studied and estimates indicate that biofilm infectionscomprise 65% to 80% of the human infections in the developed world (3). As forthe oral cavity, caries, gingivitis, and marginal periodontitis are examples of diseasescaused by bacterial biofilms in the form of supragingival or subgingival dental plaque.

Mounting evidence indicates that apical periodontitis is also a biofilm-induceddisease (4–6). In situ investigations using optical and/or electron microscopy haveallowed observations of bacteria colonizing the root canal system in primary orpersistent/secondary infections as sessile biofilms covering the dentinal walls (7–12). Apical ramifications, lateral canals, and isthmuses connecting main root canalshave all been shown to harbor bacterial cells, which are also frequently organized inbiofilm-like structures (13–15). In addition, biofilms adhered to the apical rootsurface (extraradicular biofilms) have been reported and regarded as a possiblecause of posttreatment apical periodontitis (16, 17).

Although the concept of apical periodontitis as a biofilm-induced disease has beenbuilt upon these observations, the prevalence of biofilms and their association with clin-ical and histopathologic findings have not yet been reported. Before this informationbecomes available, it may seem somewhat imprecise to generalize and categorize apicalperiodontitis as a biofilm-induced disease. The purpose of the present study was two-fold: (1) evaluate the prevalence of intraradicular and extraradicular bacterial biofilmsin untreated and treated root canals of human teeth evincing apical periodontitisthrough a histobacteriologic approach and (2) look for associations of biofilms withsome clinical conditions, radiographic size, and the histopathologic type of apical pe-riodontitis lesions.

Materials and MethodsClinical Specimens

The material for this study consisted of sequential biopsies of roots or root tipstogether with surrounding apical periodontitis lesions. Specimens were part of thehistologic collection of one of the authors (DR). The material comprised 106 rootsfrom 100 human teeth. Of these, 58 were teeth with untreated root canals (6 incisors,3 canines, 18 premolars, and 31 molars) from 52 patients (25 females, 27 males) aged18 to 75 years (mean, 42 years). In total, 64 roots from untreated teeth were available,of which 59 were extracted with apical periodontitis lesions attached while in the other

ontics, Faculty of Dentistry, Estacio de Sa University, Rio de Janeiro, Brazil.lvario, 7, 87022 Cetraro (CS), Italy. E-mail address: [email protected].

Biofilms and Apical Periodontitis 1277

mailto:[email protected]

TABLE1.

Prev

alen

ceof

Intr

arad

icul

arB

iofil

ms

atth

eAp

ical

Segm

ent

and

Extr

arad

icul

arB

iofil

ms

inU

ntre

ated

and

Roo

tCa

nal–

trea

ted

Teet

hAc

cord

ing

toth

eH

isto

path

olog

icTy

peof

Apic

alPe

riod

ontit

isan

dCl

inic

alFe

atur

es

Intr

arad

icu

lar

bio

film

Extr

arad

icu

lar

bio

film

Un

trea

ted

(%)

Trea

ted

(%)

Tota

l(u

ntr

eate

d+

trea

ted

)(%

)U

ntr

eate

d(%

)Tr

eate

d(%

)To

tal

(un

trea

ted

+tr

eate

d)

(%)

Ove

rall

pre

vale

nce

51/6

4(8

0)

31/4

2(7

4)

82/1

06

(77)

4/6

4(6

)2/4

2(5

)6/1

06

(6)

Lesi

on

typ

eG

ran

ulo

ma,ep

ith

eli

ali

zed

5/6

(83)

1/1

(100)

6/7

(86)

0/6

(0)

0/1

(0)

0/7

(0)

Gra

nu

lom

a,n

on

ep

ith

eli

ali

zed

13/2

0(6

5)

22/3

2(6

9)

35/5

2(6

7)

1/2

0(5

)1/3

2(3

)2/5

2(4

)C

yst,

tru

e10/1

0(1

00)

—10/1

0(1

00)

0/1

0(0

)—

0/1

0(0

)C

yst,

po

cket

(bay)

7/8

(87.5

)—

7/8

(87.5

)2/8

(25)

—2/8

(25)

Cys

t,u

ncl

ass

ified

2/2

(100)

—2/2

(100)

1/2

(50)

—1/2

(50)

Ab

scess

,ep

ith

eli

ali

zed

3/3

(100)

—3/3

(100)

0/3

(0)

—0/3

(0)

Ab

scess

,n

on

ep

ith

eli

ali

zed

10/1

4(7

1)

7/7

(100)

17/2

1(8

1)

0/1

4(0

)1/7

(14)

1/2

1(5

)U

ncl

ass

ified

1/1

(100)

1/2

(50)

2/3

(67)

0/1

(0)

0/2

(0)

0/3

(0)

Sin

us

tract

2/2

(100)

5/6

(83)

7/8

(87.5

)1/2

(50)

2/6

(33)

3/8

(37.5

)Sy

mp

tom

s41/5

0(8

2)

16/1

9(8

4)

57/6

9(8

3)

4/5

0(8

)2/1

9(1

0.5

)6/6

9(9

)

Clinical Research
five specimens lesions had to be removed separately. All untreated teethhad necrotic pulps and gross carious lesions and were extractedbecause they were judged unrestorable or the patient did not agreeto save the tooth. Records were made of any symptoms that the patientexperienced or was experiencing in relation to the affected tooth. Thepresence/absence of a sinus tract was also recorded. For 33 teeth,conventional periapical radiographs were available before extraction,and the largest diameter of the periradicular radiolucency wasmeasured. Teeth with periodontal pockets or longitudinal fracturesor cracks involving the root were excluded from analysis. Some ofthe examined teeth were included in previous studies (18, 19).
The 42 biopsies of roots/root tips from root canal-treated teethwere from 42 patients, 24 (12 symptomatic and 12 asymptomatic) ofwhich took part in a recent publication (10) and were re-evaluatedin this study for the presence of biofilms following the parameters es-tablished herein (see later). All 42 cases were categorized as treatmentfailures on the basis of clinical and/or radiographic follow-ups, aftera minimum recall period of 4 years for the asymptomatic cases and 1year for the symptomatic ones. For 10 of the new cases, the quality ofprevious endodontic treatment was judged as inadequate. Radiographswere not available for four of the new 18 cases. All patients had givenconsent for examination of their teeth.

Tissue ProcessingImmediately after removal (by periradicular surgery or extrac-

tion), the biopsy specimen was immersed in 10% neutral-bufferedformalin for at least 48 hours. Demineralization was performed in anaqueous solution consisting of a mixture of 22.5% (vol/vol) formicacid and 10% (wt/vol) sodium citrate for 3 to 4 weeks, with theendpoint being determined radiographically. All specimens werewashed in running tap water for 24 to 48 hours, dehydrated inascending grades of ethanol, cleared in xylene, infiltrated, andembedded in paraffin (melting point, 56�C) according to standardprocedures. To produce sections parallel to the long axis of the rootcanal, special precautions were taken. Roots in multirooted teethwere dissected free and processed separately. If curved, roots wereseparated in two pieces, one encompassing the coronal two thirdsand the other including the apical one third. These two pieces wereembedded separately, but only the apical segment was evaluated inthis study.

With the microtome set at 4 to 5 mm, meticulous longitudinal serialsections were taken until each specimen was exhausted. For some spec-imens, cross-cut sections were taken. Every fifth slide was stained withhematoxylin-eosin for screening purposes and for assessment ofinflammation. A modified Brown and Brenn technique for stainingbacteria (20) was used for selected slides. The accuracy of the bacterialstaining method was tested using the protocol described by Ricucci andBergenholtz (21).

Evaluation CriteriaSlides were examined by two evaluators. Evaluations were per-

formed separately, and whenever disagreement occurred, it wasresolved by joint discussion. The following aspects were specificallylooked for in the examination: 1) the presence and location of bacteriain the apical segment of the root canal, including the main canal, lateralcanals, apical ramifications, and isthmuses (intraradicular infection) orwithin the body of the apical periodontitis lesion or adhered to theexternal apical root surface (extraradicular infection). The parameterused for classification of bacterial community structures as biofilms fol-lowed the biofilm definition given by Hall-Stoodley et al (22): ‘‘Micro-bial biofilms are populations of microorganisms that are concentrated

1278 Ricucci and Siqueira Jr. JOE — Volume 36, Number 8, August 2010http://endodontic.ws

http://endodontic.ws

Clinical Research
at an interface and typically surrounded by an extracellular polymericsubstance matrix.’’ Also, according to these authors, aggregates or coag-gregates of bacterial cells not apparently attached to the surface wereclassified as ‘‘flocs’’; and 2) the presence and distribution of acuteand chronic inflammatory cells and epithelium in the inflamed perira-dicular tissues; lesions were diagnosed histologically as follows: apicalabscess (epithelialized or nonepithelialized), granuloma (epithelializedor nonepithelialized), or cyst (true, pocket, or unclassified; the latterdiagnosis was made when the lesion showed a cavity lined by epithe-lium, but the soft tissue did not remain attached to the root tip andhad to be removed separately).
Statistical AnalysisThe Fisher exact test or the chi-square test was used to check for

associations of intraradicular biofilms with the following parameters:radiographic size of the apical periodontitis lesion (grouped as lesions#5 mm and >5 mm), histopathologic general type of apical periodon-titis (granuloma, cysts, and abscesses with no distinction of subtypes),presence and absence of epithelial proliferation (irrespective of thelesion type), sinus tract, and symptoms. Every analysis took into consid-eration only untreated root canals, only treated canals, and all speci-mens together. The prevalence of biofilms in untreated and treatedteeth was also compared by the chi-square test.

ResultsBiofilm Overall Prevalence

Bacteria were found in all specimens, except for one asymptomaticroot canal–treated tooth in which disease emerged probably because ofa foreign body reaction. This case was reported in a previous study(10). Overall, bacterial arrangements as intraradicular biofilms wereobserved in the apical segment of 82 of 106 (77%) root canals. Of these,51 of 64 (80%) were from untreated canals and 31 of 42 (74%) fromtreated canals (Table 1). This difference was not statistically significant(c2, p = 0.6).

Figure 1. Examples of intracanal biofilms with different bacterial cell morphologieswith the root canal wall (Taylor’s modified Brown and Brenn, original magnificatiobution of bacterial cells within the extracellular material (original magnification �


Morphologic Description of Bacterial ColonizationIntraradicular bacterial biofilms were usually thick and composed

of several layers of bacterial cells. At high magnification, three basicbacterial cell morphologies could be recognized in most cases: cocci,rods, and filaments (Figs. 1a and b, 2c and d, 3c and d, 4d, and 5c andd). These morphotypes were often present together in varied propor-tions in the same biofilm. However, a single morphotype appeared todominate each biofilm (Figs. 1a and b, 2c and d, 3c, 4d, and 5c and d).

In the biofilm structure, the proportion between bacterial cells andextracellular matrix was highly variable. In some instances, bacterialcells appeared so clumped that the extracellular component was virtu-ally not visible (Figs. 2d and 5c and d). In other cases, the extracellularmatrix was abundant, and less bacterial cells were seen distributed in anuneven pattern (Figs. 1a and b, 2c, 3c, and 4c and d). In many biofilms,cells were abundant in the deepest layers (Figs. 1a and 4c). In someother cases, however, they were prevalent in the most superficial layers.

In many specimens, multilayered biofilms covered uniformly theroot canal walls to a long extent. In some cases, opposite root canalwalls covered by biofilms were faced with necrotic debris or inflamma-tory cells in the canal lumen (Figs. 3b and 5b). In teeth with severecaries destruction and longstanding pulp necrosis, the observationthat the entire apical canal was filled by a dense biofilm was notuncommon. In some instances, however, although some areas of thecanal were completely covered by biofilms, others were apparentlyfree from bacterial colonization (Fig. 2g). This pattern was morecommonly observed in treated root canals but also seen in a fewuntreated specimens.

In some instances, bacterial colonization was restricted to the rootcanal wall surface, and no deep dentinal invasion was observed. Thiswas probably because of the reduced number and small diameter oreven the lack of dentinal tubules in certain regions of the apical rootsegment (Figs. 1a and b, 2c and d, and 3c). However, when dentinaltubules were present and abundant, they usually appeared colonizedby bacteria spreading from the biofilm and penetrating at varying depths(Fig. 6a and b).

. (a) The predominance of cocci. Note the high concentration of cells in contactn �1,000). (b) Predominance of filamentous forms. Note the irregular distri-1,000).


Figure 2. (a) Mandibular molar with the crown destroyed by a gross carious process and a radiolucency around the mesial root assessed as >5 mm. Severalexacerbations developed over the previous months, but the tooth was asymptomatic at the time of extraction. (b) Sections were taken on a mesiodistal plane. Thesection passing through the apical third of the mesiobuccal canal. The lumen appears partly filled by large fuchsin-stained bodies, likely food remnants (largevegetable cells). More apically the walls appeared covered by a dense bacterial biofilm (Taylor’s modified Brown and Brenn, original magnification �100).(c) High magnification of the area of the root canal wall indicated by the left arrow in b. Rods are the prevailing bacterial morphotype at this level (original magni-fication�1,000). (d) High magnification of the area of the root canal wall indicated by the right arrow in b. The high bacterial population density seems to obscurethe extracellular matrix. Note the polymorphonuclear neutrophils in contact with the biofilm surface (original magnification�1,000). (e) Cross-cut section takenat the transition between the apical and the middle third of the mesial root. The low-magnification overview shows that the two mesial canals are connected by a wideisthmus, clogged with bacteria (original magnification �8). (f) Detail of the isthmus. Its lumen is filled by a dense biofilm (original magnification �100). (g)Magnification of the left canal in e. The majority of the root canal circumference is covered by a bacterial biofilm (original magnification �100).

Clinical Research

One of the untreated teeth exhibited an unusual pattern of intraca-nal bacterial colonization and deserves a more detailed description. It isa maxillary first premolar from a 58-year-old man. The patient reportedseveral pain episodes, and the tooth, judged nonrestorable because ofgross coronal destruction, was extracted. Sections passing through thebuccal canal showed that the apical canal lumen was clogged withnecrotic debris and bacteria (Fig. 7a). At the center of the main rootcanal, a large bacterial floc composed of ramifying filamentous bacterialcells was present and surrounded by a distinct layer of an amorphousmaterial (Fig. 7b-d). Bacteria appeared particularly condensed in thisstructure, and at the periphery they exhibited a starburst appearance

1280 Ricucci and Siqueira Jr. http://endo

typical of actinomycotic colonies (Fig. 7c and d). Serial sections re-vealed that this large colony was not contiguous with the root canalwalls. Actually, it was apparently free floating in the root canal lumen,enmeshed in the remainder of the canal content. This case has been as-sessed as showing a biofilm, not because of the unusual large floc sus-pended in the canal, which according to the definition criteria usedherein cannot be strictly assessed as a biofilm, but rather because thebacterial condensations present more apically were clearly adheredto the root canal walls forming a biofilm-like structure (Fig. 7b).

Biofilms were also seen covering the walls of apical ramifications,lateral canals, and isthmuses in both untreated and treated canals. In

JOE — Volume 36, Number 8, August 2010dontic.ws

http://endodontic.ws

Figure 3. Grossly carious single-rooted mandibular second molar extracted with the apical periodontitis lesion attached. The radiographic size of the lesion was<5 mm (inset). The tooth was symptomatic. (a) The section passing approximately at the center of the foramen. The overview shows granulomatous tissue ingrowthat the very apical canal (Taylor’s modified Brown and Brenn, original magnification�16). (b) Detail of the apical foramen region. A biofilm is present covering theroot canal walls, and a dense bacterial aggregate is evidenced more coronally. Empty spaces are shrinkage artefacts (original magnification �100). (c) Highermagnification of the area demarcated by the rectangle in b. Bacterial filamentous forms prevail, and the extracellular component is abundant at this level (originalmagnification�400). (d) Higher magnification of the bacterial aggregate indicated by the arrow in b. Different morphotypes are present. Note the concentration ofpolymorphonuclear neutrophils in contact with the biofilm surface (original magnification �400).

Clinical Research

some untreated canals, bacteria in the form of biofilms or flocs did notreach the apical foramen because vital inflamed tissue was observedoccupying the very apical canal. In these cases, the front of infectionwas located some millimeters short of the foramen.

Small bacterial flocs and planktonic cells were found in virtually allspecimens including those negative for the presence of biofilms. Flocsand planktonic cells were usually observed in the lumen of the maincanal, ramifications, and isthmuses, either apparently floating ormore commonly enmeshed in the necrotic pulp tissue.


Association with Diverse ConditionsRadiographs available for 71 specimens were used to check for an

association between intraradicular biofilms and the radiographic size ofapical periodontitis. Bacterial biofilms were visualized in 23 of 37(62%) root canals with small lesions (#5 mm), whereas in specimenswith large lesions (>5 mm), biofilm structures were present in 28 of 34(82%) (Table 2). Statistical analysis disclosed a p value very close to thelevel of significance used (c2, p = 0.059). Specifically for untreatedcanals, the prevalence of intraradicular biofilms was 59% and 87.5%


Figure 4. (a) Maxillary second premolar with a periapical radiolucency whose diameter was measured >5 mm. The tooth was extracted in the presence of clinicalsymptoms (pain and swelling). (b) The apical periodontitis lesion remained attached to the root tip. Section passing through the main wide apical foramen (Tay-lor’s modified Brown and Brenn, original magnification�16). (c) Detail of the foramen area. A biofilm is present in the most apical canal, faced with inflammatorytissue. Note the resorption of the canal walls (original magnification�100). The inset shows PMNs from the center of the lesion, one of which exhibits a cytoplasmengulfed with several bacterial fragments (original magnification �1,000). (d) Higher magnification of the area from the cementum fragment indicated by thearrow in c. Bacterial filamentous forms are dominant at this level. The biofilm is surrounded by PMNs (original magnification �400).

Clinical Research

for specimens with small and large lesions, respectively. Biofilms intreated teeth were disclosed in 65% and 78% of the canals with smalland large lesions, respectively. All five canals associated with lesionslarger than 10 mm harbored intraradicular biofilms (Table 2).

Regarding the histopathologic diagnosis, intraradicular biofilmswere significantly associated with epithelialized lesions (cysts and epi-thelialized granulomas or abscesses) (Fisher test, p < 0.001). Of the 30lesions exhibiting epithelial proliferation, 28 (93%) were associatedwith intraradicular biofilms. Although canals associated with epithelial-ized or nonepithelialized granulomas exhibited intraradicular biofilmsin 86% and 67%, respectively, this 20% difference did not reach signif-

1282 Ricucci and Siqueira Jr.

icance for the sample size evaluated (Fisher test, p = 0.4). When lesionswere grouped as granulomas (epithelialized or not), abscesses (epithe-lialized or not), and cysts (true, pocket, and unclassified), intraradic-ular biofilms were found significantly more frequently in cysts than ingranulomas (Fisher test, p = 0.03). Actually, only one out of the 20(5%) cystic lesions was negative for the presence of intraradicular bio-films. No other significant differences were observed for lesion types (p> 0.05). The overall prevalence of biofilms in granulomas was 41 of 59(69.5%), and in abscesses it was 20 of 24 (83%). In general, findingsfrom separate analysis of untreated canals were similar to the overallfindings. Because cysts were not observed in association with root


Figure 5. (a) Mandibular second premolar extracted after several pain episodes. The apical periodontitis lesion remained attached to the root tip at extraction.The section passing approximately at the center of the root canal. The overview discloses a severely resorbed root apex. The canal appears filled with tissue, andthere are two bacterial masses on the opposite root canal walls (Taylor’s modified Brown and Brenn, original magnification�16). (b) Higher magnification of theapical canal. The bacterial masses were biofilm structures. The canal lumen is filled with inflammatory cells (original magnification �100). (c and d) Highermagnification of the left and right biofilm structures respectively. Both were apparently exclusively composed of the bacterial filamentous morphotype. A severeinflammatory reaction surrounded these bacterial biofilms (original magnification �400).

Clinical Research

canal–treated teeth, statistical analysis was not performed separately fordata from treated canals.

Intraradicular biofilms were found in seven of the eight (87.5%)specimens associated with a sinus tract and in 57 of 69 (83%) of thesymptomatic cases (Table 1). However, despite the high prevalence,these values were not statistically significant when compared with caseswith no sinus tract (75/98, 76.5%) (Fisher test, p = 0.7) and no symp-toms (25/37, 68%) (c2, 0.08). No significance was found for data fromuntreated canals and treated canals when examined separately (p >0.05).

Extraradicular bacterial biofilms were observed in only six speci-mens (6%), four from untreated canals and two from treated canals. Allthese specimens were associated with clinical symptoms. Of the eightcases with sinus tracts, extraradicular biofilms were detected in three(37.5%). In all but one of the cases, the extraradicular biofilm was asso-


ciated with an intraradicular biofilm. Two of these extraradicular bio-films showed areas of mineralization with a calculus-like appearance(Fig. 8).

DiscussionDetermination that a given human infectious disease is caused by

biofilms is not an easy task. Difficulties may be related to severalreasons, including the coexistence of biofilm and planktonic bacteriain many infections, the absence of a definitive marker for bacteria form-ing biofilms, and the loss of the biofilm phenotype when subject tosampling and culturing procedures (23). By taking such difficultiesinto account, Parsek and Singh (23) proposed the following criteriato define infections caused by biofilms: (1) the infecting bacteria areadhered to or associated with a surface (‘‘associated with’’ implies


Figure 6. Maxillary lateral incisor with a large periapical radiolucency (>5 mm) (inset). (a and b) The section taken at the transition between the apical and themiddle third showing a bacterial biofilm covering the dentinal walls. The dentinal tubules subjacent to the biofilm are heavily invaded and colonized to varyingdepths (Taylor’s modified Brown and Brenn, original magnification �100 and �400).

Clinical Research

that aggregates/coaggregates do not need to be firmly attached); (2)direct examination of infected tissue shows bacteria forming clustersor microcolonies encased in an extracellular matrix, which may be ofbacterial or host origin; (3) the infection is generally confined toa particular site, and although dissemination may occur, it is a secondaryevent; and (4) the infection is difficult or impossible to eradicate withantibiotics despite the fact that the responsible microorganisms aresusceptible to killing in the planktonic state. The following criterionwas further added by Hall-Stoodley and Stoodley (24): (5) ineffectivehost clearance evidenced by the location of bacterial colonies indiscrete areas in the host tissue associated with host inflammatory cells.According to this last criterion, evidence of polymorphonuclear neutro-phils (PMNs) and macrophages surrounding bacterial aggregates/coaggregates in situ considerably increases the suspicion of biofilminvolvement with disease causation. We propose the following sixthcriterion: the elimination or significant disruption of the biofilm struc-ture and ecology leads to remission of the disease process.

Although there are recognized limitations to these criteria, it isassumed that they provide general characteristics that allow for consid-ering the role of biofilms in the pathogenesis of a certain human disease(24). The present findings showing biofilm structures in the greatmajority of cases of primary (80%) and posttreatment (74%) apical pe-riodontitis along with the observed morphological features of these bio-films seem to fulfill four of the six criteria. Although adhesion andstrength of adhesion cannot be measured by the methods used in thepresent study, the bacterial agreggates/coaggregates were observed toat least be associated with the root canal dentin surface (criterion 1).Bacterial colonies were seen in the huge majority of the specimens en-cased in an amorphous extracellular matrix whose origin was, however,not possible to determine (criterion 2). Endodontic biofilms were oftenconfined to the root canal, in a few cases extending to the external rootsurface, but dissemination through the lesion never occurred (criterion3). In the great majority of cases, biofilms were directly faced by inflam-


matory cells in the very apical canal, in ramifications, and in isthmuses(criterion 5). Criterion 4 was not assessed in this study, but it is widelyknown that intraradicular endodontic infections are not treatable bysystemic antibiotics, even though most endodontic bacteria are suscep-tible to currently used antibiotics (25). The problem with using systemicantibiotics is related to the fact that endodontic infection occurs in anavascular space with restricted access to antibiotics, but the recognitionof endodontic infections as biofilm infections still strengthens the expla-nations for antibiotic ineffectiveness. As for criterion 6 proposed in thisstudy, effects of treatment on biofilms and how they influence the treat-ment outcome were not evaluated herein and await further investiga-tions in a longitudinal experimental design. Nonetheless, the frequentobservation of biofilms in treated canals with posttreatment diseasemay at least suggest that there is a potential for fulfillment of this crite-rion.

The apical root canal can be regarded as a critical territory forpathogenetic and therapeutic reasons (26). This is because bacterialocated at the apical canal of teeth with apical periodontitis are insuch a strategic position that they may be regarded as the most impor-tant infective agents related to the disease pathogenesis. With this inmind, the present study restricted the investigation of biofilm prevalenceto the apical root canal system. The very high prevalence of intraradic-ular biofilms may be related to the fact that bacteria in the apical canalcompose the advanced front of infection and then directly face an in-flamed tissue area. Inflammatory exudate seeps into the apical canalto create a fluid phase and provide bacteria with nutrients in theform of glycoproteins and proteins. This may represent optimal condi-tions for biofilm formation and help explain the many exuberant bio-films observed in the apical canal, especially in primary endodonticinfections in which the untreated root canal may afford more spacefor exudate seepage and stagnation into the canal.

Overall, intraradicular biofilms were 20% more frequent in teethwith large radiographic lesions than in those with small lesions. All root


Figure 7. (a) Untreated maxillary first premolar extracted with the apical periodontitis lesion attached to the root tip. The longitudinal section passing approx-imately through the center of the buccal canal. Ingrowth of granulomatous tissue is evident at the apical foramen (Taylor’s modified Brown and Brenn, originalmagnification �16). (b) Detail of the apical root canal showing the bacterial content. A large bacterial floc exhibiting a high bacterial density and surrounded byamorphous material can be distinguished at the center of the canal lumen. Empty spaces are artifacts (original magnification �100). (c and d) Higher magni-fication of the upper and the lower halves of the floc in b. A great amount of intertwining filaments radiating at the periphery and projecting into a distinct extra-cellular surround is discernible. Note the totally different arrangement of the bacterial populations outside the floc (original magnification �400).

TABLE 2. Prevalence of Intraradicular Biofilms at the Apical Segment ofUntreated and Treated Canals of Teeth with Apical Periodontitis According tothe Lesion Size as Determined Radiographically

Lesion sizeUntreated

(%) TreatedTotal (untreated

+ treated)

#5 mm 10/17 (59) 13/20 (65) 23/37 (62)>5 mm 14/16 (87.5) 14/18 (78) 28/34 (82)$10 mm* 2/2 (100) 3/3 (100) 5/5 (100)

*These specimens were also included in the group >5 mm.

Clinical Research

canals associated with very large lesions (>10 mm) were found toharbor intraradicular biofilms. Large lesions have been associatedwith complex intraradicular infections characterized by bacterialcommunities with increased species richness and high populationaldensity (27, 28). Because it takes time for apical periodontitis todevelop and become radiographically visible, it is conceivable toassume that large lesions represent a longstanding pathologicprocess caused by an even ‘‘older’’ intraradicular infection. Ina longstanding infectious process, involved bacteria may have hadenough time and conditions to adapt to the environment and seta mature and organized biofilm community. The fact that infectedroot canals of teeth with large lesions harbor a large number of cellsand species almost always organized in biofilms may help explain thelong-held concept that the treatment outcome may be influenced bythe lesion size (29).


The present study revealed that intraradicular biofilms were signif-icantly more frequent in root canals of teeth with epithelialized lesions(cysts and epitheliazed granulomas or abscesses). Ninety-three percentof the lesions exhibiting some level of epithelial proliferation were in


Clinical Research
association with root canals colonized by bacterial biofilms. As for thelesion histopathologic type, intraradicular biofilms were significantlymore detected in cases diagnosed as cysts (95% as compared with69.5% in granulomas and 83% in abscesses). Because apical cystsdevelop as a result of epithelial proliferation in some granulomas(30), it is reasonable to assume that the older the apical periodontitislesion, the greater the probability of it becoming a cyst. Similar to teethwith large lesions, the age of the pathologic process may also helpexplain the higher prevalence of biofilms in association with cysts.
Extraradicular bacterial biofilms were found in only six speci-mens, and except for one case they were always associated with intra-radicular biofilms. This low prevalence of extraradicular biofilms is inaccordance with previous studies (10, 31). These findings also suggest

Figure 8. Maxillary premolar with clinically necrotic pulp and a sinus tract buccallylesion on the radiograph was >5 mm (inset). (a and b) After extraction, calculus isnot remain attached to the root tip and was removed separately. (c) The section taexternal surface is covered by a bacterial biofilm (Taylor’s modified Brown and Brenby the upper left arrow in c. Biofilm with high bacterial density (original magnificatioin c. A dense biofilm with a prevalence of filamentous morphotypes. Note the area(original magnification �1,000). (f) Higher magnification of the area from the extalized with relatively few bacteria (original magnification �1,000).


that extraradicular biofilms are usually maintained by intraradicularinfection. All cases showing an extraradicular biofilm exhibitedclinical symptoms, and three of them were associated with sinustracts. In abscesses, individual bacterial cells were seen within theinflamed periradicular tissue and commonly being phagocytosed byPMNs (Fig. 4a). These findings indicate that extraradicular infectionsin the form of biofilms or planktonic bacteria are not a common occur-rence, are usually dependent on the intraradicular infection, and aremore frequent in symptomatic teeth.

Similar to other studies, bacteria were also seen in the lumen of themain canal, ramifications, and isthmuses as flocs and planktonic cells,either intermixed with necrotic pulp tissue or possibly suspended ina fluid phase. Bacterial flocs in clinical specimens may have originated

. No periodontal pockets were disclosed at probing. The largest diameter of theobserved covering exclusively the root apex. The apical periodontitis lesion didken on a mesiodistal plane not passing through the main foramen. The apicaln, original magnification�16). (d) Higher magnification of the area indicatedn�400). (e) Higher magnification of the area indicated by the lower left arrowapparently free of bacterial cells, which may be likely a focus of calcification

ernal radicular profile indicated by the right arrow in c. The biofilm is miner-


Clinical Research
from the growth of cell aggregates in a fluid or they may have detachedfrom biofilms (24). Flocs may exhibit many of the same characteristicsas biofilms (22) and along with planktonic bacteria have been sug-gested to play a role in the pathogenesis of acute clinical forms of apicalperiodontitis (5).
The ability of endodontic bacteria to organize themselves in bio-film communities is of great therapeutic interest in endodontics.Although bacteria present as flocs and planktonic cells in the mainroot canal may be easily accessed and eliminated by instruments andsubstances used during treatment, those organized in biofilms attachedto the canal walls or located into isthmuses and ramifications are defi-nitely more difficult to reach. Many bacteria under the biofilm were seeninvading dentinal tubules (Fig. 6), which also pose a problem for disin-fection. Some biofilm-covered walls of the main canal may remainuntouched by instruments, which is especially true when the root canalis irregular, flattened, or oval in cross-section (32–34) (Fig. 2e and f).Biofilm remnants were observed on the root canal walls of treated teethin the present study. This study confirmed that isthmuses, lateral canals,and apical ramifications can be clogged with bacteria, including intreated teeth (13–15). These areas are not expected to be reachedby instruments and antimicrobial irrigants. Even in the event thatantimicrobial agents reach the biofilm, this is no guarantee ofsuccessful antimicrobial activity because bacteria arranged inbiofilms exhibit increased resistance to antimicrobials (35, 36).

Biofilms were classified morphologically as described by Hall-Stoodley et al (22) in a comprehensive review on the subject. A verysimilar definition is provided by Costerton in his ‘‘biofilm primer’’(1).‘‘A biofilm is a multicellular community composed of prokaryotic and/or eukaryotic cells embedded in a matrix composed, at least partially, ofmaterial synthesized by the sessile cells in the community.’’ There areobviously some features associated with biofilms such as differentialgenetic expression and the presence of water and nutrient channelsin the matrix that could not be evaluated by the method used in thisstudy. However, biofilms significantly differ in structure according tothe overall physical, chemical, and biological features of the environ-ment (1, 37–39). For instance, in an environment where there is lowshear force related to the passage of fluid or air, a strong adhesion tothe surface is not made necessary. Promoting such adhesionwould represent energy waste for the community. Therefore, ourmorphologic findings support the inclusion of apical periodontitis inthe set of biofilm-induced diseases. Further studies are required tocompare the main structural and physiologic features of endodonticbiofilms to other biofilms in nature.

It is reasonable to surmise that the unique root canal environ-mental conditions are expected to influence the biofilm structure andfunction to the point of giving rise to endodontic biofilms with typicalfeatures. However, although limited in resolution power, our studyfailed to show any specific morphological pattern for endodontic bio-films. Actually, endodontic biofilm morphology differed consistentlyfrom individual to individual, and the reasons for that deserve furtherinvestigations but may be conceivably related to different speciescomposition and resulting interactions, type and availability of nutri-ents, and time of infection.

Although a very high prevalence of biofilms was observed in teethwith apical periodontitis, the possibility exists that the figures reportedin this study still represent an underestimation. Some root canalcontents may have been washed away during histological processingbecause of the numerous chemical solutions, and gram-negativebacteria may sometimes be overlooked by the method used. Evenconsidering these limitations, the very high prevalence of biofilms as re-ported in this study indicates that when properly and meticulously per-formed, conventional paraffin techniques and so-called ‘‘old’’ bacterial


staining protocols are still valuable tools for studying bacterial coloni-zation of the root canal system.

In conclusion, the present study revealed a very high prevalence ofbacterial biofilms in the apical root canals of both untreated and treatedteeth with apical periodontitis. The pattern of bacterial communityarrangement in the canal, which adhered to or at least was associatedwith the dentinal walls with cells encased in an extracellular amorphousmatrix and often surrounded by inflammatory cells, is consistent withacceptable criteria to include apical periodontitis in the set ofbiofilm-induced disease. Biofilm morphologic structure varied fromcase to case, and no unique pattern for endodontic infections was deter-mined. Bacterial biofilms are still more expected to be present in asso-ciation with longstanding pathologic processes, including large lesionsand cysts.

References1. Costerton JW. The biofilm primer. Berlin, Heidelberg: Springer-Verlag; 2007.2. Marsh PD. Dental plaque: biological significance of a biofilm and community life-

style. J Clin Periodontol 2005;32(suppl 6):7–15.3. Costerton B. Microbial ecology comes of age and joins the general ecology commu-

nity. Proc Natl Acad Sci U S A 2004;101:16983–4.4. Chavez de Paz LE. Redefining the persistent infection in root canals: possible role of

biofilm communities. J Endod 2007;33:652–62.5. Siqueira JF Jr, Rocas IN. Community as the unit of pathogenicity: an emerging

concept as to the microbial pathogenesis of apical periodontitis. Oral Surg OralMed Oral Pathol Oral Radiol Endod 2009;107:870–8.

6. Svensater G, Bergenholtz G. Biofilms in endodontic infections. Endod Topics 2004;9:27–36.

7. Nair PNR. Light and electron microscopic studies of root canal flora and periapicallesions. J Endod 1987;13:29–39.

8. Siqueira JF Jr, Rocas IN, Lopes HP. Patterns of microbial colonization in primaryroot canal infections. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;93:174–8.

9. Molven O, Olsen I, Kerekes K. Scanning electron microscopy of bacteria in the apicalpart of root canals in permanent teeth with periapical lesions. Endod Dent Trauma-tol 1991;7:226–9.

10. Ricucci D, Siqueira JF Jr, Bate AL, et al. Histologic investigation of root canal-treatedteeth with apical periodontitis: a retrospective study from twenty-four patients. J En-dod 2009;35:493–502.

11. Carr GB, Schwartz RS, Schaudinn C, et al. Ultrastructural examination of failed molarretreatment with secondary apical periodontitis: an examination of endodontic bio-films in an endodontic retreatment failure. J Endod 2009;35:1303–9.

12. Schaudinn C, Carr G, Gorur A, et al. Imaging of endodontic biofilms by combinedmicroscopy (FISH/cLSM - SEM). J Microsc 2009;235:124–7.

13. Nair PN, Henry S, Cano V, et al. Microbial status of apical root canal system of humanmandibular first molars with primary apical periodontitis after ‘‘one-visit’’endodontic treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;99:231–52.

14. Ricucci D, Siqueira JF Jr. Apical actinomycosis as a continuum of intraradicular andextraradicular infection: case report and critical review on its involvement with treat-ment failure. J Endod 2008;34:1124–9.

15. Ricucci D, Siqueira JF Jr. Fate of the tissue in lateral canals and apical ramificationsin response to pathologic conditions and treatment procedures. J Endod 2010;36:1–15.

16. Tronstad L, Barnett F, Cervone F. Periapical bacterial plaque in teeth refractory toendodontic treatment. Endod Dent Traumatol 1990;6:73–7.

17. Ferreira FB, Ferreira AL, Gomes BP, et al. Resolution of persistent periapical infec-tion by endodontic surgery. Int Endod J 2004;37:61–9.

18. Ricucci D, Pascon EA, Pitt Ford TR, et al. Epithelium and bacteria in periapicallesions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;101:239–49.

19. Ricucci D, Mannocci F, Pitt Ford TR. A study of periapical lesions correlating thepresence of a radiopaque lamina with histological findings. Oral Surg Oral MedOral Pathol Oral Radiol Endod 2006;101:389–94.

20. Taylor RD. Modification of the Brown and Brenn Gram stain for the differentialstaining of gram-positive and gram-negative bacteria in tissue sections. Am J ClinPathol 1966;46:472–6.

21. Ricucci D, Bergenholtz G. Bacterial status in root-filled teeth exposed to the oralenvironment by loss of restoration and fracture or caries—a histobacteriologicalstudy of treated cases. Int Endod J 2003;36:787–802.

22. Hall-Stoodley L, Costerton JW, Stoodley P. Bacterial biofilms: from the natural envi-ronment to infectious diseases. Nat Rev Microbiol 2004;2:95–108.


Clinical Research
23. Parsek MR, Singh PK. Bacterial biofilms: an emerging link to disease pathogenesis.
Annu Rev Microbiol 2003;57:677–701.24. Hall-Stoodley L, Stoodley P. Evolving concepts in biofilm infections. Cell Microbiol

2009;11:1034–43.25. Baumgartner JC, Xia T. Antibiotic susceptibility of bacteria associated with

endodontic abscesses. J Endod 2003;29:44–7.26. Siqueira JF Jr. Reaction of periradicular tissues to root canal treatment: benefits and

drawbacks. Endod Topics 2005;10:123–47.27. Sundqvist G. Bacteriological studies of necrotic dental pulps [Odontological Disser-

tation no.7]. Umea, Sweden: University of Umea; 1976.28. Rocas IN, Siqueira JF Jr. Root canal microbiota of teeth with chronic apical perio-

dontitis. J Clin Microbiol 2008;46:3599–606.29. Strindberg LZ. The dependence of the results of pulp therapy on certain factors. Acta

Odontol Scand 1956;14(suppl 21):1–175.30. Lin LM, Ricucci D, Lin J, et al. Nonsurgical root canal therapy of large cyst-like

inflammatory periapical lesions and inflammatory apical cysts. J Endod 2009;35:607–15.

31. Siqueira JF Jr, Lopes HP. Bacteria on the apical root surfaces of untreated teeth withperiradicular lesions: a scanning electron microscopy study. Int Endod J 2001;34:216–20.


32. Siqueira JF Jr, Araujo MC, Garcia PF, et al. Histological evaluation of the effectivenessof five instrumentation techniques for cleaning the apical third of root canals. J En-dod 1997;23:499–502.

33. Wu M-K, van der Sluis LWM, Wesselink PR. The capability of two hand instrumen-tation techniques to remove the inner layer of dentine in oval canals. Int Endod J2003;36:218–24.

34. Paque F, Balmer M, Attin T, et al. Preparation of oval-shaped root canals in mandib-ular molars using nickel-titanium rotary instruments: a micro-computed tomog-raphy study. J Endod 2010;36:703–7.

35. Mah TF, O’Toole GA. Mechanisms of biofilm resistance to antimicrobial agents.Trends Microbiol 2001;9:34–9.

36. Chavez de Paz LE, Bergenholtz G, Svensater G. The effects of antimicrobials onendodontic biofilm bacteria. J Endod 2010;36:70–7.

37. Stoodley P, Dodds I, Boyle JD, et al. Influence of hydrodynamics and nutrients onbiofilm structure. J Appl Microbiol 1999;85:S19–28.

38. Purevdorj B, Costerton JW, Stoodley P. Influence of hydrodynamics and cellsignaling on the structure and behavior of Pseudomonas aeruginosa biofilms.Appl Environ Microbiol 2002;68:4457–64.

39. Paramonova E, Kalmykowa OJ, van der Mei HC, et al. Impact of hydrodynamics onoral biofilm strength. J Dent Res 2009;88:922–6.


Clinical Research

The Operating Microscope Enhances Detection andNegotiation of Accessory Mesial Canals in MandibularMolarsMeric Karapinar-Kazandag, DDS, PHD,* Bettina R. Basrani, DDS, PhD,

†

and Shimon Friedman, DMD†

Abstract
Introduction: Detection and negotiation of accessorymesial canals in mandibular molars was investigatedwith the aid of magnifying loupes or the operating micro-scope. Methods: First and second mandibular molars (n= 96) were mounted in mannequins. Three independentinvestigators (endodontists) prepared access cavitiesusing 4.5� loupes, attempting to detect and negotiateaccessory mesial canals with ultrasonic instruments. Ifdetection or negotiation was unsuccessful, the proce-dure was continued using the microscope. The locationof accessory mesial canals was mapped in relation to themain mesial canals, and their pathway shown with in-serted files. The mesial roots were cross-sectioned atthree levels to inspect for nonnegotiated accessorymesial canals. Results: With the microscope, thenumber of detected accessory mesial canals increasedfrom 8 (16%) to 9 (18%) in first molars and from 8(16%) to 11 (22%) in second molars. Negotiated acces-sory mesial canals increased from 6 (12%) to 7 (14%)and from 5 (10%) to 9 (18%) in the first and secondmolars, respectively. All 20 detected accessory mesialcanals were located in the mesial subpulpal groove,closer to the mesiolingual canal (45%), in the middle(30%), or closer to the mesiobuccal canal (25%). Allnegotiated accessory mesial canals merged with oneof the main two canals. Cross-sections of the rootsconfirmed that no accessory canals were present in addi-tion to those negotiated. Conclusions: Within the limi-tations of this study, more accessory canals weredetected and negotiated when using the microscopecompared with loupes. This improvement was morepronounced in second molars than in first molars. Allnegotiated accessory canals merged with either one ofthe main mesial canals. (J Endod 2010;36:1289–1294)
Key WordsAccessory canals, magnifying loupes, mandibularmolars, operating microscope

From the *Department of Endodontics Faculty of Dentistry, YediToronto, Toronto, Ontario, Canada.

Address requests for reprints to Dr Meric Karapinar-Kazandag,Goztepe/Istanbul/Turkey. E-mail address: [email protected]/$0 - see front matter

Copyright ª 2010 American Association of Endodontists.doi:10.1016/j.joen.2010.04.005


The recent influx of current technologies intended for endodontic treatment has beenfocused largely on improving the quality of treatment. The primary example of this

trend has been the introduction to endodontics of the operating microscope, widelyaccepted as a beneficial aid in improving clinicians’ ability to detect root canals (1,2), particularly in teeth in which accessory canals are present. Several investigationshave supported the advantage of the microscope over the use of no magnification(3–9). When compared with magnifying loupes, the microscope was eithercomparable (9) or superior (7). For the test model, researchers have used thefrequently present but often elusive accessory canal in the mesiobuccal (MB) root ofmaxillary molars (4, 6–10). According to Gorduysus et al (6), clinicians did not detectmore accessory canals in maxillary molars with the aid of the microscope, but theirability to negotiate the canals improved by over 10% when compared with no magni-fication.

One of the teeth with a complex root canal system is the mandibular molar, asshown in the early work of Hess and Zurcher (11) and in subsequent investigations(5, 12–25). The mesial root in mandibular molars is commonly considered to havetwo canals (11–13) with an isthmus in between (14, 16, 21, 24, 26–28). Withinthis system, the presence of an accessory mesial canal has been identified witha prevalence ranging from 0% to 17% (5, 12–22, 24, 25) (Table 1). The discrepancybetween the studies has been attributed to ethnicity (29), age (28, 30), and sex (23).Although the location of the accessory mesial canal orifice has been reported closer tothe mesiolingual (ML) canal (5), its pathway converges with either the ML(15) canal orthe MB canal (17, 18). When explored with the aid of the operating microscope, anaccessory mesial canal was detected in 17% of first molars and under 5% of secondmolars compared with 0% when no magnification was used (5). Apart from the prev-alence, the ability to negotiate the detected accessory mesial canals, including thetroughing depth required to enable negotiation, has not been well characterized.

The primary purpose of this study was to assess the ability to detect and negotiatethe accessory mesial canals in mandibular first and second molars, with the aid ofmagnifying loupes and the microscope. The secondary goal was to characterize the de-tected canals with regard to prevalence, location, negotiability, and pathway.

MethodsThis study methodology was modeled after a previous study on maxillary molars

(6). Mandibular first and second molars were collected from oral surgery clinics inIstanbul, Turkey, and exposed on periapical radiographs in the buccolingual direction.After exclusion of molars with previous endodontic treatment, a deficient coronal

tepe University, Istanbul, Turkey; and †Discipline of Endodontics, Faculty of Dentistry, University of

Yeditepe University, Faculty of Dentistry, Department of Endodontics, Bagdat cad No:238 34728,

Accessory Mesial Canals in Mandibular Molars 1289


TABLE 1. Summary of Studies on the Prevalence of Accessory Mesial Canals in Mandibular Molars

Number of molars (n) Mesial accessory canals (%n)

Study Methodology First Second First Second

Skidmore & Bjorndal, 1971 Plastic casts 45 40 0 0Pineda & Kuttler, 1972 Radiography 300 300 0 0Pomeranz et al, 1981 Clinical 61 39 11.4 12.8Martinez-Berna, 1983 Clinical 1,418 944 1.3 0.2Vertucci, 1984 Clearing 100 100 1 0Fabra-Campos, 1985 Clinical 145 0 2.7 —Fabra-Campos, 1989 Clinical 760 0 2.6 —Goel, 1991 Clinical 60 0 15 —Caliskan et al, 1995 Clearing 100 100 3.4 1.9de Carvalho & Zuolo, 2000 Extracted teeth 93 111 17.2 4.5Gulabivala et al, 2001 Clearing 139 134 7.1 0Gulabivala et al, 2002 Clearing 118 60 5.9 1.7Sert & Bayirli, 2004 Clearing 200 200 1.5 0Ahmed et al, 2007 Clearing 100 100 4 0Navarro et al, 2007 Micro-CT 27 0 14.8 —

Scanning electron microscope 25 0 12 —

Clinical Research

structure nonamenable to conventional endodontic access cavity prep-aration, aberrant anatomy, calcified canals, fused roots, single roots,and C-shaped canals, 48 first and 48 second molars were selectedfor the study. They were stored in 0.1% thymol solution until used.

The teeth were embedded in dentaforms and mounted in manne-quins to simulate clinical conditions as best as possible. Conventionalendodontic access cavities were completed in all teeth with the aid of4.5�magnification loupes. Each of the first and second molars groupswas randomly divided into three subgroups (n = 18). Three endodon-tists were assigned a subgroup of first and second molars each. Workingindependently, they set out to explore the mesial root canals in anattempt to detect accessory mesial canals using a standardized sequenceas follows.

In the first stage, only loupes were used for magnification. Theaccess cavity was refined with ultrasonic tips (Buc 1 and Buc 3; Sybron

Figure 1. (A-C) Radiographs of mandibular molars exposed from the mesial dirmesial canals. Files were inserted into the MB, ML, and accessory canals. (D) The dthe pulp chamber floor as a reference.

1290 Karapinar-Kazandag et al.

Endo, Orange, CA) to remove any dentin overhanging the mesial canalorifices and the isthmus. The mesial subpulpal groove was exploredwith sharp endodontic explorers (DG 16; Hu-Friedy, Chicago, IL, andStewart probe; Premier Dental Products, Norristown, PA), and thenumber of canal orifices detected was recorded. Attempts were thenmade to negotiate detected accessory mesial canals with size 06 K-type hand files (Dentsply Maillefer, Balleigue, Switzerland). When nego-tiation was unsuccessful, the isthmus was troughed apically with theultrasonic tips to pursue the accessory canal deeper into the root whilerepeating negotiation attempts. Irrigation with 1% NaOCl and a Stropkoair irrigator (Sybron Endo) were used intermittently to optimize visi-bility. Troughing was continued apically until (1) the accessory mesialcanal was negotiated, (2) it was considered too risky to continuetroughing further apically, (3) the accessory mesial canal was no longerdetectable, or (4) a perforation occurred.

ection (the distal roots were resected) showing the pathway of the accessoryepth of dentin removal while troughing to negotiate the accessory canal using


Clinical Research
In the second stage, the teeth in which no accessory mesial canal
was detected and those in which a detected accessory mesial canalcould not be negotiated were submitted to further investigation aidedwith the microscope (Protege, Global Surgical Corp, St Louis, MO).Under the microscope, further dentin was selectively removed alongthe subpulpal groove with the ultrasonic tips while repeating negotiationattempts. Again, the results were recorded in the same four categoriesdefined previously.

In the third stage, a K-type file was negotiated into each of themesial canals as far apically as possible. The teeth were removedfrom the dentaforms and then exposed on periapical radiographsfrom the mesial direction to show the pathways of the mesial canals(Fig. 1).

In the fourth stage, the files were removed, and 1% sodium hypo-chlorite was left in the pulp chambers for 24 hours in order to remove anysoft-tissue remnants. The chambers were dried and subsequently photo-graphed with the aid of a stereomicroscope (Nikon, Tokyo, Japan). Thenumber of teeth having 1, 2 (Fig. 2A), or 3 (Fig. 2B and C) canals or anopen isthmus (Fig. 2D) was recorded. The location of the accessorymesial canals orifices in the photographs was mapped with the aid ofSpot Software (Diagnostic Instruments, Sterling Heights, MI).

Figure 2. Photographs and a schematic representation of the subpulpal groove ob(D) an open isthmus.


At the fifth stage, the distal roots of the teeth were resected, andradiographs of the teeth were exposed from the mesial aspect. Theseradiographs were digitized (Sony Cyber-shot, DSC-W80, Tokyo, Japan),and the depth of dentin removal was measured with Spot Software usingthe pulp chamber floor as reference (Fig. 1D).

Finally, the mesial roots of all 96 molars were sectioned at threelevels: 1 mm, 4 mm, and 8 mm from the apex. The sectioned surfacesat each level were examined under the operating microscope at 30�magnification to detect additional accessory mesial canals beyond thosedetected during the experimental procedure and to trace the pathway ofdetected accessory mesial canals relative to the main canals.

ResultsTables 2 and 3 summarize the results of the investigation in the 96

first and second mandibular molars, regarding the effect of magnifica-tion aids on detection and negotiation of accessory mesial canals, thelocation and pathways of the accessory canals and the depth of trough-ing required to negotiate or rule out the presence of the accessorycanals.

served in mandibular molars with (A) two canals, (B and C) three canals, and


TABLE 2. Numbers and Proportion of Detected and Negotiated Accessory Mesial Canals in Mandibular Molars

First molars (n = 48) Second molars (n = 48) Total (n = 96)

Magnification aid Detected Negotiated Detected Negotiated Detected Negotiated

Loupes (%) 8 (16) 6 (12) 8 (16) 5 (10) 16 (16) 11 (11)Microscope (%) 9 (18) 7 (14) 11 (22) 9 (18) 20 (20) 16 (16)

Loupes were used first as the magnification aid followed by the use of the operating microscope for teeth where the previous attempts were unsuccessful.

Clinical Research

Effect of Magnification AidsThe total number of accessory mesial canals detected and negoti-

ated was 4% higher in the second molars than in the first molars (Table2). In both tooth types, the use of the microscope improved the resultsachieved with the loupes. The total number of detected and negotiatedaccessory mesial canals increased from 16% to 20% and from 11% to16%, respectively, with 4 of 20 accessory mesial canals (20%) detectedbut not negotiated.

Location of Accessory Mesial Canal OrificesAll 20 accessory mesial canals were located in the mesial subpul-

pal groove. The distribution pattern along the groove was rather similarin the first and second molars, with the highest proportion of accessorymesial canals (45%) located closer to the ML canals, 30% located in themiddle between the MB and ML canals, and 25% located closer to theMB canals (Table 3).

The mean distance between the MB and ML canals in the first andsecond molars was 3.2 � 0.8 mm and 2.8 � 0.9 mm, respectively.When the accessory mesial canals were closer to either the ML or MBcanals, their distance from the closest main canal was greater in the firstmolars (1.0-1.3 mm) than in the second molars (0.8-0.9 mm)(Table 3).

Pathways of Negotiated Accessory Mesial CanalsAll 16 negotiated accessory mesial canals were confluent with the

main canals, with none terminating in an independent apical foramen(Table 3). The pathways of the accessory mesial canals differed betweenthe first and second molars. In the first molars, the majority (42%)merged with the MB canals, whereas in the second molars 55% mergedwith the ML canals. Cross-sections of the roots revealed only the nego-tiated accessory mesial canals at the 8-mm level, suggesting that allaccessory mesial canals present were successfully negotiated. None of

TABLE 3. Characteristics of Detected and Negotiated Accessory Mesial Canals (AM

Observation First molars

Location (%) n = 9 detectedCloser to MB 2 (22)Closer to ML 4 (44)At middle 3 (33)

Distance (mm)MB to ML 3.21 � 0.76AMC to MB* 1.00 � 0.28AMC to ML* 1.30 � 0.14

Pathway (%) n = 7 negotiatedMerged with MB 3 (42)Merged with ML 2 (29)Merged with both 2 (29)

Depth† (mm) 1.10 � 1.14

AMC, accessory mesial canals; MB, mesiobuccal; ML, mesiolingual. Location and pathway are related to t

*Teeth where the AMC was closest to either main canal.†Relative to the pulp chamber floor.


the negotiated accessory mesial canals were observed at 4 mm fromthe apex.

Depth of Troughs RequiredThe mean depth of dentin removed in order to negotiate or rule

out the presence of accessory mesial canals in the first and secondmolars was 1.1 � 1.4 mm and 0.7� 0.6 mm, respectively (Table 3).

DiscussionOne of the challenges facing clinicians when performing

endodontic treatment in molars is the complexity of the root canalsystems. Although the high-prevalence accessory mesiobuccal canalsin maxillary molars have been well characterized (4, 6–10), thelower-prevalence accessory mesial canals in mandibular molars(12–14, 20–24) are not well recognized by clinicians. The accessorymesial canals invariably originate within the subpulpal groove oristhmus connecting the two main canals (14, 16, 21, 24, 26–28),making their detection very challenging. This study was undertakento assess the potential to facilitate the detection and negotiation ofaccessory mesial canals in mandibular molars with the aid of theoperating microscope.

Teeth were mounted in dentaforms to simulate clinical conditions(6, 8). The detection of accessory mesial canals without magnificationaids, which was assessed in several previous studies (5, 6, 8, 9), was notattempted because currently magnification is considered indispensablewhen performing endodontic treatment (1, 29). Thus, 4.5� loupeswere used first to enhance magnification, followed by the use of theoperating microscope (up to 30� power) to assess its potentialadvantage. The access cavities were refined with ultrasonic tips toallow accurate, controlled removal of dentin along the mesialsubpulpal groove in search for accessory mesial canal orifices. Themesial dentinal protuberance was removed to afford a direct view of

Cs) in Mandibular Molars

Second molars Total

n = 11 detected n = 20 detected3 (27) 5 (25)5 (45) 9 (45)3 (27) 6 (30)

2.80 � 0.90 Mean 3.00 � 0.900.87 � 0.39 Mean 0.93 � 0.320.81 � 0.58 Mean 0.95 � 0.53

n = 9 negotiated n = 16 negotiated3 (33) 6 (38)5 (55) 7 (43)1(11) 3 (19)

0.66 � 0.57 0.93 � 0.94

he main canals in the mesial root, the MB, and the ML.


Clinical Research
the line angle between the mesial wall and the pulp chamber floor(17).The investigators were three endodontists with varying levels ofexperience to eliminate the individual’s skill as a dominant variable.Accessory mesial canals were characterized for proportion of success-ful detection (5) and negotiation (14), location and depth of the orifices(6), and pathway (18). The findings were confirmed for accuracy byradiographic examination (31) and cross-sectioning of the mesial roots(7, 8, 32).
With the aid of the microscope, the number of detected accessorymesial canals increased by 4% overall in all teeth, with a greaterimprovement in the second molars (6%) than in the first molars(2%). When the use of the microscope was compared with no magni-fication in a previous study (5), the number of detected accessorymesial canals increased by 17% in first molars and by almost 5% insecond molars. Also, in maxillary molars, the microscope facilitatedthe detection of MB2 canals (4, 6, 8, 9).

Our ability to negotiate accessory mesial canals with the aid ofloupes in 12% and 10% of first and second molars, respectively, wassimilar to that reported previously without any magnification (14).Importantly, the final microscope-aided negotiation of accessory mesialcanals in 14% of the first molars matched the 14.8% observed bycomputerized tomography (25), which is the ultimate method for inves-tigating root canal morphology. Also, as suggested earlier, cross-sections through the roots confirmed that no additional accessorymesial canals were present beyond the negotiated ones. The greaterbenefit of the microscope in the second molar was again apparent,with the 8% increase in negotiated accessory mesial canals, comparedwith the 2% increase in the first molars. The second molars have beenreported to have accessory mesial canals less frequently (5, 16, 20–24)and an isthmus more frequently (21, 22) than the first molars.However, our combined use of microendodontic instruments (4)and the microscope facilitated exploration of the isthmuses (33),resulting in additional accessory mesial canals negotiated in the secondmolars.

Pomeranz et al (14) classified the accessory mesial canals intothree categories: (1) ‘‘fin,’’ allowing free instrument movement betweenthe main and accessory canals (we did not consider such fins as acces-sory mesial canals); (2) ‘‘independent,’’ having a separate orifice andapical terminus, which is known to be rare (5, 12–25) and was notobserved in our sample; and (3) ‘‘confluent,’’ having a separateorifice but merging more apically with the MB or ML canals as inall accessory mesial canals observed in our sample (15, 17, 18).Although also relating detected and negotiated accessory mesialcanals to this classification, we aimed primarily to map the locationof accessory mesial canals in relation to the main mesial canals (MBand ML). The main purpose of such mapping was to provideclinicians with a navigation guide for detection of accessory mesialcanals, such as previously reported for maxillary molars (6). All ofthe accessory mesial canals were located in the mesial subpulpal groove(5, 25, 33). In about 45% of both the first and second molars, theaccessory mesial canals were detected closer to the ML canals, whichis in agreement with previous reports (5). The mean distance fromthe ML orifice was longer in the first molars (1.3 mm) than in thesecond molars (0.8 mm), reflecting the larger buccal-lingual dimen-sion of the pulp chamber in the former. Thus, clinicians should specif-ically search for accessory mesial canals starting from the ML canalorifice and progress systematically along the subpulpal groove towardsthe MB canal.

Frequently, there was the need to explore the mesial subpulpalgroove by troughing in the apical direction to allow negotiation of a de-tected accessory mesial canal with an endodontic instrument. The meandepth of the troughs in the first molars reached 1.1 mm compared with


approximately 0.7 mm in the second molars. Although in several teeththe troughs were as deep as 2.2 mm, we found them less deep than whatis occasionally required in maxillary molars when accessory MB canalsare negotiated (6). In the maxillary molars, the accessory MB canalsdepart the chamber at a sharp mesial inclination and then bend againdistally, which makes their negotiation challenging (6); the deepertrough helps eliminate the first angled portion of the canal, allowinginsertion of instruments beyond the bend (6). In the mandibularmolars, once the mesial dentinal protuberance is eliminated, the acces-sory mesial canals are not strongly inclined mesially, and they do notbend distally after departing the chamber. Thus, a shallow trough issufficient to allow insertion of instruments into the accessory mesialcanals.

All the negotiated accessory mesial canals in this study wereconfluent with one of the main mesial canals, but the confluence patterndiffered between the first and second molars. In the former, the acces-sory mesial canals frequently crossed the midline and merged with theMB canals (17, 18), whereas in the latter they more frequently mergedwith the ML canals (15). The cross-sections through the rootsconfirmed that all negotiated accessory mesial canals were no longerobserved at 4 mm from the apex, having blended with either the isthmusor one of the main canals. This finding supported the argument that theaccessory mesial canals are not additional canals but rather an accessinto the isthmus between the main canals (33).

The isthmus is a characteristic feature of the mesial roots ofmandibular molars found in 17% to 83% of first molars, mostfrequently at the level of 3 to 6 mm from the apical foramen(26–28). It has been defined as a lateral interconnection (13), trans-verse anastomosis (12, 16), or intercanal communication (32, 21). Itcan be complete, forming a continuous connection between the maincanals, or partial, forming an incomplete or narrow communicationwith one or more patent openings (32). Despite current advances inirrigation delivery, disinfection of the isthmus is an elusive goal (34,35). The accessory mesial canals, when negotiated and shaped,comprise a pathway into the isthmus (33), which may provide accessfor irrigation solutions and root filling materials. This appears to bethe main clinical significance of detecting and attempting to negotiateaccessory mesial canals in the mandibular molars.

In conclusion, within its limitations, this study suggested that use ofthe operating microscope enhanced both the detection and negotiationof accessory mesial canals in the mandibular first and second molarsbeyond what could be achieved with the aid of loupes. This benefitwas more pronounced in the second molars than in the first molars,resulting in an 8% increase in the number of negotiated accessorycanals. Not all detected accessory canals could be negotiated. Althoughaccessory canals were negotiated in as many as 14% of first molars and18% of second molars, all of these canals merged with either one of themain mesial canals. The importance of negotiating the accessory mesialcanals in mandibular molars is in the access it provides for irrigationsolutions and filling materials into the otherwise inaccessible isthmus.

References1. Kersten DD, Mines P, Sweet M. Use of the microscope in endodontics: results of

a questionnaire. J Endod 2008;34:804–7.2. Lee M, Winkler J, Hartwell G, et al. Current trends in endodontic practice: emer-

gency treatments and technological armamentarium. J Endod 2009;35:35–9.3. Saunders WP, Saunders EM. Conventional endodontics and the operating micro-

scope. Dent Clin North Am 1997;41:415–28.4. Stropko JJ. Canal morphology of maxillary molars: clinical observations of canal

configurations. J Endod 1999;25:446–50.5. de Carvalho MC, Zuolo ML. Orifice locating with a microscope. J Endod 2000;26:

532–4.


Clinical Research
6. Gorduysus MO, Gorduysus M, Friedman S. Operating microscope improves negoti-
ation of second mesiobuccal canals in maxillary molars. J Endod 2001;27:683–6.7. Schwarze T, Baethge C, Stecher T, et al. Identification of second canals in the me-

siobuccal root of maxillary first and second molars using magnifying loupes or anoperating microscope. Aust Endod J 2002;28:57–60.

8. Baldassari-Cruz LA, Lilly JP, Rivera EM. The influence of dental operating micro-scope in locating the mesiolingual canal orifice. Oral Surg Oral Med Oral PatholOral Radiol Endod 2002;93:190–4.

9. Buhrley LJ, Barrows MJ, BeGole EA, et al. Effect of magnification on locating theMB2 canal in maxillary molars. J Endod 2002;28:324–7.

10. Sempira HN, Hartwell GR. Frequency of second mesiobuccal canals in maxillarymolars as determined by use of an operating microscope: a clinical study. J Endod2000;26:673–4.

11. Hess W, Zurcher E. The anatomy of the root canals of the teeth of the permanent anddeciduous dentitions. New York: William Wood & Co; 1925.

12. Skidmore AE, Bjorndal AM. Root canal morphology of the human mandibular firstmolar. Oral Surg Oral Med Oral Pathol 1971;32:778–84.

13. Pineda F, Kuttler Y. Mesiodistal and buccolingual roentgenographic investigation of7,275 root canals. Oral Surg Oral Med Oral Pathol 1972;33:101–10.

14. Pomeranz HH, Eidelman DL, Goldberg MG. Treatment considerations of the middlemesial canal of mandibular first and second molars. J Endod 1981;7:565–8.

15. Martinez-Berna A, Badanelli P. Investigacion clinica de molars inferiors con cincoconductos. Boletin de Informacion Dental 1983;43:27–41.

16. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral MedOral Pathol 1984;58:589–99.

17. Fabra-Campos H. Unusual root anatomy of mandibular first molars. J Endod 1985;11:568–72.

18. Fabra-Campos H. Three canals in the mesial root of mandibular first permanentmolars: a clinical study. Int Endod J 1989;22:39–43.

19. Goel N, Gill K, Taneja J. Study of root canals configuration in mandibular first perma-nent molar. J Indian Soc Pedod Prev Dent 1991;8:12–4.

20. Caliskan MK, Pehlivan Y, Sepetcioglu F, et al. Root canal morphology of humanpermanent teeth in a Turkish population. J Endod 1995;21:200–4.

21. Gulabivala K, Aung TH, Alavi A, et al. Root and canal morphology of Burmesemandibular molars. Int Endod J 2001;34:359–70.


22. Gulabivala K, Opasanon A, Ng YL, et al. Root and canal morphology of Thai mandib-ular molars. Int Endod J 2002;35:56–62.

23. Sert S, Bayirli GS. Evaluation of the root canal configurations of the mandibular andmaxillary permanent teeth by gender in the Turkish population. J Endod 2004;30:391–8.

24. Ahmed HA, Abu-bakr NH, Yahia NA, et al. Root and canal morphology of permanentmandibular molars in a Sudanese population. Int Endod J 2007;40:766–71.

25. Navarro LF, Luzi A, Garcia AA, et al. Third canal in the mesial root of perma-nent mandibular first molars: review of the literature and presentation of 3clinical reports and 2 in vitro studies. Med Oral Patol Oral Cir Bucal 2007;12:E605–9.

26. Mannocci F, Peru M, Sherriff M, et al. The isthmuses of the mesial root ofmandibular molars: a micro-computed tomographic study. Int Endod J 2005;38:558–63.

27. von Arx T. Frequency and type of canal isthmuses in first molars detected by endo-scopic inspection during periradicular surgery. Int Endod J 2005;38:160–8.

28. Gu L, Wei X, Ling J, et al. A microcomputed tomographic study of canal isthmuses inthe mesial root of mandibular first molars in a Chinese population. J Endod 2009;35:353–6.

29. Vertucci F. Root canal morphology and its relationship to endodontic procedures.Endod Topics 2005;10:3–29.

30. Peiris HR, Pitakotuwage TN, Takahashi M, et al. Root canal morphology of mandib-ular permanent molars at different ages. Int Endod J 2008;41:828–35.

31. Kulild JC, Peters DD. Incidence and configuration of canal systems in the mesiobuc-cal root of maxillary first and second molars. J Endod 1990;16:311–7.

32. Weller RN, Niemczyk SP, Kim S. Incidence and position of the canal isthmus. Part 1.Mesiobuccal root of the maxillary first molar. J Endod 1995;21:380–3.

33. Mortman RE, Ahn S. Mandibular first molars with three mesial canals. Gen Dent2003;51:549–51.

34. Burleson A, Nusstein J, Reader A, et al. The in vivo evaluation of hand/rotary/ultra-sound instrumentation in necrotic, human mandibular molars. J Endod 2007;33:782–7.

35. Carr GB, Schwartz RS, Schaudinn C, et al. Ultrastructural examination of failed molarretreatment with secondary apical periodontitis: an examination of endodontic bio-films in an endodontic retreatment failure. J Endod 2009;35:1303–9.


Clinical Research

Postoperative Pain after the Application of Two DifferentIrrigation Devices in a Prospective Randomized Clinical TrialEudes Gondim Jr., DDS, MS, PhD,* Frank C. Setzer, DMD, PhD, MS,*

Carla Bertelli dos Carmo, DDS,†

and Syngcuk Kim, DDS, PhD*

Abstract
Introduction: The extrusion of irrigation solutionsbeyond the apical constriction may result in postoperativepain. Sodium hypochlorite can cause severe tissue irrita-tion and necrosis outside the root canal system if extrudedinto the periodontal ligament (PDL) space. Differentdelivery techniques were discussed to reduce this poten-tial risk. The aim of this study was to compare the post-operative level of pain after root canal therapy usingeither endodontic needle irrigation or a negative apicalpressuredevice. Material and Methods: In a prospectiverandomized clinical trial, 110 asymptomatic single-rootedanterior and premolar teeth were treated endodonticallywith two different irrigation techniques. The teeth wererandomly assigned to two groups. In the MP group(n = 55), procedures were performed using anendodontic irrigating syringe (Max-i-Probe; DentsplyRinn, Elgin, IL). The EV group (n = 55) used an irrigationdevice based on negative apical pressure (EndoVac;Discus Dental, Culver City, CA). Postoperatively, thepatients were prescribed ibuprofen 200 mg to take every8 hours if required. Pain levels were assessed by ananalog scale questionnaire after 4, 24, and 48 hours.The amount of ibuprofen taken was recorded at thesame time intervals. Results: During the 0- to 4-, 4- to24-, and 24- to 48-hour intervals after treatment, thepain experience with the negative apical pressure devicewas significantly lower than when using the needle irriga-tion (p < 0.0001 [4, 24, 48 hours]). Between 0 and 4 and4 and 24 hours, the intake of analgesics was significantlylower in the group treated by the negative apical pressuredevice (p < 0.0001 [0-4 hours], p = 0.001 [4-24 hours]).The difference for the 24- to 48-hour period was notstatistically different (p = 0.08). The Pearson correlationcoefficient revealed a strongly positive and significantrelationship for the MP group (r = 0.851, p < 0.001)and the EV group (r = 0.596, p < 0.0001) betweenpain intensity and the amount of analgesics. Conclusion:The outcome of this investigation indicates that the use ofa negative apical pressure irrigation device can result in
From the *Department of Endodontics, School of Dental MedicSurgeons, Sao Paulo, Brazil.

Address requests for reprints to Dr Frank C. Setzer, Departmentphia, PA 19104. E-mail address: [email protected]/$0 - see front matter



a significant reduction of postoperative pain levels in comparison to conventional needleirrigation. (J Endod 2010;36:1295–1301)

Key WordsEndoVac, irrigation, negative apical pressure, postoperative pain

Postoperative pain is an unwanted yet unfortunately common sensation afterendodontic treatment. The incidence of postoperative pain was reported to range

from 3% to 58% (1). Even severe pain may occur within 24 to 48 hours after therapy(2). After the treatment was finished, 12% of patients experienced severe pain withinthis time interval according to a visual analog scale (VAS) (2). The factors for postop-erative pain are many-fold and can include microbial factors, the effects of chemicalmediators, phenomena related to the immune system, cyclic nucleotide changes,psychological factors, and changes in the local adaptation and the periapical tissuepressure (3). Irritants to the periapical tissues that can evoke pain sensation includemedications or irrigating solutions (3).

Antimicrobial debridement is a key step in root canal therapy. Bacteria playa primary role in the development of pulp necrosis, periapical pathosis, and posttreat-ment disease (4). Mechanical instrumentation alone is not enough to render canals freefrom microorganisms (5). Several studies have proven the effectiveness of sodiumhypochlorite for bacterial reduction in addition to mechanical cleaning and shaping(6). Other irrigants with similar antimicrobial effects include chlorhexidine (7) andMTAD (8). Only sodium hypochlorite, however, has also proven highly effective in tissuedissolution (9) and the removal of bacterial biofilm (10). Because tissue dissolution isa prerequisite for antimicrobial action (11), sodium hypochlorite is considered themost important antimicrobial irrigant in root canal therapy (9). Sodium hypochloriteworks because of its ability to hydrolyze and oxidize cell proteins, its release of freechlorine, and its pH of 11 to 12 (7).

Because of the strong cell toxicity, an associated risk with the use of sodium hypo-chlorite is the inadvertent injection into the periapical tissues through the apicalconstriction of the root canal, leading to severe, painful postoperative complications.Sodium hypochlorite accidents have been reported in the literature (12). Teeth withwide open foramina or with apical constrictions damaged by resorptive processes orby iatrogenic errors during instrumentation are at an elevated risk for the extrusionof sodium hypochlorite (13). Moreover, if excessive pressure is used during irrigationor the irrigation needle is bound within the root canal and prevents the safe coronaloutflow of the solution, large quantities of sodium hypochlorite may be pushed outinto the periapical tissues and subsequently lead to tissue necrosis and postoperativepain (13). This causes a dilemma because it is known that a high volume and frequency

ine, University of Pennsylvania, Philadelphia, PA, USA; and the †Sao Paulo Association of Dental

of Endodontics, School of Dental Medicine, University of Pennsylvania, 240 S 40th Street, Philadel-

Post-Operative Pain after Application of Two Irrigation Devices 1295


Clinical Research
of irrigation (14) as well as the ability to reach the apical intraradiculartissues (15) are necessary for effective disinfection.
To prevent periapical tissue damage and lessen postoperativepain, a safe irrigation delivery system is desirable. Commonly, hypo-dermic or endodontic needles are used for irrigation. Recently,a new irrigation system, the EndoVac system (Discus Dental, CulverCity, CA), was introduced to endodontics. Conventional irrigation workswith positive pressure to flush the disinfecting solution into the rootcanal and force the irrigant out again coronally by displacement withnew volumes of solution. The EndoVac system works with negative pres-sure. A detailed design and working mechanism have been describedbefore (16). Briefly, an irrigation tip is attached to a conventionalmedical syringe containing the solution. Through this tip, irrigant isreleased into the pulp chamber. Overflow is prevented by a suctiontip that is directly attached to the delivery tip and connects to thehigh-speed suction of the dental unit. A second tube, connected tothe high-speed suction, is used for the attachment of cannulas of varyingdiameter for different levels of irrigation within the root canal. A stain-less steel microcannula of size #32 with 12 small, lateral holes is usedfor the apical 0 to 3 mm. The tip is inserted to the working length andprovides a constant flow of new irrigation solution to the apical third bysucking it apically from the fresh reservoir in the pulp chamber anddisposing the used solution through the evacuation tube toward thehigh-speed suction of the dental unit.

In three recent in vitro studies, the EndoVac system showed signif-icantly better apical debridement (17) and an equal performance inantimicrobial disinfection (18, 19) in single straight canals whencompared with other irrigation techniques. Yet, no literature existsclaiming whether the irrigation with a negative apical pressure deviceprovides more or less favorable results in terms of postoperative painwhen compared with positive-pressure irrigation protocols. Thepurpose of this study was to evaluate and compare the postoperativepain after the use of two different irrigation protocols.

Materials and MethodsIn this prospective randomized clinical trial, single-visit root canal

treatments were performed. A questionnaire was given to the partici-pants to note the amount of analgesics taken postoperatively as wellas the intensity of pain. A pain scale frequently used for medical studies,the CR10 Borg list (20), was implemented to quantify the participants’individual pain experience.

Patient SelectionEighty volunteer patients with 110 teeth fitting the inclusion criteria

described later were included in this study. All patients were treated bya single operator in a private practice specializing in endodontics overa period of 25 months. Only single-rooted teeth with one canal wereselected for this investigation. Diagnoses were either asymptomatic irre-versible pulpitis caused by carious exposures or normal pulp if thepatient had been referred for intentional endodontic therapy for pros-thetic reasons. The individual diagnosis was confirmed by obtaining thedental history, periradicular radiographs, periodontal evaluation,percussion, and cold test (EndoIce; Coltene/Whaledent Inc, CuyahogaFalls, OH). The diagnostic findings were verified by comparing themwith adjacent sound teeth with vital pulps. Only patients who hada noncontributory medical history and did not take analgesic medica-tion at the initiation of the root canal treatment were asked to participatein the study. The treatment and the study design were explained to thequalifying patients. Patients were informed that participation was volun-tary and did not affect the treatment. All patients who agreed to partic-ipate in this study signed an informed consent. Although the patients

1296 Gondim Jr. et al.

were informed which irrigation devices were used in general, therewas no information for the participant which system was used for theparticular treatment.

Randomized Selection of Irrigation DeviceThe goal of the study was 100 patients, with at least 50 procedures

in each group. In order to compensate for a possible dropout rate of10%, the prospective sample size for each group was set at 55. To ensurerandomization of the process, 55 red and 55 green chips were placed ina bag at the beginning of the investigation. Before each treatment,a dental assistant of the operator randomly determined the irrigationdevice by taking out one of the colored chips without replacement untilall 110 procedures had been performed. The assistant could not see thecolor of the chip before it was removed from the bag. Group MP (red)was assigned for treatment with a conventional endodontic needlesyringe (Max-i-Probe 30G; Dentsply Rinn, Elgin, IL). Group EV (green)received treatment with the negative-pressure device (EndoVac).

Endodontic ProtocolAll patients received a topical anesthetic (Benzotop; DFL, Rio de

Janeiro, Brazil) before infiltration. Local anesthesia was achieved bylocal infiltration with 3.6 mL of lidocaine with 1:100,000 epinephrine(Alphacaine, DFL). After anesthesia, a rubber dam was placed and dis-infected with 3% hydrogen peroxide, and the tooth was accessed usingsterile carbide burs under a dental operating microscope. In cases withdeep carious lesions, the main decay was excavated before accessing thepulp to prevent the introduction of microorganisms into the root canalsystem. A glide path was established with stainless steel hand instrumentsup to a size #15. The canal was instrumented with Gates Glidden burs #4,#3, and #2 (Dentsply Maillefer, Ballaigues, Switzerland) followed bynickel-titanium rotary instruments (ProTaper; Dentsply Tulsa, JohnsonCity, TN). Patency was established and verified with #10 files. Theworking length to the apical constriction was confirmed by an electronicapex locator (Root ZX; Morita, Irvine, CA) and periapical radiographs.The established working length was checked repeatedly throughout theprocedure. Depending on the individual tooth, the final instrumentationsize was determined as three sizes larger than the first file binding at theworking length. Final preparation ended either with ProTaper F3, F4, F5,or F5 plus additional apical enlargement with nickel-titanium handinstruments to size #60. A smaller taper #35 ISO nickel-titanium handinstrument was used for the F3 preparations in group EV to verify freeaccess to the full working length for the microcannula. All teeth wereobturated in the same session using gutta-percha with warm verticalcompaction in the continuous wave technique (System B; SybronEndo, Orange, CA) and a gutta-percha backfill (Obtura II; ObturaSpartan, Earth City, MO). Depending on whether a post placementwas planned by the referring dentist, the tooth was either temporizedusing a sterile cotton pellet and Cavit (3M, St Paul, MN) or a direct adhe-sive buildup with a composite resin material (P60 Singlebond, 3M).After the treatment, all patients received postoperative instructionsand eight tablets of ibuprofen 200 mg with the instructions to takeonly one tablet if it was needed within the 0- to 4-hour time interval afterthe treatment and then one every 8 hours in the event of pain.

Irrigation ProtocolsAll teeth received the same volume of irrigants. Altogether, 130 mL

2.5% sodium hypochlorite (Formula & Acao, Sao Paulo, Brazil) and 10mL EDTA 17% (Formula & Acaol) were used. For both groups, thesodium hypochlorite was held in and dispensed from a mechanicalsyringe pump (Aladdin Pump; World Precision Instruments, Sarasota,FL) providing a constant flow. Twenty milliliters of sodium hypochlorite


Clinical Research
were used during access and initial coronal instrumentation for bothprotocols. Ten milliliters of sodium hypochlorite followed after everyuse of a rotary instrument. Twenty milliliters were reserved for the finalflush after EDTA application. The remainder of the 130 mL was used upafter the final preparation of the root canal space. In group MP, all irri-gation was performed with the 30-G Max-i-Probe needle up to 2 mmshort of the final working length, which was verified by rubber stops.In group EV various tips of the EndoVac system were used followingthe manufacturer’s recommendation. An EndoVac Master DeliveryTip was applied for the initial irrigation followed by a macrocannulain a pecking motion during the main instrumentation and a microcan-nula for the final irrigation with EDTA and sodium hypochlorite. Theinsertion of the EndoVac microcannula was to the working length ofthe prepared root canal space. During the treatment, patients were pre-vented from seeing the irrigation device.
Patient QuestionnaireAll participants received a questionnaire for the evaluation of pain

and the frequency of analgesic use for each root canal procedure at 4,24, and 48 hours after the endodontic treatment was completed. After48 hours, the participants were called by telephone and asked for theanswers to the questionnaire that they had to note on the form. Theperson calling was blinded to the irrigation device that was used duringthe treatment of the particular patient. The participants were askedabout their general feeling in the area of the root canal, with optionsfor feeling generally fine, slightly uncomfortable, having pain on chew-ing, or constant severe pain. The second question recorded the numberof ibuprofen pills that had been taken by the patient up until this point.Furthermore, after 4, 24, and 48 hours, the pain intensity was also re-corded using the CR10 Borg list or Borg scale. The participants labeledthe intensity of pain by a numeric and verbal scale. Level 10, ‘‘extremelystrong,’’ represented the strongest pain the participant had ever expe-rienced. The participant then verbally expressed the level of discomfortby choosing a number in comparison to level 10 pain and quantified thepain using the following values: level 0, ‘‘absolutely nothing’’; level 1,‘‘very weak’’; level 2, ‘‘weak (light)’’; level 3, ‘‘moderate’’; level 4,‘‘somewhat strong’’; levels 5 and 6, ‘‘strong (heavy)’’; levels 7 to 9,‘‘very strong’’; and level 10, ‘‘extremely strong, maximum pain.’’ Theparticipants also had the option to note other sensations. All partici-pants were instructed to immediately contact the office or the emer-gency number in case the analgesics protocol did not provide painrelief or any other emergency occurred.

Statistical AnalysisDescriptive analysis, means, and standard deviations were calcu-

lated using SPSS 15.0 for Windows (SPSS, Chicago, IL). The nonpara-

TABLE 1. Descriptive Analysis of Patient Distribution, Minimum and Maximum Pai

Method Min

Age 16Pain 4 h 0

Group A Pain 24 h 0Pain 48 h 0

Maxi 1 Probe Analgesics 4 h 0Analgesics 24 h 0Analgesics 48 h 0Age 17Pain 4 h 0

Group B Pain 24 h 0Pain 48 h 0

EndoVac Analgesics 4 h 0Analgesics 24 h 0Analgesics 48 h 0


metric Mann-Whitney U test (p = 0.05) and Pearson correlationcoefficient (p = 0.05) were used for statistical analysis.

ResultsAll 110 questionnaires were obtained and evaluated by statistical

analysis. The patients’ age ranged from 16 to 89 years, with a medianage of 48 years. Of a total of 80 patients, 46 were female and 34were male. Table 1 shows the population distribution according togroup MP and group EV. There were 15 lateral and central incisorsand 40 canines and premolars (of which 21 were in the maxilla and19 in the mandible) in group MP. Group EV had 17 lateral and centralincisors and 38 canines and premolars (with 19 each in maxilla andmandible). In group MP, 22 teeth (40%) were treated for dental decayand 33 teeth (60%) for prosthodontic reasons. Fourteen teeth (25.5%)in group EV were treated because of decay, and 41 teeth (74.5%) weretreated intentionally for prosthodontics. The difference in this ratio isexplained by the random assignment of teeth. The distribution of finalapical preparation dimensions were ProTaper F3 (n = 10), F4 (n =25), F5 (n = 13), and F5 plus hand instrument 60, 0.02 (n = 7) forgroup MP and F3 (n = 15), F4 (n = 28), F5 (n = 9), and F5 plushand instrument 60, 0.02 (n = 3) for group EV. The difference inthe distribution was a result of the random allocation of teeth and theindividual root canal morphology. After the 0- to 4-hour interval, twopatients reported to have taken two ibuprofen 200 mg instead of one.No patient contacted the office to change the analgesic protocol orbecause of an emergency situation.

Pain LevelsTable 2 describes the minimum and maximum pain that was expe-

rienced by the participants as well as the statistical analysis of thepatients’ pain levels. For both groups, some patients did not experienceany pain or did not take any analgesic medication, regardless of the timeinterval after treatment. In group MP, the maximum pain intensitydescribed by one patient was 7 within the 0- to 4-hour time interval aftertreatment. For group EV, the maximum pain intensity was 3 consistentlyfor all three time intervals. The maximum pain in group MP decreasedover time. For both groups, the maximum pain values became moreconsistent over time. For group MP, 34.5% of the patients (n = 19)felt no pain during the 0- to 4-hour time interval, 10.9 % (n = 6) felt1 of 10, 27.3% (n = 15) felt 2 of 10 (weak) pain, and only 9.1% (n= 5) felt strong to very strong pain. In group EV, 94.5% of the patients(n = 52) felt no to weak pain. Only 5.5% (n = 3) reported moderatepain up to 4 hours. Within the 4- to 24-hour time period, the maximumpain intensity in group MP decreased to 5 of 10 (strong) in 3.6% of thepatients (n = 2); during the 24- to 48-hour time interval, all patientsexperienced no pain or only weak pain levels. Between 24 and 48

n Levels, and Number of Analgesic Pills

Max Median SD

89 49.80 15.657 1.72 1.755 1.45 1.422 0.50 0.682 0.49 0.572 0.38 0.651 0.05 0.22

87 46.95 12.903 0.39 0.823 0.31 0.663 0.18 0.641 0.09 0.291 0.05 0.220 0.00 0.00


TABLE2.

Pain

Inte

nsity

Dis

trib

utio

nD

urin

g0-

to4-

,4-

to24

-,an

d24

-to

48-H

our

Tim

ein

terv

als

Pai

nin

ten

sity

dis

trib

uti

on

0-4

h4

-24

h2

4-

48

hrs

Met

ho

dIn

tensi

tyFr

equen

cyV

alid

per

cen

tC

um

ula

tive

per

cen

tFr

equ

ency

Val

idp

erce

nt

Cu

mu

lati

vep

erce

nt

Freq

uen

cyV

alid

Per

cen

tC

um

ula

tive

Per

cen

t

019

34.5

34.5

18

32.7

32.7

33

60.0

60.0

16

10.9

45.5

14

25.5

58.2

16

29.1

89.1

215

27.3

72.7

11

20.0

78.2

610.9

100.0

Gro

up

A3

916.3

89.1

610.9

89.1

--

-4

11.8

90.9

47.3

96.4

--

-M

axi

1Pro

be

52

3.6

94.5

23.6

100.0

--

-6

23.6

98.2

——

—-

--

71

1.8

100.0

——

--

-To

tal

55

100.0

55

100.0

55

100.0

042

76.4

76.4

43

78.2

78.2

050

90.9

Gro

up

B1

712.7

89.1

814.5

92.7

12

3.6

23

5.5

94.5

35.5

98.2

21

1.8

En

do

Vac

33

5.5

100.0

11.8

100.0

32

3.6

Tota

l55

100.0

55

100.0

55

100.0

pva

lue

(Man

n-W

hit

ney

test

)p

<0.0

001

*p

<0.0

001*

p<

0.0

001*

*Sta

tistic

ally

sign

ifica

nt.

Clinical Research


hours, respectively, one (1.8%) and two (3.6%) patients in group EVexperienced still moderate pain (3/10).

Pain intensity and analgesic intake were compared in regard to thetime intervals. A normal distribution for pain intensity and analgesicintake was not accepted for statistical analysis; therefore, the MannWhitney U test for independent samples was applied at a significancelevel of 5%. For pain intensity, differences between the MP and EVgroups were statistically significant with p < 0.0001 at all time intervals,with greater pain intensity in the MP group.

Independent from the individual time intervals, the overall painintensity was less in the EV group than in the MP group. The medianpain intensity was 1.22 (standard deviation = 1.06) in the MP groupand 0.29 (standard deviation = 0.56) in the EV group. The Mann-Whitney U test revealed a statistically significant difference betweenthe median pain intensity depending on the irrigation protocol. Therewas significantly less overall pain associated with the treatment in theEV group (EndoVac, p < 0.0001).

Analgesic IntakeTable 3 gives a detailed overview of the patients’ intake of analge-

sics in the number of pills and statistical analysis. The maximum intakeof analgesic medication was higher in the MP group than the EV groupfor all time intervals. In group MP, two patients (3.6%) took two pillswithin the first 4 hours after treatment. Both patients had reported 6 of10 (strong) pain intensity for this time interval. These two participantswere not excluded from the statistical analysis. The patients did notchange the drug and followed the analgesic protocol in the followingtwo time intervals. Also, it would more likely influence the validity ofthe results if patients who recorded pain were excluded from an analysisof occurrence and intensity of pain because they experienced morediscomfort than other participants. For both groups, the consumptionof analgesics decreased with time. In group MP, a total of 27 pills weretaken within the 0- to 4-hour time interval, 21 between 4 and 24 hours,and only 3 between 24 and 48 hours. In the EV group, five patients tookone pill within 4 hours after treatment and three between 4 and 24hours. No patient in this group required pain medication during the24- to 48-hour time interval.

The Mann Whitney U test for independent samples showed thatbetween 0 and 4 hours the number of pills consumed was significantlylower in the EV group with p < 0.0001 and p < 0.001 between 4 and 24hours after treatment. There was no statistically significant difference inanalgesic intake during the 24- to 48-hour interval between both groups(p = 0.08).

Independent from the individual time intervals, the overall numberof analgesic pills was less in the EV group than in the MP group. Themedian consumption was 0.30 pills (standard deviation = 0.37) inthe MP group and 0.04 pills (standard deviation = 0.13) in the EVgroup. The Mann-Whitney U test revealed a statistically significant differ-ence between the median number of pills taken by the participants. Thenumber of analgesics taken was significantly higher in the MP group(p < 0.0001).

Correlation of Median Pain Intensity with the MedianNumber of Pills

Calculation of the Pearson correlation coefficient for the measureof dependence between two variables revealed a strong positive andsignificant relationship (r = 0.851, p < 0.001) for pain intensity andthe number of pills for the MP group (Fig. 1) and also a positive andsignificant correlation (r = 0.596, p < 0.0001) between the variablesof pain and analgesics for the EV group (Fig. 2).


TABLE3.

Anal

gesi

cIn

take

Dis

trib

utio

nD

urin

g0-

to4-

,4-

to24

-,an

d24

-to

48-H

our

Tim

eIn

terv

als

An

alg

esic

inta

ked

istr

ibu

tio

n0

-4h

4-2

4h

24

-4

8h

rs

Met

ho

dN

um

ber

of

pills

Freq

uen

cyV

alid

per

cen

tC

um

ula

tive

per

cen

tFr

equ

ency

Val

idP

erce

nt

Cu

mu

lati

veP

erce

nt

Freq

uen

cyV

alid

Per

cen

tC

um

ula

tive

Per

cen

t

030

54.5

54.5

39

70.9

70.9

52

94.5

94.5

Gro

up

A1

23

41.8

96.4

11

20.0

90.9

35.5

100.0

Maxi

IPro

be

22

3.6

100.0

59.1

100.0

--

-To

tal

55

100.0

55

100.0

55

100.0

Gro

up

B0

50

90.9

90.9

52

94.5

94.5

55

100.0

100.0

15

9.1

100.0

35.5

100.0

--

-En

do

Vac

Tota

l55

100.0

55

100.0

55

100.0

pva

lue

(Man

n-W

hit

ney

test

)p

<0.0

001*

p<

0.0

01*

p=

0.0

8*

*Sta

tistic

ally

sign

ifica

nt.

Clinical Research


DiscussionThe purpose of this study was to compare the differences in post-

operative pain after endodontic therapy after using two different irriga-tion techniques. Mild discomfort after root canal treatment is a commonexperience for patients (21). The reasons for postoperative pain,however, can be many (22). The main causes are mechanical, chemi-cal, or microbial injuries to the periapical tissues that result in acuteinflammation (23). In a clinical investigation, it is difficult to determineif a single or multiple factors elicit pain. If a root canal system was notcleaned properly, residual infection may cause exacerbation by imbal-ances in the host-bacteria relationship, synergistic or additive microbialinteractions, or the presence of decisively pathogenic bacteria beforethe initiation of treatment (23). A mechanical reason may be overinstru-mentation; chemical factors include the extrusion of medications, fillingmaterials, or irrigants (23).

In the present study, great care was taken to rule out avoidablepreoperative factors and to minimize any unavoidable causes of postop-erative pain. Teeth with apical periodontitis, necrotic teeth, or retreat-ment cases were not incorporated, and a meticulous aseptic protocolwas maintained to reduce the risk of exacerbation by residual microor-ganisms or the introduction of bacterial contamination. Therefore, onlyteeth with the diagnosis of irreversible pulpitis or normal pulp weretreated. The study was also limited to asymptomatic teeth becausepreoperative pain is one of the most predictable indicators for postop-erative pain (24). Only teeth in which a single canal could be foundunder the microscope were incorporated to minimize the risk of iatro-genic errors because of missed or complicated root canal anatomy andto make sure the same amount of irrigation solution would pass by eachcanal. All teeth were instrumented and obturated in one session to elim-inate intracanal medication as another possible factor for postoperativeflareup. Furthermore, only patients without a contributing medicalhistory who did not take analgesic medication recently were includedso that no other pain source or drug interaction could interfere withpain resulting from the endodontic therapy.

Even with all the precautions taken, one cannot be sure in a clinicalstudy if pain is coming from the single factor under investigation. Allpossible sources of pain can never be controlled completely. Therefore,under the particular circumstances of our study, postoperative pain mayhave been related to apical trauma because of overinstrumentation orextrusion of debris, sealer, or gutta-percha rather than sodium hypo-chlorite. Bacteria may have been introduced from decay, canal anatomymay have been missed, the soft tissues may have been hurt because ofthe application of the rubberdam or injection, or the patient may havedeveloped unrelated orofacial pain. Taking into consideration that allpatients underwent the same treatment protocol, with the only differ-ence being the irrigation technique, the highly statistically significantoutcome and the strong correlation of pain and analgesic intake allowthe conclusion that, indeed, the particular irrigation protocol hadsignificant impact on the level and time of postoperative pain.

In general, the pain levels the patients experienced in our investi-gation were very low, with only 12 of 330 (3.6%) total reports exceedingmoderate pain, including one single report of very strong pain. No patientreported any other symptoms or complications like swelling or pares-thesia. All these facts underline the level of care that was given to providean atraumatic treatment protocol. However, when all forms of pain wereconsidered, ranging from level 1 (very weak) to level 5 (strong), 44.5%of all treatment were associated with pain during the 4- to 24-hour timeperiod. Of these, 67.3% were in group MP and 21.8% were associatedwith group EV. Other studies showed pain levels beween 12 and 24 hoursto be between 7.1% and 7.8% for teeth with no preoperative history ofpain, respectively, treatment of vital pulps (20).


Figure 1. The correlation of median pain intensity with the median number ofpills in the MP group (Max-i-Probe).

Clinical Research

The difference here may lie in the use of different pain scales. Afrequently used scale for the evaluation of dental pain is the visualanalog scale (VAS) (24, 25). The reliability of the VAS as a measureof pain intensity for patients postoperatively with mild to moderatepain has been shown (26). The VAS ranks pain by a visual scalefrom 0 to 100. Commonly, these values are transfered to four intensitylevels for pain: none (level 1), mild (level 2), moderate (level 3), andsevere pain (level 4) (26). Although in this investigation the highest re-ported values were 7 of 10 and 6 of 10 (once and twice out of 110reports) after 4 hours and 5 of 10 (twice out of 110 reports) after24 hours according to the Borg scale, 4 of 4 referring to the VAS isa rather frequently reported value in studies on postoperative pain(22, 25). The choice of the Borg scale for the evaluation of pain inour study provided good results for statistical analysis. The Borgscale evaluates pain in 10 full steps yet is theoretically open ended. Itwas argued that when intermediate levels are too small, differencesbetween groups might be statistically significant, but the results maynot be clinically significant (27). A benefit of the Borg scale, however,is that strong pain (levels 5-6) is covered by two and very strong pain

Figure 2. The correlation of median pain intensity with the median numberof pills in the EV group.


(levels 7-9) by three levels, which is supposed to address the logarith-mic increase of pain sensation (20). The relatively low dose ofibuprofen 200 mg was chosen to allow a better measure of analgesicintake. High doses may have obscured the outcome, especially withthe very low pain levels created by our endodontic treatment protocolin general.

The results that the irrigation protocol in the EV group resulted inless postoperative pain intensity and analgesic intake may fit with otherfindings. It was found that the use of the EndoVac system did not (28) orsignificantly less (29) result in the apical extrusion of irrigant; hence,chemical irritation of the periapical tissues leading to postoperativepain may not be likely. Because the majority of root canal irrigantsare cytotoxic to the periapical tissues, the irrigation solution shouldbe restricted to within the root canal system. In our study, both tech-niques were either used according to the manufacturer’s recommenda-tions or, if not available, according to the common protocol (28). To besafe, irrigation with Max-i-Probe was 2 mm from the working length,which is within the range of 1.5 to 3.0 mm used in comparable studieswith Max-i-Probe or identical irrigation needles (17, 19, 28, 30, 31),whereas the EndoVac negative apical pressure tip was routinely usedall the way to the working length (0 mm), thus fulfilling the claim fordirect irrigation in the apical third (15).

In conclusion, the negative apical pressure irrigation system Endo-Vac resulted in significantly less postoperative pain and necessity foranalgesic medication than a conventional needle irrigation protocolusing the Max-i-Probe. From the results of this study, it may be assumedthat it is safe to use a negative apical pressure irrigation protocol forantimicrobial debridement up to the full working length.

AcknowledgmentDiscus Dental provided the EndoVac kits for clinical use for

this study.

References1. Sathorn C, Parashos P, Messer H. The prevalence of postoperative pain and flare-up

in single- and multiple-visit endodontic treatment: a systematic review. Int Endod J2008;41:91–9.

2. Ng YL, Glennon JP, Setchell DJ, et al. Prevalence of and factors affecting post-obturation pain in patients undergoing root canal treatment. Int Endod J 2004;37:381–91.

3. Seltzer S. Pain in endodontics. 1986. J Endod 2004;30:501–3.4. Kakehashi S, Stanley HR, Fitzgerald RJ. The effects of surgical exposures of dental

pulps in germfree and conventional laboratory rats. Oral Surg Oral Med Oral Pathol1965;20:340–9.

5. Dalton BC, Orstavik D, Phillips C, et al. Bacterial reduction with nickel-titaniumrotary instrumentation. J Endod 1998;24:763–7.

6. Shuping GB, Orstavik D, Sigurdsson A, et al. Reduction of intracanal bacteria usingnickel-titanium rotary instrumentation and various medications. J Endod 2000;26:751–5.

7. Hauman CH, Love RM. Biocompatibility of dental materials used in contemporaryendodontic therapy: a review. Part 1. Intracanal drugs and substances. Int EndodJ 2003;36:75–85.

8. Torabinejad M, Shabahang S, Aprecio RM, et al. The antimicrobial effect of MTAD:an in vitro investigation. J Endod 2003;29:400–3.

9. Naenni N, Thoma K, Zehnder M. Soft tissue dissolution capacity of currently usedand potential endodontic irrigants. J Endod 2004;30:785–7.

10. Abdullah M, Ng YL, Gulabivala K, et al. Susceptibilties of two Enterococcus faecalisphenotypes to root canal medications. J Endod 2005;31:30–6.

11. Harrison JW, Hand RE. The effect of dilution and organic matter on the anti-bacterialproperty of 5.25% sodium hypochlorite. J Endod 1981;7:128–32.

12. Kleier DJ, Averbach RE, Mehdipour O. The sodium hypochlorite accident: experi-ence of diplomates of the American Board of Endodontics. J Endod 2008;34:1346–50.

13. Hulsmann M, Hahn W. Complications during root canal irrigation—literaturereview and case reports. Int Endod J 2000;33:186–93.


Clinical Research
14. Siqueira JF Jr, Rocas IN, Favieri A, et al. Chemomechanical reduction of the bacterial
population in the root canal after instrumentation and irrigation with 1%, 2.5%, and5.25% sodium hypochlorite. J Endod 2000;26:331–4.

15. Chow TW. Mechanical effectiveness of root canal irrigation. J Endod 1983;9:475–9.16. Schoeffel GJ. The EndoVac method of endodontic irrigation, part 3: system compo-

nents and their interaction. Dent Today 2008;27(106):108–11.17. Nielsen BA, Baumgartner JC. Comparison of the EndoVac system to needle irrigation

of root canals. J Endod 2007;33:611–5.18. Brito PR, Souza LC, Machado de Oliveira JC, et al. Comparison of the effectiveness of

three irrigation techniques in reducing intracanal Enterococcus faecalis popula-tions: an in vitro study. J Endod 2009;35:1422–7.

19. Miller TA, Baumgartner JC. Comparison of the antimicrobial efficacy of irrigationusing the EndoVac to endodontic needle delivery. J Endod 2010;36:509–11.

20. Borg G, Holmgren A, Lindblad I. Quantitative evaluation of chest pain. Acta MedScand Suppl 1981;644:43–5.

21. Harrison JW, Baumgartner JC, Svec TA. Incidence of pain associated with clinicalfactors during and after root canal therapy. Part 2. Postobturation pain. J Endod1983;9:434–8.

22. Genet JM, Hart AA, Wesselink PR, et al. Preoperative and operative factors associatedwith pain after the first endodontic visit. Int Endod J 1987;20:53–64.


23. Siqueira JF Jr, Barnett F. Interappointment pain: mechanisms, diagnosis, and treat-ment. Endod Topics 2004;7:93–109.

24. Glennon JP, Ng YL, Setchell DJ, et al. Prevalence of and factors affecting postpreparationpain in patients undergoing two-visit root canal treatment. Int Endod J 2004;37:29–37.

25. El Mubarak AH, Abu-bakr NH, Ibrahim YE. Postoperative pain in multiple-visit andsingle-visit root canal treatment. J Endod 2010;36:36–9.

26. Myles PS, Troedel S, Boquest M, et al. The pain visual analog scale: Is it linear ornonlinear? Anesth Analg 1999;89:1517–20.

27. Bodian CA, Freedman G, Hossain S, et al. The visual analog scale for pain: clinicalsignificance in postoperative patients. Anesthesiology 2001;95:1356–61.

28. Desai P, Himel V. Comparative safety of various intracanal irrigation systems. J En-dod 2009;35:545–9.

29. Mitchell RP, Yang SE, Baumgartner JC. Comparison of apical extrusion of NaOClusing the EndoVac or needle irrigation of root canals. J Endod 2010;36:338–41.

30. Hockett JL, Dommisch JK, Johnson JD, et al. Antimicrobial efficacy of two irrigationtechniques in tapered and nontapered canal preparations: an in vitro study. J Endod2008;34:1374–7.

31. Brito PR, Souza LC, Machado de Oliveira JC, et al. Comparison of the effectiveness ofthree irrigation techniques in reducing intracanal Enterococcus faecalis popula-tions: an in vitro study. J Endod 2009;35:1422–7.


Clinical Research

Prevalence of Three-rooted Mandibular Permanent FirstMolars among the Indian PopulationAmit Kumar Garg, BDS (Lko), MDS (Lko),* Rajendra K. Tewari, BDS (Lko), MDS (Lko),*

Ashok Kumar, BDS (Lko), MDS (Lko),* Sarwat H. Hashmi, BDS (Lko), MDS (Lko),†

Neha Agrawal,‡ and Surendra K. Mishra, BDS (Lko), MDS (Lko)*

Abstract
Introduction: The aim of this retrospective study wasto determine the prevalence of three-rooted mandibularpermanent first molars among the Indian population byusing periapical radiographs. Methods: Five hundredeighty-six patients (320 females and 266 males) wereselected, with at least 1 mandibular first molar. A totalof 1054 periapical radiographs of mandibular firstmolars, comprising 526 right side and 528 left side,were included. The radiographs were taken at 30-degree mesial angulation and were evaluated by usingthe magnifying lens. The incidence, gender, andsymmetry of three-rooted mandibular first molars wererecorded and analyzed by using the c2 test. Results:The prevalence of three-rooted mandibular first molarswas 5.97% for all patients and 4.55% for all teeth,respectively. The incidence of bilateral symmetricaldistribution was 37.14%. The incidence was 6.88% forfemale patients and 4.89% for male patients (c2 =1.02, P > .05) and 4.94% for the right side and 4.17%for the left side, respectively (c2 = 0.36, P > .05). Nostatistically significant differences were found betweenfemale and male patients and between the right-sideand left-side occurrences. Conclusions: Cliniciansshould be aware of the high racial prevalence of thisunusual root morphology in mandibular first molarsamong the Indian population before and during theroot canal treatment of three-rooted mandibular firstmolars. (J Endod 2010;36:1302–1306)
Key WordsRadix entomolaris, radix paramolaris, supernumeraryroot, three-rooted mandibular first molars

From the *Department of Conservative Dentistry andEndodontics and †Department of Oral and MaxillofacialSurgery, Dr. Z. A. Dental College, Aligarh Muslim University, Ali-garh (U.P.), India; and ‡Department of Preventive and Commu-nity Dentistry, M. S. Ramaiah Dental College and Hospital,Bangalore (Karnataka), India.

Address requests for reprints to Dr Amit Kumar Garg, Assis-tant Professor, Department of Conservative Dentistry andEndodontics, Dr. Z. A. Dental College, Aligarh Muslim Univer-sity, Aligarh 202002 (U.P.), India. E-mail address:[email protected]/$0 - see front matter


1302 Garg et al.

The main objective of root canal treatment is thorough mechanical and chemicaldebridement of all root canals and their complete obturation with an inert filling

material and a coronal filling, preventing the ingress of microorganisms (1). One ofthe main reasons for the failure of root canal treatment is the inadequate removal ofpulp tissue and microorganisms from the root canal system. Root canal anatomy andthe confounding nature of the human pulpal system pose significant challenges inrendering endodontic treatment. Therefore, it is imperative that the aberrant anatomyis identified before and during the root canal treatment of three-rooted mandibular firstmolars.

It is known that mandibular first molars might display several anatomical varia-tions, because the number of root canals and number of roots might also vary (2).The major variant in this tooth is the presence of a supernumerary root that can be founddistolingually. This macrostructure, first mentioned by Carabelli, is called radix entomo-laris (RE) (3). An RE can be found in the first, second, and third mandibular molars,occurring the least frequently in the second molar (4). An additional root at the mesio-buccal side is called radix paramolaris (RP). The RE mostly has Vertucci type I canalconfiguration (5). The RE, which in general is smaller than distobuccal and mesialroots, can be separate from or partially fused with these other roots (6, 7). De Mooret al (1) have classified the RE into 3 types according to the buccolingual variations;type I refers to a straight root, type II to an initially curved entrance that continues asa straight root, and type III to an initial curve in the coronal third of the root canal, fol-lowed by a second curve beginning in the middle and continuing to the apical third (1).

This supernumerary root in the mandibular first molar is associated with certainethnic groups as follows: European, 3.4%–4.2% (8–11); African, 3% (12); Eurasianand Indian, less than 5% (13); Europeans, 4.2% (1, 6, 14); Asians, such as Chinese,Eskimo, and American Indians have 5% to more than 30% (15–18), and the overallincidence in German patients was 1.35% (19), and among Taiwanese it was about21% (20) (Table1). Because of its high frequency in mongoloid populations, the REis considered to be a normal morphologic variant or eumorphic root morphology(6) and can be seen as the Asiatic trait (16). Among Caucasians, RE is not very common(21, 22) and is considered to be an unusual or dysmorphic root morphology.

In dysmorphic supernumerary root, its root formation is related to external factorsduring odontogenesis (6) or is due to penetration of atavistic gene or polygenic system(6, 17). In eumorphic roots, racial genetic factors influence the profound expression ofa particular gene that results in a more profound phenotypic expression (6). Midtbøand Halse (23) concluded that X chromosome deficiency influences root formation.

To the best of our knowledge few studies such as ours have been undertaken in thecontext of the Indian population. This retrospective study was done to evaluate the inci-dence of three-rooted mandibular permanent first molars, their gender and side-related differences among the Indian population, by using periapical radiographs.The results should be of interest to clinical dentists, dental morphologists, and dentalanthropologists.

Materials and MethodsA total of 586 patients’ retrospective periapical radiographs recorded in the

Department of Oral Medicine and Radiology, Dr. Z. A. Dental College, Aligarh Muslim



TABLE 1. Survey of Available Studies by Extracted Teeth and Periapical Radiographs on the Prevalence of Three-rooted Mandibular First Molars

Periapical radiographs

No. of teeth/person 3% RM1

Authors Year Area of origin Sample Gender(M/F) 3% RM1 Gender(M/F) Right Left Bilateral Total

Tratman (13) 1938 Chinese 1615 5.80Malay 475 8.60Javanese 110 10.9Indians* 453 0.20Eurasians 262 4.20Japanese 168 1.20

Laband (24) 1941 Malaysian 134 8.20Somogyl-Csizmazia and Simons (25) 1971 Canadian Indians† 250 15.6Souza-Freitas et al (9) 1971 European descent 422 3.20Skidmore and Bjorndahl (10) 1971 White 45 2.2

Japanese descent 233 135/98 17.8 1.65/1 9.87 12.88 22.75Turner (26) 1971 Aleutian Eskimo 263 32.0

American Indians† 1983 5.80Curzon and Curzon (15) 1971 Keewatin Eskimo 98 27.0Curzon (27) 1974 Baffin Eskimo 69 21.7Hochstetter (28) 1975 Guam 400 14.3 2/1Jones (29) 1980 Chinese 52 13.4

Malaysian 149 16.0Reichart and Metah (17) 1981 Thai 364 19.2Walker and Quackenbush(18) 1985 Hong Kong Chinese 213 14.6Steelman (11) 1986 Hispanic children 156 73/83 1.50/1 2.60 0.60 3.20 6.40Walker (30) 1988 Hong Kong Chinese 100 15.0Harada et al(31) 1989 Japanese 2331 18.8Loh (14) 1990 Singaporean Chinese 304 7.9Ferraz and Pecora (22) 1992 Japanese descent 105 1/1 11.4

Negro 106 2.80White 117 4.20

Yew and Chan (16) 1993 Chinese 179 21.5Gulabivala (32) 2001 Burmese 139 10.1Gulabivala (33) 2002 Thai 118 13.0Huang et al (34) 2007 Taiwanese 332 21.7 26.9Tu et al (20) 2007 Taiwanese 332 79/87 17.77 0.75 4.22 2.41 14.46 21.09Schafer et al (19) 2009 Germans 524 264/260 0.68 1.33 0.57 0.76 1.34

3% RM1, % of 3-rooted mandibular 1st molars.

*Indian subcontinent.†Native American Indians.

ClinicalResearch

JOE—

Volume

36,N

umber

8,August

2010Prevalence

ofThree-rooted

Mandibular

Permanent

FirstM

olarsam

ongthe

IndianPopulation

1303

TABLE 2. Number and Percentage of Three-rooted Mandibular First Molars

No. of three-rooted mandibular first molars

Right Left Bilateral Total

No. of patients and teeth No. % No. % No. % No. %

Female 320 10 3.13 6 1.88 6 1.88 22 6.88Male 266 3 1.13 3 1.13 7 2.63 13 4.89Total patients 586 13 2.22 9 1.54 13 2.22 35 5.97No. of all right first molars examined 526 13 2.47 — — 13 2.47 26 4.94No. of all left first molars examined 528 — — 9 1.71 13 2.46 22 4.17Total teeth 1054 13 1.23 9 0.85 26 2.47 48 4.55

Clinical Research

University, Aligarh, India from December 2008–December 2009 werescreened and examined. The bilateral eccentric periapical radiographs(30-degree mesial angulation with protractor) of patients who visitedthe Department of Conservative Dentistry and Endodontics for treatmentof either pain or caries in the mandibular molars were obtained. Each ofthese patients had at least 1 mandibular first molar and was of Indianorigin. Demographic details including age, sex, and race of all thesepatients were recorded.

The x-ray machine used for tooth identification was EndosAC (VillaSistemi Medicali Spa, Buccinasco, Italy) (70 kV and 8 mA). Periapicalradiographs were taken with Kodak Ultraspeed films (Eastman Kodak

Figure 1. Periapical radiographs of three-rooted mandibular first molars showingon the left side unilaterally, (C) the mesiobuccal root on the right side unilaterally

1304 Garg et al.

Ultra-speed film; Kodak, Rochester, NY). A total of 1054 periapicalradiographs of mandibular first molars of 586 patients (320 femalesand 266 males) were selected for the study. The radiographs wereplaced on a viewing box, and the light surrounding the radiographwas blocked. Each radiograph was independently studied by 2 authors(G.A. and T.R.) by using magnifying lens (3�). Any disagreement in theinterpretation of images was discussed by 2 endodontists, anda consensus was reached (19, 20). The criteria for the indication ofan extra root were justified by crossing the translucent lines, definingthe pulp space and the periodontal ligaments in the mandibular firstmolars (18–20). The overall incidence of three-rooted mandibular first

(A) the distolingual root on the right side unilaterally, (B) the distolingual root, and (D) the mesiobuccal root on the left side unilaterally.


Clinical Research
molars in the patients and their correlations between female and malepatients and between the right-side and left-side occurrences wereanalyzed by using the c2 test (19, 20). The bilateral incidence ofthese three-rooted mandibular first molars was also evaluated.
ResultsPeriapical radiographs of 586 patients, 320 females and 266

males, with age range of 15–75 years and average age of 30.3 �12.5 years, were studied. The periapical radiographs of 35 patients,22 females and 13 males, had three-rooted mandibular first molars.A total of 1054 periapical radiographs of mandibular first molarscomprising 526 right and 528 left molars were evaluated (Table 2).Of these three-rooted mandibular first molars, 26 were found on theright side and 22 on the left side. The prevalence of patients withthree-rooted mandibular first molar was 5.97% (35 of 586 patients),6.88% (22 of 320) for female patients and 4.89% (13 of 266) formale patients (Table 2). The prevalence of three-rooted mandibularfirst molars from all teeth examined was 4.55% (48 of 1054), 4.94%(26 of 526) for the right side and 4.17%(22 of 528) for the left sideoccurrences (Table 2). There was no statistical significant differencein the incidence of three-rooted mandibular first molars betweenfemale and male patients (c2 = 1.02, P > .05) and between theright-side and left-side occurrences (c2 = 0.36, P > .05) (Table 2,Fig. 1). The bilateral incidence of symmetrical distribution was 37.14%.

DiscussionIn the present study, the prevalence of three-rooted mandibular

first molars among the Indian population was 5.97% of all patientsand 4.55% of all teeth examined (Table 2). This figure is higher thanthe result of the study by Tratman (13) (0.20%) among Asiatic Indiansand similar to the result of the study by Turner (26) (5.8%) amongAmerican Indians and less than the study by Somogyl-Csizmazia andSimons (25) among Canadian Indians (15.6%) (Table1). In this study,there was no significant difference according to gender (P > .05),which is similar to the recent studies (19, 20). There was also nosignificant difference according to the side occurrence (right versusleft side, P > .05), which was also similar to the recent study (19).However, some studies reported that three-rooted mandibular firstmolars occurred more frequently on the right side than on the leftside (11, 20), whereas there are also studies showing that thesethree-rooted mandibular first molars occurred more frequently onthe left side (21, 32).

The bilateral occurrence of three-rooted mandibular first molarswas 37.14% (13 of 35), which was more than the recent study (19)among the German population (0%) and lower than several studieson the Asiatic descent population (56.6%–67%) (9, 11, 16, 18, 20).These contradictory variations might be explained by markeddifferences in the sample size, case selection, and the methods used.Thus, further investigations are necessary to clarify the issue. Acomparison of the previous reports is presented in Table 1.

There have been several studies of extracted permanent mandib-ular first molars (13–15, 22, 29, 32, 33), but it is impossible tocompare the results of these studies related to gender and bilateraloccurrences. The present noninvasive study used 2-dimensionalimages (periapical radiographs) of patients’ mandibular first molaras a tool for studies related to gender and side-related differences.

The presence of RE has clinical implications in endodontic treat-ment. An accurate diagnosis of these supernumerary roots can avoidcomplications or missing a canal during the root canal treatment(34). Apart from complicating the root canal procedure, RE hasbeen found to be a contributing factor to localized periodontal destruc-

JOE — Volume 36, Number 8, August 2010 Prevalence of Three

tion (35). In addition, reports correlate significantly higher probingdepths with attachment loss at the distolingual aspect of three-rootedmolars (2, 35).

According to Walker and Quackenbush (18), normally a thirdroot should readily be evident in about 90% of cases radiographically,but occasionally it might be difficult to see because of its slender dimen-sions. In addition, a file placed in such a root might give an artifactualappearance of a perforation. In such instances, an angled view (verti-cally and horizontally) is always beneficial (25). With the distolinguallylocated orifice of RE, a modification of the classic triangular openingcavity to a trapezoidal form to locate and access the root canal betteris essential; the straight line access must be established (1, 6).

ConclusionWith the frequency of occurrence of 5.97% among the Indian pop-

ulation, every possible effort should be made for locating an extra rootin mandibular first molars because it might be useful for successfulendodontic treatment. Clinicians should be aware of the high racialprevalence of this unusual root morphology in mandibular first molarswhile treating Indian patients.

AcknowledgmentsThe authors thank Prof. Aziz Khan (Department of Community

Medicine, AMU, Aligarh) and Prof. A. R. Kidwai (Director, UGCAcademic Staff College, AMU, Aligarh) for their guidance and thetime and effort they devoted to this study.

References1. De Moor RJ, Deroose CA, Calberson FL. The radix entomolaris in mandibular first

molars: an endodontic challenge. Int Endod J 2004;37:789–99.2. Schumann C. Endodontic treatment of a mandibular first molar with radix entomo-

laris: a case report. ENDO (Lond Engl) 2008;2:301–4.3. Bolk L. Bemerkungen uber Wurzelvariationen am menschlichen unteren Molaren.

Zeitschrift fur Morphologie Anthropologie 1915;17:605–10.4. Visser JB. Beitrag zur Kenntnis der menschlichen Zahnwurzelformen. Hilversum,

Netherlands: Rotting; 1948. 49–72.5. Segura-Egea JJ, Jimenez-Pinzon A, Rios-Santos JV. Endodontic therapy in a 3-rooted

mandibular first molar: importance of a thorough radiographical examination.J Can Dent Assoc 2002;68:541–4.

6. Calberson FL, De Moor RJ, Deroose CA. The radix entomolaris and paramolaris:clinical approach in endodontics. J Endod 2007;33:58–63.

7. Carlsen O, Alexandersen V. Radix entomolaris: identification and morphology.Scand J Dent Res 1990;98:363–73.

8. Taylor AE. Variations in the human tooth-form as met with in isolated teeth. J AnatPhysiol 1899;33:268–72.

9. de Souza-Freitas JA, Lopes ES, Casati-Alvares L. Anatomic variations of lower firstpermanent molar roots in two ethnic groups. Oral Surg Oral Med Oral Pathol1971;31:274–8.

10. Skidmore AE, Bjorndahl AM. Root canal morphology of the human mandibular firstmolar. Oral Surg Oral Med Oral Pathol 1971;32:778–84.

11. Steelman R. Incidence of an accessory distal root on mandibular first permanentmolars in Hispanic children. ASDC J Dent Child 1986;53:122–3.

12. Sperber GH, Moreau JL. Study of the number of roots and canals in Senegalese firstpermanent mandibular molars. Int Endod J 1998;31:117–22.

13. Tratman EK. Three-rooted lower molars in man and their racial distribution. BrDent J 1938;64:264–74.

14. Loh HS. Incidence and features of three-rooted permanent mandibular molars. AustDent J 1990;35:434–7.

15. Curzon MEJ, Curzon JA. Three-rooted mandibular molars in the Keewatin Eskimo.J Can Dent Assoc 1971;37:71–3.

16. Yew SC, Chan K. A retrospective study of endodontically treated mandibular firstmolars in a Chinese population. J Endod 1993;19:471–3.

17. Reichart PA, Metah D. Three-rooted permanent mandibular first molars in the Thai.Community Dent Oral Epidemiol 1981;9:191–2.

18. Walker RT, Quackenbush LE. Three-rooted lower first permanent molars in HongKong Chinese. Br Dent J 1985;159:298–9.

19. Schafer E, Breuer D, Janzen S. The prevalence of three-rooted mandibular perma-nent first molars in a German population. J Endod 2009;35:202–5.

-rooted Mandibular Permanent First Molars among the Indian Population 1305

Clinical Research
20. Tu MG, Tsai CC, Jou MJ, et al. Prevalence of three-rooted mandibular first molars
among Taiwanese individuals. J Endod 2007;33:1163–6.21. Curzon ME. Three-rooted mandibular permanent molars in English Caucasians.

J Dent Res 1973;52:181.22. Ferraz JAB, Pecora JD. Three-rooted mandibular molars in patients of Mongolian,

Caucasian and Negro origin. Br Dent J 1992;3:113–7.23. Midtbø M, Halse A. Root length, crown height, and root morphology in Turner

syndrome. Acta Odontol Scand 1994;52:303–14.24. Laband F. Two years’ dental school work in British North Borneo: relation of diet to

dental caries among natives. J Am Dent Assoc 1941;28:992–8.25. Somogyl-Csizmazia W, Simons AJ. Three-rooted mandibular first molars in Alberta

Indian Children. J Can Dent Assoc 1971;37:105–6.26. Turner CG 2nd. Three-rooted mandibular first permanent molars and the question

of American Indian origins. Am J Phys Anthropol 1971;34:229–41.27. Curzon MEJ. Miscegenation and the prevalence of three-rooted mandibular first

molars in the Baffin Eskimo. Community Dent Oral Epidemiol 1974;2:130–1.

1306 Garg et al.

28. Hochstetter RL. Incidence of trifurcated mandibular first permanent molars in thepopulation of Guam. J Dent Res 1975;54:1097.

29. Jones AW. The incidence of the three-rooted lower first permanent molar in Malaypeople. Singapore Dent J 1980;5:15–7.

30. Walker RT. Root form and canal anatomy of mandibular first molars in a SouthernChinese population. Dent Traumatol 1988;4:19–22.

31. Harada Y, Tomino S, Ogawa K, et al. Frequency of three-rooted mandibular firstmolars. Shika Kiso Igakkai Zasshi 1989;31:13–8.



34. Bains R, Loomba K, Chandra A, et al. The radix entomolaris: a case report. ENDO(Lond Engl) 2009;3:121–5.

35. Huang RY, Lin CD, Lee MS, et al. Mandibular disto-lingual root: a consideration inperiodontal therapy. J Periodontol 2007;78:1485–90.


Clinical Research

Biologic Markers for Odontogenic Periradicular PeriodontitisBruna Burgener, DDS,* Angelique R. Ford, DDS,* Hongsa Situ, DDS,*

Mohamed I. Fayad, DDS, MS, PhD,* Jian Jun Hao, BDS, MS, PhD,†

Christopher S. Wenckus, DDS, FICD,* Bradford R. Johnson, DDS, MHPE,* Ellen A. BeGole, PhD,*

and Anne George, PhD†

Abstract
Introduction: The diagnosis and assessment of apicalperiodontitis by traditional periapical radiographs canbe challenging and might yield false-negative results.The aim of this study was to determine whetherinterleukin-1beta (IL-1b) and dentin sialoprotein (DSP)in gingival crevicular fluid (GCF) can be used as biolog-ical markers for apical periodontitis. Methods: Fortyhealthy patients with teeth diagnosed with apical perio-dontitis of pulpal origin were included in the study. GCFsamples were obtained from the diseased tooth andfrom a healthy contralateral control tooth. Total proteinconcentration in each sample was determined by usingthe Bio-Rad protein assay. Enzyme-linked immunosor-bent assay was used to analyze the concentration ofIL-1b and DSP in the samples. Results: Protein contentof the GCF was statistically significantly higher in thedisease group compared with the control group. Thelevels of IL-1b and DSP were not statistically differentbetween disease and control groups. Conclusions:Although this study was unable to demonstrate a signif-icantly higher level of IL-1b or DSP in the GCF of teethwith apical periodontitis, the observed presence ofa significantly higher level of total protein in the GCFof diseased teeth suggests the possible role of totalprotein level as a marker for periapical disease. (J Endod2010;36:1307–1310)
Key WordsApical periodontitis, biological marker, gingival crevicu-lar fluid

From the )Department of Endodontics and †Department ofOral Biology, University of Illinois at Chicago College ofDentistry, Chicago, Illinois.

Address requests for reprints to Dr Bradford R. Johnson,University of Illinois at Chicago College of Dentistry, Departmentof Endodontics, 801 S Paulina St, Chicago, IL 60612. E-mailaddress: [email protected]/$0 - see front matter



The inflammatory periapical lesion is a common sequela of infected pulp necrosis andmanifests itself as the host defense response to microbial challenge emanating from

the pulp canal system. Numerous cell types, including polymorphonuclear leukocytes,T and B lymphocytes, macrophages, and plasma cells, have been identified in periapicallesions (1). These inflammatory cells, especially macrophages, mediate the immuno-logic response seen in apical periodontitis (2). Bone resorption seen in periradicularlesions is mainly caused by the production of interleukin-1beta (IL-1b) by macro-phages (3, 4) and tumor necrosis factor–b by T lymphocytes (5). Il-1b is commonlyfound in human periapical lesions (6).

Dentin resorption can also occur during the development of apical periodontitis(7). Dentin sialoprotein (DSP) is a dentin non-collagenous protein involved in themineralization of predentin into dentin. DSP was found in the gingival crevicular fluid(GCF) of patients presenting with external apical root resorption caused by orthodonticmovement (8). DSP was once thought to be a dentin-specific protein, but its expressionwas also demonstrated in bone tissue, although in much lower levels than in dentin (9).Therefore, it is possible that DSP detected in GCF is not exclusively from dentinresorption.

Currently, the presence or absence of apical periodontitis is determined by clinicaland radiographic examination. Because there is a relatively low incidence of clinicalsigns and symptoms associated with periradicular periodontitis (10), the diagnosisis established primarily by radiographic findings in association with pulp vitality tests.Clinical symptoms are present in only approximately 18%–24% of teeth with radio-graphic evidence of apical periodontitis (10, 11). The limitations of radiographicexamination in detecting the presence of apical periodontitis are well-known and arerelated to the amount of bone loss caused by the lesion, the spread of bone resorptioninto the cortical bone, location in the jaw, and operator variability in radiographic inter-pretation (12, 13). Recent studies demonstrated that cone beam computed tomography(CBCT) is more accurate in detecting apical periodontitis compared with conventionalradiographs (14–19). Because both the cost of CBCT and radiation exposure continueto decrease, its use for the assessment of periapical healing will likely become morecommon.

Peripheral body fluids such as GCF are often used as identity markers of acute andchronic inflammation because the composition of these fluids might change as a resultof their proximity to an inflammatory focus. GCF is the inflammatory exudate that can becollected at the gingival crevice. Collection of GCF is simple and presents minimal risk tothe patient. Biological markers such as inflammatory mediators and neuropeptideswere detected in the GCF of patients with periodontal disease (20) and root resorptioncaused by orthodontic treatment (8). Increased levels of substance P, neurokinin A,and IL-8 were found in the GCF of patients with acute irreversible pulpitis (21, 22).In the presence of active disease associated with apical periodontitis, inflammatorymediators and dentin proteins that are released in the periapical tissues mightdiffuse through the periodontal ligament and subsequently into the gingival crevice.

The objective of this study was to test the hypothesis that higher levels of IL-1b andDSP would be detected in the GCF of teeth diagnosed with apical periodontitis comparedwith healthy control teeth in the same patient. The long-term goal is to develop a reliable,inexpensive, and noninvasive test that could be used as an adjunct to currently useddiagnostic tools to detect the presence of active periapical inflammation.

Biologic Markers for Odontogenic Periradicular Periodontitis 1307


Figure 1. Filter paper strips were inserted 1–2 mm into the gingival crevice ofeach tooth for 30 seconds to collect the GCF sample. (This figure is available incolor online at www.aae.org/joe/.)

TABLE 1. Protein Concentration (mg/mL) in the Control and Disease Groups

Group N Mean Standard deviation P value

Disease group 40 66.52 50.96 .003Control group 40 36.69 34.05

TABLE 2. IL-1b Absorbance in Control and Disease Groups



Clinical Research

Materials and MethodsForty patients were recruited from the pool of patients seeking

routine or emergency treatment in the Postgraduate Endodontics Clinicof the University of Illinois at Chicago (UIC). The UIC InstitutionalReview Board approved the study. Written and verbal informed consentwas obtained from each patient.

The inclusion/exclusion criteria for this study were as follows.Participants must be 18 years old or older with an unremarkablemedical history and radiographically evident apical periodontitis ona restorable single or multirooted tooth. Patients undergoing firsttime root canal therapy (RCT), retreatment RCT, and surgical RCTwere included. Pulp vitality testing was performed on teeth that hadnot been previously treated to confirm pulpal necrosis. Patients werenot excluded on the basis of previous or current use of antibiotics oranalgesics. Patients were excluded if periodontal probing depthswere greater than 4.0 mm on the experimental or control tooth or ifbleeding on probing was detected. Patients were also excluded ifundergoing orthodontic treatment.

GCF was collected from the experimental tooth and a healthycontralateral control tooth in each subject immediately before treat-ment. Contralateral teeth were examined clinically and radiographically,and cold testing was performed to ensure the tooth had normal pulpand periradicular tissues. One investigator (B.B.) collected all samples.

Before sample collection, the tooth was washed gently with water,dried, and isolated with cotton rolls. A saliva ejector was also used toprevent saliva contamination. Filter paper strips (Periopaper; Oraflow,Plainview, NY) were then inserted 1–2 mm into the gingival crevice ofeach tooth or until mild resistance was sensed. The strips were removedfrom the gingival sulcus after 30 seconds (Fig. 1). Two additionalsamples were collected from the same tooth at 5-minute intervals.Samples were rejected if contaminated by blood or saliva. A total of 6strips were collected from each tooth. After removal from the gingivalsulcus, all the strips were immediately placed in a microcentrifugetube containing ice-cold 1X phosphate-buffered saline solution with0.1 mmol/L phenylmethylsulfonyl fluoride. All patients were treatedby endodontic residents according to established standard proceduresfor nonsurgical or surgical RCT. Patients were referred back to theirgeneral dentists for final restoration and were scheduled for a 6-monthfollow-up appointment in the Postgraduate Endodontics Clinic forassessment of healing and collection of GCF.

1308 Burgener et al.

The protein concentration of each sample was determined by theBradford method (Bio-Rad protein assay; Bio-Rad, Hercules, CA) at4�C to avoid protein degradation. Bovine serum albumin was used asthe standard. After dilution of samples to equalize the protein contentof each pair of control and experimental teeth, indirect enzyme-linked immunosorbent assay (ELISA) was used to detect and quantifythe presence of IL-1b and DSP. All samples were assayed in duplicate.Primary antibody dilutions in this assay were 1:200 for IL-1b (SantaCruz Biotech Co, Santa Cruz, CA) and 1:2000 for DSP (Dr Anne George,Chicago, IL). The secondary anti-rabbit immunoglobulin G antibodywas used at a 1:7000 dilution (Sigma-Aldrich, St Louis, MO). The opticaldensity was measured at 405 nm with a microtiter plate reader (Biotek,Winooski, VT).

An independent samples t test was used to compare the proteinconcentration and the levels of IL-1b and DSP in the GCF of experi-mental and control groups (SPSS, Chicago, IL). The significance levelwas set at P <.05.

ResultsForty diseased and 40 control teeth were tested (37 patients, 24

female and 13 male). The Bradford method for protein concentrationshowed an average of 66.52� 50.96 mg/mL in the diseased group and36.69 � 34.05 mg/mL in the control group. There was a statisticallysignificant difference between the 2 groups (t = 3.08; P = .003)(Table 1). The mean absorbance of IL-1b in the GCF in the diseasedgroup was 0.18� 0.06, which was not statistically different comparedwith the control group, 0.16� 0.05 (t = 1.62; P = 0.15) (Table 2). Themean absorbance of DSP was 0.34 � 0.11 in the diseased group and0.30� 0.09 in the control group. There was no statistically significantdifference between groups (t = 1.78; P = .09) (Table 3).

DiscussionThe outcome of RCT is influenced by a number of factors. One of

the most important determinants of success is the status of the pulp andperiapical tissues before treatment (10). Periapical radiographs are themost commonly used tool for evaluation of healing after RCT. However,the limitations associated with traditional 2-dimensional radiographicimaging to detect periapical pathosis are well-known (18, 23, 24). Inaddition, even though the majority of periapical lesions will showradiographic evidence of healing 1 year after treatment (25–27),healing progresses in a linear manner, and 3–4 years might berequired to truly evaluate healing (28, 29). A noninvasive tool tomeasure the presence of active periapical inflammation could bea useful adjunct to radiographic evaluation, particularly in cases inwhich the radiographic interpretation is uncertain. Knowing thestatus of the immune response and not its consequences could beimportant in cases with apparently slow healing lesions.


http://www.aae.org/joe/

TABLE 3. DSP Absorbance in Control and Disease Groups



Clinical Research

The use of GCF as a diagnostic aid in periodontal disease is nota new concept (30, 31). However, use of GCF as a potential tool fordiagnosis of periapical lesions of endodontic origin is a relativelynovel concept. Belmar et al (32) found significantly elevated levels ofmatrix metalloproteinases (MMP-9 and MMP-2) in the GCF of teethwith periapical lesions. Although the study by Belmar et al was not avail-able as a reference when the current study was designed and conducted,the methods were similar. MMPs in GCF might emerge as useful biolog-ical markers for monitoring apical periodontitis (32). Other investiga-tors have found elevated levels of inflammatory markers in the GCF ofteeth with a clinical diagnosis of irreversible pulpitis (21, 22).Specific organic matrix proteins and cytokines (osteopontin,osteoprotegerin, and receptor activator for nuclear factor kappa Bligand) have been found in higher levels in the GCF of teeth with rootresorption as a result of orthodontic movement (33).

The results of this study demonstrated a higher level of IL-1b andDSP in the diseased group compared with the control group, but theresults did not reach statistical significance. One possible explanationfor this finding could be the dynamics of the development of apical pe-riodontitis. This disease process is characterized by 2 distinct phases, anactive bone resorption phase and a chronic phase with little lesionexpansion (34). It is possible that IL-1b and DSP levels are elevatedonly during the active phase of the disease, and if this is the case,IL-1b and DSP might not be suitable markers for apical periodontitis.

Samples were prepared and diluted for the ELISA test in a way thatequalized the amount of total protein in control and disease samples.This step was necessary because IL-1b and DSP are not exclusivelyseen in cases of apical periodontitis. The same markers were detectedin patients with periodontal disease and root resorption. It was impor-tant to use a contralateral endodontically healthy tooth as a control ineach patient to rule out false positives as a result of other inflammatoryconditions.

Patients taking antibiotics and anti-inflammatory drugs wereincluded in this study. The rationale for this decision was that manypatients presenting for endodontic procedures are taking one orboth of these types of medication. Even though anti-inflammatory drugscould decrease the amount of inflammatory mediators such asinterleukin, these patients were included in the study.

Karapanou et al (22) showed an increased level of IL-8 (CXCL8) inthe GCF of patients with irreversibly inflamed pulps compared withhealthy contralateral teeth. An interesting finding in this study wasthat if samples were collected after the diseased tooth had received localanesthesia, the levels of CXCL8 dropped to levels similar to those foundon healthy teeth, which shows the influence of anesthesia on the levels ofCXCL8. We did not take the time of anesthesia into consideration in ourstudy. Even though most of the samples were collected before anes-thesia, some patients received anesthesia before collection, which mighthave influenced the results. Controlling for the influence of local anes-thesia and concurrent use of antibiotics or anti-inflammatory drugs arerelevant considerations for future research.

The presence of a significantly higher concentration of nonspecificprotein in the GCF of diseased teeth compared with control teeth was aninteresting finding and suggests a potential biochemical marker forperiapical disease. Although the current study was unable to demon-


strate a significant difference in IL-1b and DSP levels between diseasedand control teeth, future studies might identify other biochemicalmarkers for active apical periodontitis.

AcknowledgmentsThis research was funded in part by a grant from the American

Association of Endodontists Foundation.

References1. Barkhordar RA, Desouza YG. Human T-lymphocyte subpopulations in periapical

lesions. Oral Surg Oral Med Oral Pathol 1988;65:763–6.2. Metzger Z. Macrophages in periapical lesions. Endod Dent Traumatol 2000;16:1–8.3. Dinarello CA. Biology of interleukin 1. Faseb J 1988;2:108–15.4. Heath JK, Saklatvala J, Meikle MC, Atkinson SJ, Reynolds JJ. Pig interleukin 1

(catabolin) is a potent stimulator of bone resorption in vitro. Calcif Tissue Int1985;37:95–7.

5. Wang CY, Stashenko P. Characterization of bone-resorbing activity in human peri-apical lesions. J Endod 1993;19:107–11.

6. Barkhordar RA, Hussain MZ, Hayashi C. Detection of interleukin-1 beta in humanperiapical lesions. Oral Surg Oral Med Oral Pathol 1992;73:334–6.

7. Malueg LA, Wilcox LR, Johnson W. Examination of external apical root resorptionwith scanning electron microscopy. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 1996;82:89–93.

8. Balducci L, Ramachandran A, Hao J, Narayanan K, Evans C, George A. Biologicalmarkers for evaluation of root resorption. Arch Oral Biol 2007;52:203–8.

9. Qin C, Brunn JC, Cadena E, et al. The expression of dentin sialophosphoprotein genein bone. J Dent Res 2002;81:392–4.

10. Friedman S, Abitbol S, Lawrence HP. Treatment outcome in endodontics: theToronto Study—phase 1: initial treatment. J Endod 2003;29:787–93.

11. Pekruhn RB. The incidence of failure following single-visit endodontic therapy.J Endod 1986;12:68–72.

12. Bender IB, Seltzer S. Roentgenographic and direct observation of experimentallesions in bone: I—1961. J Endod 2003;29:702–6. discussion 701.

13. Bender IB, Seltzer S. Roentgenographic and direct observation of experimentallesions in bone: II—1961. J Endod 2003;29:707–12. discussion 701.

14. Estrela C, Bueno MR, Leles CR, Azevedo B, Azevedo JR. Accuracy of cone beamcomputed tomography and panoramic and periapical radiography for detectionof apical periodontitis. J Endod 2008;34:273–9.

15. Estrela C, Bueno MR, Azevedo BC, Azevedo JR, Pecora JD. A new periapical indexbased on cone beam computed tomography. J Endod 2008;34:1325–31.

16. Estrela C, Bueno MR, De Alencar AH, et al. Method to evaluate inflammatoryroot resorption by using cone beam computed tomography. J Endod 2009;35:1491–7.

17. Patel S. New dimensions in endodontic imaging: part 2—cone beam computedtomography. Int Endod J 2009;42:463–75.

18. de Paula-Silva FW, Wu MK, Leonardo MR, da Silva LA, Wesselink PR. Accuracy ofperiapical radiography and cone-beam computed tomography scans in diagnosingapical periodontitis using histopathological findings as a gold standard. J Endod2009;35:1009–12.

19. Cotton TP, Geisler TM, Holden DT, Schwartz SA, Schindler WG. Endodontic appli-cations of cone-beam volumetric tomography. J Endod 2007;33:1121–32.

20. Linden GJ, McKinnell J, Shaw C, Lundy FT. Substance P and neurokinin A in gingivalcrevicular fluid in periodontal health and disease. J Clin Periodontol 1997;24:799–803.

21. Awawdeh L, Lundy FT, Shaw C, Lamey PJ, Linden GJ, Kennedy JG. Quantitative anal-ysis of substance P, neurokinin A and calcitonin gene-related peptide in pulp tissuefrom painful and healthy human teeth. Int Endod J 2002;35:30–6.

22. Karapanou V, Kempuraj D, Theoharides TC. Interleukin-8 is increased in gingivalcrevicular fluid from patients with acute pulpitis. J Endod 2008;34:148–51.

23. Goldman M, Pearson AH, Darzenta N. Endodontic success: who’s reading theradiograph? Oral Surg Oral Med Oral Pathol 1972;33:432–7.

24. Patel S, Dawood A, Whaites E, Pitt Ford T. New dimensions in endodontic imaging:part 1—conventional and alternative radiographic systems. Int Endod J 2009;42:447–62.

25. Trope M, Delano EO, Orstavik D. Endodontic treatment of teeth with apical perio-dontitis: single vs multivisit treatment. J Endod 1999;25:345–50.

26. Waltimo T, Trope M, Haapasalo M, Orstavik D. Clinical efficacy of treatment proce-dures in endodontic infection control and one year follow-up of periapical healing.J Endod 2005;31:863–6.

27. Penesis VA, Fitzgerald PI, Fayad MI, Wenckus CS, BeGole EA, Johnson BR. Outcomeof one-visit and two-visit endodontic treatment of necrotic teeth with apical

Biologic Markers for Odontogenic Periradicular Periodontitis 1309

Clinical Research
periodontitis: a randomized controlled trial with one-year evaluation. J Endod 2008;34:251–7.
28. Peters LB, Wesselink PR. Periapical healing of endodontically treated teeth in oneand two visits obturated in the presence or absence of detectable microorganisms.Int Endod J 2002;35:660–7.

29. Weiger R, Axmann-Krcmar D, Lost C. Prognosis of conventional root canal treatmentreconsidered. Endod Dent Traumatol 1998;14:1–9.

30. Golub LM, Kleinberg I. Gingival crevicular fluid: a new diagnostic aid in managingthe periodontal patient. Oral Sci Rev 1976;49–61.

1310 Burgener et al.

31. Lamster IB. Evaluation of components of gingival crevicular fluid as diagnostic tests.Ann Periodontol 1997;2:123–37.

32. Belmar MJ, Pabst C, Martinez B, Hernandez M. Gelatinolytic activity in gingival crev-icular fluid from teeth with periapical lesions. Oral Surg Oral Med Oral Pathol OralRadiol Endod 2008;105:801–6.

33. George A, Evans CA. Detection of root resorption using dentin and bone markers.Orthod Craniofac Res 2009;12:229–35.

34. Stashenko P, Yu SM, Wang CY. Kinetics of immune cell and bone resorptiveresponses to endodontic infections. J Endod 1992;18:422–6.


Clinical Research

Clinical and Radiographic Evaluation of a Resin-based RootCanal Sealer: An Eight-year UpdateOsvaldo Zmener, DDS, Dr Odont,* and Cornelis H. Pameijer, DMD, MScD, DSc, PhD†

Abstract
Introduction: This retrospective clinical and radiographicstudy evaluated the 8-year outcome of one-visitendodontic treatment of root canals filled with gutta-percha and a methacrylate resin–based sealer (EndoREZ).Methods: From an initial sample size of 180 patients,subsequently 145 and 120 patients were evaluated after14–18 months and 5 years, respectively. Of the remainingpatient pool of 120 patients evaluated after 5 years, 112patients with 212 root canals responded to the 8-yearrecall. The outcome of treatments was assessed on thebasis of clinical and radiographic criteria. Endodonticsuccess was rated on the basis of absence of clinicalsymptoms, the presence of a normal or slightly widenedperiodontal ligament space, and absence or substantialreduction in size of preexisting periradicular radiolu-cencies. Teeth that did not meet these criteria wereconsidered endodontic failures. Results: The root canalshad been adequately filled to the working length in 90teeth (80.35%) and were short in 19 instances(16.96%). None of the roots showing apical extrusionof the sealer immediately postoperatively had radio-graphic evidence of the sealer in the periradicular tissuesafter 8 years. At recall, all patients were comfortable andfree of clinical symptoms. A life table analysis showeda cumulative probability of success of 86.5% after 8 years,with a 95% confidence interval of 79.0–92.0. Conclu-sions: The results of this retrospective clinical and radio-graphic study suggest that the tested methacrylate resin–based sealer used in conjunction with gutta-percha conesperformed similarly to conventional endodontic sealersduring a period of up to 8 years. (J Endod2010;36:1311–1314)
Key WordsEndodontic therapy, EndoREZ, methacrylate-basedsealers, root canal filling

From the *Postgraduate Program for Specialized Endodon-tics, Faculty of Medicine, School of Dentistry, University of ElSalvador Buenos Aires, Republica Argentina; and †ProfessorEmeritus, University of Connecticut School of Dental Medicine,Farmington, Connecticut.

Dr. Pameijer is a consultant for Ultradent Products Inc.Address requests for reprints to Dr Cornelis H. Pameijer,

Professor Emeritus, University of Connecticut School of DentalMedicine, 10 Highwood, Simsbury, CT 06070. E-mail address:[email protected]/$0 - see front matter



Success or failure rates of treatment modalities are an important part of evidence-based practice of endodontics. Numerous studies have been published evaluating

endodontic success and failure by using clinical and radiographic examination(1–5). Well-defined predetermined clinical and radiographic criteria offer a reliablemethod to evaluate the long-term results of endodontic therapy (2–4, 6–8). Apreliminary retrospective study on 180 patients (9) evaluated the results of endodontictreatment of 295 root canals filled with laterally condensed gutta-percha cones inconjunction with EndoREZ (Ultradent Products Inc, South Jordan, UT), a methacrylateresin–based endodontic sealer. After 14–24 months, 145 patients were evaluated fora follow-up examination. An overall success rate of 91% was reported. In a secondfollow-up study performed 5 years after initial therapy, 120 of 180 patients were avail-able for follow-up evaluation (10), and an overall success rate of 90% was reported.Because the outcome of endodontic treatment varies over time, the purpose of thisretrospective study was to obtain 8-year postoperative data on the same patient poolthat was previously evaluated.

Materials and MethodsOf the original patient pool (age range, 12–75 years) treated in private practice, 112

patients (44.64% male and 55.35% female) with 212 root canals were available for an 8-year follow-up examination during which they were clinically and radiographically eval-uated. Subjects were contacted by mail or telephone or e-mails were sent requesting theycome in for a follow-up examination. Preoperative radiographs were taken during theinitial treatment, and the status of pulp and periradicular areas was recorded. All treat-ments had been completed in a single visit. After administration of local anesthesia,rubber dam was placed, and the pulp chamber was accessed. The canals were hand-instrumented with a crown-down technique for radicular access combined witha step-back technique for apical preparation. The coronal two thirds were first flaredwith #1-3 Gates Glidden drills (Dentsply/Maillefer, Ballaigues, Switzerland), and theworking length was established with a #15 file, approximately 1 mm short of the radio-graphic apex. Canal preparation was made with K-type and Hedstrom files (Dentsply/Maillefer) at the apical third to a master apical #30-40 file and coronally to a #60 file.On occasion, the instrumentation sequence was modified because of difficulty in nego-tiating root canals with complex anatomy. Patency was confirmed with a #10 K-file. Irri-gation was performed after every change of instrument by using 2.0 mL of 2.5% NaOClfollowed by rinsing with 2.0 mL of sterile saline. After instrumentation, a final copiousrinse with saline was performed. The irrigation solutions were administered with sterileplastic syringes and through 30-gauge endodontic irrigation needles. Excess irrigationsolution was removed with sterile paper points; however, the canal walls were kept slightlymoist to take maximum advantage of the hydrophilic properties of the resin sealer, thusallowing for deep penetration into the dentinal tubules and promoting a better seal. Thecanals were then filled with lateral condensation of gutta-percha cones andEndoREZ as the sealer. The access cavities were temporarily sealed with IRM (Dents-ply/Caulk, Milford, DE), and the patients were instructed to see their referring dentistfor definitive restorative care.

During the follow-up evaluation, a clinical examination was performed (percus-sion), and radiographs were made. Postoperative and recall radiographs were madeby using the same x-ray unit with a film-holder attached to beam-guiding XCP instrument(Rinn Corp, Elgin, IL) and Kodak 32 � 43 mm ultraspeed films (Eastman Kodak

Clinical and Radiographic Evaluation of Resin-based Root Canal Sealer 1311


TABLE 1. Tooth Number and Location in the Maxillary or Mandibular ArchEvaluated 8 Years Postoperatively

Maxillary Mandibular Total

Central incisor 18 2 20Lateral incisor 9 1 10Canine 12 5 17First premolar 5 9 14Second premolar 9 9 18First molar 8 10 18Second molar 4 8 12Third molar — 3 3Total 65 47 112

TABLE 3. Relation of Preoperative Factors to Treatment Results in Root CanalsFilled with Gutta-Percha and EndoREZ

Clinical Research

Company, Rochester, NY). The immediate and 8-year postoperativeradiographs were compared in a darkened room by using an illumi-nated x-ray viewer with a magnifying glass. All radiographs were analyzedby 2 independent and calibrated endodontists with more than 25 yearsof clinical experience. Calibration was carried out by having the evalu-ators analyze twice a standard set of 110 individual pairs of postoperativeand recall radiographs of endodontic treatments that were randomlyselected from the files of 2 private and 1 postgraduate endodonticservices. To meet the inclusion criteria, the radiographs had to be ofhigh quality and had to clearly exhibit periapical tissues, widened peri-odontal space, loss of cortical bone, changes in trabecular patterns, oreasily discernible periapical radiolucencies. If there was a disagreementbetween the evaluators, the x-rays were assessed jointly until a consensuswas reached. If necessary, additional radiographs were made at differenthorizontal angulations to improve visualization, thus improving the reli-ability of the evaluation. The level of the root canal fillings in relation tothe working length was recorded, and the quality of the fillings wasjudged to be adequate when they were placed to the full working lengthand no voids were detected, while special attention was focused on thelast 5 mm of the root canal. Canals that did not meet these conditionswere categorized as filled short or inadequate. Failure of 1 canal in mul-tirooted teeth was considered a complete failure, regardless whetherother canals were rated successful. In cases with apical radiolucencies,the size of the lesions was estimated on the radiographs as being <2 or>2 mm. Success or failure of the endodontic treatment was determinedon the basis of radiographic findings and clinical signs and symptomsaccording to the following criteria.

For success, (1) radiographically, the contours and width of theperiodontal ligament (PDL) space were within normal limits or slightlywidened around an accidental overfill, and the patient was free of symp-toms. Slight tenderness to percussion for a brief postoperative period wasconsidered acceptable. (2) The size of a preoperative radiolucent areadecreased by at least 50% and the patient was free of symptoms, or thecontours and width of the PDL space had returned to normal. (3) Absenceof preoperative periapical radiolucency remained unchanged over time.

For failure, (1) periapical radiolucency was observed in thepreoperative radiograph and remained unchanged or increased insize over time or (2) there was a root that, in absence of preoperativeperiapical pathosis, developed a radiolucency over time.

TABLE 2. Outcome of Treatment by Gender and Age in Root Canals Filled withGutta-Percha and EndoREZ after 8 Years

Factor No. of cases (%) Success (%) Failure (%)

GenderMale 50 (44.64) 48 (96.00) 4 (8.00)Female 62 (55.35) 57 (91.93) 3 (4.83)

Age (y)12–30 19 (16.96) 17 (89.47) 2 (10.52)31–55 61 (54.46) 58 (95.08) 3 (4.91)56–75 32 (28.57) 30 (93.75) 2 (6.25)

1312 Zmener and Pameijer

The clinical and radiographic data recorded by the 2 examinerswere analyzed for interexaminer agreement. The correlation of treat-ment outcomes with respect to age, gender, and specific preoperativeand postoperative data were analyzed by the Fisher exact test (P <.05). Taking into consideration the censored data, ie, the total numberof patients who did not respond to the previous 14- to 24-month and5-year recalls (35 and 25, respectively) (9,10), a life table survivalanalysis was used to determine the cumulative probability of successof the 8-year recall. A corresponding 95% confidence interval wasdetermined.

ResultsThe examiner calibration showed an interexaminer agreement

ratio of 92%, revealing a strong interobserver agreement. Therefore,the radiographic interpretation of the results was considered reliable.The recall rate after 8 years was 62.22%. A total of 112 patients with212 treated root canals presented for follow-up evaluation. All datacollected from the 112 patients were tabulated, and the tooth locationswere noted. The number and location of teeth that were evaluated areshown in Table 1. Distribution of patients by age and gender is pre-sented in Table 2. Distribution by significant preoperative and postop-erative factors related to treatment results is shown in Tables 3 and 4,respectively. Fig. 1 is representative of the successful treatment of a lowermolar. A postoperative glass ionomer restoration was replaced some-times after 5 years with a bonded resin composite filling because thegeneral practitioner judged the glass ionomer restoration in need ofreplacement as a result of breakdown.

Ninety teeth (80.35%) were evaluated as adequately filled to theworking length. In 19 cases (16.96%) the apical limit of the root fillingmaterial was found to be short of the working length. Fifteen (13.39%)of these, which were filled flush at the time of endodontic treatment,underwent slight resorption of the sealer within the lumen of the canals.These cases showed that the end of the root fill was located at�2 mmfrom the radiographic apex. Three cases in which extrusion of the sealerwas radiographically established immediately after treatment showedno radiographic evidence of the sealer in the periradicular tissues.Forty-nine teeth (43.75%) with preoperative vital pulps were successfulin 46 cases, whereas 63 (56.25%) with preoperative nonvital pulpswere successful in 59 cases. Forty-six teeth (41.07%) with preoperativeperiapical radiolucencies revealed almost total or total healing in 43cases, whereas 3 of them were evaluated a failure clinically and radio-graphically. Sixty-six teeth (58.92%) without preoperative periapicalradiolucent areas were successful in 62 instances. In 7 of these, a slightwidening of the PDL space was noted, but the teeth were asymptomaticand the radiographs showed the presence of well-defined cortical bone.The remaining 4 teeth were considered a failure clinically and radio-graphically. Overall, after 8 years, all patients were clinically

FactorNo. of

cases (%)Success

(%)Failure

(%)

Pulp diagnosisVital 49 (43.75) 46 (93.87) 3 (6.12)Nonvital 63 (56.25) 59 (93.65) 4 (6.34)

Periapical radiolucencyPresent 46 (41.07) 43 (93.47) 3 (6.52)Absent 66 (58.92) 62 (93.93) 4 (6.06)

Lesion size<2 mm 38 (82.60) 35 (92.10) 3 (7.89)>2 mm 8 (17.39) 4 (50.00) 4 (50.0)


TABLE 4. Relation of Final Restoration to Treatment Results in Root Canals Filled with Gutta-Percha and EndoREZ

Restoration No. of teeth (%) Success (%) Failure (%)

None 2 (1.78) 2 (100)Post (with or without crown) 48 (42.85) 46 (95.83) 2 (4.16)Coronal filling (amalgam, composite, glass ionomer, etc) 62 (55.35) 59 (95.16) 3 (4.83)

Clinical Research

comfortable. The differences in the outcome of treatments related toage, gender, preoperative pulp or periapical status, the size of periapicallesions, and the type of permanent restorations were not statisticallysignificant (P > .05). The life table analysis revealed a cumulative prob-ability of success of 86.5% at the 8-year recall, with a 95% confidenceinterval of 79.0–92.0.

Figure 1. (A) Preoperative radiograph of mandibular left second molar. (B)Immediate postoperative view of root canal filling. Tooth was restored withglass ionomer cement. (C) Five-year follow-up radiograph showing no abnor-malities. (D) Eight-year recall radiograph demonstrating normal periapicalcondition. Postoperative glass ionomer restoration was replaced sometimeafter 5 years with bonded resin composite filling because the general practi-tioner judged the glass ionomer restoration in need of replacement as a resultof breakdown.

DiscussionThis retrospective 8-year clinical and radiographic cohort study of

a methacrylate-based endodontic sealer and gutta-percha was consid-ered reliable and demonstrated a stable outcome of treatment asdefined per parameters outlined by Ørstavik (11). Using a method ofevaluating consenting patients following a predetermined clinical andradiographic protocol is considered a reliable procedure when evalu-ating the outcome of endodontic treatment (2–4, 6–8, 12), especiallybecause the evaluation criteria are currently being used by clinicians. Inthis respect, 2 recent histologic investigations (7, 12) demonstrateda good correlation between radiographic success and the histologicstatus of the periapical tissues in humans.

In common with a previous report (10), the current study was de-signed to show whether EndoREZ can be recommended for routine usein clinical endodontics. The recall rate of 62.22% after 8 years waswithin the American Dental Association requirements for subject sizein clinical trials as reported by Franco et al (13) and met the requiredstandards for evidence levels (14). It was also comparable to that inprevious endodontic follow-up studies (1–5, 11, 15) and is inagreement with the study by Ørstavik (11) in that the recall rates infollow-up studies were substantially reduced as the recall periodincreased. The influence of the recall rates on the results of the currentstudy deserves some discussion. The 8 patients who were not evaluatedeither could not be located or did not respond to recall request. Thismight mean that these patients were without symptoms, they had relo-cated, or they had returned to the referring dentist when problemsoccurred. When a patient does not respond to a recall, there is alwaysthe possibility that one is dealing with an endodontic failure, and there-fore, the data that were generated might not be totally representative ofthe actual results. It should be noted, however, that the results ofendodontic treatments in patients who did not return for follow-up(censored data) are not considered representative of a particular treat-ment result category (5). It should also be pointed out that the 8 patientswho could not be evaluated at this recall were seen at the 5-year follow-up evaluation and categorized as endodontically successful (10).

Data related to the type and location of teeth were pooled becauseit has been shown that these factors did not skew the outcome ofendodontic treatment (3–6). Factors such as gender and age did notnegatively affect the results of the study. These observations are inagreement with our previous findings (10) and with those of others(5, 6, 15, 16). Furthermore, no significant differences were foundbetween teeth with vital and nonvital pulps, as has been previouslyreported by Barbakow et al (3) and Sjogren et al (6). The presenceof a preoperative apical radiolucent area did not appear to adverselyaffect the outcome of endodontic treatment. This observation is insupport of our previous findings (9, 10) but disagrees with others(1, 5, 17, 18) who found significantly lower success rates in teeth


with infected root canals and preexisting periapical pathosis.However, our results are in agreement with Sjogren et al (6) andPeak et al (19), who showed that the prognosis of teeth with nonvitalpulps and preexisting periapical radiolucent areas was as good asthat for vital teeth. We can hypothesize that factors such as early coronalflaring complemented with a careful instrumentation technique inwhich the incremental removal of the bulk of infected root dentin,thus allowing for a more effective penetration of irrigants, as well asthe previously reported tight seal provided by EndoREZ (20), mighthave contributed to a more favorable condition for periapical healing.

Of further interest is that extrusion of EndoREZ, which accidentallyoccurred in 10 cases at the initiation of the study, did not show anadverse effect on the outcome of treatments. This is in contradictionwith some authors who stated that extrusion of root filling materialmight interfere with the repair process (17, 21, 22). After 8 years,

Clinical and Radiographic Evaluation of Resin-based Root Canal Sealer 1313

Clinical Research
however, all these cases appeared radiographically normal withoutevidence of sealer in the periapical tissues. These findings suggestthat the lack of adverse effects from the extruded EndoREZ can beattributed to the good tissue compatibility of the sealer, as has beendemonstrated in previous animal studies (23, 24). In the currentstudy, all patients were treated in a single visit. Our results tend tosupport previous evidence that the single-visit endodontic therapyconstitutes a reliable procedure (25–29), even in cases with infectedroot canals and preexisting periradicular pathosis. In this respect,more recent evidence provided by Molander et al (30) and a Cochranesystematic review by Figini et al (31) showed that the outcome of treat-ment was not significantly influenced whether endodontic therapy wasperformed during a single or multiple visit protocol.
Previous studies (5, 6) reported that the type of coronalrestoration (single coronal restoration, presence or absence ofa post in the canal) did not significantly affect the outcome ofendodontic treatment. In this study, 55.35% presented with singlemetal/ceramic, amalgam, and resin composite or glass ionomercoronal fillings, whereas in 42.85%, posts were present. Two caseswere classified as failures. These cases did not show periapicalradiolucencies at the time of the initial treatment, whereas at the 5-year follow-up (10) the patients were asymptomatic, with no radio-graphic changes in the periapical tissues and with teeth showingadequate coronal fillings. Therefore, they were evaluated as successfulafter 5 years, whereas at the 8-year recall these teeth presented withoutcoronal restoration and radiographically detectable periapical radiolu-cent areas. Feedback from these patients revealed that the coronal fill-ings were lost, and the root canals were exposed to saliva fora prolonged period of time. This observation suggests that although En-doREZ offers a good adaptation to the root canal walls (32, 33–35),treatment failure might occur as a result of coronal bacterialpenetration caused by the loss of coronal protection.

In conclusion and within the limitations of this clinical and radio-graphic study, the results suggest that EndoREZ used in conjunctionwith gutta-percha constitutes an acceptable root canal filling procedure.Patients recalled after 8 years reported being comfortable with thetreated teeth, which continued to be functional. The sealer seems tobe well-tolerated by periapical tissues even in cases of accidental extru-sion beyond the apical foramen. Furthermore, the success rate wascomparable to what had been reported previously (4, 5, 19, 36–38)with different endodontic sealers.

AcknowledgmentsThe authors would like to thank Prof. R. Macchi for his invalu-

able input in the statistical analysis of the data.

References1. Grossman LI, Shepard LI, Pearson LA. Roentgenologic and clinical evaluation of

endodontically treated teeth. Oral Surg Oral Med Oral Pathol 1964;17:368–74.2. Heling B, Tamshe A. Evaluation of the success of endodontically treated teeth. Oral

Surg Oral Med Oral Pathol 1970;30:533–6.3. Barbakow FH, Cleaton-Jones P, Friedman D. An evaluation of 566 cases of root

canal therapy in general dental practice: 2—postoperative observations. J Endod1980;6:485–9.

4. Swartz DB, Skidmore AE, Griffin JA. Twenty years of endodontic success and failure.J Endod 1983;9:198–202.

5. Friedman S, Lost C, Zarrabian M, Trope M. Evaluation of success and failure afterendodontic therapy using a glass ionomer cement sealer. J Endod 1995;21:384–90.

6. Sjogren U, Hagglund B, Sundqvist G, Wing K. Factors affecting the long-term resultsof endodontic treatment. J Endod 1990;16:498–504.

7. Green TL, Walton RE, Taylor JK, Merrel P. Radiographic and histologic periapicalfindings of root canal treated teeth in cadaver. Oral Surg Oral Med Oral Pathol1997;83:707–11.

1314 Zmener and Pameijer

8. Ørstavik D, Qvist V, Stolze K. A multivariate analysis of the outcome of endodontictreatment. Eur J Oral Sci 2004;112:224–30.

9. Zmener O, Pameijer CH. Clinical and radiographic evaluation of a resin-based rootcanal sealer. Am J Dent 2004;17:19–22.

10. Zmener O, Pameijer CH. Clinical and radiographical evaluation of a resin-based rootcanal sealer: a 5-year follow-up. J Endod 2007;33:676–9.

11. Ørstavik D. Time-course and risk analyses of the development and healing ofchronic apical periodontitis in man. Int Endod J 1996;29:150–5.

12. Ricucci D, Lin LM, Spangberg LSW. Wound healing of apical tissues after root canaltherapy: a long-term clinical, radiographic and histopathologic observation study.Oral Surg Oral Med Oral Pathol 2009;108:609–21.

13. Franco EB, Benetti AR, Ishikiriama SK, et al. 5-year clinical perfomance of resincomposite versus resin modified glass ionomer restorative system in non-cariouscervical lesions. Oper Dent 2006;31:403–8.

14. Friedman S. Expected outcomes in the prevention and treatment of apical periodon-titis. In: Ørstavik D, Pitt Ford T, eds. Essential endodontology: prevention and treat-ment of apical periodontitis. 2nd ed. Frederiksberg, Denmark: BlackwellMunksgaard Ltd; 2008:408–69.

15. Selden HS. Pulpoperiapical disease: diagnosis and healing—a clinical endodonticstudy. Oral Surg Oral Med Oral Pathol 1974;27:271–83.

16. Kerekes K, Tronstad L. Long term results of endodontic treatment performed witha standardized technique. J Endod 1979;5:83–90.

17. Seltzer S, Bender IB, Turkenkopf S. Factors affecting successful repair after rootcanal therapy. J Am Dent Assoc 1963;67:651–62.

18. Seltzer S, Bender IB, Smith J, Freedman I, Nazimov H. Endodontic failures: an anal-ysis based on clinical, roentgenographic, and histologic findings. Oral Surg OralMed Oral Pathol 1967;23:517–30.

19. Peak JD, Hayes SJ, Bryant ST, Dummer PMH. The outcome of root canal treatment:a retrospective study within the armed forces (Royal Air Force). Br Dent J 2001;190:140–4.

20. Pameijer CH, Zmener O. Current status of methacrylate-based sealers and obtura-tion techniques. Pract Proced Aesthet Dent 2006;18:674–6.

21. Storms JL. Factors that influence the success of endodontic treatment. J Can DentAssoc 1969;35:83–97.

22. Seltzer S. Long term radiographic and histological observations of endodonticallytreated teeth. J Endod 1999;25:818–22.

23. Louw NP, Pameijer CH, Norval G. Histopathological evaluation of a root canal sealerin subhuman primates (abstract). J Dent Res 2001;79:654.

24. Zmener O, Banegas G, Pameijer CH. Bone tissue response to a methacrylate-basedendodontic sealer: a histological and histometric study. J Endod 2005;31:457–9.

25. Oliet S. Single–visit endodontics: a clinical study. J Endod 1983;9:147–52.26. Soltanoff W. A comparative study of the single-visit and the multiple-visit endodontic

procedure. J Endod 1978;4:278–81.27. Pekruhn RB. The incidence of failure following single-visit endodontic therapy.

J Endod 1986;12:68–72.28. Weiger R, Rosendahl R, Lost C. Influence of calcium hydroxide intracanal dressing

on the prognosis of teeth with endodontically induced periapical lesions. Int Endod J2000;33:219–26.

29. Field JW, Gutmann JL, Solomon ES, Rauskin H. A clinical radiographic retrospectiveassessment of the success rate of single–visit root canal treatment. Int Endod J 2004;37:70–82.

30. Molander A, Warfvinge J, Reit C, Kvist T. Clinical and radiographic evaluation of one-and-two visit endodontic treatment of asymptomatic necrotic teeth with apical perio-dontitis: a randomized clinical trial. J Endod 2007;33:1145–8.

31. Figini L, Lodi G, Gorni F, Gagliani M. Single versus multiple visits for endodontic treat-ment of permanent teeth: a Cochrane systematic review. J Endod 2008;34:1041–7.

32. Zmener O, Pameijer CH, Macri E. Evaluation of the apical seal in root canalsprepared with a new rotary system and obturated with a methacrylate-basedendodontic sealer: an in vitro study. J Endod 2005;31:392–5.

33. Gillespie WT, Loushine RJ, Weller RN, et al. Improving the performance of EndoRezroot canal sealer with a dual-cured two-step self-etch adhesive. II—apical andcoronal seal. J Endod 2006;32:771–5.

34. Zmener O, Pameijer CH, Alvarez Serrano S, Vidueira M, Macchi RL. Significance ofmoist root canal dentin with the use of methacrylate-based endodontic sealers: an invitro coronal dye leakage study. J Endod 2008;34:76–9.

35. Herbert J, Bruder M, Braunsteiner J, Altenburger MJ, Wrbas K-T. Apical quality andadaptation of Resilon, EndoRez, and Guttaflow root canal fillings in combinationwith a noncompaction technique. J Endod 2009;35:261–4.

36. Ørstavik D, Kerekes K, Eriksen HM. Clinical perfomance of three endodontic sealers.Endod Dent Traumatol 1987;3:178–86.

37. Augsburger RA, Peters DD. Radiographic evaluation of extruded obturation mate-rials. J Endod 1990;16:492–7.

38. Huumonen S, Lenander-Lumikari M, Sigurdsson A, Ørstavik D. Healing of apical pe-riodontitis after endodontic treatment: a comparison between a silicone-based andzinc oxide-eugenol-based sealer. Int Endod J 2003;36:296–301.


Clinical Research

Influence of a Passive Sonic Irrigation System on theElimination of Bacteria from Root Canal Systems: A ClinicalStudyS. Kirk Huffaker, DMD, MDS, Kamran Safavi, DMD, MEd, Larz S.W. Spangberg, DDS, PhD,and Blythe Kaufman, DMD, MDS

Abstract
Introduction: The present investigation evaluated theability of a new passive sonic irrigation (sonic group)system (EndoActivator) to eliminate cultivable bacteriafrom root canals in vivo and compared it with that ofstandard syringe irrigation (control group). Methods:Data were obtained by using bacteriologic sampling ofroot canals treated by endodontic residents. Samplingresults from 1 session of treatment were then comparedwith results obtained after intervisit calcium hydroxidedisinfection and a second session of treatment. Results:There was no significant difference in the ability of sonicgroup and control group to eliminate cultivable bacteriafrom root canals (P > .05). A second session and inter-visit calcium hydroxide disinfection were able to elimi-nate cultivable bacteria from significantly more teeththan a single session of treatment (P < .05). Conclu-sions: These in vivo results strengthen the case fora multi-visit approach to the treatment of apical perio-dontitis. (J Endod 2010;36:1315–1318)
Key WordsBacteria, culture, EndoActivator, endodontic treatment,sonic irrigation

From the Division of Endodontology, University of Connect-icut Health Center, Farmington, Connecticut.

Address requests for reprints to Dr Blythe Kaufman, Divi-sion of Endodontology, University of Connecticut HealthCenter, 263 Farmington Ave, Farmington, CT 06030-1715.E-mail address: [email protected]/$0 - see front matter


JOE — Volume 36, Number 8, August 2010 In

Apical periodontitis is the defense mechanism the human body has developed tokeep microbial infection of the root canal system from spreading beyond the apical

foramen. Studies have shown that periradicular inflammation will not occur withoutinvasion of the root canal system by microorganisms (1, 2). The goal of endodontictreatment is, therefore, the elimination of viable bacteria from the root canal system (3).

The timing of when to complete nonsurgical root canal treatment of teeth withapical periodontitis is controversial (4, 5). In controlled, clinical settings, it hasbeen shown that treatment in at least 2 visits with calcium hydroxide disinfectionresults in improved healing rates (6). On the other hand, meta-analysis suggests thereis no statistical difference in the healing rates of single-visit and multi-visit treatments(7). Regardless of when treatment is completed, healing of apical periodontitis ismore likely to occur if, before obturation, the root canal system has been disinfectedto a level in which bacteria can no longer be cultured (8, 9).

The quest to completely eliminate bacteria from root canal systems has resulted insome novel treatment modalities (10, 11). The phenomenon of acousticmicrostreaming and cavitation inside of irrigant-filled root canals has been investigated(12). When cavitation bubbles are produced by acoustic waves, they eventuallycollapse, and the energy released is transferred to the root canal wall, liberating anydebris found thereon (13). Microstreaming then carries the debris coronally so thatit can be removed from the canal (14). Acoustic cavitation has been shown to removeand destroy biofilm (15).

The EndoActivator (EA) (Dentsply/Tulsa Dental Specialties, Tulsa, OK) is a cord-less, battery-powered handpiece with a sonic motor. It has been developed with thehope of ‘‘safer, better, and faster.debridement and disruption of the smear layerand biofilm’’ (16). This statement is based on the proposed ability of EA to producecavitation and acoustic streaming inside of root canal systems. Clinical, peer-reviewed research is warranted to substantiate this claim.

The purposes of this study were (1) to evaluate whether the addition of EA to stan-dard chemomechanical instrumentation results in a greater elimination of cultivablebacteria from root canals compared with standard irrigation (control group) and(2) to compare the ability of one-session treatment to eliminate cultivable bacteriawith that of a second session with calcium hydroxide disinfection.

Materials and MethodsApproval for the project was obtained from the Institutional Review Board of the

University of Connecticut Health Center. A power analysis (17) before beginning thestudy resulted in a sample size of 42 for each independent sample (sonic group orcontrol group) for a total patient recruitment of 84. The 2 groups were randomizedby using a computer program (www.randomizer.com), and patients were assigned ac-cording to the randomization sequence. Patients with any tooth with apical periodonti-tis, verified with a radiograph and negative cold test, were recruited for the study. Teethwith apical periodontitis previously treated were excluded. Patient consent was obtainedbefore recruitment. All treatment was performed by 1 of 10 endodontic residents.

Each tooth was isolated with rubber dam and disinfected with 30% hydrogenperoxide and 5% iodine tincture to eliminate surface contaminants (18). All caries

fluence of Passive Sonic Irrigation System on Elimination of Bacteria from Root Canal Systems 1315

http://www.randomizer.com


Clinical Research
and previous restoration were then removed, and the tooth was disin-fected again as before. The tooth was then irrigated with 5% sodiumthiosulfate to inactivate the iodine, and a bacteriologic sample was takenby using sterile paper cones and liquid thioglycolate broth enrichedwith vitamin K1 and hemin in culture tubes. All samples were obtainedby the same operator (S.K.H.) throughout the duration of the study tostandardize the culture technique.
Following disinfection protocol, access was gained to the rootcanal system, and the canals were preflared slightly to allow spacefor paper cones to enter. The canals were then filled with sterile saline.The contents were absorbed into sterile paper cones placed at the mostapical extent of the canal until the canal was dry; the saturated coneswere deposited into a culture tube. This sample was taken to confirmthe presence of bacteria in the system. In cases with multiple canals,each canal was sampled for viable bacteria.

Standard clinical instrumentation protocol followed the secondbacterial culture. This involves obtaining working length (WL) approx-imately 1 mm short of the radiographic apex. This was established withan apex locator (Raypex 4, Johnson City, TN) and confirmed by radio-graph. Root canals were then chemomechanically instrumented withhand (Kontrolflex K-file; Brasseler USA, Savannah, GA) and EndoSe-quence rotary instruments (Brasseler USA) to WL by using 0.5% sodiumhypochlorite (NaOCl). The size of the master apical file (MAF) wasdetermined by each of the 10 clinicians. Needle irrigation was per-formed with a 27-gauge side-vented monojet needle (Kendall, MansfieldMA). When chemomechanical instrumentation was complete, a treat-ment card with either ‘‘EndoActivator’’ or ‘‘Standard Irrigation’’ printedon it was brought to the resident performing treatment. Both the resi-dent and the operator taking the bacteriologic sample were blindedas to which treatment group the tooth would be assigned before thispoint.

If the card indicated ‘‘EndoActivator,’’ the following protocol wasfollowed. Each canal was filled with NaOCl, and then EA was insertedinto the canal and activated for 30 seconds at 10,000 cpm. Time waskept with a timer. After 30 seconds, a fresh solution of NaOCl was intro-duced into the canals, and EA was again activated for 30 more secondsfor a total of 1 minute. This is in accordance with manufacturer recom-mendation of 1-minute activation per solution per canal. After activa-tion, canals were then flushed with sterile saline and dried withsterile paper cones. If the card indicated ‘‘Standard Irrigation,’’ thesame protocol was followed but without the use of EA in the canal.

Each canal was then flushed with 5% sodium thiosulfate, driedwith paper cones, and then filled with sterile saline. A hand file equalin size to the MAF was then inserted into the canals and scraped againstthe canal walls to remove any debris and bacteria from the canal walls(18). The contents of the canal were then absorbed into paper conesand placed into the culture tubes as before. All samples were then incu-bated for 7 days and observed for growth daily by the study coordinator.No growth was assigned if turbidity was absent on the seventh day.

Following this third sample, each canal was filled with a slurry ofcalcium hydroxide (Henry Schein, Melville, NY) applied with a lentulo

TABLE 1. Culture Results for EndoActivator (PSI) and Standard Irrigation (SI) Gr

Protocol PSI Count% within protocol

SI Count% within protocol

Total Count% within protocol

1316 Huffaker et al.

spiral or hand file, and the tooth was temporarily restored with a 4-mmlayer of Cavit (3M ESPE, St Paul, MN) covered by Fuji IX (GC Company,Tokyo, Japan). The patient was then scheduled for a second appoint-ment at least 2 weeks later.

At the second session, a surface disinfection sample was againtaken before reaccessing the tooth. The teeth were then reaccessed,and the canals were irrigated copiously along with any additional instru-mentation to remove calcium hydroxide and remaining debris. The EAwas not used during the second visit in either of the treatment groups.Needle irrigation with NaOCl was the sole means of irrigation. A postme-dication bacteriologic sample was then obtained, and the treatment wascompleted with root fillings placed.

Six teeth with no signs of radiographic periapical inflammationand vital pulps confirmed by the presence of vital tissue on entry intothe pulp chamber were used for negative controls. Three teeth were as-signed to the sonic group and 3 to the control group. The entire exper-imental protocol as described above was performed on these 6 teeth.

Data were analyzed (SPSS Statistical pack 17.0; SPSS Inc, Chicago,IL) by using independent and paired-sample t tests. Hypotheses weretested at the .05 level of significance.

ResultsA negative culture was obtained for surface disinfection in 96.5%

of the samples. All controls tested negative for growth at the end of thefirst and second visits.

Bacteria were initially present in the root canals of all 84 teethtreated. Ten patients refused to appear for the second session andwere excluded from the data comparing one- and two-session treat-ment. The data comparing sonic group and control group at the firstsession are found in Table 1. After activation of NaOCl with EA, 25 teeth(60%) still harbored cultivable bacteria compared with 22 teeth (52%)for the control group. These differences were not significant (P > .05).

The data comparing one- and two-visit treatment are found inTable 2. At the end of the first session, 47 teeth (56%) still harboredcultivable bacteria. After calcium hydroxide disinfection and a secondsession of treatment, 20 teeth (27%) still harbored cultivable bacteria.This difference was significant (P < .05).

There was no significant difference found at the second visitbetween teeth assigned to sonic group or control group. At the endof the second session, 9 of 36 and 11 of 38 teeth still harbored bacteriain the sonic group and control group, respectively.

DiscussionWhen dealing with nonsurgical root canal treatment, clinicians are

usually faced with 2 situations. The first is a tooth with a vital pulp inwhich the tissue found in the root canal is inflamed but not completelyinfected by microorganisms. When these types of cases present them-selves, the treatment is relatively straightforward. In vital cases, all ofthe clinician’s effort is spent removing the sterile tissue asepticallyand not introducing microorganisms into the root canal. This type of

oups

Culture result, 1st visit

Negative Positive Total

17 25 4240.5 59.5 100.020 22 4247.6 52.4 100.037 47 8444.0 56.0 100.0


TABLE 2. Culture Results for One- and Two-visit Treatment Groups

Culture result

Negative Positive Total

Treatment group One-visit Count 37 47 84% within visit result 44.0 56.0 100.0

Two-visit Count 54 20 74% within visit result 73.0 27.0 100.0

Total Count 91 67 158% within visit result 57.6 42.4 100.0

Clinical Research

treatment can usually be completed in 1 treatment session (19). Theother situation is more complex. When a patient presents with a tooththat has been infected by bacteria causing periapical bone breakdown,the clinician must use all the means he/she has to kill and remove theinvading bacteria and their inflammatory by-products from the canalsystem.

In recent years, it has been suggested that files attached to ultra-sonic handpieces be used to aid in the irrigation and debridement ofinfected root canals (20, 21). Recently, the EA has beenrecommended to enhance the cleaning efficacy of irrigation of rootcanal systems. Its proposed ability to create sonic waves in irrigatingsolutions deposited inside of the root canal might aid in the killing ofbacteria and debridement of necrotic tissue. In the current study itwas not shown that EA improved the ability to eliminate cultivablebacteria from root canal systems. Our study found no significantdifference between the number of negative cultures obtained bystandard irrigation and the number obtained with use of EA. This isconsistent with the results of a recent in vitro study evaluatingbacterial removal in simulated canals by using both EA and needleirrigation (22).

One reason for this might be that EA produces only sonic waves.Ultrasonic instruments have been shown to enhance the cleaning effi-ciency of irrigation (23, 25). Stamos et al (24) compared the use ofsonically and ultrasonically activated instruments and found that theultrasonically activated instruments removed significantly more debristhan those that were activated sonically. Node production along acti-vated files is an important part of acoustic streaming (12), resultingin a strong current produced along the activated instrument (14). Ifthe instrument touches the canal wall, the node in the immediate vicinitywill be diminished (26). Because it is inevitable that the file will touchthe canal wall, it is important to create several nodes along the instru-ment being activated. Ultrasonic energy has the ability to create severalnodes along the length of the file (27). Sonic energy only has the powerto produce 1 node along the length of the instrument, so any constraintof the instrument will significantly decrease, if not eliminate, theacoustic streaming necessary to dislodge and carry away necrotic debris(27).

EA might not be powerful enough to disrupt bacterial biofilms. Ah-mad et al (28) showed that ultrasonically activated instruments couldnot disrupt bacteria but simply dispersed it to other areas of the canal.Even with ultrasonics, Mayer et al (25) found that only the coronal thirdof the canal was being cleaned. In that study there was no significantdifference between syringe irrigation and ultrasonic irrigation in theapical third of the canal. This might be due to the activated file touchingthe canal wall in the apical third and not being able to produce thenecessary nodes for acoustic streaming and cavitation (14). If ultra-sonic instruments with their constant power supply and increasednode production cannot effectively clean in the apical third, it is likelythat EA with its battery power and sonic engine will have the sameproblem.

JOE — Volume 36, Number 8, August 2010 Influence of Passive

The ability of a second session of instrumentation and irrigationtogether with an interappointment medication of calcium hydroxidewas compared with a single session of instrumentation and irrigation.Studies have shown that when treatment of necrotic pulps is performedin 2 sessions with calcium hydroxide disinfection, there is a reduction inintracanal bacteria and a greater likelihood of obtaining a negativeculture (29). Calcium hydroxide raises the pH of the root canal systemto a level at which many microorganisms cannot survive (30). It has alsobeen found that when necrotic tissue has been in direct contact withcalcium hydroxide, it becomes more soluble and susceptible to disso-lution by NaOCl (31). In the present study it was shown that the additionof calcium hydroxide together with another round of instrumentationand irrigation was effective at eliminating cultivable bacteria from signif-icantly more teeth.

In the current study, two-visit treatment with calcium hydroxidedisinfection was able to eliminate cultivable bacteria from about 75%of teeth exhibiting apical periodontitis. This is comparable with somestudies (29) and is a lower number than other studies (32). It is knownthat some bacteria are more resistant to the high pH of calciumhydroxide (33). Bacteria living in biofilms are also more resistant toNaOCl and the alkaline stress of calcium hydroxide, even when directlyexposed in vitro (34).

A very important factor in the effectiveness of calcium hydroxide isthe ability of the operator to place it effectively. It is essential that thecalcium hydroxide be placed in the instrumented canal as a thick, moistpaste that completely obturates the canal space (30, 35). The manner inwhich the calcium hydroxide was mixed and placed was not observed orstandardized in this study, and this could be a confounder.

If a resistant species of bacteria is present, if bacteria have estab-lished themselves as a biofilm inside of the canal, or if the operator is notcareful about his/her placement of calcium hydroxide, the effectivenessof the dressing will be compromised. One of the greatest benefits ofmulti-visit treatment might be the benefit of a second or third opportu-nity to disrupt biofilms and irrigate them out of the root canal (36).Calcium hydroxide is not the only benefit in a multi-session approachto apical periodontitis.

Our study supports the recommendation to render treatment ofteeth exhibiting signs and symptoms of apical periodontitis in at least2 sessions with the use of calcium hydroxide as an interappointmentmedicament. The EA did not enhance the ability of standard needle irri-gation to eliminate cultivable bacteria from root canals in this study.Other in vivo studies comparing the ability of sonic and ultrasonicinstruments to eliminate bacteria from root canals are warranted.

References1. Kakehashi S, Stanley HR, Fitzgerald RJ. The effects of surgical exposures of dental

pulps in germ-free and conventional laboratory rats. Oral Surg Oral Med Oral Pathol1965;20:340–9.

2. Sundqvist G. Bacteriological studies of necrotic dental pulps (dissertation). Sweden:Umea University; 1976.

Sonic Irrigation System on Elimination of Bacteria from Root Canal Systems 1317

Clinical Research
3. Abbott PV. The periapical space: a dynamic interface. Aust Endod J 2002;28:
96–107.4. Figini L, Lodi G, Gorni F, Gaglianai M. Single versus multiple visits for endodontic

treatment of permanent teeth: a Cochrane systematic review. J Endod 2008;34:1041–7.

5. Mohammadi Z, Farhad A, Tabrizizadeh M. One-visit versus multiple-visit endodontictherapy: a review. Int Dent J 2006;56:289–93.

6. Katebzadeh N, Sigurdsson A, Trope M. Radiographic evaluation of periapicalhealing after obturation of infected root canals: an in vivo study. Int Endod J2000;33:60–6.

7. Ng YL, Mann V, Rahbaran S, Lewsey J, Gulabivala K. Outcome of primary root canaltreatment: systematic review of the literature: part 1—effects of study characteristicson probability of success. Int Endod J 2007;40:921–39.

8. Sjogren U, Figdor D, Persson S, Sunqvist G. Influence of infection at the time of rootfilling on the outcome of endodontic treatment of teeth with apical periodontitis. IntEndod J 1997;30:297–306.

9. Molander A, Warfvinge J, Reit C, Kvist T. Clinical and radiographic evaluation of one-and two-visit endodontic treatment of asymptomatic necrotic teeth with apicalperiodontitis: a randomized clinical trial. J Endod 2007;33:1145–8.

10. Meire MA, De Prijck K, Coenye T, Nelis Hj, De Moor RJ. Effectiveness of differentlaser systems to kill Enterococcus faecalis in aqueous suspension and in an infectedtooth model. Int Endod J 2009;42:351–9.

11. Nielsen BA, Baumgartner JC. Comparison of the EndoVac system to needle irrigationof root canals. J Endod 2007;33:611–5.

12. Roy RA, Ahmad M, Crum LA. Physical mechanisms governing the hydrodynamicresponse of an oscillating ultrasonic file. Int Endod J 1994;27:197–207.

13. Leighton TG. The acoustic bubble. New York: Academic Press; 1994. chapters 1 and 2.14. Ahmad M, Pitt Ford TR, Crum LA. Ultrasonic debridement or root canals: acoustic

streaming and its possible role. J Endod 1987;14:490–9.15. Ohl CD, Arora M, Ikink R, et al. Sonoporation from jetting cavitation bubbles. Bio-

phys J 2006;91:4285–95.16. Advanced Endodontics: Endoactivator System web page. Available at: http://www.

endoruddle.com/?name=endoactivatord (2009). Accessed June 22, 2010.17. Faul F, Erdfelder E, Lang A-G, Buchner A. G)Power 3: a flexible statistical power

analysis program for the social, behavioral, and biomedical sciences. Behav ResMethods 2007;39:175–91.

18. Moller A JR. Microbiological examination of root canals and periapical tissues ofhuman teeth: methodological studies. Odontol Tidsk 1966;74(Suppl):1–380.

19. Trope MB, Bergenholtz G. Microbial basis for endodontic treatment: can a maximaloutcome be achieved in one visit? Endod Topics 2002;1:40.

20. Weller RN, Brady JM, Berneir WE. Efficacy of ultrasonic cleaning. J Endod 1980;6:740–3.

1318 Huffaker et al.

21. Goodman A, Reader A, Beck M, Melfi R, Meyers W. An in vitro comparison of theefficacy of the step-back technique versus a step-back/ultrasonic technique inhuman mandibular molars. J Endod 1985;11:249–56.

22. Townsend C, Maki J. An in vitro comparison of new irrigation and agitation tech-niques to ultrasonic agitation in removing bacteria from a simulated root canal. JEndod 2009;35:1040.

23. Sjogren U, Sundqvist G. Bacteriologic evaluation of ultrasonic root canal instrumen-tation. Oral Surg Oral Med Oral Pathol 1987;63:366–70.

24. Stamos DE, Sadehi EM, Haasch GC, Gerstein H. An in vitro comparison study toquantitate the debridement ability of hand, sonic, and ultrasonic instrumentation.J Endod 1987;13:434–40.

25. Mayer BE, Peters OA, Barbakow F. Effects of rotary instruments and ultrasonic irri-gation on debris and smear layer scores: a scanning electron microscopic study. IntEndod J 2002;35:582–9.

26. Stock CJR. Current status of the use of ultrasound in endodontics. Int Nat Dent J1991;41:175–82.

27. Lumley PJ, Blunt L, Walmsey AD, Marquis PM. Analysis of the surface cut in sonicfiles. Endod Dent Traumatol 1996;12:240–5.

28. Ahmad M, Pitt Ford TR, Crum LA, Wilson RF. Effectiveness of ultrasonic files in thedisruption of root canal bacteria. Oral Surg. Oral Med Oral Pathol 1990;70:328–32.

29. Shuping GB, Ørstavik D, Sigurdsson A, Trope M. Reduction of intracanal bacteriausing nickel-titanium rotary instrumentation and various medications. J Endod2000;26:751–5.

30. Teixeira FB, Leving LG, Trope M. Investigation of pH at different dentinal sites afterplacement of calcium hydroxide dressing by two methods. Oral Surg Oral Med OralPathol Oral Radiol Endod 2005;99:511–6.

31. Andersen M, Lund A, Andreasen JO, Andreasen FM. In vitro solubility of human pulptissue in calcium hydroxide and sodium hypochlorite. Endod Dent Traumatol 1992;8:104–8.

32. Bystrom A, Claesson R, Sundqvist G. The antibacterial effect of camphorated para-monochlorophenol, camphorated phenol and calcium hydroxide in the treatment ofinfected root canals. Endod Dent Traumatol 1985;1:170–5.

33. Figdor D, Davies JK, Sundqvist G. Starvation survival, growth and recovery of Entero-coccus faecalis in human serum. Oral Microbiol Immunol 2003;18:234–9.

34. Chavez de Paz LE, Bergenholtz G, Dahlen G, Svensater G. Response to alkaline stressby root canal bacteria in biofilms. Int Endod J 2007;40:344–55.

35. Figdor D, Sundqvist G. A big role for the very small: understanding the endodonticmicrobial flora. Aust Dent J 2007;52(Suppl):S38–51.

36. Nair PN, Henry S, Cano V, Vera J. Microbial status of apical root canal system ofhuman mandibular first molars with primary apical periodontitis after ‘‘one-visit’’endodontic treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;99:231–52.


http://www.endoruddle.com/?name=endoactivatord

http://www.endoruddle.com/?name=endoactivatord

Clinical Research

Root and Canal Morphology of Mandibular Second Molars inan Indian PopulationPrasanna Neelakantan, MDS,* Chandana Subbarao, BDS,*

Chandragiri Venkata Subbarao, MDS,* and Mithun Ravindranath, MDS†

Abstract
Introduction: There are no reports on the root canalanatomy of Indian mandibular second molars. The aimof this study was to investigate the root and canalmorphology of Indian mandibular second molars bya canal staining and tooth clearing technique. Method-ology: Mandibular second molars (345) were collectedfor analyzing the morphology of the roots and root canalsystems. The teeth were subjected to a canal stainingand clearing technique; after which, the followingfeatures were examined under magnification: numberand morphology of roots, number of root canals, rootcanal system configurations (Vertucci’s classificationand Gulabivala’s additional classes), number of apicalforamina, and intercanal communications. Results:Most of the second molars had two separate roots(87.8%) with three canals. C-shaped canal morphologywas observed in 7.5% of the teeth examined. Both themesial and distal roots of two rooted molars showedwide variations in canal number and configuration.Type IV and type I canal anatomies were most commonin the mesial and distal roots of two rooted secondmolars, respectively. Approximately 54.84% of the teethshowed two apical foramina, and one specimen (3.8%)of the C-shaped roots showed three apical foramina.Conclusion: The most common root morphology inIndian second molars is the two rooted morphologywith three canals. Both the mesial and distal rootsshowed wide variations in canal anatomy with type IVand type I canal configuration predominating in themesial and distal roots, respectively. (J Endod2010;36:1319–1322)
Key WordsIndian, mandibular, molar, root canal, staining andclearing

From the *Department of Conservative Dentistry andEndodontics, Saveetha Dental College and Hospitals, SaveethaUniversity, Chennai, India; and †Private Practice, Cochin,Kerala, India.

Address requests for reprints to Dr. Prasanna Neelakantan,Plot 1500, 16th Main Road, Anna Nagar West, Chennai, TamilNadu, India. E-mail address: [email protected]/$0 - see front matter



The knowledge of root canal anatomy has a major influence on the success rate ofendodontic treatment. Root canal anatomy and root morphology may have definitive

racial influences, thereby necessitating the identification of root canal morphologies ofdifferent races (1). Studies on the root canal anatomy of mandibular first and secondmolars have been performed on several populations (2–5). The Indian race is generallyconsidered to be a hybrid of several races with characteristics of Caucasian, Mongoloid,and Negroid races, which is generally referred to as the Dravidian race (6). An extensiveliterature search showed us that there is only one study on the root canal morphology ofIndian mandibular first molars (7). There are no reports, however, on the root andcanal morphology of Indian mandibular second molars. The aim of this investigationwas to study the root and canal morphology of mandibular second molars of an Indianpopulation using a canal staining and clearing technique.

Materials and MethodsThree hundred forty-five mandibular second molars were collected from dental

practitioners across the Indian subcontinent. It was ensured that the teeth belongedto indigenous Indians, and no teeth from other minority ethnicities were included.The process of collection was performed by a team of practitioners who were madeto understand the aims of the study, and the collection of every tooth was accompaniedby a case record stating and confirming the ethnicity of the patients. The teeth werewashed under tap water immediately after extraction and stored in distilled waterwith thymol iodide crystals (Titan Pharma, Mumbai, India) until the collection wascomplete. The entire collection process was completed in 2 months. The sampleswere washed thoroughly under tap water and immersed in 2.5% sodium hypochlorite(Prime Dental Products, Mumbai, India) for 30 minutes to remove adherent soft tissue.The staining and clearing protocol was adopted from a previously described method (8).

The teeth were immersed in 2.5% sodium hypochlorite for 24 hours followed byultrasonication for 30 minutes (Ultrasonic Bath; Confident Dental Systems, Bangalore,India). After drying of the teeth, Indian ink was injected into the root canal systems usinga 27-G monoject needle assisted by vacuum suction apically. The teeth were subjected tonegative pressure for 2 minutes (24 torr or 3,199 Pa), and a reapplication of negativepressure was performed after 3 minutes. After 12 hours of drying, the teeth were decal-cified in 10% nitric acid (Merck Limited, Mumbai, India) for 28 to 30 hours. The acidwas changed after 24 hours and was stirred every 8 hours. The endpoint of decalcifi-cation was determined by examining the relative loss or decrease of radiopacity ofenamel and dentin (by taking periodic radiographs) of five sample teeth. After thoroughwashing of the decalcified teeth in running tap water for 4 hours, the samples were de-hydrated in ascending concentrations of ethanol (70%, 80%, 95%, and 100%; StarChem, Chennai, India) for 1 day, and the samples were rendered transparent by immer-sion in methyl salicylate (Star Chem) for 2 days. Digital images of the samples weretaken in mesiodistal and buccolingual dimensions and analyzed under 5� magnifica-tion by two calibrated endodontists.

The following features were analyzed: the number of root canals per root (basedon the number of root canal orifices in the pulp chamber); the root canal configuration(Vertucci’s classification, Fig. 1) (9); and additional classes based on the number oforifices, canals, and apical foramina (Gulabivala’s classification, Fig. 2) (4), the pres-ence of intercanal communications (communications within different canals in thesame root) and the number of apical foramina.

Morphology of Mandibular Second Molars in an Indian Population 1319


Figure 1. Vertucci’s classification of canal configurations.

Figure 2. Gulabivala’s additional classes of canal configurations.

Clinical Research

ResultsThe number of roots, root morphology, root canals, canal config-

uration, apical foramina, and intercanal communications are summa-rized later. Unless mentioned, the values are in percentage of thetotal number of teeth examined. Both the endodontist examinerswere consistent in reporting their findings (100% intrarater and inter-rater agreement), and a test to evaluate examiner reliability was deemedunnecessary.

Number of Roots and Root MorphologyThe most common morphology was two separate roots (83.4%).

C-shaped root morphology was observed in 7.5% of the teeth examined(Table 1).

Number of Root Canals and Canal ConfigurationsThe mesial roots of two-rooted molars commonly showed two

canals (86.1% of two-rooted molars), whereas the distal roots oftwo-rooted molars (77.7% of two-rooted molars) and all roots ofthree-rooted molars commonly showed one canal (Tables 1 and 2).In molars with two roots, both the mesial and distal roots showedwide variations in canal configuration. The most commonly found canal

TABLE 1. Root Morphology and Number of Roots and Root Canals in Mandibular

Specimen Number of specimens Root O

Two separate roots 288 (83.4) MD 2

Three separate roots 31 (8.98) MBMLD

C-shaped root 26 (7.53)

Values indicate number of teeth in which each of the features was identified. Values within parentheses a

M, mesial; D, distal; MB, mesiobuccal; ML, mesiolingual.

1320 Neelakantan et al.

configuration in the mesial (M) root (75.6% of two-rooted molars) wastype IV, whereas the distal (D) root had predominantly type I canalconfiguration (77.7% of two-rooted molars). The type I and IV canalsystems were identified in 24.2% and 16.1% of the three-rooted molars,respectively. Additional canal types were identified in 4% of the teethexamined.

Intercanal CommunicationsThe highest incidence of intercanal communications was found in

the distal roots (81.1%) of two-rooted molars followed by mesial roots(36.8%) of two-rooted molars.

Apical ForaminaAll roots exhibiting type IV, V, VII, 3-2, and 4-2 canal configuration

showed two separate apical foramina (98.6% of two-rooted secondmolars, 22.5% of three-rooted second molars, and 46.1% of C-shapedroots). Three apical foramina were identified in 3.8% of the teeth (onespecimen) with C shaped roots. The other roots showed only one apicalforamen.

Second Molars (N = 345)

ne canal Two canals Three canals Four canals

29 (8.4) 248 (71.8) 11 (3.18) —24 (64.9) 62 (17.9) 02 (0.57) —28 (8.1) 03 (0.86) — —30 (8.69) 01 (0.28) — —26 (7.5) 05 (1.4) — —10 (2.89) 13 (3.76) 01 (0.28) 02 (0.57)

re percentages (% of total number of teeth examined).


TABLE2.

The

Confi

gura

tion

ofR

oot

Cana

lSy

stem

san

dIn

terc

anal

Com

mun

icat

ions

inM

andi

bula

rSe

cond

Mol

ars

(N=

345)

Spec

imen

Ro

ot

Typ

eI

Typ

eII

Typ

eII

ITy

pe

IVTy

pe

VTy

pe

VI

Typ

eV

IITy

pe

VII

IA

dd

itio

nal

can

alty

pe

Inte

rcan

alco

mm

un

icat

ion

s

Two

sep

ara

tero

ots

(288)

M29

(8.4

)7

(2.0

2)

5(1

.44)

218

(63.1

)18

(5.2

)—

——

7(2

.02)*

4(1

.15)†

127

(36.8

)

D224

(64.9

)16

(4.6

3)

2(0

.57)

38

(11)

6(1

.73)

——

2(0

.57)

—280

(81.1

)Th

ree

sep

ara

tero

ots

(31)

MB

28

(8.1

)1

(0.2

8)

—2

(0.5

7)

——

——

—1

(0.2

8)

ML

30

(8.6

)—

—1

(0.2

8)

——

——

——

D26

(7.5

)1

(0.2

8)

—2

(0.5

7)

2(0

.57)

——

——

2(0

.57)

Csh

ap

ed

roo

t(2

6)

10

(2.8

)—

2(0

.57)

7(2

.02)

1(0

.28)

—3

(0.8

6)

—2

(0.5

7)‡

1(0

.28)§

12

(3.4

7)

Valu

esin

dica

tenu

mbe

rof

teet

hin

whi

chea

chof

the

feat

ures

was

iden

tified

.Va

lues

with

inpa

rent

hese

sar

epe

rcen

tage

s(%

ofto

tal

num

ber

ofte

eth

exam

ined

).

M,

mes

ial;

D,

dist

al;

MB

,m

esio

bucc

al;

ML,

mes

iolin

gual

.

*Typ

e3-

1ca

nal

syst

em.

†Ty

pe3-

2ca

nal

syst

em.

‡Ty

pe4-

2ca

nal

syst

em.

§Ty

pe2-

3ca

nal

syst

em.

Clinical Research


DiscussionThe methods used in analyzing the root canal morphology are

canal staining and tooth clearing (10–12); conventional radiographs(13, 14); alternative radiographic techniques (15–17); radiographicassessment enhanced with contrast media (18); and, more recently,computed tomographic techniques (19, 20). The most commonlyused technique is canal staining and clearing because of its accuracy(4, 10–12, 21). The root canal anatomy of Indian teeth by canalstaining and the tooth clearing technique has not been studied todate except for a study on premolars (21). Fine details like intercanalcommunications could be visualized even without ink penetration insome areas, showing that the quality of clearing is a critical factor toensure the visualization of intricate details of root canal anatomy. Instudies evaluating the root canal anatomy, it is very important to havea sufficient sample size in order to apply the results to the general pop-ulation (4, 22). Hence, it was ensured that molars were collected frompractitioners across the country, ensuring ethnicity of the populationfrom which the teeth was extracted.

The most commonly observed root morphology was the two sepa-rate rooted mandibular second molars (83.4%), which is higher thanthe prevalence in the Burmese (58.2%) and Thai (54%) populations(4, 5). These two-rooted second molars had one distal canal andtwo mesial canals, which either had a separate course (76%) orcombined in the middle third of the root (5.54%) or apically(0.69%). Mandibular second molars with three roots were observedin 8.98% of the teeth examined in contrast to the report of Gulabivalaet al (4) in which no mandibular second molars with three rootswere observed. The prevalence of three-rooted second molars is higherthan in Thai population and white populations (5, 23). The position ofthis third root was lingual, which was in agreement with the findings ofWalker (2) and Gulabivala et al (5) who identified the distolingual rootin first molars and proposed that this should be considered as a genetictrait and not a developmental anomaly. We speculate that this consider-ation may be applicable to this mesiolingual root of the second molarsas well. The presence of this root definitely has clinical importance; itmay not be visualized radiographically and may give an artifactualimpression of a perforation (5). The distolingual root or radix entomo-laris was not identified in the cohort investigated in this study. Thisfinding is in agreement with other reports from Chinese, Korean, andwhite cohorts (23–25).

C-shaped canal configuration has been shown to have a high prev-alence (14%-52%) in the mandibular second molars of Chinese,Japanese, and Lebanese populations (25–29). Our study showed thatC-shaped canal morphology was noted in 7.5% of the teethexamined, which was similar to the prevalence in the Mongoloidgroup (10%) but much less than reported in the Burmese (22.4%)populations (5, 20). The root canal configuration of C-shaped canalshas been claimed to be complex by some authors (24, 26). Ourinvestigation showed that 38.4% of the C-shaped roots (10specimens) had a single canal, which was similar to the observationmade in Thai population (5) but higher than in Burmese molars (4).C-shaped roots showed wide variations in canal configuration (typesI, III, IV, V, VII, 4-2, and 2-3) in accordance with other reports (4,5, 24–26). The presence of a high incidence of transverseanastomoses, lateral canals, and apical deltas in C-shaped canalsmakes it difficult to clean and seal the root canal system (30). Mostof the C-shaped roots (57.6% of the teeth with C-shaped morphology)showed two separate apical foramina, which was in contrast to the find-ings of Gulabivala et al (4).

The most commonly observed morphology was the two-rootedmorphology, with type IV canals and type I canal systems predominating

Morphology of Mandibular Second Molars in an Indian Population 1321

Clinical Research
in the mesial and distal roots, respectively. This has been reported to bea Mongoloid trait (5).
ConclusionsThe most common morphology in Indian mandibular second

molars was the two-rooted teeth with three canals (two mesial andone distal). C-shaped canals were found in 7.5% of the teeth, most ofwhich had single canals. The observations made in this study showthat Indian mandibular second molars exhibit both Mongoloid andCaucasian traits.

AcknowledgmentsThe authors wish to thank the Department of Oral and Maxil-

lofacial Surgery, Saveetha University, and the private practitionersacross the country for the collection of teeth for this study.

References1. Sperber GH. The phylogeny and odontogeny of dental morphology. In: Sperber GH,

ed. From Apes to Angels. New York: Wiley-Liss; 1999:215–9.2. Walker RT. Root form and canal anatomy of mandibular second molars in

a southern Chinese population. J Endod 1988;14:325–9.3. Salwa AY, Abdullah RA, Mohammad FF. Three-rooted permanent mandibular first

molars of Asian and Black groups in the Middle East. Oral Surg Oral Med OralPathol 1990;69:102–5.



6. Reich D, Thangaraj K, Patterson N, et al. Reconstructing Indian population history.Nature 2009;24:489–94.

7. Reuben J, Velmurugan N, Kandaswamy D. The evaluation of root canal morphologyof the mandibular first molar in an Indian population using spiral computed tomog-raphy scan: an in vitro study. J Endod 2008;34:212–5.

8. Robertson D, Leeb IJ, McKee M, et al. A clearing technique for the study of root canalsystems. J Endod 1980;6:421–4.

9. Vertucci FJ. Root canal morphology of mandibular premolars. J Am Dent Assoc1978;97:47–50.

10. Alavi AM, Opasanon A, Ng YL, et al. Root and canal morphology of Thai maxillarymolars. Int Endod J 2002;35:478–85.

11. Awawdeh L, Abdullah H, Al-Qudah A. Root form and canal morphology of Jordanianmaxillary first premolars. J Endod 2008;34:956–61.

1322 Neelakantan et al.

12. Pineda F, Kuttler Y. Mesiodistal and buccolingual roentgenographic investigation of7275 root canals. Oral Surg Oral Med Oral Pathol 1972;33:101–10.

13. Pattanshetti N, Gaidhane M, Al Kandari AM. Root and canal morphology of the me-siobuccal and distal roots of permanent first molars in a Kuwait population—a clin-ical study. Int Endod J 2008;41:755–62.

14. Fan W, Fan B, Gutmann JL, et al. Identification of a C-shaped canal system in mandib-ular second molars. Part III. Anatomic features revealed by digital subtraction radi-ography. J Endod 2008;34:1187–90.

15. Patel S, Dawood A, Whaites E, et al. New dimensions in endodontic imaging: part1. Conventional and alternative radiographic systems. Int Endod J 2009;42:447–62.

16. de Oliveira SH, de Moraes LC, Faig-Leite H, et al. In vitro incidence of root canalbifurcation in mandibular incisors by radiovisiography. J Appl Oral Sci 2009;17:234–9.

17. Scarfe WC, Fana CR, Farman AG. Radiographic detection of accessory/lateral canals:use of RadioVisioGraphy and Hypaque. J Endod 1995;21:185–90.

18. Plotino G, Grande NM, Pecci R, et al. Three dimensional imaging using micrcom-puted tomography for studying tooth macromorphology. J Am Dent Assoc 2006;137:1555–61.

19. Sberna MT, Rizzo G, Zacchi E, et al. A preliminary study of the use of peripheralquantitative computed tomography for investigating root canal anatomy. Int EndodJ 2009;42:66–75.

20. Ng YL, Aung TH, Alavi A, et al. Root and canal morphology of Burmese maxillarymolars. Int Endod J 2001;34:620–30.

21. Velmurugan N, Sandhya R. Root canal morphology of mandibular first premolars inan Indian population: a laboratory study. Int Endod J 2009;42:54–8.

22. Wasti F, Shearer AC, Wilson NHF. Root canal systems of the mandibular and maxillaryfirst permanent molar teeth of South Asian Pakistanis. Int Endod J 2001;34:263–6.

23. Tratman EK. Three-rooted lower molars in man and their racial distribution. BrDent J 1939;64:264–74.


25. Song JS, Choi HJ, Jung IY, et al. The prevalence and morphologic classification ofdistolingual roots in the mandibular molars in a Korean population. J Endod2010;36:653–7.

26. Vertucci FJ, Williams R. Root canal anatomy of mandibular first molar. J N J DentAssoc 1974;45:27–8.

27. Fan B, Min Y, Lu G, et al. Negotiation of C-shaped canal systems in mandibularsecond molars. J Endod 2009;35:1003–8.

28. Haddad GY, Nehme WB, Ounsi HF. Diagnosis, classification, and frequency of C-shaped canals in mandibular second molars in the Lebanese population. J Endod1999;25:268–71.

29. Yang ZP, Yang SF, Lin YC, et al. C-shaped root canals in mandibular second molarsin a Chinese population. Endod Dent Traumatol 1988;4:160–3.

30. Melton DC, Krell KV, Fuller MW. Anatomical and histological features of C-shapedcanals in mandibular second molars. J Endod 1991;17:384–8.


Basic Research—Biology

Periapical Bone Regeneration after Endodontic Microsurgerywith Three Different Root-end Filling Materials: Amalgam,SuperEBA, and Mineral Trioxide AggregateSeung-Ho Baek, DDS, PhD,* Woo Cheol Lee, DDS, PhD,* Frank C. Setzer, DMD, PhD, MS,

†

and Syngcuk Kim, DDS, PhD†

Abstract
Introduction: The purpose of this study was to deter-mine the bone regeneration potential to different root-end filling materials by evaluating the distance betweenthe materials and newly regenerated bone after root-endsurgery. Material and Methods: Periapical lesionswere induced in premolars and molars of five femalebeagle dogs. The teeth were treated endodontically afterthe development of the lesions. After 1 week, the teethunderwent root-end surgery using modern microsurgicaltechniques. Three different root-end filing materialswere used: amalgam (Tytin; Kerr Mfg Co, Romulus,MI), SuperEBA (Bosworth, Skokie, IL), and mineraltrioxide aggregates (MTA; Dentsply, York, PA). After 4months, the dogs were sacrificed, and the jaws wereprepared for histological sectioning. The distances fromthe root-end filling materials to the regenerated bonewere determined by the evaluation of microradiographicimages of the sections with imaging software (SigmaScan/Image; Jandel Scientific Software, San Rafael,CA). The results were statistically analyzed with analysisof variance using Sigma Stat software (Jandel ScientificSoftware, San Rafael, CA). Results: The mean distancesfrom the newly regenerated bone were 0.397 � 0.278mm in the MTA group, 0.756 � 0.581 mm in the Super-EBA group, and 1.290 � 0.386 mm in the amalgamgroup. There was a statistically significant differencebetween the amalgam and MTA groups (p < 0.05). Nosignificant differences existed for amalgam versusSuperEBA and SuperEBA versus MTA. Conclusion:MTA showed the most favorable periapical tissueresponse. The distance from MTA to the regeneratedbone was similar to the normal average periodontal liga-ment thickness in dogs. (J Endod 2010;36:1323–1325)
Key WordsAmalgam, apical bone regeneration, microradiograph,mineral trioxide aggregate, root-end seal, SuperEBA

From the *Department of Conservative Dentistry, School of DenMedicine, University of Pennsylvania, Philadelphia, PA, USA.

Address requests for reprints to Dr Syngcuk Kim, Department of EPA 19104-6030. E-mail address: [email protected]/$0 - see front matter


JOE — Volume 36, Number 8, August 2010 Pe

Many established root-end filling materials are being used, and, in addition, severalnewer materials are now also used for apical surgery. The ideal root-end filling

material should be biocompatible, bacteriocidal, or at least bacteriostatic; should beneutral to neighboring tissues; and should provide excellent sealing. Furthermore, itshould promote regeneration of the original tissues (1).

Amalgam was considered the root-end filling material of choice until the 1990s.However, in recent years, many have questioned the safety and integrity of amalgam ingeneral and as a root-end filling material in particular because it has many disadvan-tages including the release of ions, mercury toxicity, corrosion and electrolysis, delayedexpansion, marginal leakage, and leaving tissue tattoos (2–4).

SuperEBA (Bosworth, Skokie, IL) cement as a root-end filling material was sug-gested by Oynick and Oynick (5). SuperEBA, a composition of zinc oxide and aluminumoxide mixed with o-ethoxybenzoic acid and eugenol (2, 5, 6), was shown to be superiorthan amalgam in terms of sealing ability, apical tissue compatibility, and theirregeneration potential (7, 8). In a clinical study, the healing success after root-endsurgery was 96.4% after 1 year when SuperEBA was used as a root-end filling materialin conjunction with microsurgical techniques (9).

Mineral trioxide aggregate (MTA) as a root-end filling material was introduced toendodontics by Torabinejad et al in 1993 (10). MTA was shown to have excellent seal-ing ability (11–14) and promoted osteoblast activity (15, 16). It was less cytotoxic thanamalgam, IRM, or SuperEBA (11, 17) and had an antimicrobial effect (18). Results ofMTA studies in dogs and monkeys showed that MTA caused significantly less inflamma-tion than amalgam. More importantly, cementum bridges formed directly over the MTAroot-end fillings confirming the tissue friendliness of this material and its potential ce-mentogenic property (1, 19, 20).

Microradiography is the process of producing enlarged images of the interior ofthin, usually small specimens by the penetration of low-energy (0.1-10 keV) x-rays.Microradiography has been used since 1950 to study the structures and compositionof biological material, mainly bone tissue, and is specifically designed for boneevaluation. Because this study is about bone regeneration in response to differentroot-end filling materials, the microradiograph technique was chosen. Histologicevaluation is less powerful than microradiographic evaluation for bone regeneration.Uneven distribution of one mineral was revealed by this technique, and quantitativeinformation was obtained on the mineral content of different areas of bone (21).This study was conducted to compare the distance from root-end filling materialsto regenerated bone in contact with amalgam, SuperEBA, and MTA in dog teeth usingmicroradiography.

tistry, Seoul National University, Seoul, Korea; and †Department of Endodontics, School of Dental

ndodontics, School of Dental Medicine, University of Pennsylvania, 240 S 40th Street, Philadelphia,

riapical Bone Regeneration after Endodontic Microsurgery with Amalgam, SuperEBA, or MTA 1323


Figure 1. An enlarged view of a microradiograph of the MTA filling. The prox-imity of a root-end filling material to the regenerated bone measured. C,center; L, lingual; B, buccal.


Material and MethodsFive healthy female beagle dogs weighing between 7 and 9 kg were

used in this study, which was conducted in accordance with a Universityof Pennsylvania Institutional Review Board–approved animal protocol.Each animal was anesthetized with an intramuscular injection of 0.7 mg/kg of acepromazine (Aveco Co Inc, Fort Dodge, IA) as a preanestheticand an intravenous injection of 0.7 mg/kg of propofol for short-termanesthesia. Subsequently, the general inhalation anesthetic isoflurane(2%-3%) was administered via an endotracheal tube throughout thesurgical procedures. During the surgical phase, a dose of 2% xylocainewith 1:50,000 epinephrine was injected into the surgical site to achievemaximum hemostasis. After recovery from the general anesthesia, eachdog was given 0.3 mg/kg butorphenol tartrate (Fort Dodge LaboratoriesInc, Fort Dodge, IA) as an analgesic and was kept in a recovery area forobservation.

The Induction of Periapical LesionsThe pulp tissues of molars and premolars were removed from the

canals, and plaque contaminated paper points were placed into thecanals and sealed with IRM (Dentsply, York, PA) for 2 weeks. Periapicallesions formed between 4 to 6 weeks and were verified radiographi-cally. At this point, the teeth were treated by nonsurgical root canal treat-ment, and the access cavities were restored with IRM. Because of thelimited time available during short-term anesthesia, the surgical phasecould not be performed in the session.

In the surgical stage, 1 week after the verification of the periapicalradiolucencies and the nonsurgical root canal therapy, full mucoperios-teal triangular flaps were made with vertical incisions at the mesial lineangle of the cuspids, and intrasulcular incisions were extended to themesial of the second molars. The cortical and cancellous bones atthe apices were removed with a water-cooled high-speed rotary instru-ment, creating osseous cavities of about 4 � 4 mm in diameter. Afterresection of each root end and removal of the radicular lesion,a 3-mm deep root-end cavity was prepared with an ultrasonic KiS tip(Obtura/Spartan, Fenton, MO). All surgical procedures and root-endpreparations were performed under an operating microscope at 8 to24� magnification. The resected roots of 24 teeth were randomlydivided into three groups. In group A root ends were filled withamalgam, in group B the filling material was SuperEBA, and in groupC the root-end filling was gray MTA.

Four months after the last surgical procedure, the dogs were sacri-ficed with intravenous overdoses of sodium pentobarbital (Nembutal;Abbot Lab, North Chicago, IL); 4 months was selected because a completeregeneration of bone had taken place, as evidenced radiographically.The jaws were perfused with a mixture of 10% buffered formalin and80% ethanol via the carotid arteries. The jaws were then prepared forhistological evaluation using a sawing and grinding technique developedby Donath and Breuner (22) for the examination of normal bone andteeth with attached soft tissue. After fixation for 1 month, the demineral-ized specimens were processed for embedding with methylmethacrylatemonomer (No.8060061; Merck, Darmstadt, Germany). After polymeri-zation of the methacrylate, sections were cut from the block with anIsomet low-speed saw (No. 11-1180; Beuhler Ltd, Lake Buff, IL).Subsequent grinding and polishing were performed with sandpaper(silicon carbide or alumina grits of 220, 400, 600, or 1,200 grit, Struers,Cleveland, OH) resulting in approximately 80-mm thickness of eachspecimen. Subsequently, contact microradiographs were prepared.

The distance from root-end fillings to the regenerated bone wascalculated with image software at three points: the buccal margins,the center, and the lingual margins of the root-end fillings (Fig. 1).The distance from filling materials to the regenerated alveolar bone

1324 Baek et al.

was determined using Sigma Scan/Image software (Jandel ScientificSoftware, San Rafael, CA), whereas the statistical analysis was performedusing Sigma Stat software (Jandel Scientific Software, San Rafael, CA).Statistical values for each group of data were subsequently calculatedwith analysis of variance.

ResultsOf the 24 original specimens, 23 were subjected to the microra-

diographic evaluation. One specimen had to be excluded from the studybecause the amalgam root-end filling disassociated from the root-endcavity during the healing period. New bone formation on the root-end filling materials was determined by the distance from filling mate-rials to the regenerated alveolar bone. The distance from the center ofthe root-end filling materials to regenerated bone was calculated at0.397� 0.278 mm in the MTA group, 0.756� 0.581 mm in the Super-EBA group, and 1.290 � 0.386 mm in the amalgam group (Table 1).The MTA group showed superior bone regeneration to the other tworoot-end filling materials, and there is a statistically significant differ-ence between the MTA and amalgam groups (p < 0.05). There is nosignificant difference in amalgam versus SuperEBA and SuperEBAversus MTA.

DiscussionThe ultimate success of periapical surgery depends on the regen-

eration of a functional periodontal attachment apparatus (23). Regen-eration has been defined as the replacement of tissue components intheir appropriate locations in the correct amounts and the correct rela-tionship to each other (24). However, biocompatibility and sealingability after placement of most root-end filling materials are problem-atic. At present, there is no material that satisfies each criterion forthe ideal root-end filling material. Amalgam, SuperEBA, and MTA asroot-end filling materials were chosen in this study because they areused most frequently by clinicians.

Amalgam has been and still is the most widely used root-end fillingmaterial. Severe inflammatory responses to amalgam fillings have beenreported in dogs (1, 19, 25). Fibrous tissue capsules were found inclose proximity to most amalgam root-end fillings by Torabinejad et al(20) and Baek et al (1). These findings clearly show that amalgam isnot biologically suitable as a root-end filling material.


TABLE 1. Distance from Filling Material to Regenerated Bone (mm) (Mean�Standard Deviation)

Buccal side Center Lingual side

Amalgam(n = 5)

1.358 � 0.135 1.290 � 0.385 0.997 � 0.227

SuperEBA(n = 9)

0.783 � 0.599 0.756 � 0.581 0.684 � 0.394

MTA (n = 9) 0.241 � 0.063 0.397 � 0.278 0.376 � 0.156


SuperEBA was very popular in the 1990s. Oynick and Oynick (5)found collagen fibers around the material and actually growing into it,which suggested that the EBA was well tolerated by the tissue. Becauseboth SuperEBA and IRM contain eugenol, concern has been expressedabout possible harmful effects on the periapical tissues. On subsequentexaminations, the SuperEBA group showed fewer inflammatory cellinfiltrates than the amalgam group but more than the MTA group (1, 7).

MTA was a relatively new material that became available in the late1990s. This material appears to be the most promising to date because itcomes closest to being the ideal material for root-end filling and theresults of reported studies are indeed impressive (1, 15, 16, 20, 26).It was listed as one of the recommended root-end filling materials forendodontic microsurgery (27). Many studies reported MTA asa biocompatible material that induces osteogenesis and odontogenesis(28–30). When MTA directly contacts fibroblast, cementoblast, andosteoblast cells of the periodontal ligament, MTA was morebiocompatible and less toxic than SuperEBA (11, 17, 31).

This study was limited to the determination of the space betweenthe filling material and the newly formed bone. The distances from thecenter of the resected root surface containing the MTA filling to the re-generated bone were measured with the expectation that a biologicallyacceptable material would have a closer proximity to the newly formedbone. From the center of the resected root surface to the bone, the MTAgroup had the closest distance with 0.397 mm. This distance was two tothree times wider in the amalgam and SuperEBA groups (Table 1). Thedistance between the regenerated bone and the mineral trioxide aggre-gate in the MTA group of our study was comparable to published datafor the normal average PDL thickness in beagle dogs, which was 0.386� 0.025 mm (32).

ConclusionFrom the center of the resected root surface to the bone, the MTA

group had the closest distance. The periodontal ligament (PDL) thick-ness in the MTA group of our study could be considered a normalaverage PDL thickness. The gap between the SuperEBA and the amalgamgroups were two times and three times wider, respectively, than in theMTA group, suggesting that MTA promotes bone and PDL regenerativeability.

References1. Baek SH, Plenk H, Kim S. Periapical tissue responses and cementum regeneration

with amalgam, SuperEBA, and MTA as root-end filling materials. J Endod 2005;31:444–9.

2. Dorn SO, Gartner AH. Retrograde filling materials: A retrospective success-failurestudy of amalgam, EBA and IRM. J Endod 1990;16:391–3.

JOE — Volume 36, Number 8, August 2010 Periapical Bone Rege

3. Shahi S, Rahimi S, Lofti M, et al. A comparative study of the biocompatibility of threeroot-end filling materials in rat connective tissue. J Endod 2006;32:776–80.

4. Friedman S. Retrograde approaches in endodontic therapy. Endod Dent Traumatol1991;7:97–107.

5. Oynick J, Oynick T. A study of a new material for retrograde fillings. J Endod 1978;4:203–6.

6. Trope M, Lost C, Schmitz HJ, et al. Healing of apical periodontitis in dogs afterapicoectomy and retrofilling with various filling materials. Oral Surg Oral MedOral Pathol 1996;81:221–8.

7. Pitt Ford TR, Andreasen JO, Dorn SO, et al. Effect of SuperEBA as a root-end fillingon healing after replantation. J Endod 1995;21:13–5.

8. Torabinejad M, Hong CU, Pitt Ford TR, et al. Tissue reaction to implanted super-EBAand mineral trioxide aggregate in the mandible of guinea pigs: a preliminary report.J Endod 1995;21:569–71.

9. Rubinstein RA, Kim S. Short-term observation of the results of endodontic surgerywith the use of a surgical operation microscope and Super-EBA as root-end fillingmaterial. J Endod 1999;25:43–8.

10. Torabinejad M, Watson TF, Pitt Ford TR. The sealing ability of a mineral trioxideaggregate as a root end filling material. J Endod 1993;19:591–5.

11. Torabinejad M, Smith PW, Kettering JD, et al. Comparative investigation of marginaladaptation of mineral trioxide aggregate and other commonly used rootend fillingmaterials. J Endod 1995;21:295–9.

12. Torabinejad M, Higa RK, McKendry DJ, et al. Dye leakage of four root end fillingmaterials: effects of blood contamination. J Endod 1994;20:159–63.

13. Torabinejad M, Rastegar AF, Kettering JD, et al. Bacterial leakage of mineral trioxideaggregate as a root-end filling material. J Endod 1995;21:109–12.

14. Bates CF, Carnes DL, del Rio CE. Longitudinal sealing ability of mineral trioxideaggregate as a root-end filling material. J Endod 1996;22:575–8.

15. Koh ET, McDonald F, Pitt-Ford TR, et al. Cellular response to mineral trioxide aggre-gate. J Endod 1998;24:543–7.

16. Torabinejad M, Pitt-Ford TR, Abedi HR, et al. Tissue reaction to implanted root-endfilling materials in the tibia and mandible of guinea pigs. J Endod 1998;24:468–71.

17. Keiser K, Johnson C, Tipton D. Cytotoxicity of mineral trioxide aggregated usinghuman periodontal ligament fibroblasts. J Endod 2000;26:288–91.

18. Torabinejad M, Hong CU, McDonald F, et al. Physical and chemical properties ofa new root end filling material. J Endod 1995;21:349–53.

19. Torabinejad M, Hong CU, Lee SJ, et al. Investigation of mineral trioxide aggregate forroot-end filling in dogs. J Endod 1995;21:603–8.

20. Torabinejad M, Pitt Ford TR, McKendry D, et al. Histologic assessment of mineraltrioxide aggregate as a root-end filling in monkeys. J Endod 1997;23:225–8.

21. Boivin G, Munier PJ. The degree of mineralization of bone tissue measured bycomputerized quantitative contact microradiography. Calcif Tissue Int 2002;70:503–11.

22. Donath K, Breuner G. A method for the study of undecalcified bones and teeth withattached soft tissues. J Oral Pathol 1982;11:318–26.

23. Andreasen JO, Rud J. Modes of healing histologically after endodontic surgery in 70cases. Int J Oral Surg 1972;1:148–60.

24. Aukhil I. Biology of tooth-cell adhesion. Dent Clin North Am 1991;35:459–67.25. Kimura JT. A comparative analysis of zinc and non-zinc alloys used in retrograde

endodontic surgery. Part 1. Apical seal and tissue reaction. J Endod 1982;8:359–63.

26. Regan JD, Gutmann JL, Witherspoon DE. Comparison of Diaket and MTA when usedas root-end filling materials to support regeneration of the periradcular tissues. Int JEndod 2002;35:840–7.

27. Kim S, Kratchman S. Modern endodontic surgery concepts and practice: a review.J Endod 2006;32:601–23.

28. Koh ET, Torabinejad M, Pitt Ford TR, et al. Mineral trioxide aggregate stimulatesa biological response in human osteoblasts. J Biomed Mater Res 1997;5:432–9.

29. Holland R, Souza V, Nery MJ, et al. Reaction of rat connective tissue to implanteddentin tubes filled with mineral trioxide aggregate or calcium hydroxide. J Endod1999;25:161–6.

30. Holland R, Souza V, Nery MJ, et al. Reaction of dog’s teeth to root canal filling withmineral trioxide aggregate or a glass ionomer sealer. J Endod 1999;25:728–30.

31. Tani-Ishii N, Hamada N, Watanabe K, et al. Expression of bone extracellular matrixproteins on osteoblast cells in the presence of mineral trioxide. J Endod 2007;33:836–9.

32. Lindhe J, Erickson I. The influence of trauma from occlusion on reduced but healthyperiodontal tissues in dogs. J Clin Periodontol 1976;11:279–89.

neration after Endodontic Microsurgery with Amalgam, SuperEBA, or MTA 1325


The Role of Heme Oxygenase-1 in the Proliferation andOdontoblastic Differentiation of Human Dental Pulp CellsSun-Ju Kim, PhD,*† Kyung-San Min, DDS, PhD,‡ Hyun-Wook Ryu, DDS, MSD,‡ Hwa-Jeong Lee, PhD,* and Eun-Cheol Kim, DDS, PhD*

Abstract
Introduction: It was recently reported that hemeoxygenase-1 (HO-1) activity is related to stem cell differ-entiation; however, the involvement of HO-1 in pulp cellgrowth and differentiation has not been well explored.The purpose of this study was to investigate the roleof HO-1 in the growth and differentiation of humandental pulp cells (HDPCs). Methods: We evaluatedcell growth by MTT assay, mineralization by alizarinred staining, and differentiation marker mRNA expres-sion by reverse transcriptase polymerase chain reaction.Results: HO-1 induction by cobaltic protoporphyrin IX(CoPP) in HDPCs increased cell growth and mineraliza-tion and up-regulated the messenger RNA expressionof such odontoblastic markers as alkaline phosphatase,osteopontin, bone sialoprotein, dentin matrix protein-1,and dentin sialophosphoprotein. Carbon monoxidescavenger, iron chelator, HO-1 inhibitor, and HO-1 smallinterfering RNA (siRNA) attenuated HDPC growth anddifferentiation. Conclusions: CoPP treatment resultsin dental pulp cell proliferation and odontoblast differ-entiation that appears partly mediated by HO-1. Ourresults suggest that odontoblastic differentiation andgrowth are positively regulated by HO-1 induction andnegatively regulated by HO-1 inhibition. Thus, pharma-cologic HO-1 induction might represent a potent thera-peutic approach for pulp capping and the regenerationof HDPCs. (J Endod 2010;36:1326–1331)
Key WordsDifferentiation, growth, heme oxygenase-1, humandental pulp cells

From the Departments of *Oral and Maxillofacial Pathologyand ‡Conservative Dentistry, School of Dentistry, WonkwangUniversity, Iksan, South Korea; and †Department of DentalHygiene, Cheongju University, Cheongju, South Korea.

This study was supported by a grant from the Korea Health-care Technology R&D Project, Ministry for Health, Welfare &Family Affairs, Republic of Korea (A084458).

Address requests for reprints to Dr Eun-Cheol Kim, Depart-ment of Oral and Maxillofacial Pathology, Dental College,Wonkwang University, Sinyoungdong 344-2, Iksan City, Jeon-buk, 570-749, South Korea. E-mail address: [email protected]/$0 - see front matter


1326 Kim et al.

Human dental pulp (HDP) healing and repair are the result of successive and inter-related processes, including the proliferation, chemotaxis, and differentiation of

dental pulp cells into odontoblasts leading to reparative dentin formation (1). Thus,differentiated and undifferentiated cells within dental pulp may contribute to thedentinal regeneration process (2). Odontoblasts secrete several collagenous and non-collagenous proteins, including osteonectin, osteopontin (OPN), bone sialoprotein,dentin matrix protein-1 (DMP-1), and dentin sialophosphoprotein (DSPP) (3), whichhave been used as mineralization markers for the odontoblast-/osteoblast-like differen-tiation of HDP cells (HDPCs) (4). However, the molecular control mechanism under-lying the effect of the inductive signal on odontoblastic growth and differentiationremains to be elucidated (5).

Heme oxygenase-1 (HO-1, heat shock protein 32) is the inducible isoform of therate-limiting enzyme responsible for the breakdown of heme into carbon monoxide(CO), biliverdin, and free iron (6). Previously, we reported that the HO-1 pathway playsa key role in the adaptation of cells to stressful conditions and the recovery of HDPCsand periodontal ligament cells (PDLCs) from injurious events (7–12). Moreover,several studies have shown that HO-1 expression is related to adipogenesis in humanmesenchymal stem cells (MSCs) (13), osteoblastic differentiation in PDLCs (7, 12),and neuronal differentiation in MSCs (14).

Cobaltic protoporphyrin IX (CoPP), or hemin, has been shown to strongly induceHO-1 expression both in vivo and in vitro (15). Hemin promotes proliferation anddifferentiation in endothelial progenitor cells (16) and erythroid differentiation inhuman myeloid leukemia cells (17). In addition, we previously showed that the induc-tion of HO-1 by hemin enhanced the expression of such osteoblastic differentiationmarkers as OPN, osteonectin, OCN, and bone sialoprotein on PDLCs (7, 12).

However, no information is available regarding the effects of HO-1 induction byCoPP on the odontogenic potential of HDPCs. To elucidate the role of HO-1 in HDPCs,we investigated the effects of CoPP (an HO-1 inducer), tin protoporphyrin (SnPP, anHO-1 inhibitor), HO-1 siRNA, and HO-1 metabolites on the growth and differentiationof HDPCs.

Materials and MethodsReagents

Dulbecco modified Eagle medium, fetal bovine serum, and the other tissue culturereagents were obtained from Gibco BRL (Grand Island, NY). CoPP, SnPP, and heminwere obtained from Porphyrin Products (Logan, UT). Hemoglobin (Hb) and desfer-rioxamine (DFO) were purchased from Sigma (St Louis, MO).

Cell CultureWe used the HDPCs lines immortalized by transfection with the telomerase catalytic

subunit hTERT gene (18). Cells were cultured in Dulbecco modified Eagle medium sup-plemented with 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin in a humid-ified atmosphere of 5 % CO2 at 37�C. For mineralization experiments, cells werecultured in osteogenic media (OM) including 50 mg/mL ascorbic acid, 10�7 mol/Ldexamethasone, and 10 mmol/L b-glycerophosphate as described previously (19).




TABLE 1. RT-PCR Primers Sequence

Gene Sequence (5’ -3’) Size (bp)

Heme oxygenase-1 (HO-1) Forward: AAGATTGCCCAGAAAGCCCTGGReverse: AACTGTCGCCACCAGAAAGCTGAG 399

Alkaline phosphatase (ALP) Forward: ACGTGGCTAAGAATGTCATCReverse: CTGGTAGGCGATGTCCTTA 475

Osteopontin (OPN) Forward: CCAAGTAAGTCCAACGAAAGReverse: GGTGATGTCCTCGTCTGTA 347

Dentin-matrix protein-1 (DMP-1) Forward: CAGGAGCACAGGAAAAGGAGReverse: CTGGTGGTATCTTCCCCCAGGAG 213

Dentin sialophosphoprotein (DSPP) Forward: CAGGAGCACAGGAAAAGGAGReverse: CTGATTTGCTGCTGTCTGAC 488GAPDH Forward: CGGAGTCAACGGATTTGGTCGTATReverse: AGCCTTCTCCA TGGTGGTGAAGAC 306


Cell Viability AssayAfter treatment of HDPCs with CoPP or other drugs, 50 mL 3-(4,5-

dimethylthiazol-2-yl)-2.5 diphenyltetrazolin bromide (MTT, 2 mg/mL)was added to each well, and then the samples were incubated for 4hours and centrifuged (200 g for 10 minutes). After aspiration of super-natant, the cells were lysed and solubilized by the addition of 50 mLdimethyl sulfoxide (DMSO). The absorbance of each sample wasanalyzed at 540 nm using the microtiter plate reader. Cytotoxicity(%) was calculated relatively to the control.

Semiquantitative Reverse Transcriptase PolymeraseChain Reaction

Total RNA was extracted from the cells by using TRIzol reagent(Invitrogen, Carlsbad, CA) according to the manufacturer’s instruc-

Figure 1. The effects of CoPP on (A) growth, (B) mineralized nodule formation, anconcentrations ranging from 0 to 40 mmol/L for 3, 7, and 14 days in OM with 10 mcorbic acid. (A) Cell viability was evaluated by using an MTT assay. (B) Mineralizatiosuch odontoblastic differentiation markers as DMP-1, OPN, and DSPP were assessediments. )Statistically significant difference as compared with the control, p < 0.05


tions. Then, 1 mg RNA was reverse transcribed for first-strand comple-mentary DNA synthesis (Gibco BRL, Rockville, MD). The cDNA wasamplified in a final volume of 20 mL containing 2.5 mmol/L magnesiumdicholoride, 1.25 U Ex Taq polymerase (Bioneer, Daejeon, Korea), and1 mmol/L specific primers. The sequences of the specific primers usedin this study are detailed in Table 1. Reverse transcriptase polymerasechain reaction products were electrophoresed on 1.5 % agarose gelwith 0.5 mg/mL ethidium bromide. Bands were detected by ultravioletillumination of ethidium bromide–stained gels.

Alizarin Red StainAfter 2 weeks of differentiation induction, cells were rinsed with

phosphate buffered saline, air dried, and fixed in ice-cold 95% ethanolfor 30 minutes at �20�C. Subsequently, the cells and the matrix were

d (C) odontoblastic differentiation in HDPCs. Cells were cultured with CoPP atmol/L b-glycerophosphate, 10�7 mol/L dexamethasone, and 50 mg/ml L-as-

n was analyzed by alizarin red staining. (C) The mRNA expression of HO-1 andby RT-PCR analysis. These data are representative of three independent exper-.

HO-1 in Differentiation of Pulp Cells 1327

stained with 40 mmol/L Alizarin red-S (pH = 4.2) for 1 hour at roomtemperature, washed extensively five times with deionized water, andrinsed with PBS (without Mg2+ or Ca2+) for 15 minutes.
Transfection of siRNACells in 6-well plates at a density of 5� 105 cells/well were grown

for 16 hours and then transfected with 20 pmol siRNA duplexes (Dae-jon, South Korea) by using Lipofectamine RNAiMAX (Invitrogen, Carls-bad, CA) according to the manufacturer’s instructions. SilencerNegative Control siRNA (Invitrogen) was used as a negative controland was introduced into the cells using the same protocol. After trans-fection, cells were cultured in six-well plates at 37�C until needed.

Statistical AnalysisValues were calculated as the mean and standard deviation. Statis-

tical significance was evaluated by one-way analysis of variance using theSPSS (Version 11.0; SPSS, Chicago, IL) computer program.

ResultsEffects of HO-1 Induction on the Proliferation andDifferentiation of HDPCs

We first evaluated the effects of exogenous CoPP on HDPCs growthby using an MTT assay. As shown in Figure 1A, CoPP stimulated cellularviability in a dose-dependent manner after 3 and 7 days of treatmentcompared with the control. Significant growth stimulatory effects

Figure 2. The effects of a CO scavenger, iron chelator, and HO-1 inducer on the (A)Hb (80 mmol/L), DFO (0.5 mmol/L), hemin (20 mmol/L), and CoPP (20 mmol/L) fFigure 1 was followed. These data are representative of three independent experim

1328 Kim et al.

were observed at 40 mmol/L CoPP on days 3 and 7. However, after expo-sure to CoPP for 14 days, pulp cells showed no difference in cell viabilitycompared with controls.

To investigate the potential of HDPCs for odontoblast-like differen-tiation after CoPP treatment, we used alizarin red staining and assessedthe messenger RNA (mRNA) expression of several differentiationmarkers. In HDPCs, exposure to 40 mmol/L CoPP for 14 days resultedin the stimulation of extracellular mineral deposition (Fig. 1B). Treat-ment of cells with control (OM) increased mRNA expression of differ-entiation markers in a time-dependent fashion, with a maximal increasefor 14 days. The effect of CoPP on odontogenic gene (eg, DMP-1 andDSPP) and osteoblastic gene (eg, ALP and OPN) mRNA expressionwas upregulated in a dose- and time-dependent manner (Fig. 1C).

Effects of HO-1 Metabolites and Hemin on HDPCProliferation Differentiation

To determine whether the CO and iron released during hemedegradation by HO-1 were responsible for the growth and differentia-tion, we examined the effects of Hb (a CO scavenger, 80 mmol/L)and desferrioxamine (DFO, an iron chelator, 0.5 mmol/L) on theviability and mRNA expression of various differentiation markers inHDPCs. Hb and DFO were used at noncytotoxic doses and had no signif-icant effect on cellular viability at days 3, 7, and 14 (Fig. 2A). Thepretreatment of HDPCs with Hb and DFO inhibited HO-1 and themRNA expression of several differentiation markers, including ALP,OPN, DMP-1, and DSPP, compared with the control (Fig. 2B). To prove

viability and (B) odontoblastic differentiation of HDPCs. Cells were treated withor 7 and 14 days in OM. The same procedure as that described in the legend toents. )Statistically significant difference as compared to the control, p < 0.05.


that the differentiation inductive action of CoPP in HDPCs is relevant toHO-1 activation, we examined whether another HO-1 inducer, hemin,also increases HDPC proliferation and differentiation. Hemin increasedcell growth after 7 days (Fig. 2A) and the expression of several odon-toblastic differentiation markers (Fig. 2B).
Effects of HO-1 Inhibition by SnPP and siRNA on Growthand Odontoblastic Differentiation in HDPCs

The involvement of HO-1 in HDPC growth and differentiation wastested using HO-1 siRNA and the specific HO-1 inhibitor SnPP. SnPP

Figure 3. (A and D) The effects of the HO-1 inhibitor SnPP and HO-1 siRNA on godontoblastic differentiation markers in HDPCs. (A-C) Experimental group were trevector and HO-1 siRNA for 7 days. )Statistically significant difference as compared wthe OM-treated group, p < 0.05. These data are representative of three independe


suppressed cell viability (Fig. 3A) and mineralized nodule formation(Fig. 3B) and the expression of osteogenic supplement medium–induced differentiation mRNAs (Fig. 3C) in a dose-dependent manner.We next used RNA interference to knockdown HO-1 expression and as-sayed cell growth and differentiation after CoPP treatment. The siRNA-induced knockdown of HO-1 treatment alone was found to be associ-ated with a significant decrease in cell viability in HDPCs. The treatmentof HDPCs with HO-1 siRNA resulted in a 35% decrease in CoPP-inducedcell viability (Fig. 3D) and blocked the CoPP-induced up-regulation ofALP, OPN, DMP-1, and DSPP mRNA expression, whereas the

rowth, (B) mineralized nodule formation, and (C and E) mRNA expression ofated with OM media. (D and E) Cells were transiently transfected with controlith the control, p < 0.05. #Statistically significant difference as compared with

nt experiments.


transfection of HDPCs with the same amount of nonspecific siRNA wasnot effective (Fig. 3E).
DiscussionPredictable pulp capping procedures remain problematic,

possibly because of the lack of appropriate stimulatory factors fordentin formation. Recently, we showed that simvastatin (20) andmineral trioxide aggregate plus enamel matrix derivative (21) stimu-lated odontoblastic differentiation in HDPCs. Thus, the activation ofstem cells in human pulp represents a potential cellular approach topulp capping or dentin regenerative treatment (4, 22). Ourhypothesis is that the differentiation of HDPCs into odontoblasts canbe accelerated by signaling molecules containing HO-1 as a pulp-capping strategy. The aim of this study was to investigate whether HO-1 induction by CoPP or hemin promotes growth and differentiationand whether HO-1 inhibition by SnPP, an HO-1 metabolite, and HO-1siRNA could suppress cell growth and differentiation.

CoPP has been shown to be a potent inducer of HO-1 expressionand activity in vivo and in vitro (23–25). CoPP treatment has anti-inflammatory effects in macrophages (23) and immunosuppressiveeffects in monocyte-derived dendritic cells (24). Recently, the upregu-lation of HO-1 expression was shown to promote osteoblastic differen-tiation in MSCs (25) and endothelial progenitor cells (16). Our resultsshow that HO-1 induction by CoPP increases cell growth, calciumnodule formation, and mRNA expression of HO-1 and some odonto-blastic markers (Fig. 1). These results are consistent with our previousdata from PDLCs (7, 12). Thus, the beneficial effect of CoPP results inincreased growth of HDPCs and odontoblastic differentiation that maypartially involve HO-1.

Increasing evidence suggests that HO-1 functions in the modula-tion of differentiation (13, 14); however, the mechanism has not yetbeen clearly delineated. Ferrous iron released as a result of HO-1 activityrapidly induces the expression of ferritin, thereby protecting cells underoxidizing conditions by sequestering free cytosolic iron (26). CO acts asan activator of guanylyl cyclase in a manner similar to nitric oxide–likeretrograde messenger (27) and carries out an anti-inflammatory andcytoprotective function (28). In the present study, we showed that COscavenger and iron chelator blocked the differentiation-inducing effectsin OM-stimulated HDPCs (Fig. 2). These results suggest that the CO andiron derived from heme degradation mediate the differentiation ofHDPCs.

During the course of the study, there was a variation of the dosezero expression of odontoblast-like mRNA markers in Figure 1. Asimilar pattern was observed in the control groups of Figures 2 and 3although this was not significant. This variation could be caused bythe fact that different preparation or populations of cells are presentin each experiment (29). More work is needed to understand the basicdynamics and differences of human protein or mRNA levels and its vari-ation in individual cells.

To determine whether the proliferation and differentiation ofHDPCs afforded by CoPP and hemin are associated with HO-1 expres-sion, we used HO-1 siRNA and SnPP. The down-regulation of HO-1expression by HO-1 siRNA transfection or the arrest of HO activity bythe HO inhibitor SnPP blocked OM-induced HO-1, OPN, ALP, DSPP,and DMP-1 mRNA expression (Fig. 3). These expressions appear tobe partially blocked, and this partial blockage suggests that odonto-blastic differentiation may be partly mediated by an HO-1 dependentpathway in HDPCs. Furthermore, we were able to show that thesiRNA-induced knockdown of HO-1 is associated with growth inhibitionin HDPCs cells. These data provide evidence for the functional role ofHO-1 as an important survival factor in HDPCs.

1330 Kim et al.

To our knowledge, this study is the first to show that growth andodontoblastic differentiation in HDPCs is activated by HO-1 inductionand attenuated by HO-1 inhibition. Thus, HO-1 is an important cellulartarget of HDPCs, with clinical implications for pulp capping materials orthe regeneration of human dental tissues for tissue engineering.

References1. Gronthos S, Brahim J, Li W, et al. Stem cell properties of human dental pulp stem

cells. J Dent Res 2002;81:531–53.2. Agata H, Kagami H, Watanabe N, et al. Effect of ischemic culture conditions on the

survival and differentiation of porcine dental pulp-derived cells. Differentiation2008;76:981–93.

3. Butler WT, Ritchie H. The nature and functional significance of dentin extracellularmatrix proteins. Int J Dev Biol 1995;39:169–79.

4. Gronthos S, Mankani M, Brahim J, et al. Postnatal human dental pulp stem cells(DPSCs) in vitro and in vivo. Proc Natl Acad Sci U S A 2000;97:13625–30.

5. Nakashima M. Bone morphogenetic proteins in dentin regeneration for potentialuse in endodontic therapy. Cytokine Growth Factor Rev 2005;163:369–76.

6. Choi AM, Alam J. Heme oxygenase-1: function, regulation, and implication of a novelstress-inducible protein in oxidant-induced lung injury. Am J Respir Cell Mol Biol1996;15:9–19.

7. Lee SK, Choi HI, Yang YS, et al. Nitric oxide modulates osteoblastic differentiationwith heme oxygenase-1 via the mitogen activated protein kinase and nuclearfactor-kappaB pathways in human periodontal ligament cells. Biol Pharm Bull2009;32:1328–34.

8. Min KS, Hwang YH, Ju HJ, et al. Heme oxygenase-1 mediates cytoprotection againstnitric oxide-induced cytotoxicity via the cGMP pathway in human pulp cells. OralSurg Oal Med Oral Pathol Oral Radol Endod 2006;102:803–8.

9. Min KS, Kwon YY, Lee HJ, et al. Effects of proinflammatory cytokines on the expres-sion of mineralization markers and heme oxygenase-1 in human pulp cells. J Endod2006;32:39–43.

10. Pi SH, Kim SC, Kim HT, et al. Defense of heme oxygenase-1 against cytotoxic andreceptor of NF kappa-B ligand inducing effects of hydrogenin human periodontalligament cells. J Periodontal Res 2007;42:331–9.

11. Lee SK, Pi SH, Kim SH, et al. Substance P regulates macrophage inflammatory protein3a/CCL20 with hemein human periodontal ligament cells. Clin Exp Immunol 2007;150:567–75.

12. Kook YA, Lee SK, Son DH, et al. Effects of substance P on osteoblastic differentiationand heme oxygenase-1 in human periodontal ligament cells. Cell Biol Int 2009;33:424–8.

13. Kim DH, Burgess AP, Li M, et al. Heme oxygenase-mediated increases in adiponectindecrease fat content and inflammatory cytokines tumor necrosis factor-alpha andinterleukin-6 in Zucker rats and reduce adipogenesis in human mesenchymalstem cells. J Pharmacol Exp Ther 2008;325:833–40.

14. Barbagallo I, Tibullo D, Di Rosa M, et al. A cytoprotective role for the hemeoxygenase-1/CO pathway during neural differentiation of human mesenchymalstem cells. J Neurosci Res 2008;86:1927–35.

15. Johns DG, Zelent D, Ao Z, et al. Heme-oxygenase induction inhibits arteriolar throm-bosis in vivo: effect of the non-substrate inducer cobalt protoporphyrin. Eur J Phar-macol 2009;606:109–14.

16. Wang JY, Lee YT, Chang PF, et al. Hemin promotes proliferation and differentiationof endothelial progenitor cells via activation of AKT and ERK. J Cell Physiol 2009;219:617–25.

17. Chou CC, Hsu CY. Involvement of PKC in TPA-potentiated apoptosis induction duringhemin-mediated erythroid differentiation in K562 cells. Naunyn SchmiedebergsArch Pharmacol 2009;379:1–9.

18. Kitagawa M, Ueda H, Iizuka S, et al. Immortalization and characterization of humandental pulp cells with odontoblastic differentiation. Arch Oral Biol 2007;52:727–31.

19. Yasuda Y, Ogawa M, Arakawa T, et al. The effect of mineral trioxide aggregate on themineralization ability of rat dental pulp cells: an in vitro study. J Endod 2008;34:1057–60.

20. Min KS, Lee YM, Hong SO, et al. Simvastatin promotes odontoblastic differentiationand expression of angiogenic factors via heme oxygenase-1 in primary culturedhuman dental pulp cells. J Endod 2010;36:447–52.

21. Min KS, Yang SH, Kim EC. The combined effect of mineral trioxide aggregate andenamel matrix derivative on odontoblastic differentiation in human dental pulpcells. J Endod 2009;35:847–51.

22. Shi S, Bartold PM, Miura M, et al. The efficacy of mesenchymal stem cells to regen-erate and repair dental structures. Orthod Craniofac Res 2005;8:191–9.

23. Lin HY, Shen SC, Lin CW, et al. Cobalt protoporphyrin inhibition of lipopolysaccha-ride or lipoteichoic acid-induced nitric oxide production via blocking c-JunN-terminal kinase activation and nitric oxide enzyme activity. Chem Biol Interact2009;180:202–10.


24. Listopad J, Asadullah K, Sievers C, et al. Heme oxygenase-1 inhibits T cell-dependent
skin inflammation and differentiation and function of antigen-presenting cells. ExpDermatol 2007;16:661–70.

25. Vanella L, Kim DH, Asprinio D, et al. HO-1 expression increases mesenchymal stemcell-derived osteoblast but decreases adipocyte lineage. Bone 2010;46:236–43.

26. Ward RJ, Kuhn LC, Kaldy P, et al. Control of cellular iron homeostasis by iron-responsive elements in vivo. Eur J Biochem 1994;220:927–31.


27. Verma A, Hirsch DJ, Glatt CE, et al. Carbon monoxide: a putative neural messenger.Science 1993;259:381–4.

28. Otterbein LE, Bach FH, Alam J, et al. Carbon monoxide has anti-inflammatoryeffects involving the mitogen-activated protein kinase pathway. Nat Med 2000;6:422–8.

29. Raser JM, O’Shea EK. Noise in gene expression: origins, consequences, and control.Science 2005;309:2010–3.



Transforming Growth Factor b1–Induced Heat ShockProtein 27 Activation Promotes Migration of Mouse DentalPapilla–derived MDPC-23 CellsSeong-Min Kwon, MS,* Soo-A. Kim, PhD,

†Jung-Hoon Yoon, DDS, PhD,* and Sang-Gun Ahn, PhD*

Abstract
Introduction: Transforming growth factor b1 (TGFb1)regulates cellular functions including cell growth, differ-entiation, angiogenesis, migration, and metastasis. TheTGFb1 signal transduction pathways are mostly unde-fined in mouse dental papilla-derived MDPC-23 cells.In this study, we investigated TGFb1-induced migrationfocusing on heat shock protein 27 (Hsp27) activation.Methods: Cellular responses mediated by TGFb1 inMDPC-23 cells were measured by Western blot andMTT assays. Cell migration was determined by countingmigrated cells using the chemotaxis cell migrationassay. Results: TGFb1 induced cell migration andincreased the phosphorylation of Hsp27 and p38MAPK in MDPC-23 cells. However, TGFb1 did not affectAkt/NF-kB signaling to regulate the migration of MDPC-23 cells. Inhibiting p38 MAPK with SB203580 blockedTGFb1-induced Hsp27 activation and cell migration.Conclusion: Hsp27 phosphorylation followed by p38MAPK activation was required for TGFb1-inducedmigration, and Hsp27 itself contributed to MDPC-23cell migration. (J Endod 2010;36:1332–1335)
Key WordsHeat shock protein 27, MDPC-23, p38 MAPK, transform-ing growth factor b1

From the *Department of Pathology, School of DentistryChosun University, Gwangju, Korea; and †Department ofBiochemistry, Oriental Medicine Dongguk University,Gyeongju, Korea.

Supported by the Korea Science and Engineering Founda-tion (KOSEF) grant funded by the Korea government (MOST)(no. R13-2008-010-01001-0), the Korean Research FoundationGrant funded by the Korean Government (MOEHRD, BasicResearch Promotion Fund) (KRF-2007-331- C00208), and theNational R&D Program for Cancer Control, Ministry of Health& Welfare, Republic of Korea (0720430).

Address requests for reprints to Dr Sang-Gun Ahn or Dr.Jung-Hoon Yoon, Department of Pathology, School of DentistryChosun University, Gwangju 501-759, Korea. E-mail address:[email protected] or [email protected]/$0 - see front matter


1332 Kwon et al.

Transforming growth factor b (TGFb) is a multifunctional cytokine that regulatesa variety of cellular processes such as proliferation, migration, differentiation,

apoptosis, and immune responses in numerous cell types (1–3). Recent studies havesuggested that TGFb-induced migration and invasion is mediated by several factorssuch as phosphatidylinositol 3-kinase (PI3K) and mitogen-activated protein kinases(p38 MAPK and ERK) in cancer and smooth muscle cells (4, 5). Furthermore, TGFb1induces the secretion of several dentin matrix proteins associated with primarydentinogenesis through smooth muscle actin (SMA)- and mothers againstdecapentaplegic (MAD)-related protein (SMAD) signaling in dental pulp cells (6, 7).

Previously, the expression of heat shock protein 27 (Hsp27) has been reportedin odontoblast and ameloblasts during tooth development (8, 9). Hsp27 is anadenosine triphosphate-independent molecular chaperone that protects cells fromexternal stimuli and chemotherapeutic agents (10). Phosphorylation of Hsp27 wasshown to be necessary for actin formation, stabilization of focal adhesions, andpromotion of cell migration through signal transducer and activator of transcription3 (STAT3) signaling and focal adhesion kinase signaling (11, 12).

Cell migration is critical for a variety of processes, including angiogenesis, inflam-mation, development, wound healing, and tumor metastasis. It has been shown that cellmigration is regulated by various signaling pathways, including b1 integrin, Akt, extra-cellular signal-regulated kinase (ERK), and NF-kB–dependent pathways in humancancer cells (5, 13). Although the roles of TGFb1 and Hsp27 in cell migration havebeen studied in some cancer cell lines, the roles of these proteins in dental pulpand/or papilla cells remain largely unknown.

In this study, we investigated whether TGFb1 is involved in the migration of mousedental papilla derived MDPC-23 cells. We showed that TGFb1 induces the phosphory-lation of Hsp27 through p38 MAPK activation, which mediates the migration of MDPC-23 cells.

Materials and MethodsReagents and Antibodies

TGFb1 was purchased from R&D System (Minneapolis, MN). The p38 inhibitorSB203580 was purchased from Sigma-Aldrich Corporation (St Louis, MO). Theprimary antibodies used were mouse anti-p38 MAPK monoclonal antibody,antiphospho-p38MAPK (Thr180/Tyr182) polyclonal antibody, anti-Hsp27 monoclonalantibody, and antiphospho-Hsp27 (Ser82) polyclonal antibody (Santa Cruz, CA).

Cell CultureWe used an MDPC-23 cell line derived from 18- to 19-day-old CD-1 fetal mouse

molar dental papilla (14). The mouse MDPC-23 cells were maintained in Dulbeccomodified Eagle medium supplemented with 10% fetal bovine serum, 1 � modifiedEagle medium nonessential amino acids, 100 U/mL penicillin, and 100 mg/mL strepto-mycin in a humidified atmosphere containing 5% CO2 and 95% air at 37�C.

MTT AssayBriefly, cells (1 � 105 per well) were seeded into 12-well plates. Before testing,

3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyl-tetrazolium bromide (MTT) solution




Figure 1. The effect of TGFb1 on cell proliferation and migration of MDPC-23 cells. (A) MDPC-23 cells were incubated with 1 ng/mL of TGFb1 for 24 or 48 hours.Cell proliferation was analyzed by MTT assay. (B) MDPC-23 cell migration induced by TGF-b1 (1 ng/mL) was measured by a transwell assay. The graph representsthe relative number of cells from three separated experiments. A p value <0.001 was regarded as statistically significant when compared with the control.

Figure 2. The effect of TGFb1 on the phosphorylation of p38 MAPK andHsp27 in MDPC-23 cells. Cells were treated with 1 ng/mL of TGFb1 for theindicated time. (A) Western blot analysis of p38MAPK/Hsp27 phosphorylation.(B) Integrin b1/Akt signaling on TGFb1-treated MDPC-23 cells.


(5 mg/mL in PBS) was added, and cells were incubated at 37�C for 3hours. The culture medium was then aspirated, and acid isopropanol(0.04 mol/L hydrogen chloride (HCl) in isopropanol) was added todissolve the dark blue crystals. The optical density value of the dissolvedsolute was then measured using a Microplate Autoreader (Bio-TekInstruments Inc., Winooski, VT) at a wavelength of 570 nm.

Cell Migration AssayThe cell migration assay was performed using a Chemotaxis Cell

Migration Assay kit (CHEMICON, Millipore Corporation, Billerica,MA) according to the manufacturer’s instructions. The cells (1 �104) were allowed to migrate for 24 hours at 37�C in a humidifiedatmosphere containing 5% CO2. The cells that migrated to the lowersurface of the membrane were fixed with methanol and stained withhematoxylin for 5 minutes. Images were captured using an OlympusBX41 (Tokyo, Japan) inverted microscope, and migrated cells werecounted. The number of migrated cells was counted from three inde-pendent experiments.

Western Blot AnalysisThe total protein was extracted from control and TGFb1-treated

MDP-C23 cells and then extracted using lysis buffer (1% TritonX-100 (Bio-Rad, Quarry Bay, Hong Kong), 150 mmol/L NaCl, 5mmol/L EDTA, and protease inhibitors). Equal amounts of proteinwere electrophoresed on 12% sodium dodecyl sulfate polyacrylamidegel electrophoresis (SDS-PAGE) and then transferred to a polyvinylidenefluoride (PVDF) membrane. All primary antibodies were used ata 1:1,000 dilution. Actin was used as a control. Protein detection wasperformed using the Super Signal West Femto Maximum SensitivitySubstrate (Pierce, Rockford, IL).

Statistical AnalysisStatistical analysis was performed with the data obtained from

three independent experiments. Data were analyzed and presented asthe mean� standard error of the mean. A p value <0.05 was regardedas statistically significant.

ResultsTGFb1-Induced Migration of Mouse Dental Papilla-derived MDPC-23 Cells

To investigate whether TGFb1 induces cell growth and migrationof MDPC-23 cells, the cells were treated with TGFb1 for 24 or 48 hours.

JOE — Volume 36, Number 8, August 2010 TGFb1-Induce

As shown in Figure 1A, TGFb1 did not affect the cell growth. We evalu-ated the effects of TGFb1 on the migration of MDPC-23 cells. A clearlya number of cells treated with TGFb1 migrated than did the control cells(Fig. 1B).

TGFb1-Induced Migration Involves p38 MAPK/Hsp27Activation

Because the p38 MAPK/Hsp27 pathway and integrin b1/Aktsignaling are involved in migration, we examined the effect of TGFb1on migration pathway in MDPC-23 cells. As shown in Figure 2A, phos-phorylation of p38 MAPK and Hsp27 was induced after 6 hours ofTGFb1 treatment and continued for up to 24 hours. We then examinedwhether TGFb1 also enhances integrin b1/Akt signaling activation. Wefound that the expression of integrin b1 was not altered by TGFb1.TGFb1 did not affect the pathway up- and down-stream of integrinb1 including focal adhesion kinase, Akt, and IkB (Fig. 2B).

TGFb1-Induced Migration Induced Phosphorylation ofHsp27 via p38 MAPK Activation

To confirm that the TGFb1-induced migration of MDPC-23 cellswas dependent on Hsp27, an Hsp27-specific antibody was used. Wefound that TGFb1-induced migration was significantly inhibited byHsp27 antibody treatment (Fig. 3A and B). The results indicated that

d Heat Shock Protein 27 Activation Promotes Migration of MDPC-23 Cells 1333

Figure 3. The effect of Hsp27 and p38 MAPK inhibition on TGFb1-induced migration. MDPC-23 cells were incubated with or without TGFb1 (1 ng/mL) and Hsp27antibody (3 mg/mL) for 24 hours. (A) Migrated cells were counted by a transwell assay. (B) Cell migration was quantified by counting the number of cells thatmigrated into the inner membrane. The graph represents the relative number of cells from three separated experiments. A p value <0.001 was regarded as statis-tically significant. (C) MDPC-23 cells were pretreated with a p38 MAPK inhibitor, SB203580 (10 mmol/L), for 1 hour and then treated TGFb1 (1 ng/mL) for 24hours. (D) Cell migration was quantified by counting the number of migrated cells. The graph represents the relative number of cells from three separated exper-iments. A p value <0.01 was regarded as statistically significant. (This figure is available in color online at www.aae.org/joe/.)


TGF-b1-induced migration was dependent on Hsp27 activation inMDPC-23 cells.

To investigate whether the phosphorylation of Hsp27 was depen-dent on p38MAPK activation, cells were pretreated with a p38MAPKinhibitor, SB203580, before TGFb1 stimulation. Western blot analysisshowed that SB203580 inhibited TGFb1-induced p38 MAPK phosphor-ylation. The phosphorylation of Hsp27 was also inhibited by SB203580(Fig. 3C). We found that SB203580 significantly reduced TGFb1-induced migration (Fig. 3D). These results indicate that p38MAPKwas required for TGFb1-induced Hsp27 phosphorylation and migra-tion of MDPC-23 cells.

DiscussionTGFb is an important mediator of extracellular matrix biosynthesis

and is involved in the regulation of cell growth, differentiation, andmigration (1–3). In TGFb signaling, the role of TGFb1 inodontoblast differentiation and dentin mineralization during primarydevelopment and dentinogenesis is well established (4, 5). Severalrecent publications reported that TGFb1 regulates growth anddifferentiation of pulp cells via MAPK, matrix metalloproteinases,cathepsins, and the Smad2/3 pathway (6, 7, 15, 16). In addition,TGFb2 modulates the pulpal functions at specific stages ofdifferentiation processes via activation of the MEK/ERK1/2 andALK5/Smad2/3 pathways (17, 18).

In this study, we examined the role of TGFb1 in mouse dentalpapilla-derived MDPC-23 cells. We detected the phosphorylation ofp38 MAPK and Hsp27 in response to TGFb1 treatment. Several studieshave shown that Hsp27 directly participate in TGF-b–regulated physio-

1334 Kwon et al.

logical processes. For example, invasive and metastatic characteristicsinduced by TGFb1 correlate with Hsp27 expression in gastric cancercells (19). Hatakeyama et al (20) showed that TGFb-stimulatedHsp27 induction is associated with proliferation in osteoblast-like cells.However, until recently, the mechanisms of action of TGFb and Hsp27have been unknown in dental papilla-derived MDPC-23 cells.

We show for the first time that TGFb1 is involved in the migration ofMDPC-23 cells. Furthermore, TGFb1-induced migration was blockedby an Hsp27-specific antibody in MDPC-23 cells. These results sug-gested that TGFb1-induced Hsp27 phosphorylation was able topromote cell migration in MDPC-23 cells. Previous studies have shownthat the expression of Hsp25 (homologous to Hsp27 in the mouse) issuggested to act as a switch between cell proliferation and early-stagedifferentiation in developing odontoblast- and ameloblast-lineage cells(21). Therefore, our findings raise the possibility that Hsp27 may playan important role in migration toward the dentin before differentiation.

It has been reported that p38 MAPK is involved in the phosphor-ylation of Hsp27 and the migration of vascular smooth muscle andhepatocellular carcinoma cells (22, 23). In addition, invasion andmigration in cancer cells are also involved in integrin b1 signaling(24). We showed that TGFb1 induced Hsp27 phosphorylation via theactivation of p38 MAP kinase in MDPC-23 cells. A p38 MAPK-specificinhibitor, SB203580, blocked the phosphorylation of Hsp27 and themigration of MDPC-23 cells. However, there was no significant changein the activation of intergrin b1/Akt signaling in TGFb1-stimulated cells.Based on these findings, it is possible that TGFb1-induced p38 MAPK/Hsp27 pathway signaling may regulate MDPC-23 cell migration (Fig. 4).

It is well recognized that Hsp27 acts as a molecular chaperonesand improves survival under stressful conditions. It has been shown



Figure 4. A model of TGFb1-mediated cell migration in mouse dental papilla-derived MDPC-23 cells. The regulation of cell migration and phosphorylationof Hsp27 and p38 MAPK by TGFb1. See Discussion section for additionaldetails. (This figure is available in color online at www.aae.org/joe/.)


that Hsps including Hsp70 and Hsp27 are expressed during osteoblastsdifferentiation (25). In our study, TGFb1 did not affect the level of otherHsps such as Hsp40, Hsp60, Hsp70, Hsp90, and Hsp110 in MDPC-23cells but did enhance phosphorylation of Hsp27 (data not shown).These findings lead us to speculate that TGFb1-induced Hsp27 activa-tion may play an important role in MDPC-23 functions.

One study has reported the effect of TGFb1 on apoptosis in MDPC-23 cells (26). The reasons for the difference of TGFb1-mediatedbiological effect in MDPC-23 cells are unclear but may be related toindividual differences in culture conditions (serum or other additionalfactors) and differentiation stages of cultured cells. Further investigationof the relationship between Hsp27 and TGFb1 responsive migratoryfactors, such as matrix metalloproteinases, on cell migration mediatedby TGFb1, will better define the molecular mechanisms of TGFb1during dentinogenesis.

References1. Ignotz RA, Massague J. Cell adhesion protein receptors as targets for transforming

growth factor-beta action. Cell 1987;51:189–97.2. Brenmoehl J, Miller SN, Hofmann C, et al. Transforming growth factor-beta 1

induces intestinal myofibroblast differentiation and modulates their migration.World J Gastroenterol 2009;15:1431–42.

3. Yoo KS, Nastiuk KL, Krolewski JJ. Transforming growth factor beta 1 inducesapoptosis by suppressing FLICE-like inhibitory protein in DU145 prostate epithelialcells. Int J Cancer 2009;124:834–42.

4. Yeh YY, Chiao CC, Kuo WY, et al. TGF-beta1 increases motility and alphavbeta3 in-tegrin up-regulation via PI3K, Akt and NF-kappaB-dependent pathway in humanchondrosarcoma cells. Biochem Pharmacol 2008;75:1292–301.

5. Kim HP, Lee MS, Yu J, et al. TGF-beta 1 (transforming growth factor-beta1)-mediatedadhesion of gastric carcinoma cells involves a decrease in Ras/ERKs (extracellular-

JOE — Volume 36, Number 8, August 2010 TGFb1-Induce

signal-regulated kinases) cascade activity dependent on c-Src activity. Biochem J2004;379:141–50.

6. Hwang YC, Hwang IN, Oh WM, et al. Influence of TGF-beta1 on the expression ofBSP, DSP, TGF-beta 1 receptor I and Smad proteins during reparative dentinogen-esis. J Mol Histol 2008;39:153–60.

7. He WX, Niub ZY, Zhao SL, et al. TGF-b activated Smad signaling leads to a Smad3-mediated down-regulation of DSPP in an odontoblast cell line. Arch Oral Biol 2004;49:911–8.

8. Ohshima H, Ajima H, Kawano Y, et al. Transient expression of heat shock protein(Hsp) 25 in the dental pulp and enamel organ during odontogenesis in the ratincisor. Arch Histol Cytol 2000;63:381–95.

9. Ohtsuka Y, Nakakura-Ohshima K, Noda T, et al. Possible role of heat shock protein(Hsp) 25 in the enamel organ during amelogenesis in the rat molar. Arch HistolCytol 2001;64:369–78.

10. Ciocca DR, Oesterreich S, Chamness GC, et al. Biological and clinical implicationsof heat shock protein 27,000 (Hsp27): a review. J Natl Cancer Inst 1993;85:1558–70.

11. Hirano S, Shelden EA, Gilmont RR. HSP27 regulates fibroblast adhesion, motility,and matrix contraction. Cell Stress Chaperones 2004;9:29–37.

12. Lee JW, Kwak HJ, Lee JJ, et al. HSP27 regulates cell adhesion and invasion via modu-lation of focal adhesion kinase and MMP-2 expression. Eur J Cell Biol 2008;87:377–87.

13. Wei YY, Chen YJ, Hsiao YC, et al. Osteoblasts-derived TGF-beta 1 enhance motilityand integrin upregulation through Akt, ERK, and NF-kappaB-dependent pathwayin human breast cancer cells. Mol Carcinog 2008;47:526–37.

14. Hanks CT, Sun ZL, Fang DN, et al. Cloned 3T6 cell line from CD-1 mouse fetal molardental papillae. Connect Tissue Res 1998;37:233–49.

15. Palosaari H, Wahlgren J, Larmas M, et al. The expression of MMP-8 in human odon-toblasts and dental pulp cells is down-regulated by TGF-beta 1. J Dent Res 2000;79:77–84.

16. Tai TF, Chan CP, Lin CC, et al. Transforming growth factor beta 2 regulatesgrowth and differentiation of pulp cells via ALK5/Smad2/3. J Endod 2008;34:427–32.

17. Tersariol IL, Geraldeli S, Minciotti CL, et al. Cysteine cathepsins in human dentin-pulp complex. J Endod 2010;36:475–81.

18. Tai TF, Chan CP, Lin CC, et al. Transforming growth factor beta2 regulatesgrowth and differentiation of pulp cells via ALK5/Smad2/3. J Endod 2008;34:427–32.

19. Wang K, Li J, Zhen C, et al. Enhanced invasive and metastatic potential induced bytransforming growth factor-beta 1 might be correlated with glutathione-S-transferase-pi, cofilin and heat shock protein 27 in SGC-7901 gastric cancer cells.Acta Biochim Biophys Sin 2007;39:520–6.

20. Hatakeyama D, Kozawa O, Niwa M, et al. Upregulation by retinoic acid of transform-ing growth factor-L-stimulated heat shock protein 27 induction in osteoblasts:involvement of mitogen-activated protein kinases. Biochimica Biophysica Acta2002;1589:15–30.

21. Nakasone N, Yoshie H, Ohshima H. An immunohistochemical study of theexpression of heat-shock protein-25 and cell proliferation in the dental pulpand enamel organ during odontogenesis in rat molars. Arch Oral Biol 2006;51:378–86.

22. Chen HF, Xie LD, Xu CS. The signal transduction pathways of heat shock protein 27phosphorylation in vascular smooth muscle cells. Mol Cell Biochem 2010;333:49–56.

23. Guo K, Liu Y, Zhou H, et al. Involvement of protein kinase C beta-extracellular signal-regulating kinase 1/2/p38 mitogen-activated protein kinase-heat shock protein 27activation in hepatocellular carcinoma cell motility and invasion. Cancer Sci 2008;99:486–96.

24. Mon NN, Ito S, Senga T, et al. FAK signaling in neoplastic disorders: a linkagebetween inflammation and cancer. Ann N Y Acad Sci 2006;1086:199–212.

25. Shakoori AR, Oberdorf AM, Owen TA, et al. Expression of heat shock genes duringdifferentiation of mammalian osteoblasts and promyelocytic leukemia cells. J CellBiochem 1992;48:277–87.

26. He WX, Niu ZY, Zhao SL, Smith AJ. Smad protein mediated transforming growthfactor beta 1 induction of apoptosis in the MDPC-23 odontoblast-like cell line.Arch Oral Biol 2005;50:929–36.

d Heat Shock Protein 27 Activation Promotes Migration of MDPC-23 Cells 1335



Effects of Cryopreservation of Intact Teeth on the IsolatedDental Pulp Stem CellsSheng-Yang Lee, DDS, PhD,* Pao-Chang Chiang, DDS, MS,* Yu-Hui Tsai, PhD,†

Shih-Ying Tsai, PhD,†

Jiiang-Huei Jeng, DDS, PhD,‡

Toshitsugu Kawata, DDS, PhD,§

andHaw-Ming Huang, PhDk

Abstract
Introduction: Human dental pulp stem cells (DPSCs)have been reported to be useful material for futureregenerative medicine. Clinically, cryopreservation ofintact teeth can successfully preserve the periodontalligament for future autotransplantation; however, theeffects of cryopreservation procedure on the propertiesof DPSCs are still unclear. The aim of this study was totest whether DPSCs isolated from cryopreserved teethcan express stem cell–specific markers. Methods: Inthis study, a novel programmable freezer coupled toa magnetic field was used to perform the cryopreserva-tion experiments. The tested DPSCs were isolated frommagnetically cryopreserved and non-cryopreservedfresh teeth with an enzyme digestion procedure. Thesuccess rate of isolation, growth curves, morphology,stem cell–specific markers, and the differentiationcapacity of the isolated cells were evaluated andcompared. Results: The isolation rate of dental pulpcells from magnetically cryopreserved teeth was 73%.After culture for 5 generations, there was no significantdifference in cell viability between cells isolated frommagnetically cryopreserved teeth and those isolatedfrom fresh teeth. There were also no visible differencesbetween the 2 groups of dental pulp cells inmorphology, expression of stem cell markers, or osteo-genic and adipogenic differentiations. Conclusions:The results suggest that cryopreserved whole teethcan be used for autotransplantation and provide a viablesource of DPSCs. (J Endod 2010;36:1336–1340)
Key WordsCryopreservation, dental pulp stem cells, differentiation

From the *School of Dentistry, †Graduate Institute of MedicalSciences, and kGraduate Institute of Biomedical Materials andEngineering, Taipei Medical University, Taipei, Taiwan; ‡Schoolof Dentistry, National Taiwan University, Taipei, Taiwan; and§Department of Orthodontics, Hiroshima University, Hiroshima,Japan.

Drs Lee and Chiang contributed equally to this work.Address requests for reprints to Dr Haw-Ming Huang, Grad-

uate Institute of Biomedical Materials & Engineering, TaipeiMedical University, 250 Wu-Hsing Street, Taipei, Taiwan.E-mail address: [email protected]/$0 - see front matter


1336 Lee et al.

Several studies have indicated that postnatal stem cells are present in bone marrow,neural tissue, skin, and retina (1). These cells exhibit capacities for differentiation

into various cells and development into diverse tissue. These self-renewal capabilitiesmake stem cells become an effective material for regenerative medicine.

Human dental pulp stem cells (DPSCs) were first reported in 2000 (2). Unlikebone marrow–derived stem cells, DPSCs can be isolated with noninvasive procedures.In addition, unlike the use of embryonic stem cells, the use of DPSCs in research andtherapy is not considered to be controversial (3). In 2007, Jo et al (1) isolated postnatalstem cells from human dental tissues such as dental pulp, periodontal ligament,periapical follicle, and the surrounding mandibular bone marrow and found that thosestem cells were able to differentiate into osteoblasts and adipocytes. Because of theirmultipotent differentiation ability and immunosuppressive activity, dental stem cellsprovide an alternative to stem cells from other sources for use in regenerative medicine(4, 5). It is believed that DPSCs will play an important role in regenerative endodonticsin the near future (6). Although isolation of DPSCs from fresh teeth is possible (7),contamination and damage during long-term cryostorage might have an effect on theviability of dental pulp stem cells.

Cryopreservation on various tissue and cells has been investigated for severaldecades and has become an important issue for tissue engineering (8, 9). Laureys etal (10) cryopreserved pulpless teeth in liquid nitrogen at –196� for 7 days. Althoughtheir experiments demonstrated that pulpless teeth can revascularize after being cryo-preserved in a tooth bank for 1 week, DPSC viability and functions were not assessed(10). Recently, recovery of DPSCs from cryopreserved intact teeth was achieved byWoods et al (11). However, stem cell–specific markers CD-44 and STRO-1 were notexamined in their study.

The aim of this study was to confirm whether DPSCs isolated from cryopreservedteeth would survive and function normally after thawing.

Materials and MethodsCryopreservation and Cell Culture

Normal human premolars were collected from adults (aged 18–30 years) at theDepartment of Orthodontics, Wan-Fang Medical Center, Taipei Medical University, Taipei,Taiwan. Immediately after extraction, teeth were cleaned with phosphate-buffered salineand stored in cryoprotectant (BAMBANKER, Lymphotec, Tokyo, Japan). The teeth weredivided into 2 groups. Teeth in the magnetically cryopreserved group were cryopreservedin a program freezer (ABI, Chiba, Japan) supplied with a slight magnetic field. Briefly, theteeth were transported at 4�C and then placed in a freezer at –5�C. Teeth were maintainedat that temperature for 15 minutes and then cooled at a rate of –0.5�C/min until thetemperature reached –32�C. After the freezing procedure, the experimental teeth weretransferred to a freezer (MDF-11561; Sanyo, Osaka, Japan) and stored at –152�C for7 days. The time period was chosen according to a previous report that demonstratedno significant difference in the amount of revascularization between teeth stored in a toothbank for 7 days and those immediately transplanted without freezing (10). The controlgroup was composed of fresh teeth that had been extracted from the contralateral side ofeach patient. Those teeth were not subjected to the cryopreservation procedure. Biologictests performed on the control teeth were done immediately after extraction and cleaning.



Figure 1. (A) and (B) are inverted microscope images of pulp cells from fresh and cryopreserved teeth, respectively. Cells with fibroblast-like morphology can befound in both groups. (C) Cell growth viability was assessed by the MTT method. The growth curve shows no statistical difference between fresh and magneticallycryopreserved teeth. (D) One example of flow cytometry histogram demonstrated STRO-1 expression of DPSCs isolated from cryopreservation.


After 7 days of cryopreservation, the teeth were thawed, and dentalpulp cells were isolated with a modified enzyme digestion method (2).Briefly, minced pulp tissue was digested in an enzyme mixture of 4 mg/mL collagenase type I (Sigma, St Louis, MO) and 2 mg/mL dispase(Sigma) in a 37�C water bath. Cultures were then incubated at 37�Cin a humidified atmosphere of 95% air and 5% CO2.

To assess the success rate of culturing pulp cells after cryopreser-vation, the cell numbers on the 30th day after primary culture werecounted. In this study, we defined the average cell number countedfrom non-cryopreserved teeth on the 30th day as the threshold (6 �104 cells/mL). A successful culture was defined as one in which thecell numbers on day 30 were larger than that threshold. In this exper-iment, at least 8 samples from each experimental group were used. Allexperimental protocols were approved by the Committee on HumanResearch, Taipei Medical University. This information was also providedto the patients whose teeth were collected, and an agreement was signedby patients before the experiment.

Cell Viability and Morphology ExaminationIn the following experiments, 5–10 passages of cultured dental

pulp cells were used. Viability of DPSCs in the magnetically cryopre-served and non-cryopreserved groups was evaluated by a modified3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyl tetrazolium (MTT) (Sigma)assay. After incubation for 24, 48, 72, 96, and 120 hours, MTT workingsolution was added. The formazan salt was lysed with dimethyl


sulfoxide, and the absorbance at 570 nm/690 nm was measured(n = 4). Student t test was used to analyze the significance betweenthe 2 groups. The level of significance was set at .05. Morphologicchanges in the cultured cells were assessed when the cells wereincubated for 3 days by using an inverted microscope. For each sample,9 random fields within a sample were examined. To confirm the DPSCpopulation in the cultured dental pulp cells, STRO-1 marker was exam-ined by flow cytometry. Cultured pulp cells of both cryopreserved andnon-cryopreserved groups were stained with fluorescein isothiocyanate(FITC)–conjugated antibody against STRO-1 (sc-47733; Santa CruzBiotechnology Inc, Santa Cruz, CA) and were analyzed by usinga FACSCalibur instrument and CellQuest software (Becton-Dickson,Franklin Lakes, NJ).

Examination of Stem Cell–specific MarkersFor CD-34 examination, mouse anti-human CD-34 monoclonal

immunoglobulin G (sc-7324; Santa Cruz Biotechnology Inc) andFITC-conjugated goat anti-mouse immunoglobulin G (H+L; JacksonImmuno Research Laboratories Inc, West Grove, PA) were used asprimary and secondary antibodies, respectively. For CD-44 andSTRO-1 double staining, mouse monoclonal antibody against CD-44(sc-7297; Santa Cruz Biotechnology Inc) and FITC-conjugated antibodyagainst STRO-1 (sc-47733; Santa Cruz Biotechnology Inc) were used.After 48 hours of incubation, the cells of cryopreserved and non-cryopreserved groups were fixed in 4% paraformaldehyde for 20

Effects of Cryopreservation of Intact Teeth on Isolated DPSCs 1337

Figure 2. Double stains of the DPSCs isolated from fresh (A) and magnetically cryopreserved (D) teeth. Expression of CD44 (B, E) and STRO-1 (C, F) was found inthe fresh and magnetically cryopreserved teeth. Expressions of CD44 and STRO-1 were stained as red and green, respectively. (This figure is available in color onlineat www.aae.org/joe/.)


minutes and then incubated overnight with primary antibodies (1:500dilutions). The samples were subsequently incubated with secondaryantibodies for 1 hour at 37�C. Finally, the samples were examined byconfocal microscopy (TCS SP5; Leica Microsystems CMS GmbH,Mannheim, Germany).

Ability of the Multi-lineage DifferentiationThe medium of cultured cells was changed 2 times a week until

confluence was achieved. Then the medium was replaced bydifferentiation-inducing medium. The differentiation-inducing mediumwas supplemented with 0.5 mmol/L isobutylmethylxanthine, 60 mmol/Lindomethacin, 0.5 mmol/L hydrocortisone, and 10 mg mL insulin toinduce adipogenesis and was supplemented with 0.01 mmol/Ldexamethasone, 1.8 mmol/L KH2PO4 to induce osteogenesis. Thesamples were then incubated for another 28 days, with 2 mediumchanges per week. As a control, the noninduced cells were incubatedwith inducing chemical-free medium. At the end of the cultivationperiod, the cells were fixed with 4% formaldehyde and then stainedwith oil-red-O and alizarin-red-S to stain lipid droplets and calciumdeposition, respectively.

ResultsWe found that the culture rate of isolated pulp cells from fresh

teeth was 100%, and the culture rate of cells from cryopreserved teethwas 73%. Images taken with an inverted microscope were used toobserve the changes in morphology of the cultured cells after cryopres-

1338 Lee et al.

ervation. Cells from fresh and cryopreserved teeth showed fibroblast-like morphology. There were no visible differences in morphologybetween the 2 groups (Fig. 1A, B).

Fig. 1C demonstrates the growth curves of the cultured pulp cells.After culturing for 5 days, the viability of cells isolated from magneticallycryopreserved teeth had increased by 3.15 times, and that of cells fromfresh teeth had increased by 3.31 times. There was no significant differ-ence in cell viability between the 2 groups. Flow cytometry tests showedthat the fluorescent intensity of STRO-1 stained dental pulp cellscultured from fresh and cryopreserved teeth is slightly larger thancontrol. Quantified analysis indicated that the population of DPSCs inthe cultured pulp cells was 9% (Fig. 1D).

Immunostaining of CD34, CD44, and STRO-1 was performed toidentify the dental pulp stem cells. Cells from both groups showed nega-tive expression for CD34 (data not shown) but positive expression forCD44 and STRO-1 (Fig. 2). To test the multi-differentiation ability of thecryopreserved DPSCs, adipogenic and osteogenic differentiations weretested. As shown in Fig. 3, DPSCs from both groups were able to differ-entiate into adipocytes and osteocytes. Adipogenesis was confirmed bythe presence of fat droplets (Fig. 3C, D), and osteogenesis wasconfirmed by the presence of calcium deposition (Fig. 3G, H). Neitheradipogenesis (Fig. 3A, B) nor osteogenesis (Fig. 3E, F) was found in thenoninduced cells.

DiscussionThe aim of this study was to evaluate whether DPSCs could be

preserved and then isolated from teeth that had been subjected to



Figure 3. DPSCs from fresh (C, G) teeth and those from magnetically cryopreserved (D, H) teeth show the ability to differentiate into adipocytes and osteocytes.Adipogenesis (C, D) was confirmed by the presence of fat droplets (black arrow), and osteogenesis (G, H) was confirmed by the presence of calcium deposition(red color). (This figure is available in color online at www.aae.org/joe/.)


a cryopreservation process. Our results showed that the rate of cellculture from cryopreserved teeth was 73%. In addition, the cellsisolated from cryopreserved teeth not only maintained their growthpotential but also demonstrated a high efficiency in osteogenic andadipodenic differentiations (Fig. 3D, H).


In previous studies, the morphology of DPSCs was described asbeing similar to that of fibroblast-like cells (2) or bone marrow stemcells with spindle shape (12). In this study, the morphology datashowed similar variety of cells in both groups (Fig. 1A, B). Magneticcryopreservation in this study had no effect on the morphology of the

Effects of Cryopreservation of Intact Teeth on Isolated DPSCs 1339


isolated cells. Postnatal adult stem cells were reported to have greattherapeutic potential because of their self-renewal and their potentialto differentiate into multiple cell lineages (3), including odontoblasts(13), adipocytes, chondrocytes (14), osteocytes (3), and neuron-like cells (15). However, none of the cells from either group ex-pressed the neuron cell–specific markers GAP43 and CRMP-2(data not shown).
According to Shi et al (16), DPSCs do not express the hematopoi-etic stem cell marker CD34 but do express CD44 and STRO-1. We foundsimilar findings in our study, in which DPSCs isolated from magneticallycryopreserved and those from fresh teeth were negative for CD34 butpositive for CD44 (Fig. 2B, E) and STRO-1 (Fig. 2C, F). Those datasuggest that the surface markers of DPSCs are not influenced by thecryopreservation procedures used in this study.

The results of this study indicate that dental pulp stem cells isolatedfrom cryopreserved teeth maintain their growth potential, surfacemarkers, and, most importantly, their osteogenic and adipogenic differ-entiation ability. Although a previous report indicated that teeth can re-vascularize after autotransplantation only when the original tissue isremoved at the moment of extraction (10), we found that DPSCs canbe isolated from a preserved state after thawing. These results couldbe a useful reference for expanding the applications of tooth bankingfrom cryopreservation for autotransplantation to storage of DPSCs.

AcknowledgmentsThe authors would like to thank ABI Ltd (Chiba, Japan) and Dr

Geroge T.-J Huang (Boston University Henry M. Goldman School ofDental Medicine) for freezing and DPSC isolation techniquesupport, respectively.

References1. Jo YY, Lee HJ, Kook SY, et al. Isolation and characterization of postnatal stem cells

from human dental tissues. Tissue Eng 2007;13:767–73.

1340 Lee et al.

2. Gronthos S, Mankani M, Brahim J, Robey PG, Shi S. Postnatal human dental pulpstem cells (DPSCs) in vitro and in vivo. Proc Natl Acad Sci U S A 2000;97:13625–30.

3. Huang AH, Snyder BR, Cheng PH, Chan AW. Putative dental pulp-derived stem/stromal cells promote proliferation and differentiation of endogenous neural cellsin the hippocampus of mice. Stem Cells 2008;26:2654–63.

4. Pierdomenico L, Bonsi L, Calvitti M, et al. Multipotent mesenchymal stem cells withimmunosuppressive activity can be easily isolated from dental pulp. Transplantation2005;80:836–42.

5. Gebhardt M, Murray PE, Namerow KN, Kuttler S, Garcia-Godoy F. Cell survival withinpulp and periodontal constructs. J Endod 2009;35:63–6.

6. Murray PE, Garcia-Godoy F, Hargreaves KM. Regenerative endodontics: a review ofcurrent status and a call for action. J Endod 2007;33:377–90.

7. Zhang W, Walboomers XF, Shi S, Fan M, Jansen JA. Multilineage differentiationpotential of stem cells derived from human dental pulp after cryopreservation.Tissue Eng 2006;12:2813–23.

8. Sumida S. Transfusion and transplantation of cryopreserved cells and tissues.Cell Tissue Bank 2006;7:265–305.

9. Andreasen JO, Paulsen HU, Yu Z, Bayer T, Schwartz O. A long-term study of 370autotransplanted premolars: part II–tooth survival and pulp healing subsequentto transplantation. Eur J Orthod 1990;12:14–24.

10. Laureys W, Beele H, Cornelissen R, Dermaut L. Revascularization after cryopreser-vation and autotransplantation of immature and mature apicoectomized teeth.Am J Orthod Dentofacial Orthop 2001;119:346–52.

11. Woods EJ, Perry BC, Hockema JJ, Larson L, Zhou D, Goebel WS. Optimized cryopres-ervation method for human dental pulp-derived stem cells and their tissues of originfor banking and clinical use. Cryobiology 2009;59:150–7.

12. Huang AH, Chen YK, Lin LM, Shieh TY, Chan AW. Isolation and characterization ofdental pulp stem cells from a supernumerary tooth. J Oral Pathol Med 2008;37:571–4.

13. Couble ML, Farges JC, Bleicher F, Perrat-Mabillon B, Boudeulle M, Magloire H.Odontoblast differentiation of human dental pulp cells in explant cultures. CalcifTissue Int 2000;66:129–38.

14. Kawazoe Y, Katoh S, Onodera Y, Kohgo T, Shindoh M, Shiba T. Activation of the FGFsignaling pathway and subsequent induction of mesenchymal stem cell differentia-tion by inorganic polyphosphate. Int J Biol Sci 2008;4:37–47.

15. Iohara K, Zheng L, Ito M, Tomokiyo A, Matsushita K, Nakashima M. Side populationcells isolated from porcine dental pulp tissue with self-renewal and multipotency fordentinogenesis, chondrogenesis, adipogenesis, and neurogenesis. Stem Cells 2006;24:2493–503.

16. Shi S, Bartold PM, Miura M, Seo BM, Robey PG, Gronthos S. The efficacy of mesen-chymal stem cells to regenerate and repair dental structures. Orthod Craniofac Res2005;8:191–9.



Root Canal Morphology of Permanent Three-rootedMandibular First Molars: Part II—Measurement of RootCanal CurvaturesYongchun Gu, DDS, MS, Qun Lu, DDS, MS, PhD,Ping Wang, DDS, MS, PhD, and Longxing Ni, DDS, MS, PhD

Abstract
Introduction: The distolingual (DL) roots of three-rooted mandibular molars often challenge cliniciansduring root canal therapy. This study investigated canalcurvatures in permanent three-rooted mandibular firstmolars by using micro–computed tomography (micro-CT) scans. Methods: Twenty three-rooted (group 1)and twenty-five two-rooted mandibular first molars(group 2) were scanned by micro-CT. The specimenswere reconstructed 3-dimensionally by the softwareMimics 10.01 and shown in a parallel projectionmode. The images of the root canals in clinical view(CV) and proximal view (PV) were analyzed by the soft-ware Image-Pro Plus. Schneider method and a modifiedPruett method were used to measure the angles andradius of canal curvatures. Results: In the three-rooted molar group in a CV, the average angles ofprimary curvatures were 24.34 degrees for the mesio-buccal, 22.39 degrees for the mesiolingual, 13.71degrees for the distobuccal (DB), and 13.81 degreesfor the DL canal. In a PV, the average angles were16.60 degrees for the DB and 36.06 degrees for theDL canal, respectively. Secondary curvatures werefrequently seen in a CV (60%) for the DB canals, withan average angle of 26.94 degrees. In a PV, the averagecentral angle of curvature was 59.04 degrees for the DLcanal, and the average radius and curve length were6.17 and 5.73 mm, respectively. In general, no statisti-cally significant difference was found for canal curva-tures in the mesial roots between the three-rootedand two-rooted molar groups (P > .05). Conclusions:A better understanding of the canal curvatures is essen-tial for successful endodontic treatment of three-rootedmandibular first molars. (J Endod 2010;36:1341–1346)
Key WordsDistolingual root, micro-computed tomography scan,permanent three-rooted mandibular first molar, rootcanal curvature

From the Department of Operative Dentistry & Endodontics, SchAddress requests for reprints to Professor Longxing Ni, Departme

169 Changlexi Road, Xi0an 710032, China. E-mail address: nilongxi0099-2399/$0 - see front matter



The presence of the distolingual (DL) root in permanent mandibular first molars wasfirst mentioned in the literature by Carabelli in 1844 (1) and later named as radix

entomolaris (2). Its frequency is lower than 5% in white and African populations (3–6).In Asian populations, the prevalence of three-rooted mandibular first molars lies in therange of 5%–40%, and it has been considered as a racial characteristic (7–10).Recently, this anatomical variation was detected in a Chinese population in Taiwanby using cone beam computed tomography (11). They reported a prevalence of33.33% (41 of 123 individuals), with a bilateral incidence of a symmetrical distributionof 56.65%. This root variation presents challenges for the clinician. Failure to locate andtreat the canal in the third root can result in treatment failure. It is recommended thatpractitioners modify the traditional triangular opening to a trapezoidal opening toimprove identifying and accessing the DL canal (12).

Another challenge related to three-rooted mandibular first molars is the root canalcurvature. Conventional canal instrumentation of a curved canal with stiff steel filesmight produce ledges, zips, elbows, apical transportation, loss of working length, orperforations (13). Nickel-titanium rotary system can reduce the occurrence of theseerrors, because it is superelastic and more flexible in the canal curvature. However,it might undergo unexpected fracture as a result of cyclic fatigue (14). Therefore, accu-rate assessment of canal curvatures in three-rooted mandibular first molars has greatclinical significance. The information would be valuable in formulating a treatment planand determining prognosis.

There are many techniques to evaluate the canal curvature. The first and mostcommon method was reported by Schneider (15) in 1971. The degree of canal curva-ture was defined as the acute angle between the long axis of the canal and a line from thepoint of initial curvature to the apical foramen. In 1982, Weine (16) proposed anothermethod that defined the angle of curvature differently. The acute angle between linespassing through the apical and coronal portions was measured. Pruett et al (14)pointed out that the shape of any root canal curvature could be more accuratelydescribed by using 2 parameters, angle of curvature and radius of curvature. The radiuswas defined as the one from a circle that coincides with the path taken by the area of themost abrupt curvature, representing abruptness of curvature. Cunningham and Senia(17) studied canal curvature in mesial roots of mandibular first and second molarsby a radiographic technique. They pointed out that the root canal curvature was 3-dimensional, and curvature in proximal view (PV) could not be predicted by examininga clinical view (CV) radiograph. The secondary curvatures were more frequently seen inPV than in CV (30% versus 2.5%).

In the literature, the DL root of three-rooted mandibular first molar is typicallydescribed as a conical and small root whose apex swings toward buccal (18). De

ool of Stomatology, Fourth Military Medical University, Xi0an, China.nt of Operative Dentistry & Endodontics, School of Stomatology, Fourth Military Medical University,[email protected].

Measurement of Root Canal Curvatures in Three-rooted Mandibular First Molars 1341


Moor et al (12) classified this DL root into 3 types according to the canalcurvature: type I refers to a straight root, type II to an initially curvedentrance and then continuation as a straight root, and type III to curva-ture in the coronal third and buccal curvature from the middle third orapical third of the root. Chen et al (19) recorded the location of thecurvature as the coronal one third, the middle one third, or the apicalone third according to the location of the point of initial curvature.Nevertheless, few studies have shown quantitative measurements oncanal curvatures in three-rooted mandibular first molars.
In recent years, micro–computed tomography (micro-CT) hasbeen used to evaluate the root canal anatomy because of its high reso-lution and nondestruction of specimens (20–22). The 2-dimensional(2D) cross-sectional images and 3D models generated by micro-CTcould precisely demonstrate the complex root canal system in differentviews and magnifications. Internal and external anatomy of the teeth canbe visualized simultaneously or separately (23). At this time there areseveral analytical programs available to evaluate these mathematicalmodels efficiently and accurately. However, to date, investigation ofpermanent three-rooted mandibular first molars by this methodologyhas not been reported.

The purpose of this study was to evaluate the canal curvature in the3 roots of three-rooted mandibular first molars, focusing on the DLroot. A micro-CT system was used to reconstruct 3D models of the teethand root canals. The canal curvatures in both views (CV and PV) weremeasured by Schneider method and a modified Pruett method.

Materials and MethodsA total of 122 permanent mandibular first molars were collected

from a native Chinese population, and any attached soft tissue andcalculus were removed before the experiment. From all specimens,a DL root was present in 39 teeth (31.93%). After excluding teeth

Figure 1. Measurements of angles of canal curvatures by Schneider method. a is(This figure is available in color online at www.aae.org/joe/.)

1342 Gu et al.

with fractured roots or other major defects, 20 three-rooted (group1) and 25 two-rooted mandibular first molars (group 2) with intactroots were selected for investigation by using a micro-CT scanner(eXplore Locus SP; GE HealthCare, London, Ontario, Canada). Thechosen teeth comprised 8 left and 12 right three-rooted first molarsand 13 left and 12 right two-rooted first molars. Each specimen wasscanned along the teeth axis with voxel sizes of 21 � 21 � 21 mm.The data sets (DICOM format) were transferred to Mimics 10.01 (Mate-rialise, Leuven, Belgium) software. The anatomy of the teeth and rootcanal system was reconstructed 3-dimensionally with a semiautomaticthreshold-based segmentation approach combined with manual editingof the slices. By adjusting the transparency of the objects, the root canalsystems were made opaque, and the teeth were made transparent.

At the interface of Mimics, 3D objects are displayed in a parallelprojection mode. Therefore, the principle of imaging is the same asthat of radiographic approach. The models of the teeth and root canalsystems were viewed in CV and PV, and the screen shots were saved inBMP format. They were then analyzed in software Image-Pro Plus 6.0(Media Cybernetics, Silver Spring, MD), with a resolution of 1280 �770 pixels. The images were calibrated according to the teeth lengths(the length along the axis of the tooth), which were previously measuredin Mimics. Then the canal curvatures were measured for both views byusing the technique described by Schneider (15) (Fig. 1). Point a wasmarked at the center of the canal orifice. A line was drawn with a straightedge aligned parallel to axis from point a to a point where the long axisdeviated from the straight edge, point b. A third point (point c) wasmade at the apical foramen, and a line was drawn from this point topoint b. The acute angle formed by the intersection of the 2 lines ismeasured to evaluate the canal curvature. The canal curvatures wereseparated into 3 groups on the basis of the angles: straight (10 degreesor less), moderate (10–20 degrees), and severe (20 degrees or more).

angle of primary curvature; b is angle of secondary curvature. (A) CV; (B) PV.



If the root canal contained more than 1 curve, the primary and
secondary curvatures were measured as described by Cunninghamand Senia (17) (Fig. 1).

The canal curvature in the DL root was further evaluated by modi-fying the method of Pruett et al (14) (Fig. 2). A straight line was drawnalong the long axis of the coronal portion of the canal. A second line wasdrawn along the long axis of the apical portion of the canal. There isa point on each of these lines at which the canal deviates from the begin-ning (point b) or the end (point c) of the canal curvature. The curvedportion between point b and c is represented by a section of a circle. Theradius (r) and central angle (q) of this arc are the 2 parameters that thePruett method suggests to measure (Fig. 2).

According to the Pruett method, angle of curvature is defined as theangle formed by perpendicular lines drawn from the points of deviation(points a and b). They intersect at the center of the circle. The length ofthese lines is the radius. In this study, by using the software Image-ProPlus, we were able to measure these parameters more accurately andconveniently. The arc could be defined by 3 mathematical points: pointsb, c, and the midpoint of the curved portion. Choosing 2 terminal pointsand a midpoint might guarantee the best coincidence between thedefined circle and the canal curvature. The values of the arc length(AL), q, and r were obtained directly. The lengths of the straight portions(coronal line section L1 and apical line section L2) were also measured.

All the parameters were measured 3 times, and average valueswere used for data analysis. Student t test was used to compare themeans between groups. A paired t test was used to compare the meansbetween groups. The level of statistical significance was set at P <.05.

ResultsThe degrees of primary and secondary canal curvatures measured

by Schneider method in both views (CV and PV) are summarized inTable 1 and Fig. 3.

Figure 2. Measurements of radius, central angles, and arc lengths of canal curvatustraight portion; q is central angle; qw is angle of Weine method; r is radius. (A) L1org/joe/.)

JOE — Volume 36, Number 8, August 2010 Measu

In CV, the images of the mesiolingual (ML) canals werefrequently overlapped by the mesiobuccal (MB) canal, whereas only2 DL canals in three-rooted first molar were totally overlapped by thedistobuccal (DB) canal. The mean degrees of curvature in the MBand ML canals were very close (24.34 versus 22.39 degrees in thethree-rooted group), although the difference had statistical significanceaccording to a paired t test (P = .030). Secondary curvature was rare inthe mesial root. Only 1 two-rooted first molar exhibited an S-shaped MBcanal, whereas it was frequently seen in the DB roots of three-rootedand distal roots of two-rooted first molars. The frequency was on60% (12 of 20) of the DB canals of the three-rooted first molars,and the location was within 2 mm of the root apex. The mean angleof the second curvature was approximately twice that of the primaryone (26.94 versus 13.71 degrees in three-rooted group, P = .000;30.81 versus 12.54 degrees in two-rooted group, P = .000).

In PV, the DL canal in the three-rooted first molars exhibited thegreatest degrees of curvature among the 3 roots. The angle of curvature(by Schneider method) ranged from 14–57 degrees, with a mean of32.06 degrees. Some two-rooted first molars (7 of 25) also had DLcanals. However, the mean angle was 18.48 degrees, significantly lowerthan that of the three-rooted first molars (P < .01). Five of 18 (27.79%)MB or ML canals in the three-rooted molars and 8 of 23 (34.78%) MBor ML canals in the two-rooted molars had a secondary curvature,whereas neither DB nor DL canals in the three-rooted first molars ex-hibited this curvature. Only 2 DB canals and 4 DL canals in the two-rooted first molars exhibited a secondary curvature.

Table 2 presents measurement results of curvatures in the DLcanals of the three-rooted first molars by using the modified Pruettmethod. The mean central angle was 59.04 degrees and the meanradius was 6.17 mm in PV versus 26.17 degrees (central angle) and20.99 mm (radius) in CV. The mean arc length was 5.73 and 6.13mm in PV and CV, respectively. The lengths of L1 and L2 were frequently

res by modified Pruett method (PV). L1 is coronal straight portion; L2 is apicals 0, L2 s 0; (B) L2 = 0. (This figure is available in color online at www.aae.

rement of Root Canal Curvatures in Three-rooted Mandibular First Molars 1343



TABLE 1. Angle of canal curvature in CVs and PVs (degrees) (by Schneider method)

Three-rooted Two-rooted t test

Canal n Mean SD Range n Mean SD Range t P value

Clinical viewMB 18 24.34 7.20 12.61–37.94 23 24.37 5.57 12.68–35.79 0.015 .988ML 18 22.39 8.06 6.86–37.94 23 21.87 7.04 7.28–37.30 0.222 .826DB + D (P) 20 13.71 5.34 4.20–24.10 25 12.54 6.55 0–30.52 0.643 .524DB + D (S) 12 26.94 8.55 15.18–42.18 15 30.81 10.08 13.35–46.92 1.06 .299DL (P) 20 13.81 9.58 0–35.72 7 11.76 8.40 0–26.43 0.502 .620DL (S) 4 25.32 8.33 16.19–34.70 5 29.12 10.87 15.22–42.84 0.575 .583

Proximal viewMB (P) 18 17.50 11.46 0–44.14 23 19.51 9.01 8.38–42.94 0.629 .533MB (S) 5 26.87 7.81 16.97–37.86 8 17.04 6.35 8.29–28.46 2.494 .030*ML (P) 18 19.22 7.89 10.46–41.07 23 20.87 8.79 0–33.86 0.622 .537ML (S) 5 21.20 5.73 12.02–27.47 8 20.23 3.10 13.32–23.78 0.399 .697DB 20 16.60 9.92 0–42.20 7 7.60 10.82 0–30.45 2.019 .054DL 20 32.06 11.20 14.00–57.00 7 18.48 5.17 10.05–22.14 3.065 .005†

Total of 18 two-rooted first molars have a single canal in the distal; 2 three-rooted first molars and 2 two-rooted first molars have a single canal in the mesial root.

CV, clinical view; D, single distal canal in the distal root; DB, distobuccal; DL, distolingual; MB, mesiobuccal; ML, mesiolingual; P, primary curvature; PV, proximal view; S, secondary curvature; SD, standard deviation.

*P < .05.†P < .01.


below 2 mm and shorter than the arc length. In a few three-rooted firstmolars, the length of L1 and/or L2 in the DL canals was zero (Fig. 2B).

To verify the method of canal curvature determination, the anglesand radius of canal curvatures in both views of the 20 three-rooted firstmolars were measured by an independent investigator. The results werefound to be within�1.5 degrees for determination of the angle and�1mm for measurement of the length and radius of the canal curvature.

Figure 3. Angles of curvatures in roots of three-rooted first molars (bySchneider method). P is primary curvature, and S is secondary curvature.

1344 Gu et al.

DiscussionSuccessful root canal instrumentation requires considerable

knowledge of the canal anatomy. The information missed in a CV radio-graph could be seen in a PV radiograph (17, 18). By using the softwareMimics 10.01, we generated 3D mathematical models of the teeth basedon the micro-CT data sets. The isotropic resolution reached as high as21 mm, and the 3D models could be displayed on the screen witha perfect parallel projection. In the parallel projection mode, the depthcoordinate is ignored, and objects are assigned to the screen space ac-cording to their actual geometrical size, regardless of their distance tothe viewpoint (24). By modifying the transparency of the models, rootcanal systems can be exhibited completely, and the external contours ofthe teeth can be kept as an important point of reference. This guaran-teed better accuracy of the measurements and better image quality thanconventional radiographs, while the physical principle of imaging is thesame. The 3D modeling of root canal systems also could be a valuableteaching tool for dental students and clinicians. Most of the studiesbased on radiographic examination have used files to determine theroot canal axis, but here we did not because the files could not remaincentered in the canal, and this would introduce measurement errors.Besides this, for certain specimens with more complex root canalsystems (eg, tiny canal with ramification, severe apical secondary curva-ture, or partial calcification), attempting to insert a file to the canallength could result in a failure.

The results of the present study indicate that the canal curvaturesare distributed differently in the 3 roots of three-rooted mandibular firstmolars as shown in the CVs and PVs.

In CV, the ML and DL canals might be totally or partially overlappedby the corresponding buccal canals. Therefore, careful reading ofmulti-angled radiographs is recommended to improve the accuracyof clinical diagnosis (25). The average angles of curvature in the MBand ML canals agree with the results reported by Schafer et al (26).The latter study found that the mean was 25 degrees for MB canalsand 22 degrees for ML canals. These values were also close to thoseof Cunningham and Senia (17). They reported a mean angle of 28.7and 27.2 degrees for MB and ML canals, respectively, although thepermanent mandibular first and second molars were combined in theirsample. The curvature in the mesial root starts immediately after itleaves the canal orifice and initially progresses mesially and then distallyafter curve. In the three-rooted group, 14 of 18 MB canals and 10 of 18


TABLE 2. Measurement results of curvature in DL root by modified Pruett method (n = 20)

PV CV Paired t test

Parameters Mean SD Range Mean SD Range t P value

r (mm) 6.17 2.51 2.54–11.54 20.99 19.35 2.78–68.21 3.422 .003*AL (mm) 5.73 1.25 3.81–8.49 6.13 2.22 2.58–10.48 0.678 .506L1 (mm) 2.09 1.07 0–3.87 1.20 2.06 0–5.98 1.657 .114L2 (mm) 0.84 1.10 0–3.53 0.40 0.90 0–3.17 1.276 .217q (degrees) 59.04 19.05 31.10–97.97 26.17 13.30 6.72–53.13 6.606 .000*

AL, arc length; CV, clinical view; DL, distolingual; L1, coronal line section; L2, apical line section; PV, proximal view; SD, standard deviation.

Four DL roots exhibit a second curvature in CV. In these cases, L2 is the distance from ending point of first curve to starting point of second curve.

*P < .01.


ML canals exhibit a severe curve according to the classification bySchneider (15) (Fig. 3). This anatomical feature is closely related tooccurrence of strip perforation during canal instrumentation. Adequatecoronal flaring is advocated to reduce the magnitude of curvature and toachieve a straight access to the apical portion of the canal (17). Pres-ervation of the distal wall of the coronal canal portion is critical to avoidthe possible outcome of strip perforation or vertical fracture of themesial root (13, 17).

In PV, 27.79% of MB or ML canals in the three-rooted molars and34.78% of MB or ML canals in the two-rooted molars have a secondarycurvature. No significant difference was found between average anglesof primary and secondary curvature (P > .05). In the mesial root,confluence and divergence of the MB and ML canals can form severecanal curvatures, and the chance of exhibiting a second canal curvaturemight increase. The present study demonstrates that in all but onemesial root parameter, no statistically significant difference was foundwhen comparing the three-rooted and two-rooted first molar groups (P> .05). Only the angle of the secondary curvature of MB canal in PV wasan exception (P = .03). Therefore, if only canal curvature is considered,there is little difference in instrumentation of mesial roots between two-rooted and three-rooted mandibular first molars.

The present study showed that 12 of 20 DB canals in three-rootedfirst molars exhibit secondary curvature in a CV. Fifteen of 20 primarycurvatures belong to moderate type, whereas 9 of 12 secondary curva-tures belong to severe type (Fig. 3). The latter curve distally and arelocated at 0–2 mm from the root apex. Such apical curvatures mightgreatly increase the risk of instrument separation or procedural error.In PV, only 5 of 20 DB canals in three-rooted first molars belonged tostraight type, and secondary curvature was not observed in this study.Therefore, for DB canals, attention should be focused on the secondarycurvature in a CV.

The curvature of the DL canal is mostly present in a PV. Fig. 3shows in a CV most of the DL canals are straight (9 of 20) or havemoderate curvatures (7 of 20), whereas in PV all but one (19 of 20)DL canals exhibit severe curvature, and the mean angle of curvatureis the greatest among the 3 roots. These findings were confirmed byChen et al (19). They analyzed radiographs of 21 extractedpermanent three-rooted mandibular first molars in a Taiwanese popu-lation and reported that the mean angle of the DL root was 36.35degrees in PV versus 9.24 degrees in CV (by Schneider technique).However, Chen et al did not report the radius of the DL canal.

To measure the radius of canal curvature, the most curvedportion of the canal is always hypothesized to be an arc (14, 26).Pruett et al (14) pointed out that the angle and radius of canal curva-ture were independent of each other. Canals can have the same angleof curvature while they have different radii. This point of view can bestbe understood through geometry. In geometry, a length of arc ofa circle is proportional to its radius and the corresponding centralangle, and any 2 of above 3 parameters can define an arc and deter-

JOE — Volume 36, Number 8, August 2010 Measu

mine the remaining third parameter; 3 mathematical points can alsodefine an arc or a circle. Our measurement of curvature used thesegeometrical principles.

The present study showed that the central angle (q) is alwayslarger than the angle of Schneider method (a). For the DL canals inPV, the mean value of q was 59.68 degrees, whereas the mean valueof a was as low as 32.05 degrees. Fig. 2A shows the angle of Schneidermethod (a) where the angle is formed by the tangent line L1 andstraight line bd. Therefore, the length of L2 might influence the valueof a. Unlike a, q is independent value of L2 or L1. It only reflects thegeometrical feature of the curved portion. If L2 increases, a willincrease accordingly, while q will not change. If L2 = 0, anglea becomes the tangent chord angle (the angle formed by a tangent toa circle and a chord). In geometry, theorem about tangent chord anglestates that an angle formed by a chord and a tangent is equal to the in-scribed angle or half the corresponding central angle subtended by thechord. Consequently, equation q = 2a can be derived. In this study, themean value of L2 was as low as 0.84 mm, which was much lower thanthat of AL (5.73 mm) and r (6.17 mm). Consequently, the mean value ofq was approximately twice that of a (59.68 versus 32.05 degrees).

Pruett et al (14) reported that for nickel-titanium, engine-drivenrotary instruments, the radius of the canal curvature, the angle of curva-ture, and instrument size were important for predicting separation.However, few clinicians have realized the important role of the curva-ture length (AL) in explaining the mechanism of instrument fatigue.The microstructure and surface analysis of unused nickel-titaniumrotary instruments demonstrated that there were distortions in thelattice structure of the alloy, machining and milling marks, as well asmetal strips and microcracks on their surfaces (27). If the 2 curvatureshave the same radius but different central angles, the curvature witha larger angle will have a longer AL. This means a larger portion ofthe instrument in the canal is under stress, and more microflaws andmicrocracks in the instrument were under repeated or cyclic load. Ifthe stress level exceeds certain threshold, which is known as fatiguelimit or endurance limit, the probability of occurrence of instrumentseparation would increase accordingly. Therefore, we believe that ALis a more direct factor than q in explaining the cyclic fatigue of rotaryinstruments, and they have equal geometrical and clinical significance.The angle of Schneider method is not a pure parameter in describing thecurved portion. It is, however, used by most investigators because themeasurement of this parameter is convenient and simple.

This study showed that the curvature in the DL canals of the three-rooted first molars has a more severe angle and smaller radius in the PV,and the length of the curved portion is relatively long. The curvaturemight distribute over a large or entire portion of the canal. Theseanatomical variations imply that instrument separation can easily occurat any levels during canal preparation. A CV radiograph will miss consid-erable information about curvature in buccolingual dimension. Angledradiographic technique has been advocated to increase the accuracy of

rement of Root Canal Curvatures in Three-rooted Mandibular First Molars 1345

clinical diagnosis (18, 28). Clinicians should practice with more careand caution. To gain access to the apical portion, coronal flaring isrecommended to decrease the angle of curvature. Avoidinginstrument fatigue failure and overinstrumentation is extremelyimportant for successful root canal instrumentation of three-rooted firstmolars.
References1. Carabelli G. Systematisches Handbuch der Zahnheilkunde. Wien: Braumuller und

Seidel; 1844. 114.2. Bolk L. Bemerkungen uber Wurzelvariationen am menschlichen unteren Molaren.

Zeiting fur Morphologie und Anthropologie 1915;17:605–10.3. Curzon ME. Three-rooted mandibular permanent molars in English Caucasians. J

Dent Res 1973;52:181.4. Schafer E, Breuer D, Janzen S. The prevalence of three-rooted mandibular perma-

nent first molars in a German population. J Endod 2009;35:202–5.5. Sperber GH, Moreau JL. Study of the number of roots and canals in Senegalese first

permanent mandibular molars. Int Endod J 1998;31:117–22.6. Younes SA, al-Shammery AR, el-Angbawi MF. Three-rooted permanent mandibular

first molars of Asian and black groups in the Middle East. Oral Surg Oral Med OralPathol 1990;69:102–5.

7. Gulabivala K, Aung TH, Alavi A, Ng YL. Root and canal morphology of Burmesemandibular molars. Int Endod J 2001;34:359–70.


9. Turner CG 2nd. Three-rooted mandibular first permanent molars and the questionof American Indian origins. Am J Phys Anthropol 1971;34:229–41.

10. Curzon ME, Curzon JA. Three-rooted mandibular molars in the Keewatin Eskimo.J Can Dent Assoc (Tor) 1971;37:71–2.

11. Tu MG, Huang HL, Hsue SS, et al. Detection of permanent three-rooted mandibularfirst molars by cone-beam computed tomography imaging in Taiwanese individuals.J Endod 2009;35:503–7.

12. De Moor RJ, Deroose CA, Calberson FL. The radix entomolaris in mandibular firstmolars: an endodontic challenge. Int Endod J 2004;37:789–99.

1346 Gu et al.

13. Peters OA. Current challenges and concepts in the preparation of root canal systems:a review. J Endod 2004;30:559–67.

14. Pruett JP, Clement DJ, Carnes DL Jr. Cyclic fatigue testing of nickel-titaniumendodontic instruments. J Endod 1997;23:77–85.

15. Schneider SW. A comparison of canal preparations in straight and curved rootcanals. Oral Surg Oral Med Oral Pathol 1971;32:271–5.

16. Weine F. Endodontic therapy. 3rd ed. St Louis, MO: CV Mosby; 1982. 256–340.17. Cunningham CJ, Senia ES. A three-dimensional study of canal curvatures in the

mesial roots of mandibular molars. J Endod 1992;18:294–300.18. Jerome CE, Hanlon RJ Jr. Dental anatomical anomalies in Asians and Pacific

Islanders. J Calif Dent Assoc 2007;35:631–6.19. Chen YC, Lee YY, Pai SF, Yang SF. The morphologic characteristics of the distolingual

roots of mandibular first molars in a Taiwanese population. J Endod 2009;35:643–5.

20. Dowker SE, Davis GR, Elliott JC. X-ray microtomography: nondestructive three-dimensional imaging for in vitro endodontic studies. Oral Surg Oral Med Oral PatholOral Radiol Endod 1997;83:510–6.

21. Nielsen RB, Alyassin AM, Peters DD, Carnes DL, Lancaster J. Microcomputed tomog-raphy: an advanced system for detailed endodontic research. J Endod 1995;21:561–8.

22. Yu DC, Tam A, Schilder H. Root canal anatomy illustrated by microcomputed tomog-raphy and clinical cases. Gen Dent 2006;54:331–5.

23. Plotino G, Grande NM, Pecci R, Bedini R, Pameijer CH, Somma F. Three-dimensionalimaging using microcomputed tomography for studying tooth macromorphology.J Am Dent Assoc 2006;137:1555–61.

24. Foley JD, van Dam A, Feiner SK, Hughes JF. Computer graphics: principles and prac-tice. 2nd ed. Reading, MA: Addison-Wesley; 1995. 231–6.

25. Slowey RR. Root canal anatomy: road map to successful endodontics. Dent ClinNorth Am 1979;23:555–73.

26. Schafer E, Diez C, Hoppe W, Tepel J. Roentgenographic investigation of frequencyand degree of canal curvatures in human permanent teeth. J Endod 2002;28:211–6.

27. Alapati SB, Brantley WA, Svec TA, Powers JM, Mitchell JC. Scanning electron micro-scope observations of new and used nickel-titanium rotary files. J Endod 2003;29:667–9.

28. Brynolf I. Roentgenologic periapical diagnosis: II—one, two or more roentgeno-grams? Sven Tandlak Tidskr 1970;63:345–50.



Host–Mineral Trioxide Aggregate Inflammatory MolecularSignaling and Biomineralization AbilityJessie F. Reyes-Carmona, DDS, MS, PhD,*† Adair S. Santos, PhD,‡ Claudia P. Figueiredo, PhD,‡

Cristiane H. Baggio, PhD,‡

Mara C. S. Felippe, DDS, MS, PhD,* Wilson T. Felippe, DDS, MS, PhD,*

and Mabel M. Cordeiro, DDS, PhD§

Abstract
Introduction: The biological processes underlying theability of mineral trioxide aggregate (MTA) to promotehard-tissue deposition and wound healing remainunclear. To further study these processes, specificsignaling molecules related to the inflammatoryresponse and the biomineralization process wereanalyzed to assess host-MTA interactions in vivo.Methods: For cytokine level quantification and immu-nohistochemical analysis, human dentin tubes werefilled with ProRoot MTA (Dentsply, Tulsa Dental, OK)or kept empty and were implanted in subcutaneoustissues in the backs of mice. Dentin tubes were retrievedand subsequently observed using a scanning electronmicroscope. Results: MTA induced a time-dependentproinflammatory cytokine up-regulation up to 3 days.Immunohistochemical analyses showed an up-regulated expression of myeloperoxidase, nuclearfactor-kappa B, activating protein-1, cyclooxygenase-2,inducible nitric oxide synthase, and vascular endothelialgrowth factor on day 1. Scanning electron microscopicexamination revealed the presence of apatite-like clus-ters on collagen fibrils over the surface of tubes contain-ing MTA. With the increase in time after implantation,a more extensive mineralization showing a compactlayer of apatite was observed. Conclusion: MTAinduced a proinflammatory and pro–wound healingenvironment. The biomineralization process occurredsimultaneously at the biomaterial-dentin-tissue inter-face, with the acute inflammatory response. Thispromoted the integration of the biomaterial into theenvironment. (J Endod 2010;36:1347–1353)
Key WordsApatite, bioactivity, biomineralization, inflammation,mineral trioxide aggregate, wound healing

From the *Postgraduate Dentistry Program of the Federal UniveCosta Rica, San Jose, Costa Rica; ‡Department of Physiological Scielogical Sciences, Federal University of Santa Catarina, Florianopolis,

Supported in part by Grants in Aid for Scientific Research fromAddress requests for reprints to Dr Jessie Reyes-Carmona, Depa

address: [email protected]/$0 - see front matter



A bioactive material should be capable of stimulating specific biological responsesvia biochemical and biophysical reactions that result in the formation of an apatite

layer (1, 2). The ability to induce the formation of apatite allows the integration of thebiomaterial into the environment (2). However, host responses to biomaterials aredependent on the innate and nonspecific immune responses that occur in thesurrounding tissues (3). Biomaterials may elicit an inflammatory cascade comprisingneutrophil and macrophage recruitment and adhesion, foreign body reaction, andfibrous encapsulation (4). Cytokines and growth factors secreted by inflammatory cellsare the molecular messengers that promote inflammatory events and wound healing(5).

Inflammatory cytokines such as, interleukin (IL)-1b, tumor necrosis factora (TNF-a), and prostaglandins play an important role in the development of the inflam-matory response. The expression of these proteins is controlled by some transcriptionfactors, such as activating protein-1 (AP-1) and nuclear factor-kappa B (NF-kB) (6).NF-kB consists of a group of proteins, including p50, p65, and p105, which are seques-tered in the cytoplasm in their resting state. When activated by agents such as cytokines,NF-kB undergoes phosphorylation, leading to its nuclear translocation and binding tospecific sequences of DNA, which, in turn, results in gene transcription (7). At the onsetof inflammation, a cytokine-mediated activation of NF-kB in macrophages results in ni-tric oxide production by the inducible nitric oxide synthase (iNOS) enzyme (8, 9).Experimental evidence suggests that a relationship exists between nitric oxide andprostaglandin E2 biosynthesis, the production of which is regulated bycyclooxygenase-2 (COX-2) (9, 10). Moreover, myeloperoxidase (MPO), a leukocyte-derived enzyme that catalyzes the formation of a number of reactive oxidant species,is linked to the acute phase of inflammation (11).

Vascular endothelial growth factor (VEGF) is a glycoprotein with the ability toincrease the permeability of blood vessels, an important vascular change observedduring inflammation (12). Therefore, VEGF also plays a critical role in angiogenesisand neovascularization by participating in critical biological processes, such as tissuerepair (13).

Mineral trioxide aggregate (MTA) has been extensively used as a promisingbiomaterial for stimulating dentinogenesis and cementogenesis. Despite its widespreaduse in clinical practice, the mechanism by which it induces hard-tissue depositionremains unknown. Previous studies have shown that MTA releases calcium hydroxide,which interacts with a phosphate-containing fluid to produce calcium-deficient apatitevia an amorphous calcium phosphate phase (14–16). A preliminary study providedcompelling evidence of the biomineralization process promoted by the interaction of

rsity of Santa Catarina, Florianopolis, SC, Brazil; †Department of Restorative Sciences, University ofnces, Federal University of Santa Catarina, Florianopolis, SC, Brazil; and §Department of Morpho-SC, Brazil. Dr Reyes-Carmona is a fellow of University of Costa Rica.

the University of Costa Rica and Federal University of Santa Catarina.rtment of Endodontics, School of Dentistry, University of Costa Rica, San Jose, Costa Rica. E-mail

Host-MTA Inflammatory Molecular Signaling 1347


MTA and dentin (15). The apatite formed by the MTA–phosphate-buffered saline (PBS) system was deposited among collagen fibrils.This promoted controlled mineral nucleation on dentin, which trig-gered the formation of an interfacial layer with an intratubular miner-alization at the biomaterial-dentin interface (15, 16).
Although MTA has been studied in numerous clinical and histolog-ical studies, there is little consensus regarding the mechanism involvedduring the inflammatory reaction and its correlation with the repairprocess and hard tissue formation. Therefore, in this study, we evalu-ated specific signaling molecules related to the inflammatory processand the biomineralization ability of MTA to assess host-biomaterialinteractions in vivo.

Materials and MethodsEthical Concerns

All experimental protocols used in this study were approved by theAnimal Ethics Screening Committee and the Ethics Committee forResearch with Human Beings of the Federal University of Santa Catarina,Santa Catarina, Brazil.

Preparation of SpecimensEighty dentin tubes were prepared from extracted human tooth

roots. The crowns and the apical thirds of the roots were removed usinga low-speed water-cooled ISOMET diamond saw (Buehler, Lake Bluff,NY). Each root canal was enlarged to obtain 1.3-mm-diameter stan-dardized cavities. Tube length was 5 mm, and their outer walls wereabraded with a diamond bur to thin the walls to a 2-mm thickness.The dentin tubes were washed in distilled water and then autoclaved.

Before implantation, the tubes were thoroughly irrigated with 17%EDTA followed by a 1% sodium hypochlorite solution, dried, and thenfilled with tooth-colored ProRoot MTA (Dentsply, Tulsa Dental, OK) orkept empty (negative control).

Experimental ProtocolMale Swiss mice (35-40 g) were anesthetized with ketamine

hydrochloride (Dopalen; Division Vetbrands Animal Health, Jacareı,SP, Brazil) and xylazine (Anasedan; Agribrands do Brasil Ltda, Paulınia,SP, Brazil). Four separate 1-cm incisions were made in the backs ofmice at a 1-cm interval. The skin was deflected to create a pocket bya blunt dissection on one side of each incision. Each mouse receivedthree dentin tubes: two filled with MTA and one empty; no specimenwas inserted in the fourth pocket (sham) to ensure a rotation of sites.

After 12 hours and 1, 3, and 7 days after implantation, the animalswere euthanized, and the tubes together with surrounding tissues wereremoved. Half of the samples were fixed in 4% paraformaldehyde at 4�Cfor histological and immunohistochemical staining. To determine theprotein level expression of IL-1ß, TNF-a and IL-10, the remaininghalf of the samples and surrounding tissues were excised and processedfor tissue homogenate. Dentin tubes were retrieved and processed forscanning electron microscopic (SEM) analysis.

Histological and Immunohistochemical AnalysesFor hematoxylin-eosin and immunohistochemistry staining, tissues

were embedded in paraffin, sectioned at a 3-mm thickness, andprepared on conventional glass slides. Tissue sections were deparaffi-nized, and immunohistochemistry was performed using the followingprimary antibodies and respective dilution ratios: rabbit polyclonal anti-myeloperoxidase (MPO, 1:300; Dako Cytomation, Carpinteria, CA),mouse monoclonal anti-VEGF (1:200, C-1; Santa Cruz BiotechnologyInc, Santa Cruz, CA), rabbit polyclonal anti–COX-2 (1:200; Cell Signaling

1348 Reyes-Carmona et al.

Technology, Beverly, MA), rabbit polyclonal antiphospho-p65 NF-kB(1:50, Cell Signaling Technology), rabbit monoclonal antiphospho-c-jun AP-1 (1:50, Cell Signaling Technology), and rabbit polyclonalanti-iNOS (1:100, Cell Signaling Technology). High-temperature antigenretrieval was applied by immersing the slides in a water bath at 95�C to98�C in 10 mmol/L trisodium citrate buffer (pH = 6.0) for 45 minutes.The nonspecific binding was blocked by incubating sections for 1 hourwith goat normal serum diluted with PBS. After overnight incubationwith primary antibodies at 4�C, the slides were washed with PBS andincubated with the ready-to-use secondary antibody EnVision Plus (Da-koCytomation EnVision Doublestain System, Carpinteria, CA) for 1 hourat room temperature. The sections were washed again in PBS, and visu-alization was completed using 3,30-diaminobenzidine (DAB) (DakoCy-tomation) and counterstained lightly with Harris’ hematoxylin solution.Both control and experimental samples were placed on the same glassslide and processed under the same conditions.

Images of the stained tissue sections were acquired using a digitalcamera (Canon A620, Lake Success, NY) connected to a light micro-scope (Axiostar Plus; Carl Zeiss, Oberkochen, Germany). Settings forimage acquisition were identical for both control and experimentaltissues. Four consecutive images per sample, captured at 40� magni-fication, were taken of the tissues in contact with the material on the tu-be’s opening or on an empty opening. The threshold optical density wasobtained using NIH ImageJ 1.36b imaging software (National Institutesof Health, Bethesda, MD). The total pixel intensity was determined, anddata were expressed as optical density.

Determination of Cytokine LevelsBriefly, full-thickness tissue samples were homogenized in phos-

phate buffer containing 0.05% Tween 20, 0.1 mmol/L phenylmethylsul-phonyl fluoride, 0.1 mmol/L benzethonium chloride, 10 mmol/L EDTA,and 20 KIU aprotinin A. The levels of IL-1b, TNF-a, and IL-10 wereevaluated using DuoSet ELISA kits according to the manufacturer’srecommendations (R&D Systems, Minneapolis, MN). The resultswere expressed as picogram per milligram (pg/mg) of tissue proteinconcentration.

SEM AnalysisAfter the experimental periods, the retrieved dentin tubes were

briefly washed in distilled water and sputter coated with gold for SEMobservation (Philips SEM XL 30; Philips, Eindhoven, The Netherlands)at an accelerating voltage of 10 kV.

Statistical AnalysisData of cytokine measurement (pg/mg) and optical densities of

immunohistochemistry staining were expressed as mean � standarderror of the mean. Two-way analysis of variance followed by a Bonfer-roni posttest was performed to analyze differences between the groups(p< 0.05).

ResultsInflammatory Cytokine Expression Profile

Cytokine expression profile is shown in Figure 1. For all groups,the total amount of cytokines expressed decreased after day 3. AlthoughMTA induced a proinflammatory cytokine up-regulation during the first3 days, there was no apparent material-dependent effect on the classesof cytokines produced. However, a time-dependent manner wasobserved. In all the experimental groups, the expression of TNF-a and IL-1b peaked at 12 and 24 hours, respectively. The expressionlevels of IL-1b and TNF-a in the MTA group at 12 hours, 1 day, and


Figure 1. Cytokine levels (pg/mg protein) of (A) TNF-a, (B) IL-1b, and (C) IL-10 on tissue homogenates. Each column represents the mean� standard error ofthe mean. )p < 0.05 versus the empty tube group. #p < 0.05 versus the sham group.


3 days were significantly up-regulated when compared with those in theempty tubes and sham (p < 0.05). By day 7, no significant differenceswere found in any group. Meanwhile, IL-10 expression was up-regulated between days 1 and 3, and the expression peaked on day 1in all the experimental groups (Fig. 1C). The MTA group showeda significant increase in IL-10 expression compared with the emptytube and sham groups at 12 hours, 1 day (p < 0.05), and 3 days(p < 0.01). On day 7, the experimental groups continued to exhibitIL-10 expression. However, the magnitude of IL-10 expressiondecreased, and no statistically significant difference was observedbetween the groups.

Inflammatory Response Assessment:Histomorphological and Immunohistochemical Findings

After 12 hours, the tissue surrounding all experimental groupscontained primarily neutrophils. In all groups, neutrophil recruitmentdecreased between days 1 and 3, and cellular populations of macro-phages and lymphocytes increased and remained elevated from days3 to 7. On day 7, the inflammation intensity had diminished, anda chronic inflammatory cell infiltration consisting primarily of macro-phages, fibroblasts, lymphocytes, and few giant cells was present ina thin fibrous capsule. These findings show the transition from an acutephase to a moderate chronic response.

Immunoreactivity analyses for MPO, AP-1, NF-kB, iNOS, COX-2,and VEGF are shown in Figure 2. These analyses revealed the expressionof the different proteins in a time-dependent manner. MPO expressionpeaked at 24 hours in all experimental groups, whereas a significantincrease was observed in tissues in contact with MTA between 12 hoursand day 1 when compared with the control groups (p < 0.001).

AP-1 expression was slightly elevated in all groups. However, MTAshowed a significant up-regulation of the transcriptional factor on day 1


(p < 0.001) (Figs. 2 and 3). All experimental groups inducedpronounced phosphorylation of NF-kB, which peaked at 12 hours.As expected, MTA significantly up-regulated NF-kB expressioncompared with empty tubes and shams (p < 0.01). In all groups, theexpression of COX-2 and iNOS peaked at 12 and 24 hours, respectively(Fig. 2). However, MTA caused a significant increase in the expressionof COX-2 and iNOS when compared with the empty tubes and sham(p < 0.01). The expression of VEGF was increased at all time periods(Fig. 2). During the acute phase of inflammation, VEGF was mainly ex-pressed in the presence of neutrophils and macrophages (Fig. 3). Onday 7, VEGF expression by fibroblasts was also observed.

In Vivo Biomineralization Ability of MTASEM examination of dentin tubes showed the presence of apatite-

like clusters (Fig. 4). SEM-EDAX indicated that the precipitates mainlycontained calcium and phosphorus with Ca/P molar ratios of 1.60 to1.64 (Fig. 4). It was possible to observe numerous apatite-like clustersdeposited on collagen fibrils all over the surface of dentin tubes contain-ing MTA in as early as 12 hours from implantation. With the increase inthe implantation time, a more extensive mineralization was observed;many of these precipitates formed agglomerates. After 7 days, a compactapatite layer was observed all over the surface of the dentin tubes (Fig. 4).

DiscussionDespite the progress made in understanding the molecular biology

that controls the mechanism of action of MTA (17–19), the exactmechanism of wound healing and the nature of hard-tissue formationremain unclear and, consequently, a matter of extensive research.Therefore, our study focused on the inflammatory reaction and its


Figure 2. Immunoreactivity analysis for MPO, AP-1, NF-kB, iNOS, COX-2, and VEGF. Staining intensity and stained area of antibodies immunoreaction are ex-pressed as optical density. Each column represents the mean � standard error of the mean. )p < 0.05 versus the empty tube group. #p < 0.05 versus sham.


correlation with the biomineralization process to better understand themechanisms underlying host responses to MTA.

The inflammatory reaction is closely related to the healing process(20, 21). The body’s defense reactions involve several regulatoryfunctions and numerous molecular mediators (20). Because repairbegins at the onset of inflammation, it is necessary to further understandthe inflammatory process.

Initially, during the inflammatory response, IL-1b and TNF-a havea proinflammatory effect followed by a regulatory effect in the laterstages of inflammation, reducing immune activity (22, 23). This factmay explain the findings of our study in which overexpressedcytokine levels in the acute phase tended to decrease over time.Probably, the regulatory effect of these cytokines signaled theresolution of the inflammatory reaction during the chronic phase.Additionally, our data suggest that the presence of MTA in the dentintubes had an effect on the intercellular signaling that occurs at theimplantation site. It is interesting to note that there wasa simultaneous overexpression of IL-10 in MTA specimens. BecauseIL-10 has been shown to down-regulate cytokine production, IL-10up-regulation may have signaled the decrease in cytokine productionobserved at later time points (20). As suggested in previous studies,our data support the idea that MTA has an anti-inflammatory effect(17, 22).

MPO immunostaining showed that neutrophils were the predom-inant cells at the implantation site during the first day. Neutrophilrecruitment decreased from day 1 to day 3; after which, mostly macro-phages and lymphocytes migrated into the tissue. As a result, by day 3,the number of inflammatory cell numbers was diminished, indicatingthe beginning of the resolution phase in the inflammatory process.This histomorphological change may explain the expected transition


from an acute proinflammatory phase to an anti-inflammatory andpro–wound healing chronic environment. Biomaterials might elicitseveral signaling pathways to trigger the inflammatory cascade (3,24, 25). Therefore, we analyzed the expression of selected NF-kB-regulated gene products (iNOS and COX-2), AP-1 and VEGF by immu-nohistochemical staining.

Our results showed that NF-kB was involved in MTA-stimulatedsignal transduction in the early stages of inflammation. Therefore, wesuggest that MTA induces phosphorylation of IkBa, which leads tothe freeing of NF-kB complexes. Activated NF-kB complexes translocateto the nucleus and stimulate the expression of COX-2. Released prosta-glandins, in turn, induce the transcription of iNOS in an autocrinemanner (26).

Recent evidence shows that the production of prostanoids by COX-2 promotes the expression of VEGF and subsequent angiogenesis (11).Moreover, AP-1 may induce VEGF expression. However, we observedthat VEGF overexpression remained stable during the various timeperiods. During the acute phase, VEGF upregulation was attributed toits ability to increase the permeability of blood vessels - an importantvascular change observed during the early stages of inflammation. Onday 7, VEGF was mainly expressed by fibroblasts, suggesting sustainedangiogenesis for the induction of a repair process.

Our study provides compelling evidence of the in vivo biominer-alization process promoted by MTA. SEM analysis showed the presenceof the deposition of apatite-like clusters on collagen fibrils as early as 12hours after implantation. SEM-EDAX indicated that the precipitates weremainly composed of calcium and phosphorus. Previously, we showedthat the interaction of MTA with dentin in a phosphate-containing fluidproduces an amorphous calcium phosphate phase, which acts asa precursor during the formation of carbonated apatite (15, 16). It is


Figure 3. Representative images for tissue in contact with MTA at day 1 (40�).Scale in 50 mm. (A) Hematoxylin and eosin staining. Immunohistochemical reac-tion for (B) MPO, (C) AP-1, (D) NF-kB, (E) iNOS, (F) COX-2, and (G) VEGF. (H) Negative control of the immunohistochemical reaction. (This figure is available incolor online at www.aae.org/joe/.)


important to highlight that SEM-EDAX analysis showed similar resultsfor precipitates formed by MTA after subcutaneous implantation. Ourfindings corroborate those of previous in vitro studies that suggestthat calcium ions released by MTA react with phosphate, yieldingcarbonate apatite precipitates (14, 27).

It is well known that the organic matrix possesses properties thatcan initiate and regulate the formation of mineral crystals. Thus, crystalnucleation and controlled growth are considered to be matrix-mediatedor matrix-regulated processes (27–29). Type I collagen is the templatefor the controlled deposition of calcium phosphate, but by itself it doesnot have the capacity to induce matrix-specific mineralization (27, 30).Noncollagenous proteins present in the mineralized dentin matrix, such


as dentin matrix protein 1, have been implicated as having a regulatoryfunction in dentin formation (27). Recombinant dentin matrix protein 1molecules have been thought to perform specific molecular recognitionin conjunction with the apatite surface in order to guide calcium phos-phate clusters through the collagen matrix during recruitment (15, 27).This process has been described as controlled biomineralization (15).

To our knowledge, our study is the first to provide evidence thatthe biomineralization process occurs simultaneously with the initialacute inflammatory response. Therefore, we hypothesize that togetherwith the alkalinity of the material, the precipitation of apatite by MTAduring the acute phase of inflammation may contribute to the signalingof several unrecognized pathways in different cell types. Apatites may



Figure 4. SEM photomicrographs of the surface of dentin tubes filled with MTA after subcutaneous implantation. (A) Formation of several apatite-like precipitatesat 12 hours (3,000�). (C) Energy Dispersive X-Ray Analysis (EDAX) spectrum revealing the chemical composition and the Ca/P ratio of the precipitate shown in B(8,000�). Observe the deposition of precipitate on the collagen fibers (D, 2,500�) (E and F, 4,000�). Increased deposition of precipitates at day 1 (G, 2,500�)(H and I, 5,000�), and in areas of higher mineralization, collagen fibers exhibited a typical ‘‘corn-on-the-cob’’ appearance (arrows). Mineralization areas weremore extensive as the implantation time increased (J and K, 3,000�), forming a compact layer at 7 days after implantation (L, 100�). (This figure is available incolor online at www.aae.org/joe/.)


induce changes in gene expression and subsequently in cell functionalactivity. These changes are likely to contribute to repair and biominer-alization process.

Recent studies showed that MTA induced mineralization andmineralized tissue proteins messenger RNA expression of cemento-blasts and bone cells, which play a crucial role in cemental and osseousrepair and regeneration (31, 32). Therefore, we suggest that severalbiological mechanisms in combination with the bioactivity of MTAmay explain its ability to induce mineralized tissue deposition. Ourdata provide scientific background to develop novel biomaterialsaimed at exploiting the natural regenerative potential of pulp andbone tissues.

In summary, we showed that MTA induces a proinflammatory andpro–wound healing environment. The biomineralization process


occurs simultaneously with the acute inflammatory response. Wesuggest that when MTA is implanted, a series of biochemical andbiophysical reactions occurs at the MTA-dentin-tissue interface. Subse-quently, this activates cellular and tissue events in the inflammatory andbiomineralization processes and culminates in the formation of anapatite-like layer that allows the integration of the biomaterial intothe environment.

References1. Zhao W, Wang J, Zhai W, et al. The self-setting properties and in vitro bioactivity of

tricalcium silicate. Biomaterials 2005;26:6113–21.2. Bohner M, Lemaitre J. Can bioactivity be tested in vitro with SBF solution? Bioma-

terials 2009;30:2175–9.



3. Anderson JM. Biological responses to materials. Annu Rev Mater Res 2001;31:
81–110.4. Williams DF. On the mechanisms of biocompatibility. Biomaterials 2008;29:

2941–53.5. Schutte RJ, Xie L, Klitzman B, et al. In vivo cytokine-associated responses to bioma-

terials. Biomaterials 2009;30:160–8.6. Hsiang CY, Chen YS, Ho TY. Nuclear factor-kB bioluminescense imaging-guided

transcriptomic analysis for the assessment of host-biomaterial interaction in vivo.Biomaterials 2009;30:3042–9.

7. Hayden MS, West AP, Ghosh S. NF-kappaB and the immune response. Oncogene2006;25:6758–80.

8. Ho TY, Chen YS, Hsiang CY. Noninvasive nuclear factor-kB bioluminescence imagingfor the assessment of host–biomaterial interaction in transgenic mice. Biomaterials2007;28:4370–7.

9. De Couto Pita, Borda E, Ganzinelli S, et al. Cholinoceptor modulation on nitric oxideregulates prostaglandin E2 and metalloproteinase-3 production in experimentallyinduced inflammation of rat dental pulp. J Endod 2009;35:529–36.

10. Smith WL, DeWitt DL, Garavito RM. Cyclooxygenases: structural, cellular, and molec-ular biology. Annu Rev Biochem 2000;69:145–82.

11. Akarasereenont PC, Techatraisak K, Thaworn A, et al. The expression of COX-2 inVEGF-treated endothelial cells is mediated through protein tyrosine kinase. Media-tors Inflamm 2002;11:17–22.

12. Guven G, Altun C, Gunhan O, et al. Co-expression of cyclooxygenase-2 and vascularendothelial growth factor in inflamed human pulp: an immunohistochemical study.J Endod 2007;33:18–20.

13. Ferrara N. Molecular and biological properties of vascular endothelial growthfactor. J Mol Med 1999;77:527–43.

14. Tay FR, Pashley DH, Rueggeberg FA, et al. Calcium phosphate phase transformationproduced by the interaction of the Portland cement component of white mineraltrioxide aggregate with a phosphate-containing fluid. J Endod 2007;33:1347–51.

15. Reyes-Carmona J, Felippe MS, Felippe WT. Biomineralization ability and interactionof mineral trioxide aggregate and white Portland cement with dentin in a phosphate-containing fluid. J. Endod 2009;35:731–6.

16. Reyes-Carmona J, Felippe MS, Felippe WT. The biomineralization ability of mineraltrioxide aggregate and Portland cement on dentin enhances the push-out strength.J Endod 2010;36:286–91.

17. Huang TH, Yang C, Ding S, et al. Inflammatory cytokines reaction elicited by root-end filling materials. J Biomed Mater Res B Appl Biomater 2005;73:123–8.


18. Thomson PL, Grover LM, Lumley PJ, et al. Dissolution of bio-active dentine matrixcomponents by mineral trioxide aggregate. J Dent 2007;35:636–42.

19. Gomes AC, Gomes-Filho J, Oliveira SH. MTA-induced recruitment: a mechanismdependent on IL-1ß, MIP-2 and LTB4. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2008;106:450–6.

20. Larsen GL, Henson PM. Mediators of inflammation. Annu Rev Immunol 1983;1:335–59.

21. Teixeira de Moraes M, De Oliveira SH, Gomes-Filho JE. Mechanims of CH-inducedneutrophil migration into air-pouch cavity. Oral Surg Oral Med Oral Pathol 2008;105:814–21.

22. Silva MJ, Vieira LQ, Sobrinho AR. The effects of mineral trioxide aggregate on cyto-kine production by mouse pulp tissue. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2008;105:70–6.

23. Zakharova M, Ziegler HK. Paradoxical anti-inflammatory actions of TNF-alpha: inhi-bition of IL-12 and IL-23 via TNF receptor 1 in macrophages and dendritic cells.J Immunol 2005;175:5024–33.

24. Dayer JM, Isler P, Nicod LP. Adhesion molecules and cytokine production. Am RevResp Dis 1993;148:S70–4.

25. Newton R, Kuitert LM, Bergmann M, et al. Evidence for involvement of NF-kB in thetranscriptional control of COX-2 gene expression by IL-1. Biochem Biophys ResCommun 1997;237:28–32.

26. Minamikawa H, Deyama Y, Nakamura K, et al. Effect of mineral trioxide aggregate onrat clonal dental pulp cells: expression of cyclooxygenase-2 mRNA and inflammation-related protein via nuclear factor kappa B signaling system. J Endod 2009;35:843–6.

27. Tay FR, Pashley DH. Guided tissue remineralisation of partially demineralizedhuman dentine. Biomaterials 2008;29:1127–37.

28. LeGeros RZ. Calcium phosphates in oral biology and medicine. In: Monographs inoral science 15. Basel, Switzerland: Karger; 1991:4–66.

29. Tay FR, Pashley DH. Biomimetic remineralization of resin-bonded acid-etcheddentin. J Dent Res 2009;88:719–24.

30. Hao J, Zou B, Narayanan K, et al. Differential expression patterns of the dentin matrixproteins during mineralized tissue formation. Bone 2004;34:921–32.

31. Chen CL, Huang TH, Ding SJ, et al. Comparison of calcium and silicate cement andmineral trioxide aggregate biologic effects and bone markers expression in MG63cells. J Endod 2009;35:682–5.

32. Hakki SS, Bozkurt SB, Hakki EE, et al. Effects of mineral trioxide aggregate on cellsurvival, gene expression associated with mineralized tissues, and biomineralizationof cementoblasts. J Endod 2009;35:513–9.


Basic Research—Technology

Electropolishing Enhances the Resistance of Nickel-TitaniumRotary Files to Corrosion–Fatigue Failure in HypochloriteChonrada Praisarnti, DDS, MDS (Endo), AdvDipEndodont, Jeffrey W.W. Chang, BDS, MDS (Endo),and Gary S.P. Cheung, PhD, MDS, MSc, FCDSHK(Endo), FHKAM, FAMS, FRACDS, MRACDS(Endo)

Abstract
Introduction: The aim of this study was to examine thefatigue behavior, especially at the low-cycle fatigue(LCF) region, of an experimentally electropolishedFlexMaster and a commercial electropolished nickel-titanium (NiTi) instrument (RaCe) in a corrosive environ-ment. Methods: A total of 90 NiTi rotary instrumentswere subjected to rotational bending at various degreesof curvatures while immersed in 1.2% sodium hypochlo-rite solution until broken. The maximum surface strainamplitude, calculated from the curvature of the instru-ment and the diameter of the cross section at break,was plotted against the LCF life. The results werecompared with data for a non-electropolished commer-cial product tested by using the same methodology.Results: The fatigue life of both instruments generallydeclined with increasing surface strain amplitude; therewas a significant difference between the 2 instruments.Comparing the surface-treated FlexMaster with itscommercially available non-electropolished counterpart,an improved resistance to fatigue breakage as a result ofelectropolishing was noted (P < .05). Conclusions: TheLCF life of a NiTi instrument rotating with a curvature ina corrosive environment is enhanced by electropolish-ing. The design, both cross-sectional and longitudinal,appears to have an effect on the fatigue behavior ofNiTi rotary instruments. (J Endod 2010;36:1354–1357)
Key WordsBreakage, corrosion fatigue, defects, electropolishing,fracture, low-cycle fatigue, nickel-titanium, surfacefinish

From the Area of Endodontics, Comprehensive Dental Care,Faculty of Dentistry, University of Hong Kong, Hong Kong.

Address requests for reprints to Dr Gary S. P. Cheung, Areaof Endodontics, Floor 3A, Prince Philip Dental Hospital, 34Hospital Road, Saiyingpun, Hong Kong. E-mail address:[email protected]/$0 - see front matter


1354 Praisarnti et al.

Since its introduction in the 1990s, the remarkable mechanical properties have madenickel-titanium (NiTi) rotary instruments increasingly popular among clinicians for

root canal instrumentation. Unfortunately, these engine-driven files can fracture in usewithout undergoing any permanent deformation or other visible warning signs (1–3).Both flexural (cyclic) fatigue and torsional (shear) stresses are acting on a rotaryinstrument and might both be responsible for the breakage of NiTi rotary filesduring clinical use (3, 4). The instrument might fail in either or a combination ofthe 2 mechanisms. On the other hand, many studies have indicated that fatigueseems to be the predominant mechanism for a majority of instrument breakagesduring clinical use (3, 5–9).

The fatigue process begins with crack initiation at the material surface (4, 10).Given the importance of surface defects (eg, the machining grooves left by thegrinding process during manufacture), various treatments to improve the surfacesmoothness have been introduced for delaying the crack initiation process for animproved fatigue life of the material (10). Electropolishing, which is a methodcommonly used for surface finishing of many metallic medical appliances (11), hasbeen incorporated for 1 brand of NiTi rotary instrument (RaCe; FKG Dentaire, LaChaux-de-Fonds, Switzerland) by its manufacturer for some time. Some recent addi-tions (eg, EndoSequence; Brasseler, Savannah, GA) have also claimed to have beenelectropolished by its manufacturer. The process is carried out by immersing the device(usually connected as the anode) together with another electrode in a temperature-controlled bath of electrolyte(s) and then passing a DC electric current throughthem. The metal at the anode (especially at peaks or sharp edges where the currentdensity becomes highest) is dissolved into the solution, whereas a reduction reactionwill take place at the cathode. This process alters the surface composition and texture ofthe work piece (11, 12), renders the surface oxide layer (protective film) morehomogeneous, with less defects and residual surface stress (13), and improves thecorrosion resistance of the metal (14). It has been considered to be the best methodto improve the surface characteristics of any medical devices made of NiTi alloy (11).The underlying electrochemical process involved is rather complex and might beaffected by many parameters (15). For instance, some electrolytic species, such ashalogen ions, might become incorporated into the oxide film, leading to susceptibilityof this passive film to break down and hence initiation of corrosion pits (15). Themovement of the charged (electrolyte) ions has been shown to be ‘‘controllable’’ byan externally applied magnetic field (16). An improved result of the electropolishingprocess for the final surface has been demonstrated (15, 16). However, thismagnetoelectropolishing process has not been attempted on dimensionally smallNiTi devices that will be subjected to high operating stresses, such as the NiTi rootcanal instruments.

Many authors hold the opinion that electropolishing would bring about animproved resistance to fatigue failure, compared with the non-electropolished counter-part (17–19). It has been reported that the RaCe instrument (electropolished by itsmanufacturer) demonstrates an increased resistance to fatigue failure, comparedwith ProFile (Dentsply Maillefer, Ballaigues, Switzerland), K3 (SybronEndo, Orange,CA), HERO (Micro-Mega, Besancon, France), and Mtwo (VDW, Munich, Germany)(18). However, these various brands of instruments all have a distinctly differentcross-sectional configuration, a factor that can affect the stress and strain acting onthe instrument rotating with a curvature (20). Thus, it is difficult to conclude whether



the improved fatigue resistance has been due to the effect of electropo-lishing per se. On the other hand, some studies have indicated that elec-tropolishing (performed proprietarily by the manufacturer) did notseem to provide sufficient protection against low-cycle fatigue(LCF) failure (21, 22), especially in the presence of a corrosiveenvironment (22). Thus, the purported advantage of electropolishingin enhancing the fatigue resistance of NiTi instruments remains contro-versial. Meanwhile, the great majority of laboratory studies of the fatiguebehavior of NiTi rotary files have only adopted 1 or, at most, 2 curvaturesettings in those tests that had been carried out in air (or with somelubricant oil sprayed on the instrument) (23–26). Other strainamplitudes have not been examined, and the effect of theenvironment was not considered. The purpose of this study was todetermine the corrosion-fatigue behavior of 2 NiTi engine-files, eachhaving been electropolished with a different regime, subjected to rota-tional bending in a corrosive environment.
Figure 1. Relationship between the number of revolutions to failure and themaximum surface strain amplitude of prototype FlexMaster and RaCe instru-ments on logarithmic scales tested at various curvatures: (A) scatter diagramfor all instruments and (B) specimens that failed at the LCF region.

Materials and MethodsA strain-controlled, rotational bending, fatigue test was performed

on the commercial RaCe (FKG Dentaire) and a specially treated Flex-Master (VDW, Munich, Germany) rotary instruments. The latter waselectropolished under the influence of a magnetic field, with a patentedmethod known as magnetoelectropolishing (15), by a company notinvolved in the manufacture or in this study, whereas RaCe was electro-polished by its manufacturer by using a proprietory regime (withoutmagnetic flux). The fatigue testing machine was performed, on the basisof a method described elsewhere (26, 27), in which the instrument wasconfined by 3 smooth, high-hardness stainless steel pins into a curve(ie, 3-point bending) before setting to rotate at a rate of 250 rpm untilbroken. The test was carried out with the instrument immersed in a bathof 1.2% hypochlorite solution (Clorox, Oakland, CA); the stainless steelpins were replaced once signs of wear or rusting were discernible. Apre-test digital photograph of the curvature was taken at a fixed distanceand magnification for measurement of the radius of curvature in soft-ware (ImageJ 1.36b; NIH, Bethesda, MD) for each individual instru-ment (27). The number of revolutions to failure was recorded byusing an optical counter. After the test, the broken fragment wasmeasured for length by using a stereomicroscope (Leica DMLB; LeicaMicrosystems, Wetzlar, Germany) at 25� magnification, and then thefracture surface was examined with a scanning electron microscope(SEM) (XL-30cp; Phillips, Eindhoven, The Netherlands). The experi-ment was repeated with the instrument set up to rotate at various curva-ture settings.

The diameter of the fracture cross section was determined on thescanning electron photomicrograph for each broken fragment in thesame image analyzer software (ImageJ 1.36b). The effective surfacestrain amplitude, 3a, imposed on the rotating instrument was definedby the unit deformation of the material at its outermost surface (26, 28):

3a ¼d

2Rc

A total of 90 instruments (n = 50 for magnetoelectropolished Flex-Master and n = 40 for RaCe) were tested at various curvatures. Thenumber of revolutions or cycles to fatigue (NCF) was plotted againstea in software (SigmaPlot 10.0; Systat Software, San Jose, CA). A regres-sion line was fitted to the LCF region, where appropriate, and the datafor the 2 instruments were compared by using a two-way analysis ofcovariance (ANCOVA) (SPSS for Windows 14.0; SPSS, Chicago, IL).Data for the untreated (commercially available) FlexMaster tested byusing the same methodology were obtained from a previous study


(29) and were compared with the magnetoelectropolished ones statis-tically (ANCOVA). The significance level was set at P = .05.

ResultsThe relationship between the total fatigue lifetime and the effective

surface strain amplitude was determined for the 2 instruments exam-ined (Fig. 1). The transition (in the slope of the data trend) fromhigh-cycle fatigue (HCF) to LCF was obvious. Failure taking place atabout 2000 cycles or lower was deemed to be within the LCF region.A regression line could be fitted for the LCF lives on logarithmic scales(Fig. 1B), suggesting that the LCF life declined in a power functiondependence on the strain amplitude (P < .05). A significant differencein the rate of decline (ie, slope of the fitted line) of fatigue life was notedbetween the magnetoelectropolished FlexMaster and the RaCe instru-ment (P < .05) (Table 1). Similar slope of the fitted lines was notedfor the magnetoelectropolished and the untreated FlexMaster (Table1 and Fig. 2), with the former being significantly more resistant tofatigue fracture than the latter (P < .05). The fatigue life for RaCe instru-ment was intermediate between the magnetoelectropolished and theuntreated FlexMaster at various strain amplitudes (Fig. 2).

DiscussionCorrosion of NiTi on prolonged immersion in hypochlorite has

been demonstrated by various authors (30, 31). A negative effect on

Effect of Electropolishing on Fatigue Failure of NiTI Rotary Files 1355

TABLE 1. Slope of Fitted Line and Y-intercept Derived from the LCF Lives (Nf < 2000) for Each Group

Group InstrumentTotal no.

testedNo. failed inLCF region

Slope of fitted line(± standard error) Y-intercept

Coefficient ofdetermination (r2)

FM_E MagnetoelectropolishedFlexMaster

50 37 –1.2346 (� 0.3752) 3.18 0.54

FM_NE Commercial FlexMaster 35 29 –1.3377 (� 0.4714) 2.96 0.44RaCe RaCe 40 35 –0.7934 (� 0.1401) 3.16 0.12

LCF, low-cycle fatigue.


the fatigue resistance has also been reported for NiTi rotary files that hadbeen immersed for up to 2 hours in 5.25% hypochlorite solution beforefatigue testing the instrument in air afterwards (32). The corrosionprocess involves selective removal of nickel (the more reactive elementof the alloy) from the material surface, creating corrosion pits (33).These microstructural defects on the surface can lead to areas of stressconcentration and early crack formation on cyclic loading (10, 34).The importance of the environment on the measured fatigue lifetimeof NiTi rotary files has been reported (29, 35). Although a controlgroup of instruments fatigued in a bland solution is not includedfor comparison (for the effect of hypochlorite), it has beendemonstrated that a reduction of about 25%–30% of the total fatiguelifetime is to be expected for NiTi files rotating in hypochlorite(1.2%) solution, compared with those in water (35). Thus, testing ina noncorrosive environment could result in an overestimation of thefatigue resistance of the NiTi instruments. In this study, sodium hypo-chlorite was chosen as the medium in which the test was performedto simulate the clinical situation. A concentration of 1.2% was usedin this study because this strength is commonly used in clinical practiceand, indeed, is adopted in the polyclinic of a teaching hospital (wherethis study was performed). Higher concentrations (3%–6%) werefound to be highly corrosive to the stainless steel pins in a pilot study,necessitating very frequent changes of these pins if the test were to becarried out in such condition. A high concentration of hypochlorite isanticipated to have a negative effect on the fatigue behavior as a resultof an increased amount of available chlorine that attacks the metal.

The fatigue life of both instruments decreased as the strain ampli-tude increased (Fig. 1), with an ‘‘elbow’’ in the data trend signifyinga transition from HCF to LCF region of the material. This trend wouldnot be revealed if the test had been done at 1 or 2 curvature settingsonly. Indeed, the elbow might occur at different NCF values, dependingon the environment, as a result of corrosion fatigue (34).

It has been suggested that the LCF lives of various brands of NiTiinstrument are similar to each other, provided that they are expressed

Figure 2. Relationship between fatigue lives in LCF region (Nf < 2000) ofprototype and a commercial FlexMaster instruments [pooled data froma previous study (12)] and the maximum surface strain amplitude tested atvarious curvatures. (This figure is available in color online at www.aae.org/joe/.)

1356 Praisarnti et al.

as a function of the imposed surface strain (36). However, that doesnot seem to apply to the RaCe instrument in this present study, whichshowed a large scatter in its fatigue life, even for similar strain ampli-tudes (Fig. 1); the regression line showed a relatively low coefficientof correlation. This might be attributable, at least in part, to thespecial design of the instrument, with regions of the triangular shaftalternating with regions of a (spiraling) fluted part. That mighthave led to uneven pattern of stress concentration, possibly at thejunction between the 2 said regions, and hence a shortened fatiguelife with the test methodology in this study (FKG, personal communi-cation). It is possible that this peculiar (longitudinal) design mighthave confounded the beneficial effect of having a smooth surfacethat should help withstand the fatigue loading. A just comparisonbetween RaCe instrument and other brands (with regularly spaced,spiraling flutes) is not feasible because of the peculiar design ofthe former. Thus, although the plot of LCF lives of RaCe instrumentis positioned in between that of the nontreated and magnetoelectro-polished FlexMaster instrument (Fig. 2), it is not sure whether theeffect had been due to the special surface treatment regime or tothe design of the instrument. Despite the potential (theoretical) stressconcentration due to the design of the RaCe instrument, this electro-polished file has had a better resistance to LCF breakage thana commercial, ground, non-electropolished rotary instrument.

Comparing the results for the FlexMaster instruments with andwithout surface treatment, a promising effect of the magnetoelectropo-lishing process for improving the fatigue resistance is noted. There isa significant increase in the LCF lives for the treated versus the commer-cial (untreated) FlexMaster at all strain amplitudes. This is in directcontrast to the results of a previous study with a similar test methodology(22), which reported a lack of effect as a result of electropolishing foranother brand of NiTi rotary file that was electropolished by its manu-facturer (HERO Shaper; Micro-Mega). Unfortunately, parametersrelated to the electropolishing process (eg, strength and compositionof the electrolytes, current density, and any other special arrangementfor the electrolytic cell) in that study were not revealed by the manufac-turer, and thus the underlying mechanism responsible for the differencecould not be deduced. There were not any FlexMaster instruments thathad been electropolished by an ‘‘ordinary’’ process available forcomparison. On the other hand, the results of this study clearly indi-cated an improvement in the fatigue resistance of NiTi rotary file thathad been treated with magnetoelectropolishing, although more studiesare necessary to examine whether such surface treatment would benefitinstruments of various designs and dimensions and under various testconditions.

ConclusionThe resistance to LCF failure of a NiTi instrument rotating with

a curvature in a corrosive environment is enhanced by a magnetoelec-tropolishing process. The instrument design, both cross-sectional andlongitudinal, seems to have an effect on the fatigue behavior of NiTirotary instruments.




AcknowledgmentsThe authors would like to thank Dr May Wong, Associate

Professor in Dental Public Health of the Faculty of Dentistry,University of Hong Kong for assistance in the statistical analysis.The free supply of FlexMaster and RaCe instruments by their respec-tive manufacturers and the magnetoelectropolishing treatment byELECTROBRIGHT, Macungie, PA are gratefully acknowledged. Thestudy was funded in part by a research grant of the University ofHong Kong (account no. 10208970.12058.08008.312.01).

References1. Sattapan B, Nervo GJ, Palamara J, Messer HH. Defects in nickel titanium endodontic

rotary files after clinical usage. J Endod 2000;26:161–5.2. Peng B, Shen Y, Cheung GS, Xia TJ. Defects in ProTaper S1 instruments after clinical

use: longitudinal examination. Int Endod J 2005;38:550–7.3. Parashos P, Gordon I, Messer HH. Factors influencing defects of rotary nickel-

titanium endodontic instruments after clinical use. J Endod 2004;30:722–5.4. Cheung GSP. Instrument fracture: mechanisms, removal of fragments, and clinical.

Endodontic Topics 2009;16:1–26.5. Cheung GS, Peng B, Bian Z, Shen Y, Darvell BW. Defects in ProTaper S1 instruments

after clinical use: fractographic examination. Int Endod J 2005;38:802–9.6. Shen Y, Cheung GS, Bian Z, Peng B. Comparison of defects in ProFile and ProTaper

systems after clinical use. J Endod 2006;32:61–5.7. Spanaki-Voreadi AP, Kerezoudis NP, Zinelis S. Failure mechanism of ProTaper Ni-Ti

rotary instruments during clinical use: fractographic analysis. Int Endod J 2006;39:171–8.

8. Cheung GS, Bian Z, Shen Y, Peng B, Darvell BW. Comparison of defects in ProTaperhand-operated and engine-driven instruments after clinical use. Int Endod J 2007;40:169–78.

9. Wei X, Ling J, Jiang J, Huang X, Liu L. Modes of failure of ProTaper nickel-titaniumrotary instruments after clinical use. J Endod 2007;33:276–9.

10. Suresh S. Fatigue of materials. 2nd ed. Cambridge, UK: Cambridge University Press;1998.

11. Shabalovskaya S, Anderegg J, Van Humbeeck J. Critical overview of nitinol surfacesand their modifications for medical appliances. Acta Biomaterialia 2008;4:447–67.

12. Miao W, Mi X, Zhu M, Guo J, Kou Y. Effect of surface preparation on mechanicalproperties of a NiTi alloy. Mater Sci Forum 2002;394-395:173–6.

13. Su Y-Y, Raman V. The quest for Nitinol wire surface quality for medical applications.In: SMST-97: Proceedings of the Second International Conference on Shape Memoryand Superelastic Technologies, 1997:389-394

14. Bonaccorso A, Tripi TR, Cantatore G, Condorelli GG. Surface properties of nickel-titanium rotary instruments. Endod Pract Today 2007;1:45–52.

15. Rockicki R, Hryniewicz T. Nitinol surface finishing by magnetoelectropolishing.Transaction of the Institute of Metal Finishing 2008;86:280–5.


16. Kim J-D, Jin D-X, Choi M-S. Study on the effect of a magnetic field on an electrolyticfinishing process. Int J Med Tools Manufacturing 2008;37:401–8.

17. Parashos P, Messer HH. Rotary NiTi instrument fracture and its consequences. J En-dod 2006;32:1031–43.

18. Tripi TR, Bonaccorso A, Condorelli GG. Cyclic fatigue of different nickel-titaniumendodontic rotary instruments. Oral Surg Oral Med Oral Pathol Oral Radiol Endod2006;102:e106–14.

19. Anderson ME, Price JW, Parashos P. Fracture resistance of electropolished rotarynickel-titanium endodontic instruments. J Endod 2007;33:1212–6.

20. Kim HC, Cheung GS, Lee CJ, Kim BM, Park JK, Kang SI. Comparison of forces gener-ated during root canal shaping and residual stresses of three nickel-titanium rotaryfiles by using a three-dimensional finite-element analysis. J Endod 2008;34:743–7.

21. Herold KS, Johnson BR, Wenckus CS. A scanning electron microscopy evaluation ofmicrofractures, deformation and separation in EndoSequence and Profile nickel-titanium rotary files using an extracted molar tooth model. J Endod 2007;33:712–4.

22. Cheung GSP, Shen Y, Darvell BW. Does electropolishing improve the low-cyclefatigue behavior of a nickel-titanium rotary instrument in hypochlorite? J Endod2007;33:1217–21.

23. Pruett JP, Clement DJ, Carnes DL Jr. Cyclic fatigue testing of nickel-titaniumendodontic instruments. J Endod 1977;23:77–85.

24. Haıkel Y, Serfaty R, Bateman G, Senger B, Allemann C. Dynamic and cyclic fatigue ofengine-driven rotary nickel-titanium endodontic instruments. J Endod 1999;25:434–40.

25. Peters OA, Kappeler S, Bucher W, Barbakow F. Engine-driven preparation of curvedroot canals: measuring cyclic fatigue and other physical parameters. Aust Endod J2002;28:11–7.

26. Larsen CM, Watanabe I, Glickman GN, He J. Cyclic fatigue analysis of a new gener-ation of nickel titanium rotary instruments. J Endod 2009;35:401–3.

27. Cheung GSP, Darvell BW. Fatigue testing of a NiTi rotary instrument: part 1—strain-life relationship. Int Endod J 2007;40:612–8.

28. Huston RL. Principles of biomechanics. Boca Raton, FL: CRC Press; 2009. 79–140.29. Cheung GSP, Darvell BW. Low-cycle fatigue of rotary NiTi endodontic instruments in

hypochlorite solution. Dent Mater 2008;24:753–9.30. Stokes OW, Fiore PM, Barss JT, Koerber A, Gilbert JL, Lautenschlager EP. Corrosion

in stainless-steel and nickel-titanium files. J Endod 1999;25:17–20.31. O’Hoy PY, Messer HH, Palamara JE. The effect of cleaning procedures on fracture

properties and corrosion of NiTi files. Int Endod J 2003;36:724–32.32. Peters OA, Roehlike JO, Baumann MA. Effect of immersion in sodium hypochlorite

on torque and fatigue resistance of nickel-titanium instruments. J Endod 2007;33:589–93.

33. Sarkar NK, Redmond W, Schwaninger B, Goldberg AJ. The chloride corrosionbehaviour of four orthodontic wires. J Oral Rehabil 1983;10:121–8.

34. ASM Handbook, vol. 19: fatigue and fracture. Materials Park, OH: ASM International;1996.

35. Cheung GSP, Shen Y, Darvell BW. Effect of environment on low-cycle fatigue ofa nickel-titanium instrument. J Endod 2007;33:1433–7.

36. Cheung GSP, Darvell BW. Low-cycle fatigue of NiTi rotary instruments of variouscross-sectional shapes. Int Endod J 2007;40:626–32.

Effect of Electropolishing on Fatigue Failure of NiTI Rotary Files 1357


Comparative Evaluation of the Antimicrobial Efficacy of a 5%Sodium Hypochlorite Subsonic-activated SolutionDamiano Pasqualini, DDS,* Anna Maria Cuffini, BSc, PhD,† Nicola Scotti, DDS,*

Narcisa Mandras, PhD,†

Daniela Scalas, PhD,†Francesco Pera, DDS,* and Elio Berutti, MD, DDS*

Abstract
Introduction: The study evaluated the efficacy ofsubsonic agitation of sodium hypochlorite (NaOCl) inreducing bacterial load in the root canal. Methods:Root canals of 112 extracted human single-root teethwere preflared using K-Flexofiles (Dentsply Maillefer,Ballaigues, Switzerland) up to #20 and then shapedusing ProTaper S1-S2-F1-F2-F3 (Dentsply Maillefer) atthe working length. Irrigation was performed with 33mL of 5% NaOCl, alternating with 10 mL of 10%EDTA. After ethylene oxide sterilization, the root canalswere infected with 30 mL of Enterococcus faecalisculture and randomly assigned to four groups (n = 25)of different irrigation regimens plus positive and nega-tive controls. Irrigation was performed with 2 mL of5% NaOCl. In the NaOCl 15 group, the irrigant wasleft in place for 15 seconds, and in the NaOCl 30 groupit was left in place for 30 seconds. In the EndoActivator(EA; Dentsply Tulsa Dental Specialties, Tulsa, OK) 15 andEA 30 groups, NaOCl was subsonically agitated with EAfor 15 and 30 seconds, respectively. The residual bacte-rial count was then evaluated. Differences amonggroups were analyzed with one-way analysis of varianceand the post hoc Bonferroni test (p < 0.05). Results: Astatistically significant difference was evidenced amonggroups (F3 = 9.01, p < 0.001). The standard irrigationgroups (NaOCl 15 and 30) showed higher microbialcounts than the EA 30 group (p < 0.05). Conclusion:Thirty seconds of NaOCl subsonic agitation with En-doActivator appears to be slightly more effective inreducing bacterial load in the root canal comparedwith NaOCl irrigation alone. (J Endod 2010;36:1358–1360)
Key WordsDisinfection, EndoActivator, endodontic irrigants,sodium hypochlorite, subsonic

From the Departments of *Endodontics and †Public Healthand Microbiology, University of Turin, Turin, Italy.

Address requests for reprints to Dr Damiano Pasqualini, viaBarrili, 9–10134 Torino, Italy. E-mail address: [email protected]/$0 - see front matter


1358 Pasqualini et al.

Bacteria and their byproducts play a relevant role in the onset and perpetuation ofpulpal and periradicular disease (1). Root canal treatment aims to eliminate

remnants of pulp tissue, bacteria, and microbial toxins from the infected canal systemand to prevent reinfection in order to achieve long-term success (2–4). Clinical studieshave shown a more favorable long-term prognosis of specimens that were culture nega-tive before obturation versus culture-positive specimens (94% vs 68%) (5), whereasother studies have failed to show any significant difference concerning healing (6).However, there is general consensus that successful elimination of the causative agentsfrom the root canal system is the key to health (7).

Chemical-mechanical treatment of the root canal system has shown its efficacy inreducing bacterial load (8), even though bacteria may persist despite these efforts (9)because of the complexity of the root canal system (10–12). Sodium hypochlorite(NaOCl) has been widely used as an irrigant since its introduction in endodontics byWalker in 1936, and it is still considered an effective disinfectant agent (13). NaOClused at concentrations ranging from 0.5% to 6% is a potent antimicrobial agent and effec-tively dissolves organic debris. Numerous irrigation regimens have been proposed toenhance the effectiveness of NaOCl in disinfecting the root canal system, including incombination with sonic and ultrasonic instrumentation (14). Both cavitation and acousticstreaming may help to enhance debridement and disinfection (15) of complex root canalsystems (16). However, ultrasonic instrumentation with metal active tips may lead to canaltransportation, ledges, zipping, and stripping (17), especially in very curved canals (18).

Recently, a device known as the EndoActivator (Dentsply Tulsa Dental Specialties,Tulsa, OK) (19) has been introduced; it is designed to enhance hydrodynamicphenomena by means of the subsonic activation of a passive smooth polymer tip, whichis inserted into the root canal full of irrigating solution. The objective of this study was toevaluate the efficacy of subsonic activation of NaOCl in reducing bacterial load in theroot canal.

Materials and MethodsOne hundred twelve extracted human single-root teeth with a fully formed apex

(upper central incisors and canines with substantially equal canal curvature andmorphology) that had not undergone prior endodontic treatment were used. After de-briding the root surface, specimens were immersed in a 5% solution of NaOCl (Niclor 5;OGNA, Muggio, Italy) for 1 hour and then stored in saline solution until preparation.Each specimen was sectioned to obtain a residual root length of 15 mm. Each root canalwas preflared using K-Flexofiles (Dentsply Maillefer, Ballaigues, Switzerland) up to #20and then shaped using ProTaper S1-S2-F1-F2-F3 (Dentsply Maillefer) at the workinglength. The working length was established under microscopic vision (OPMI ProErgo; Carl Zeiss, Oberkochen, Germany) at 10� magnification when the tip of theinstrument was visible at the apical foramen. Irrigation was performed with a 30-gaugeneedle syringe using 33 mL of 5% sodium hypochlorite at 50�C (Niclor 5; OGNA, Mug-gio, Italy) and alternating with 10 mL of 10% EDTA (Tubuliclean, OGNA); the total irri-gation time was 10 minutes per specimen. After drying with paper points, the roots wereinspected under the microscope at 10� magnification to verify the absence of cracksand canal cleanliness. Root surfaces were sealed with varnish and sticky wax; each spec-imen was fixed with cyanoacrylic cement onto an Eppendorf tube, which was placed ina plastic support box. Specimens were placed in envelopes and sterilized with ethyleneoxide. This is a volatile gas that does not alter the structure of materials with which it




comes into contact and does not produce a temperature increase. Itleaves no residue at the end of the sterilization cycle, even inside thedentin tubules, not influencing the growth or vitality of bacteria inocu-lated subsequently (20–23). The procedure was as follows: 6 hours at40�C, 3 hours at 70% to 75% humidity, a 6-hour application of 10%ethylene oxide, and total removal of the gas from the envelope byrepeated replacement of the air content.
The sterilized roots were placed under a laminar flow biohazardcabinet (CLANLAF VFR 1206; Capriolo, Brescia, Italy). The root canalswere infected with a standard volume of 30 mL of a pure culture ofEnterococcus faecalis ATCC 29212, which was previously grown inbrain-heart infusion (Oxoid, Milan, Italy) medium broth for 24 hoursand adjusted spectrophotometrically to an optical density of 0.15 at620 nm (Genesys 20 Spectrophotometer; Thermo Electron Corporation,Madison, WI) to match the turbidity of 3 � 107 CFU as confirmed bycolony counts in triplicate. Specimens were further incubated aerobicallyat 37�C for 2 hours to allow penetration of E. faecalis into the root canaldentine. Two additional specimens were used as negative controls and 10as positive controls. The remaining 100 samples were randomly subdi-vided into four groups (n = 25) using a random numbers table.

Irrigation Protocols and Microbe CountSpecimens in the NaOCl 15 group (n = 25) were irrigated for 40

seconds with 2 mL of a 5% NaOCl solution at room temperature witha 30-gauge needle syringe 2 mm short of the apex. NaOCl was left inthe root canal for 15 seconds before removal with 5 mL of saline solution.Specimens in group NaOCl 30 (n = 25) followed the same procedure,but the NaOCl was left in the root canal for 30 seconds before removal.

Specimens in the EndoActivator (EA; Dentsply Tulsa DentalSpecialties, Tulsa, OK) 15 group (n = 25) were irrigated for 40 secondswith 2 mL of a 5% NaOCl solution at room temperature with a 30-gaugeneedle syringe 2 mm short of the apex. NaOCl was left in the root canaland immediately activated subsonically for 15 seconds, inserting the EA15/.02 polymer tip into the root canal 2 mm short of the apex; the irri-gant was then removed with 5 mL of saline solution. The EA driver wasset at 10.000 cpm. Specimens in the EA 30 group (n = 25) followed thesame procedure, except that the NaOCl was activated with EA for 30seconds. Positive controls (n = 10) were irrigated for 40 secondswith 2 mL of sterile water.

Subsequent to each irrigation treatment, the root canals weredried at working length and sampled with sterile paper points. Thepaper points were transferred to tubes containing 1 mL of 0.85% salinesolution and vortexed for 1 minute. After 10-fold serial dilutions,aliquots of 0.1 mL were plated onto brain-heart infusion mediumagar and incubated at 37�C under aerobic conditions for 24 hours.The colony-forming units (CFUs) grown were counted and then trans-formed into actual counts based on the known dilution factors.

Statistical MethodsThe Kolmogorov-Smirnov test for normality revealed a normal

data distribution. Statistical analysis was conducted with a model of

TABLE 1. Descriptive Statistics of the Postirrigation Microbe Count (105CFUs) and

Group N Mean STD Median Min

NaOCl-15’’ 25 3.75 3.00 2.2 1.4NaOCl-30’’ 25 3.47 2.17 3.4 0.53EA-15’’ 25 2.34 1.20 2.25 1.12EA-30’’ 25 1.01 0.84 0.67 0.43

CFU, colony-forming units; CI, confidence interval; EA, Endoactivator.


one-way analysis of variance test and a post hoc Bonferroni test formultiple comparisons. Differences were considered statistically signifi-cant when p < 0.05. All statistical analyses were performed using theSPSS for Windows 12.0 software package (SPSS Inc, Chicago, IL).

ResultsDescriptive statistics of the postirrigation microbe count and the

percentage of bacterial load reduction are summarized in Table 1.The inferential analysis revealed a statistically significant differenceamong groups (F3 = 9.01, p < 0.001). The multiple comparisonspost hoc analysis evidenced a statistically significant difference betweenstandard NaOCl irrigation groups 15 and 30 and group EA 30 in whichthe NaOCl was activated with the subsonic device for 30 seconds (p <0.05). The bacterial load reduction compared with positive controls(mean 1.26 � 1.05 � 107 CFU, 61.5% reduction) ranges from98.6% (NaOCl 15) to 99.6% (EA 30) with slight differences amonggroups.

DiscussionThe need to improve root canal disinfection is increasingly attract-

ing interest because even modern nickel-titanium rotary instrumenta-tion only act on the central portion of the root canal system, leavingpotential niches untreated (21–24). Thus, in recent decades,endeavors have been made to enhance the efficacy of irrigantsolutions through innovative irrigant delivery devices and agitationtechniques, both manual and machine assisted (19).

Sonic activation has been shown to be an effective method to re-move the oral biofilm and enhance root canal disinfection (25).However, the performance of subsonic agitation appears to be lesseffective compared with ultrasonic activation of irrigant solutions(26). This may be attributed to the different acoustic streaming velocityand frequency, which positively influence debris removal from the qual-itative standpoint. However, other studies found no difference betweenthe 2 systems (27) and reported similar penetration of the solution intoextracted teeth accessory canals (28), whereas the EA promoted lessextrusion of the irrigant over the apex (29).

The advantages of sonic agitation of the irrigant solution have beenanalyzed, reporting significantly better debridement of the root canalwalls compared with manual agitation with endodontic files (27).The EA system has been reported to effectively clean debris from lateralcanals, remove the smear layer, and dislodge clumps of simulated bio-film within the curved canals of molar teeth (30). Another recent study(31) compared the effects of different ultrasonic tips and the EA systemon necrotic pulp dissolution and transportation of the main canal usingepoxy resin–modified models with simulated accessory canals and2.5% NaOCl irrigant. The results showed that ultrasonic activation dis-solved more tissue than did sonic activation or passive irrigation; the EAsonic system with passive polymer tip and ultrasonically activatednickel-titanium tips caused no detectable canal transportation.However, these studies did not consider the influence of the type of irri-gation on root canal disinfection.

Bacterial Load Reduction (%)

95% CI

Max Lower Upper Bacterial load reduction (%)

8.5 2.47 5.04 98.66.2 2.46 4.47 98.75.08 1.83 2.85 99.13.17 0.67 1.35 99.6

Antimicrobial Efficacy of a 5% NaOCl Subsonic-activated Solution 1359

The test hypothesis of this study was that sonic activation of NaOCl
associated with a standard irrigation regimen enhances disinfection.The potential of the system, used in combination with a 5% sodiumhypochlorite, to reduce bacterial load in the root canal was investigated.It was tested on clean root canal systems. This was achieved by rootcanal chemomechanical instrumentation and debridement, exploitingthe well-known efficacy of standard irrigation protocols alternatingNaOCl and EDTA (20) in removing smear layer and organic debrisfrom the root canal system. It was hoped to suggest a possible improve-ment of disinfection because otherwise untreated niches may be open tothe hydrodynamic action of the activated solution.

The bacterial model used E. faecalis to test the efficacy of the irri-gation protocols under comparison. E. faecalis is not particularlydemanding from the nutritional standpoint, is resistant to extreme chal-lenges, and has frequently been isolated in cases of endodontic failure(32, 33) because it can penetrate the dentine tubules and escapechemomechanical treatment of the root canal system (34).

None of the protocols tested in this study completely eradicatedmicroorganisms. The results show a significant improvement of rootcanal disinfection in the EA 30 group in which 30 seconds of agitationwas applied compared with irrigation alone. For the EA 15 group, inwhich activation was only for 15 seconds, there was no difference versusirrigation alone. The EndoActivator driver was always used at themaximum power setting of 10,000 cpm; thus, comparative data con-cerning the efficacy of the device at lower power settings are not avail-able. This point remains to be investigated.

A recent study using an E. faecalis infection model (35) investi-gated the intracanal disinfection performance of three different irriga-tion techniques: conventional irrigation with NaviTip needles(Ultradent, South Jordan, UT), the EndoActivator system, and the En-doVac system (Discus Dental, Culver City, CA). The importance of che-momechanical preparation in reducing bacterial load was confirmed,but no significant differences were found in the three techniques, whichperformed similarly.

In conclusion, within the limits of this study, sonic activation for 30seconds of a 5% NaOCl solution appears to be slightly more efficaciousin disinfecting the root canal compared with a standard irrigationregimen with needles and also compared with sonic activation foronly 15 seconds. However, in the study conditions, the difference inbacterial load reduction among groups did not appear to be impressiveenough to allow clinical extrapolation of the results. In our opinion, theinteresting potential of sonic activation systems should be further inves-tigated through clinical studies aimed to establish a correct irrigationprotocol.

AcknowledgmentsThe authors gratefully thank Mario Alovisi and Francesco Co-

ero Borga for their valuable support.

References1. Kakehashi S, Stanley HR, Fitzgerald RJ. The effects of surgical exposures of dental

pulps in germ-free and conventionally laboratory rats. Oral Surg Oral Med OralPathol 1965;20:340–8.

2. Siqueira JF, Rocas IN. Clinical implications and microbiology of bacterial persis-tence after treatment procedures. J Endod 2008;34:1291–301.

3. Wong R. Conventional endodontic failure and retreatment. Dent Clin North Am2004;48:265–89.

4. Basmadijan-Charles CL, Farge P, Bourgeois DM, et al. Factors influencing the long-termresults of endodontic treatment: a review of the literature. Int Dent J 2002;52:81–6.

5. Sjogren U, Figdor D, Persson S, et al. Influence of infection at the time of root fillingon the outcome of endodontic treatment of teeth with apical periodontitis. Int EndodJ 1997;30:297–306.

1360 Pasqualini et al.

6. Peters LB, Wesselink PR. Periapical healing of endodontically treated teeth in oneand two visits obturated in the presence or absence of detectable microorganisms.Int Endod J 2002;35:660–7.

7. Chugal NM, Clive JM, Spangberg LS. A prognostic model for assessment of theoutcome of endodontic treatment: effect of biologic and diagnostic variables.Oral Surg Oral Med Oral Pathol 2001;91:342–52.

8. Sjogren U, Figdor D, Spangberg LS, et al. The antimicrobial effect ofcalcium hydroxide as a short-term intracanal dressing. Int Endod J 1991;24:119–25.

9. Peters LB, van Winkelhoff AJ, Buijs JF, et al. Effects of instrumentation, irrigation anddressing with calcium hydroxide on infection in pulpless teeth with periapical bonelesions. Int Endod J 2002;35:13–21.

10. Hess W. The anatomy of the root canals of the teeth of the permanent dentition: partI. New York: William Wood & Co; 1925:1–47.

11. Skidmore AE, Bjorndal AM. Root canal morphology of the human mandibular firstmolar. Oral Surg Oral Med Oral Pathol 1971;32:778–84.

12. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral MedOral Pathol 1984;58:589–99.

13. Bergenholtz G, Spangberg L. Controversies in Endodontics. Crit Rev Oral Biol Med2004;15:99–114.

14. Martin H. Ultrasonic disinfection of the root canal. Oral Surg Oral Med Oral Pathol1976;42:92–9.

15. Martin H, Cunningham W. Endosonics—the ultrasonic synergistic system ofendodontics. Endod Dent Traumatol 1985;1:201–6.

16. Archer R, Reader A, Nist R, et al. An in vivo evaluation of the efficacy of ultra-sound after step-back preparation in mandibular molars. J Endod 1992;18:549–52.

17. Calhoun G, Montgomery S. The effect of four instrumentation techniques on rootcanal shape. J Endod 1988;14:273–7.

18. Schulz-Bongert U, Weine FS, Schulz-Bongert J. Preparation of curved canals usinga combined hand-filing, ultrasonic technique. Compend Cont Educ Dent 1995;16:272–4.

19. Gu L, Kim JR, Ling J, et al. Review of contemporary irrigant agitation techniques anddevices. J Endod 2009;35:791–804.

20. Berutti E, Marini R, Angeretti A. Penetration ability of different irrigants into dentinaltubules. J Endod 1997;23:725–7.

21. Walton RE. Histologic evaluation of different methods of enlarging the pulp canalspace. J Endod 1976;2:304–11.

22. Peters OA. Current challenges and concepts in the preparation of root canal system.J Endod 2004;30:559–67.

23. Wu MK, Wesselink PR. A primary observation on the preparation and obturation ofoval canals. Int Endod J 2001;34:137–41.

24. Schafer E, Zapke K. A comparative scanning electron microscopic investigation ofthe efficacy of manual and automated instrumentation of root canals. J Endod2000;26:658–64.

25. Pitt WG. Removal of oral biofilm by sonic phenomena. Am J Dent 2005;18:345–52.

26. Sabins RA, Johnson JD, Hellstein JW. A comparison of the cleaning efficacy of short-term sonic and ultrasonic passive irrigation after hand instrumentation in molarroot canals. J Endod 2003;29:674–8.

27. Jensen S, Walker T, Hutter J, et al. Comparison of the cleaning efficacy of passivesonic activation and passive ultrasonic activation after hand instrumentation inmolar root canals. J Endod 1999;25:735–8.

28. De Gregorio C, Estevez R, Cisneros R, et al. Effect of EDTA, sonic and ultrasonic acti-vation on the penetration of sodium hypochlorite into simulated lateral canals: an invitro study. J Endod 2009;35:891–5.

29. Desai P, Himel V. Comparative safety of various intracanal irrigation systems. J En-dod 2009;35:545–9.

30. Caron G. Cleaning efficiency of the apical millimeters of curved canals using threedifferent modalities of irrigant activation: an SEM study [masters thesis]. Paris: ParisVII University; 2007.

31. Al-Jadaa A, Paque F, Attin T, et al. Acoustic hypochlorite activation in simulatedcurved canals. J Endod 2009;35:1408–11.

32. Siren EK, Haapasalo MP, Ranta K, et al. Microbiological findings and clinical treat-ment procedures in endodontic cases selected for microbiological investigation. IntEndod J 1997;30:91–5.

33. Kajaoglu G, Ørstavik D. Virulence factors of Enterococcus faecalis: relationship toendodontic disease. Crit Rev Oral Biol Med 2004;15:308–20.

34. Waltimo TMT, Ørstavik D, Siren EK, et al. In vitro yeast infection of human dentine. JEndod 2000;26:207–9.

35. Brito PRR, Souza LC, Machado de Oliveira JC, et al. Comparison of the effectivenessof three irrigation techniques in reducing intracanal Enterococcus faecalis popula-tions: an in vitro study. J Endod 2009;35:1422–7.



Effectiveness of Different Final Irrigant Activation Protocolson Smear Layer Removal in Curved CanalsGregory Caron, DDS,* Khan Nham, DDS,† Francois Bronnec, DDS,* and Pierre Machtou, DDS*

Abstract
Introduction: A final flush with chelating agents andantiseptic irrigating solutions is needed to remove thesmear layer. The improvement of these protocols ispossible by using specific delivery and agitation tech-niques. This study examined the effect of different finalirrigation regimens and methods of activation on smearlayer removal in curved canals after root canal instru-mentation. Methodology: Mesial root canals of 50 ex-tracted mandibular molars were prepared usingProTaper rotary files (Dentsply Maillefer, Ballaigues,Switzerland) and 3% NaOCl. Teeth were then allocatedto two control groups and four experimental groups (n= 10) for final irrigation as follows: no-activation group(final rinse with a 27-gauge needle and 17% EDTA/3%NaOCl), manual-dynamic activation group (final rinse17% EDTA/3% NaOCl + gutta-percha agitation),automated-dynamic activation group (final rinse 17%EDTA/3% NaOCl + RinsEndo [Durr Dental GmbH & CoKG, Bietigheim-Bissingen, Germany]), and sonic-activation group (final rinse 17% EDTA/3% NaOCl + En-doactivator [Advanced Endodontics, Santa Barbara,CA]). All mesial roots were split with a new approachto allow visualization of every third of the canal, partic-ularly the apical third. The samples were prepared forscanning electron microscopic observation to assessthe smear layer removal. Blind scoring was performedby two calibrated observers using a five-score scale.The differences in smear layer scores between the exper-imental groups were analyzed with the Kruskal-Wallistest and the Mann-Whitney U test. The level of signifi-cance was set at p = 0.05. Results: Very high levelsof root canal cleanliness (#score 3) were found foreach test group with activation. For the middle andapical third, the no-activation group was significantlyless effective than the three other activation groups (p< 0.05). The manual-dynamic activation group (finalrinse 17%EDTA/3%NaOCl + gutta-percha agitation)and the sonic-activation group (final rinse 17%EDTA/3%NaOCl + Endoactivator) showed significantly bettersmear layer removal (p < 0.05) in comparison with the
From the )Department of Endodontics and Restorative DentistENSCP, Paris, France.

Supported in part by the Association de formation en EndodontAddress requests for reprints to Dr Gregory Caron, 5 rue Garan

0099-2399/$0 - see front matterCopyright ª 2010 American Association of Endodontists.

doi:10.1016/j.joen.2010.03.037


other test groups in the apical third. Conclusion: Root canal cleanliness benefitsfrom solutions activation (especially sonic activation and manual-dynamic activation)in comparison with no activation during the final irrigation regimen. (J Endod2010;36:1361–1366)

Key WordsAutomated-dynamic activation, final irrigation, manual-dynamic activation, smearlayer, sonic activation

The ultimate goal of endodontic treatment is to control the microbial factor incomplex root canal anatomy, especially in the apical one third (1). This objective

is achieved by combining instrument-based preparation (manual or mechanical)with antiseptic irrigating solutions followed by three-dimensional obturation of theroot canal system. The gold standard irrigant is still sodium hypochlorite, which canbe associated with EDTA to offer bactericidal, solvent, and chelating actions all inone. This combination offsets the drawbacks of the instrument-based preparation,particularly the creation of debris (2) and the smear layer (3). The smear layer is poten-tially infected, and its removal allows more efficient penetration of intracanal medica-tions into the dentinal tubules and a better interface between the filling material and theroot canal walls (4).

The literature reports generally show that regardless of the instrumentation andirrigation techniques, the effectiveness of irrigating solutions remains limited in theapical one third of a prepared canal. This is particularly true for curved root canals(5, 6) and even on single-rooted teeth (7, 8). Therefore, the improvement ofirrigating protocols is essential during root canal treatment in order to achievebetter cleaning efficiency especially in the very complex apical area.

Currently, several techniques and systems are available and reported to improvefinal irrigation before obturation (9). Among these protocols passive ultrasonic irriga-tion has shown promising results on debris (10) and smear layer removal (11).

However, there are little published scientific data comparing the new andemerging devices and methods for disinfection with a conventional syringe irrigation.First, a fully tapered and apically trimmed nonstandardized gutta-percha master conecould be used in a well-prepared canal as a cost-effective mechanical agitator. A gentlepumping with short vertical strokes has been shown to promote disinfection (12, 13).Another recently released device, the RinsEndo irrigation system (Durr Dental GmbH &Co KG, Bietigheim-Bissingen, Germany), delivers solutions at a flush-through rate of 6.2mL/min using pressure-suction technology for intracanal activation (1.6 Hz). Thisdevice generates a mechanical action that is able to produce a hydrodynamic exchangecircuit (13, 14). Finally, the Endoactivator system (Advanced Endodontics, SantaBarbara, CA) has been purported to improve disinfection. This device uses

ry, School of Dentistry, University, Paris, France; and †Department of Surface Physico-Chemistry,

ie Appliquee.ciere 75006 Paris , France. E-mail address: [email protected].

Final Irrigant Activation Protocols for Smear Layer Removal in Curved Canals 1361


a cordless sonic handpiece to activate strong, highly flexible polymertips. Noncutting tips have tapers and terminal diameters that closelymatch the dimensions of the final root canal preparation (15). Mechan-ical oscillations are produced mainly at the tip of the activator witha frequency ranging from 1 to 10 kHz.
The purpose of this study was to assess smear layer removal effi-ciency after using gutta-percha master cone or RinsEndo or the Endoac-tivator in comparison to conventional final irrigation using a 27-gaugeneedle. The null hypothesis was that there is no difference in smear layerremoval between final irrigation with no activation and final irrigationwith activation.

Materials and MethodsThe study was conducted on 50 freshly extracted mature human

mandibular molar teeth with two separate mesial canals. None of theteeth had received restorative or endodontic treatment before extrac-tion. After extraction, the teeth were conserved in a solution of physio-logic saline to avoid damaging the pulp tissue and then stored at 4�C(16). Each individual tooth was then photographed and x-rayed to visu-alize the root canal anatomy and confirm that each canal curved at morethan 20� (17).

After cutting a four-wall access cavity, the full lengths of the mesio-buccal and mesiolingual canals were determined when a #08 K-type filecould be visualized at the apical foramen. Both the mesial and distalroots were sealed with melted wax to close the apical foramen (18).The aim was to prevent the irrigants from escaping through the apexin order to simulate in vivo conditions (19). The two mesial root canalswere prepared using the Protaper Universal rotary files system (Dents-ply Maillefer, Ballaigues, Switzerland) following the protocol describedby Machtou and Ruddle (20) in which the apical one-third taper of thefinished preparation is approximately 10%. All the canals wereprepared so that the finished size of each apical foramen rangedbetween 0.20 mm and 0.30 mm in diameter. A medium nonstandar-dized gutta-percha master cone (Henry Schein, Melville, NY) was fittedin each canal to the full working length, and then the tooth was x-rayed.After each instrument, the expending preparation was flooded bypassively irrigating 0.5 mL of 3% sodium hypochlorite (Parcan; Septo-dont, Saint-Maur-des-Fosses, France) into the canal using a 27-gaugeneedle (Monoject; Tyco Kendall, Hampshire, UK) loosely inserted asfar as possible without binding. A #10 K-type patency file was used tomaintain apical patency and move debris into suspension followed byflushing the canal again with 0.5 mL of fresh irrigant. All procedureswere performed by the same operator (GC).

The teeth were randomly divided into four experimental groups (n= 10) and two control groups. The negative controls (n = 5) receivedno final irrigation regimen after fitting of the master cone. The positivecontrols (n = 5) were immersed for 5 minutes in a bath of 17% EDTAfollowed by an immersion for 5 minutes in a bath of 3% NaOCl after thesplitting process.

Final Irrigation ProtocolsNo-activation Group. After suctioning away the intracanal surplusof NaOCl with the 27-gauge needle, 1 mL of 17% EDTA (Largal Ultra;Septodont, Saint-Maur-des-Fosses, France) was flushed into each canaland was left in place for 1 minute per canal. All canals were then flushedwith 3 mL of 3% sodium hypochlorite, which was left in place for 30seconds per canal.

Manual-dynamic Activation Group. After suctioning away theintracanal surplus of NaOCl with the 27-gauge needle, 1 mL of 17%EDTA was flushed into each canal. This solution was activated by usinga gutta-percha cone as previously described for 1 minute in each canal.

1362 Caron et al.

The frequency of activation used was 100 push-pull strokes per minute(12). The canals were then flushed with 3 mL of a 3% solution ofsodium hypochlorite. This solution was then activated for 30 secondsper canal using the pumping master cone method.

Automated-dynamic Activation Group. After optimallypreparing the canal, surplus NaOCl was suctioned away with the27-gauge needle. Following the manufacturer’s instructions, each canalwas flushed with 1 mL of 17% EDTA using the RinsEndo system for 1minute per canal. Each canal was then flushed with 3 mL of a 3% solu-tion of sodium hypochlorite delivered via the RinsEndo system for 30seconds in each canal.

Sonic-activation Group. After optimally preparing the canal,surplus NaOCl was suctioned away with the 27-gauge needle. Each canalwas then irrigated with 1 mL of 17% EDTA using the 27-gauge needle.This intracanal solution was activated with either a red (25/04) or blue(35/04) EndoActivator tip at a speed of 10 kHz for 1 minute per canal.Each canal was then flushed with 3 mL of 3% sodium hypochlorite. Thissolution was then activated using either the red or blue EndoActivator tipfor 30 seconds per canal.

After activation, the action of the sodium hypochlorite was stoppedby syringing in 3 mL of physiologic saline solution per canal (ie, 6 mL foreach tooth in the four test groups). All the samples were then placed ina solution of physiologic saline and stored at 4�C until proceeding withthe sectioning protocol.

Sectioning of the Teeth and Preparation for SEMThe teeth were sectioned in two halves; only the mesial root was

kept for further study. Two horizontal grooves were made using a Friosdiamond-cutting disk (Microsaw; Dentsply Friadent, Mannheim,Germany) mounted on a surgical dental handpiece to separate themesial root into thirds (the apical third, the middle third, and thecoronal third). This step was performed using a surgical microscope.Colored gutta-percha cones were fitted and used as markers to bestgauge groove depth. The objective was to avoid any intrusion of thecutting disc into the canals, which would pollute the samples by splat-tering cutting debris into the root canal system. To avoid any contami-nation, coronal thirds were discarded because there was a bigger gapbetween the gutta-percha markers and the prepared walls of the canal.This gap compromised vision and increased the possibility of the cuttingdisc inadvertently introducing debris into this region of the canal.

The apical and middle one thirds of the canal were thensectioned in the longitudinal plane with a precision diamond bur(889 Model; Komet, Paris, France). A continuous supply of air wasdelivered to improve vision and cutting precision, which eliminatedthe potential of introducing debris into this region of the canal.Each third was vertically split by applying slight pressure to an enamelchisel into the longitudinal groove. Each sample was dehydrated ingraded series of ethanol solutions, critical point dried, coated withgold, and viewed with a scanning electron microscope (HitachiS2500,Verrieres-le-buisson, France) at 15 kV (Fig. 1).

SEM Evaluation and Statistical AnalysisEach fragment was first viewed at low magnification (�30) by the

operator (GC) and another trained dentist with SEM studies (KN) inorder to gain an overview of the sample. Image acquisition on themost typical zones of the sample was performed at a magnification of�1,000 to assess the presence of smear layer. The images were blindlyassessed by two practitioners with no inside knowledge of the operativeprocedures and who were fully conversant with qualitative analysis onroot canal images produced by scanning electron microscopy (PM,FB). Analysis began using the scale described in Hulsmann et al


Figure 1. Four different apical sample fragments highlighting the reproducibility and preservation of the apical one-third samples using our experimental protocol.Magnification: �30.


(21), but the significant lack of sensitivity in the best scores promptedus to refine the system, as follows (Fig. 2): score 1: no smear layer anddentinal tubules open, score 2: small amounts of scattered smear layersand dentinal tubules open, score 3: thin smear layer and dentinaltubules partially open (characteristic image of crescent), score 4:partial covering with a thick smear layer, and score 5: total coveringwith a thick smear layer.

First, the full set of samples was independently evaluated by twoobservers (PM and FB). If there were conflicting results between thesetwo observers, then a final evaluation was made with the lower scorechosen every time. Nonparametric data were analyzed by using theKruskal-Wallis test and the Mann-Whitney rank sum test for pairwisecomparisons. The significance level for all statistical analyses was set

JOE — Volume 36, Number 8, August 2010 Fin

at a = 0.05. All statistical analyses were performed with the SPSS forWindows 12.0 software package (SPSS Inc, Chicago, IL).

ResultsAfter consensus was reached for each group, mean scores for

smear layer removal in the apical third and the middle third were listed(Table 1). The full set of negative control samples scored a 5 witha complete covering of a thick smear layer. All the positive controlsamples scored a 1 with no visible smear layer. In the middle third,comparisons between each group showed a statistically significantdifference (p < 0.005). When comparing each test group, only the‘‘no-activation group’’ scored a 3 with a thin smear layer and showed

al Irrigant Activation Protocols for Smear Layer Removal in Curved Canals 1363

Figure 2. A new fine-tuned scale used to evaluate sample cleanliness. Magnification: �1,000.


a statistical difference with the three other activation groups (p < 0.05)in which smear layer scores were always inferior to 3.

Comparisons between each group showed a statistically significantdifference (p < 0.005) in the apical third. When comparing each test

1364 Caron et al.

group, the ‘‘sonic group’’ (final irrigation + Endoactivator) showeda statistical difference compared with all the test groups (p < 0.05).The exception was the ‘‘manual-dynamic activation group’’ (final irriga-tion + gutta-percha agitation) in which no statistical diffence was


TABLE 1. Mean Score � Standard Deviation Comparing the Smear Layeramong the Four Final Irrigation Regimens in the Apical and Middle Thirds

Group N Apical third Middle third

Final irrigation withoutactivation

10 3.16 �0.958 3.47�0.874

Final irrigation +Master cone

10 2.21 �1.032 2.05 �1.05

Final irrigation +Rinsendo

10 3 �1.414 2.5 �1

Final irrigation +Endoactivator

10 1.75 �0.55 1.88 �0.857


detected. However, only the Endoactivator group showed smear layerremoval lower than a score 2, which equates to a high level of cleanli-ness in all apical one-third samples.

DiscussionThe endodontic community is now unanimous concerning the

positive benefit of irrigation during the root canal preparation phase(22). The chemomechanical preparation should ideally result in a fullycleansed and disinfected root canal system. Literature has shown thatapical enlargement and deeper positioning of the irrigation needleare required to clean the apical third (5, 23, 24). However, very littledata exist regarding smear layer removal, especially in molar teethexhibiting curved canals where emphasis was placed on preparinga fully tapered and well-shaped canal with narrow apical diameters.

Only one recent SEM study by Khademi et al (25) has shown theimportance of the canal taper in curved canals. When the root canaltaper was 0.06 mm/mm, there was comparable smear layer removalwhen the apical diameters were between 0.30 mm and 0.40 mm. Ithas been shown that for more narrow apical diameters the taper needsto be increased to 0.10 mm/mm in order to eliminate a maximumamount of debris (26, 27).

A recent study on curved roots has shown a correlation betweencreating sufficient taper and propagation improvement of irrigants inroot canal during the shaping process (28). In the present study, allthe shaping procedures complied with the criteria of tapered canalsand maintaining apical foramen as small as practical. In funnel-shaped canals, both the tapered gutta-percha master cones and thetapered EndoActivator tips, when activated, provided cleaner resultsthan syringe delivery systems (final irrigation with no activation and Rin-sEndo).

In this study, the four-wall access cavity provided a strategic reser-voir to hold a more effective volume of irrigant for exchange during acti-vation. This is in direct contrast with many studies in which the crownhad been removed from the samples. Desirably, the irrigating solutionsare apically exchanged each time the activator system is inserted into thecanal. When the activator tip moves toward length, the reagent is dis-placed. When the activator is partially withdrawn, there is an effectiveexchange of solution into the apical one third of the canal. The efficiencyof this hydrodynamic circuit is further enhanced when combined withsonic oscillating movements. A pumping action synergisticallycombined with mechanical agitation explains the better results achievedwith the EndoActivator. Recently, Uroz-Torres et al (29) found no differ-ence in smear layer removal between the Endoactivator and conven-tional Max-I-Probe (Dentsply Rinn, Elgin, IL) irrigation using NaOCland EDTA. The difference in results is attributed to differences in theactivation protocols used. In our present study, a second activationwas performed after the final flush with NaOCl, and all apical one-third samples were shaped to 10% to enhance irrigation exchangeand efficiency (30).

JOE — Volume 36, Number 8, August 2010 Fin

So far, activation of irrigants without concomitant instrumentationof the root canal walls has been defined as passive ultrasonic irrigation(PUI) (9). PUI has been shown to be effective in removing debris (10)and smear layer in straight canals (31). Nevertheless, there are conflict-ing results about the performance of PUI in terms of smear layer elim-ination (9, 32). Even when ultrasonically driven metal canula orinstruments are precurved, there are obvious considerations andlimitations using these devices in curved canals and to length (9,33). There is always a risk of touching the walls, which wouldautomatically trigger the formation of a new undesirable smear layer.Of greater concern, contacting a dentinal wall with an activatedultrasonic instrument invites iatrogenic events. This should becontrasted with the nonmetal surfaces of the activation systems usedin this study (gutta-percha or polymer), which cannot generatea smear layer, internal ledge, or external transportation of the foramen.

It should be noted that all of the samples tested by manual-dynamicirrigation or by sonic activation yielded very little smear layer. This levelof cleanliness is an important finding because it was achieved on curvedcanals using activation systems other than metal ultrasonic files.Although these systems generate lower frequencies compared withthe ultrasonically driven files (25-30 kHz), the vertical-stroke pumpingmotion used as part of the protocol promotes dynamic coronoapicalcirculation of the irrigating solutions. The benefit of irrigant renewaland activation needs to be researched in greater detail, comparingthis irrigation dynamic with the streaming pattern observed with PUI.

Our experimental model used final irrigation times (34) and finalirrigation volumes that proved effective and efficient (35) while at thesame time avoiding peritubular and intertubular erosion. However,the timeframe given in the protocol may not be long enough for a systemsuch as RinsEndo, which may require more time to achieve the samescores as the other activation systems. During the shaping procedures,the volume of 1 mL of irrigant dispensed after the use of each file wassufficient to get a clear solution inside the pulp chamber without visibledebris.

The data gathered through this experiment are transposable toclinical practice because the experimental model proposed closelymirrors the actual conditions encountered in routine molar treatment,but results should be interpreted with precaution because of the largestandard deviation observed. One possible explanation could be thechallenge of standardizing activation procedures in complex anatomyof mandibular molars where the two mesial canals commonly commu-nicate along their lengths.

ConclusionThe experimental design implemented in this study prompted the

following observations: the activation of irrigating solutions yieldedcleaner canals compared with no activation, and a tapered activatorthat closely adapts to the dimensions of a shaped canal is the most effec-tive (ie, the master gutta-percha cone and Endoactivator). Furtherinvestigations will be required to confirm this preliminary data, partic-ularly in terms of biofilm removal and apical disinfection results.

AcknowledgmentsThe authors are grateful to Durr Dental for loaning the Rin-

sEndo device used in this experiment. The authors thank Dr C.Ruddle and Advanced Endodontics for loaning the Endoactivatordevice used in this experiment. The authors also thank Septodontfor providing the irrigating solutions Parcan and Largal Ultraused in this experiment.

al Irrigant Activation Protocols for Smear Layer Removal in Curved Canals 1365


References1. Nair PNR, Henry S, Cano V, et al. Microbial status of apical root canal system of

human mandibular first molars with primary apical periodontitis after ‘‘one-visit’’endodontic treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;99:231–52.

2. Guppy DR, Curtis RV, Pitt Ford TR. Dentine chips produced by nickel-titanium rotaryinstruments. Endod Dent Traumatol 2000;16:258–64.

3. Jeon IS, Spangberg LSW, Yoon TC, et al. Smear layer production by 3 rotary reamerswith different cutting blade designs in straight root canals: a scanning electronmicroscopic study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:601–7.

4. Torabinejad M, Handysides R, Khademi AA, et al. Clinical implications of the smearlayer in endodontics: a review. Oral Surg Oral Med Oral Pathol Oral Radiol Endod2002;94:658–66.

5. Sedgley CM, Nagel AC, Hall D, et al. Influence of irrigant needle depth in removingbioluminescent bacteria inoculated into instrumented root canals using real-timeimaging in vitro. Int Endod J 2005;38:97–104.

6. Senia ES, Marshall FJ, Rosen S. The solvent action of sodium hypochlorite on pulptissue of extracted teeth. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1971;31:96–103.

7. Gambarini G, Laszkiewicz. A scanning electron microscopic study of debris andsmear layer remaining following use of GT rotary instruments. Int Endod J 2002;35:422–7.

8. Yamada RS, Armas A, Goldman M, et al. A scanning electron microscopic compar-ison of a high volume final flush with several irrigating solutions: part 3. J Endod1983;9:137–42.

9. Gu LS, Kim JR, Ling J, et al. Review of contemporary irrigant agitation techniques anddevices. J Endod 2009;35:791–804.

10. Jiang LM, Verhaagen B, Versluis M, et al. Evaluation of a sonic device designed toactivate irrigant in the root canal. J Endod 2010;36:143–6.

11. Paragliola R, Franco V, Fabiani C, et al. Final rinse optimization: influence ofdifferent agitation protocols. J Endod 2010;36:282–5.

12. Huang T-Y, Gulabivala K, Ng YL. A bio-molecular film ex-vivo model to evaluate theinfluence of canal dimensions and irrigation variables on the efficacy of irrigation.Int Endod J 2008;41:60–71.

13. McGill S, Gulabivala K, Mordan N, et al. The efficacy of dynamic irrigation usinga commercially available system (RinsEndo�) determined by removal of a collagen‘biomolecular film’ from an ex vivo model. Int Endod J 2008;42:602–8.

14. Hauser V, Braun A, Frentzen M. Penetration depth of a dye marker into dentine usinga novel hydrodynamic system (RinsEndo�). Int Endod J 2007;40. 664–52.

15. Ruddle CJ. Endodontic disinfection: tsunami irrigation. Endod Pract 2008;11:7–15.16. Gambarini G. Shaping and cleaning the root canal system: a scanning electron

microscopic evaluation of a new instrumentation and irrigation technique. J Endod1999;25:800–3.

17. Schneider SW. A comparison of canal preparations in straight and curved canals.Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1971;32:271–5.

1366 Caron et al.

18. Tay FR, Gu LS, Schoeffel GJ, et al. Effect of vapor lock on root canal debridement byusing a side-vented needle for positive-pressure irrigant delivery. J Endod 2010;36:745–50.

19. Schoeffel GJ. The EndoVac method of endodontic irrigation: safety first. Dent Today2007;26:92–6.

20. Machtou P, Ruddle CJ. Advancements in the design of endodontic instruments forroot canal preparation. Alpha Omegan 2004;97:8–15.

21. Hulsmann M, Rummelin C, Schafers F. Root canal cleanliness after preparation withdifferent endodontic handpieces and hand instruments: a comparative SEM. J Endod1997;23:301–6.

22. Bystrom A, Sundqvist G. Bacteriologic evaluation of the effect of 0.5 percent sodiumhypochlorite in endodontic therapy. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 1983;55:307–12.

23. Chow TW. Mechanical effectiveness of root canal irrigation. J Endod 1983;9:475–9.

24. Falk KW, Sedgley CM. The influence of preparation size on the mechanical efficacy ofroot canal irrigation in vitro. J Endod 2005;31:742–5.

25. Khademi A, Yazdizadeh M, Feizianfard M. Determination of the minimum instrumen-tation size for penetration of irrigants to the apical third of root canal systems. J En-dod 2006;32:417–20.

26. Abou-Rass M, Piccinino MV. The effectiveness of four clinical irrigation methods onthe removal of root canal debris. Oral Surg Oral Med Oral Pathol Oral Radiol Endod1982;54:323–8.

27. Albrecht LJ, Baumgartner JC, Marshall JD. Evaluation of apical debris removal usingvarious sizes and tapers of ProFile GT Files. J Endod 2004;30:425–8.

28. Bronnec F, Bouillaguet S, Machtou P. Ex vivo assessment of irrigant penetration andrenewal during the cleaning and shaping of root canals: a digital subtraction radio-graphic study. Int Endod J 2010;43:275–82.

29. Uroz-Torres D, Gonzales-Rodriguez MP, Ferre-Luque CM. Effectiveness of the En-doactivator system in removing the smear layer after root canal instrumentation.J Endod 2010;36:308–11.

30. Yana Y. An in vivo comparative study of the penetration of sodium hypochlorite inroot canal systems during cleaning and shaping procedures using the B.U. techniqueand sonic instrumentation [masters thesis]. Boston, MA: Boston University; 1989.

31. Kuah HG, Lui JN, Tseng PS, et al. The effect of EDTA with and without ultrasonics onremoval of the smear layer. J Endod 2009;35:393–6.

32. Mayer BE, Peters OA, Barbakow F. Effects of rotary instruments and ultrasonicirri-gation on debris and smear layer scores: a scanning electron microscopic study. IntEndod J 2002;35:582–9.

33. Ahmad M, Pitt Ford TR, Crum LA. Ultrasonic debridement of root canals: acousticstreaming and its possible role. J Endod 1992;13:490–9.

34. Saito K, Webb TD, Imamura GM, et al. Effect of shortened irrigation times with 17%ethylene diamine tetra-acetic acid on smear layer removal after rotary canal instru-mentation. J Endod 2008;34:1011–4.

35. Crumpton BJ, Goodell GG, McClanahan SB. Effects on smear layer and debrisremoval with varying volumes of 17% REDTA after rotary instrumentation. J Endod2005;31:536–8.



In Vitro Comparisons of Debris Removal of theEndoActivatorTM System, the F FileTM, Ultrasonic Irrigation,and NaOCl Irrigation Alone after Hand-rotaryInstrumentation in Human Mandibular MolarsSteven L. Klyn, DDS,* Timothy C. Kirkpatrick, DDS,

†and Richard E. Rutledge, DDS

†

Abstract
Introduction: The purpose of this in vitro study wasto compare the debris removal efficacy of the EndoActi-vatorTM system, the F fileTM, ultrasonic irrigation, or 6%NaOCl irrigation alone in human mandibular molarsafter hand-rotary instrumentation. Methods: A custombrass cube (K-Kube) was used to create a sealed canalsystem, allowing each tooth to serve as its own control.Forty extracted mandibular molars were randomlydivided into 4 equal experimental groups. Each toothwas mounted, sectioned at 1, 3, and 5 mm from theworking length, and then reassembled into the K-Kube, and the mesial roots were similarly prepared byusing hand-rotary instrumentation. For final debride-ment, group 1 used F file for 30 seconds, group 2used EndoActivator system for 30 seconds, group 3used ultrasonic irrigation for 30 seconds, and group4 used irrigation with 6% NaOCl within 1 mm ofworking length. All groups received a final irrigationwith 6% NaOCl in each canal. Specimens were evalu-ated at 1, 3, and 5 mm from the working length forcleanliness by capturing a digital image with a stereomi-croscope. All specimens had the percent cleanliness foreach canal and isthmus calculated both before and afterfinal debridement. Statistical analysis was completedby using a repeated-measures analysis of variancewith Tukey post hoc tests. Results and Conclusions:The results showed no statistically significant differencein canal or isthmus cleanliness among the 4 groups, butthere was a statistically significant difference (P < .001)in canal cleanliness between the 1-mm level versus the3-mm and 5-mm levels for all of the groups. (J Endod2010;36:1367–1371)
Key WordsEndoActivator, F file, irrigation, K-Kube, ultrasonics

From the *Department of Endodontics, 10th Dental Squadron, UCenter, Lackland Air Force Base, Texas.

Address requests for reprints to Timothy C. Kirkpatrick, DDS, ProSGDTN, 2450 Pepperrell St, Lackland AFB, TX 78236. E-mail addres0099-2399/$0 - see front matter

Copyright ª 2010 Published by Elsevier Inc. on behalf of the Adoi:10.1016/j.joen.2010.03.022


The prevention or treatment of apical periodontitis is the ultimate goal of endodontictherapy (1). Complete debridement of the root canal system is complicated by the

presence of a complex system of isthmuses, accessory canals, fins, and deltas that canprovide ideal locations for harboring debris (2). The residual debris within the canalsystem can be composed of bacteria, other microorganisms and their by-products, vitaland necrotic pulp tissue, smear layer, and biofilm. Although mechanical instrumenta-tion and the use of irrigants within the canal have shown effectiveness, complete clean-liness of these inaccessible areas is difficult to achieve (3–6). Studies have shown thatincomplete canal debridement can lead to a decrease in endodontic success (7, 8).

To aid in the removal of debris and the disinfection of the canal system, the use ofvarious intracanal irrigants and techniques has been advocated (9, 10). No singlesolution or technique has been found to achieve complete canal debridement, butthe use of ultrasonics as an adjunct to cleaning and shaping has shown increasedcanal cleanliness. Archer et al (11) compared step-back instrumentation alone withstep-back instrumentation followed by ultrasonic irrigation and showed significantlycleaner canals and isthmuses at 1, 2, and 3 mm from the apex with the use of ultra-sonics. Other studies have confirmed the effectiveness of ultrasonic irrigation on theremoval of debris from within the canals and isthmuses (12–16).

Ultrasonic irrigation has the potential for continued prepping or cutting of thecanal walls during its use after cleaning and shaping. Canal deviation, apical zipping,and even root perforations can occur while using an ultrasonically activated file withina curved canal during irrigation. These potential problems have led researchers to thedevelopment of new techniques based on ultrasonic technology. By using a modifiedultrasonically charged irrigation needle, Burleson et al (17) showed a significantimprovement in canal cleanliness at the apex and within the isthmuses of the mesialroots of mandibular molars without undue canal deviation.

Two systems to aid in the debridement of the canal system were recently intro-duced, the EndoActivator System (Advanced Endodontics, Santa Barbara, CA) andthe F file (Plastic Endo, Buffalo Grove, IL). Both systems use noncutting plastic or poly-mer tips to enhance root canal debridement after instrumentation. The tips do notactively engage the dentin walls, thus preventing further enlargement of the canals.The purpose of this in vitro study was to compare the effectiveness of the F file, theEndoActivator System, ultrasonic irrigation, and 6% NaOCl irrigation alone in removingcanal and isthmus debris in human mandibular molars.

nited States Air Force Academy, Colorado; and †Department of Endodontics, Wilford Hall Medical

gram Director, Endodontics Residency, Wilford Hall Medical Center, 59th Dental Training Squadron/s: [email protected].

merican Association of Endodontists.

Comparison of Debris Removal by 4 Different Systems 1367



Materials and MethodsSpecimen Preparation

Forty extracted human mandibular molars with mesial root curva-tures less than 25 degrees were selected and stored in 0.5%chloramine-T before use. The buccal and lingual cusps were flattenedto provide a reproducible reference point during instrumentation(Fig. 1C).

After standard access openings were made, working length (WL)was determined by inserting a #10 Flex-R file (Miltex, York, PA) untilthe tip of the file was visible at the apical foramen and subtracting 1mm. Each mesial canal was instrumented until a Profile GT #20/.06file (Dentsply-Tulsa Dental, York, PA) could be inserted to the WL. Irri-gation was with saline only. The access opening was sealed with a moistcotton pellet and Cavit (3M ESPE, St Paul, MN), making sure the Cavitextended into the orifice of the distal canal. The distal root was ampu-tated, and Triad gel (Dentsply Trubyte, York, PA) was used to seal theapical foramina on the mesial roots and the distal root orifice to preventmounting resin from entering these areas.

To allow more precise visualization of the root apex duringsectioning, the apical 2 mm of the mesial roots was dipped in methyleneblue. The teeth were then embedded into a custom-made metal cube (K-Kube) at the level of the cementoenamel junction by using VariKleerresin (Buehler, Lake Bluff, IL) (Fig. 1A, B, E, F). Each specimen wascured by using a pressure pot of warm water at 20 psi for 30 minutes.After the resin had set, the embedded specimens were removed from thecube and stored in 100% humidity.

Specimen SectioningThe mounted specimens were sectioned at 2, 4, and 6 mm from

the apex (Fig. 1D) of the root by using an Isomet low-speed saw witha 0.30-mm-thick diamond blade (Buehler). The blade was irrigatedwith Isocut Plus Fluid (Buehler) and water according to the manufac-turer’s recommendations.

Figure 1. Specimen preparation. (A) K-Kube disassembled; (B) K-Kube assembled;measured; (E) sectioned specimen partially reassembled in K-Kube; (F) specimen fuin color online at www.aae.org/joe/.)

1368 Klyn et al.

Three 2-mm-thick sections were used for evaluation and scoring:section 1, apex to 1 mm coronal to WL; section 2, 1–3 mm coronal toWL; and section 3, 3–5 mm coronal to WL.

Canal PreparationAfter sectioning, the specimens were reassembled into the K-Kube,

and all external hex bolts were firmly tightened. The Cavit and cottonpellet were removed, and a hand file was used to verify WL and properassembly. Following coronal flaring with Gates Glidden drills (Dentsply,York, PA), the canals were prepared with ProFile 0.04 rotary files(Dentsply-Tulsa Dental) using a crown-down technique to a masterapical file size #40. Between each rotary file, 0.5 mL of 6% NaOClwas used to irrigate each canal by using a 30-gauge Max-i-ProbeTM

(Dentsply). The Max-i-Probe was inserted until resistance was feltand then backed up approximately 0.5 mm. After final instrumentation,each canal was irrigated with 2 mL of 6% NaOCl and dried with a capil-lary tip (Ultradent, South Jordan, UT) and paper points. Each canal wasthen irrigated with 2 mL of 17% ethylenediaminetetraacetic acid anddried before final irrigation with 2 mL 6% NaOCl. All canals werethen dried and the access was sealed with a moist cotton pellet and Cavit.

Method of EvaluationEach specimen was then disassembled and images of the coronal

aspect of each section were made by using a digital camera (OlympusDP71; Olympus, Tokyo, Japan) attached to a stereomicroscope(Olympus SZX16) at the highest magnification to allow complete viewof the canals and isthmus. The full color images were viewed on a Cintiq21 UX (Wacom Co Ltd, Saitama, Japan) monitor, and an interactive penwas used to trace the outline of the root canal, isthmuses, and remainingdebris (Fig. 2). Debris was defined as any material present on the canalwalls and in the canal lumen or isthmus. The software program Image J(National Institutes of Health, v1.39a) was used to calculate the area ofthe root canals, isthmuses, and amount of debris present. In order to

(C) specimen preparation; (D) specimen mounted in resin and 2-mm sectionslly mounted in K-Kube with access for instrumentation. (This figure is available



Figure 2. Digital images of a specimen section at the 5-mm level demonstrating canal and isthmus debris. (A) Initial canal access only; (B) post cleaning andshaping; (C) post experimental treatment. An interactive pen was used to trace the outline of the canals, isthmuses, and remaining debris as shown in (D). Image Jsoftware was used to compute the canal cleanliness. (This figure is available in color online at www.aae.org/joe/.)


calculate the canal cleanliness, the area of remaining debris was dividedby the total area of the canal or isthmus to yield a percentage. Thepercentage of canal debris present was subtracted from 1 to determinethe percent of canal cleanliness.

Final DebridementThe specimen sections were reassembled into the K-Kube, and

each tooth was randomly assigned to 1 of 4 experimental groups.Each group was treated according to the manufacturer’s directionsand then dried with a capillary tip.

In group 1, the canals and chamber were filled with 2 mL of 6%NaOCl, and the F file (size #20/0.04 taper) was passively insertedinto the canal with an electric slow-speed handpiece set at 600 rpmand a torque of 132 g/cm. The file was circumferentially worked alongthe dentinal walls with a cyclic axial motion (up and down) for30 seconds.

In group 2, the canals and chamber were filled with 2 mL of 6%NaOCl before treatment. The EndoActivator sonic handpiece was setat 10,000 cpm, and a size #15/0.02 taper activator tip was passively in-serted to within 2 mm of the WL and used in a pumping action to movethe EndoActivator tip in short, 2–3 mm vertical strokes for 30 seconds.

In group 3, the canals and chamber were filled with 2 mL of 6%NaOCl, and a 30K PEC Endosonic size #20 file (Dentsply) was passivelyinserted into the canals. The file was circumferentially worked along thedentinal walls with a cyclic axial motion (up and down) for 30 secondsusing an ultrasonic unit set at a consistent low power and waterirrigation.

In group 4, each specimen had a Max-i-Probe inserted to within 1mm of WL, and both the canals and chamber were filled and irrigatedwith 2 mL (1 mL in each canal) of 6% NaOCl.


All groups received a final irrigation with 2 mL 6% NaOCl in eachcanal. After drying with a capillary tip and paper points, the specimenswere removed from the K-Kube, disassembled, and evaluated as beforefor debris. Percent cleanliness was calculated for each canal andisthmus immediately after instrumentation and after using the experi-mental final debridement technique. The percent change in debriswas statistically analyzed by using a repeated-measures analysis ofvariance with Tukey post hoc tests (significance level, P < .05).

ResultsA comparison of canal and isthmus cleanliness is shown in Fig. 3.

There was no statistically significant difference in canal or isthmuscleanliness with the F file, EndoActivator, or ultrasonics when used asan adjunct to aid in canal debridement compared with irrigation alonewith NaOCl. All 4 treatment groups demonstrated a statistically signifi-cant difference (P < .001) in canal cleanliness at 3 and 5 mm fromthe WL (>99.4%) than at 1 mm from the WL (>97.3%).

DiscussionThis study introduced the custom-designed K-Kube (Fig. 1A, B),

which was based on the technique of Bramante et al (18) with the addi-tion of a compression component. Compressing the parallel sectionswithin each specimen effectively eliminated the 0.3-mm kerf createdby each saw blade cut. The K-Kube allows a tooth to be sectionedand then reassembled to recreate an intact root canal system. Thisaffords the opportunity to evaluate not only canal and isthmus anatomybut also the effect of irrigation on residual debris while using each toothas its own control. Thus, the K-Kube enabled the evaluation of experi-mental irrigation adjuncts in a sectioned tooth for the first time.



Mean Percentage of Canal Cleanliness Comparing Experimental Groups (+1 SD)

85.0%87.0%89.0%91.0%93.0%95.0%97.0%99.0%

101.0%103.0%105.0%

1 mm 3 mm 5 mm 1 mm 3 mm 5 mm1 mm 3 mm 5 mm 1 mm 3 mm 5 mm 1 mm 3 mm 5 mm 1 mm 3 mm 5 mm

1 mm 3 mm 5 mm 1 mm 3 mm 5 mm 1 mm 3 mm 5 mm 1 mm 3 mm 5 mm1 mm 3 mm 5 mm 1 mm 3 mm 5 mm

Post ExperimentalPost Clean & Shape

Post ExperimentalPost Clean & Shape

F™ FileEndoActivator™

UltrasonicsMax-i-Probe™

BA

Mean Percentage of Isthmus Cleanliness Comparing Experimental Groups (+1 SD)

60.0%

70.0%

80.0%

90.0%

100.0%

110.0%

F™ FileEndoActivator™

UltrasonicsMax-i-Probe™

F™ File EndoActivator™ Ultrasonics Max-i-Probe™

F™ File EndoActivator™ Ultrasonics Max-i-Probe™

C Mean Percentage of Isthmus Cleanliness Comparing Levels (+1 SD)

60.0%

70.0%

80.0%

90.0%

100.0%

110.0%

Post Clean & Shape Post Experimental

D

Mean Percentage of Canal Cleanliness

Comparing Levels (+1 SD)

85.0%

90.0%

95.0%

100.0%

105.0%

Post Clean & Shape Post Experimental

**

* * ** * *

Figure 3. Comparison of canal and isthmus cleanliness post cleaning and shaping versus post experimental treatment. (A) Percentage of canal cleanliness of eachexperimental group at the 1-, 3-, and 5-mm levels. (B) Percentage of canal cleanliness of all 4 experimental groups individually at the 1-, 3-, and 5-mm levels. (C)Percentage of isthmus cleanliness of each experimental group at the 1-, 3-, and 5-mm levels. (D) Percentage of isthmus cleanliness of all 4 experimental groupsindividually at the 1-, 3-, and 5-mm levels. *Statistically significant differences between levels (P < .001).


In this study, conventional cleaning and shaping alone resulted ingreater than 94% canal cleanliness and greater than 74% isthmus clean-liness. The addition of an irrigation adjunct resulted in improved canaland isthmus cleanliness at all levels, regardless of the technique used.Most of the remaining debris was found in the apical 1 mm of the canalor isthmus. These results were consistent with previous studies inregards to canal cleanliness but were better than previous studies inregards to isthmus cleanliness after conventional cleaning and shapingalone (11–13, 15, 17, 19).

There was a higher overall standard deviation concerning isthmuscleanliness as compared with canal cleanliness. This was probably dueto the variation in isthmus width not only within each tooth but alsobetween 2 different samples. The narrow isthmuses consistently demon-strated the most residual debris both after cleaning and shaping andafter treatment with the experimental irrigation adjuncts.

Recent studies with various ultrasonic, sonic, and passive ultra-sonic irrigation devices and techniques have shown improved tissueremoval (20), more vigorous irrigation of lateral canals (21), and addi-tional removal of canal bacteria (22). Other recent studies have alsoshown that the use of various irrigating solutions not only improvessmear layer removal (23), but alters the physicochemical propertiesof dentin that influence the adherence and biofilm formation of Entero-coccus faecalis to dentin (24). In this study, there was no statistically

1370 Klyn et al.

significant difference in canal or isthmus cleanliness with the F file,EndoActivator, or ultrasonics when used as an adjunct to aid in canaldebridement compared with irrigation alone with NaOCl. The additionalirrigation with NaOCl produced the same statistical improvement incanal and isthmus cleanliness and might best be explained by the factthat a 30-gauge Max-i-Probe can be inserted to WL when the canal isprepared to an ISO size 40 (25). This ‘‘needle deep’’ irrigation wasmore important in improving canal and isthmus cleanliness than theuse of any of the adjunct devices tested (25–29). Although this studyevaluated debris removal and not the actual removal of bacteria,improved debris removal should correlate with improved bacterialcontrol.

In conclusion, the present in vitro study demonstrated that ‘‘nee-dle deep’’ irrigation with NaOCl was as effective as 3 irrigation adjunctsin canal and isthmus cleanliness utilizing the K-Kube technique. Clinicaltrials to investigate a correlation between irrigation adjunct devices andimproved clinical outcomes are needed to validate the use of thesedevices (30).

AcknowledgmentsThe authors gratefully acknowledge Dr Anneke Bush for her

statistical support and interpretation and Griffin M. Perry for his


technical expertise. This article is the work of the United Statesgovernment and may be reprinted without permission. Opinionsexpressed herein, unless otherwise specifically indicated, are thoseof the authors. They do not represent the views of the Department ofthe Air Force or any other department or agency of the United Statesgovernment.
References1. Ørstavik D, Pitt Ford T. Essential endodontology: prevention and treatment of apical

periodontitis. 2nd ed. Ames, IA: Blackwell Munksgaard Ltd; 2008:1.2. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral Med

Oral Pathol 1984;58:589–99.3. Gulabivala K, Patel B, Evans G, Ng YL. Effects of mechanical and chemical proce-

dures on root canal surfaces. Endodontic Topics 2005;10:103–22.4. Williamson AE, Sandor AJ, Justman BC. A comparison of three nickel titanium rotary

systems Endosequence, ProTaper universal, and Profile GT for canal-cleaningability. J Endod 2009;35:107–9.

5. Mandel E, Machtor P, Friedman S. Scanning electron microscope observation ofcanal cleanliness. J Endod 1990;16:279–83.

6. Zmener O, Pameijer CH, Banegas G. Effectiveness in cleaning oval-shaped rootcanals using anatomic endodontic technology, ProFile and manual instrumentation:a scanning electron microscopic study. Int Endod J 2005;38:356–63.

7. Gutmann JL. Clinical, radiographic, and histologic perspectives on success andfailure in endodontics. Dent Clin North Am 1992;36:379–92.

8. Siqueira JF Jr. Aetiology of root canal treatment failure: why well-treated teeth canfail. Int Endod J 2001;34:1–10.

9. Card SJ, Sigurdsson A, Orstavik D, Trope M. The effectiveness of increased apicalenlargement in reducing intracanal bacteria. J Endod 2002;28:779–83.

10. Siqueira JF Jr, Rocas IN, Santos SR, Lima KC, Magalhaes FA, de Uzeda M. Efficacy ofinstrumentation techniques and irrigation regimens in reducing the bacterial pop-ulation within root canals. J Endod 2002;28:181–4.

11. Archer R, Reader A, Nist R, Beck M, Meyers WJ. An in vivo evaluation of the efficacyof ultrasound after step-back preparation in mandibular molars. J Endod 1992;18:549–52.

12. Goodman A, Reader A, Beck M, Melfi R, Meyers W. An in vitro comparison of theefficacy of the step-back technique versus a step-back/ultrasonic technique inhuman mandibular molars. J Endod 1985;11:249–56.

13. Haidet J, Reader A, Beck M, Meyers W. An in vivo comparison of the step-back tech-nique versus a step-back/ultrasonic technique in human mandibular molars. J En-dod 1989;15:195–9.


14. Jensen SA, Walker TL, Hutter JW, Nicoll BK. Comparison of the cleaning efficacy ofpassive sonic activation and passive ultrasonic activation after hand instrumentationin molar root canals. J Endod 1999;25:735–8.

15. Lev R, Reader A, Beck M, Meyers W. An in vitro comparison of the step-back tech-nique versus a step-back/ultrasonic technique for 1 and 3 minutes. J Endod 1987;13:523–30.

16. Metzler RS, Montgomery S. Effectiveness of ultrasonics and calcium hydroxide forthe debridement of human mandibular molars. J Endod 1989;15:373–8.

17. Burleson A, Nusstein J, Reader A, Beck M. The in vivo evaluation of hand/rotary/ultrasound instrumentation in necrotic, human mandibular molars. J Endod2007;33:782–7.

18. Bramante CM, Berbert A, Borges RP. A methodology for evaluation of root canalinstrumentation. J Endod 1987;13:243–5.

19. Gutarts R, Nusstein J, Reader A, Beck M. In vivo debridement efficacy of ultrasonicirrigation following hand-rotary instrumentation in human mandibular molars. J En-dod 2005;31:166–70.

20. Al-Jadaa A, Paque F, Attin T, Zehnder M. Acoustic hypochlorite activation in simu-lated curved canals. J Endod 2009;35:1408–11.

21. de Gregorio C, Estevez R, Cisneros R, Heilborn C, Cohenca N. Effect of EDTA, sonic,and ultrasonic activation on the penetration of sodium hypochlorite into simulatedlateral canals: an in vitro study. J Endod 2009;35:891–5.

22. Townsend C, Maki J. An in vitro comparison of new irrigation and agitation tech-niques to ultrasonic agitation in removing bacteria from a simulated root canal. JEndod 2009;35:1040–3.

23. Gu XH, Mao CY, Kern M. Effect of different irrigation on smear layer removal afterpost space preparation. J Endod 2009;35:583–6.

24. Kishen A, Sum CP, Mathew S, Lim CT. Influence of irrigation regimens on the adher-ence of Enterococcus faecalis to root canal dentin. J Endod 2008;34:850–4.

25. Zehnder M. Root canal irrigants. J Endod 2006;32:389–98.26. Ram Z. Effectiveness of root canal irrigation. Oral Surg Oral Med Oral Pathol 1977;

44:306–12.27. Kahn FH, Rosenberg PA, Gliksberg J. An in vitro evaluation of the irrigating charac-

teristics of ultrasonic and subsonic handpieces and irrigating needles and probes. JEndod 1995;21:277–80.

28. Sedgley CM, Nagel AC, Hall D, Applegate B. Influence of irrigant needle depth inremoving bioluminescent bacteria inoculated into instrumented root canals usingreal-time imaging in vitro. Int Endod J 2005;38:97–104.

29. Zmener O, Pameijer CH, Serrano SA, Palo RM, Iglesias EF. Efficacy of the NaviTip FXirrigation needle in removing post instrumentation canal smear layer and debris incurved canals. J Endod 2009;35:1270–3.

30. Gu LS, Kim JR, Ling J, Choi KK, Pashley DH, Tay FR. Review of contemporary irrigantagitation techniques and devices. J Endod 2009;35:781–804.



Influence of the Oscillation Direction of an Ultrasonic File onthe Cleaning Efficacy of Passive Ultrasonic IrrigationLei-Meng Jiang, DMD,* Bram Verhaagen, MSc,† Michel Versluis, PhD,† and Lucas W.M. van derSluis, DDS, PhD*

Abstract
Introduction: The cleaning mechanisms and character-istics of passive ultrasonic irrigation (PUI) are not yetcompletely understood. The aim of this study was toinvestigate whether the oscillatory direction of the ultra-sonically driven file had an influence on dentin debrisremoval from artificially made grooves in standardizedroot canals. Methods: Each of 20 ex vivo root canalmodels with a standard groove in the apical portion ofone canal wall filled with dentin debris received PUIrepeatedly, either with file oscillation toward the grooveor with file oscillation perpendicular to the groove. Aftereach irrigation procedure, the amount of dentin debris inthe groove was evaluated by photographs of the grooveand by scoring. The oscillations of the ultrasonic filewere also visualized in vitro by using high-speedimaging at a time scale relevant to the cleaning process,order 10 microseconds. Results: A nonparametric anal-ysis showed significantly more dentin debris reductionwhen the file oscillated toward the groove (P = .002).High-speed imaging showed that the oscillation of thefile is in a single plane, resulting in high-velocity jetsemanating from the file tip in the direction of the oscil-lations. Conclusions: Oscillation of the ultrasonicallydriven file toward the groove is more effective inremoving dentin debris from the groove than oscillationperpendicular to the groove, which can be related tothe fact that there is a high-velocity jet from the filetip in a single direction following the file oscillation anda relatively slow inflow in the perpendicular direction.(J Endod 2010;36:1372–1376)
Key WordsDentin debris, oscillation, passive ultrasonic irrigation

From the *Department of Endodontology, Academic Centre oNetherlands; and †Physics of Fluids Group, Faculty of Science and TMedicine MIRA, University of Twente, Enschede, The Netherlands.

Address requests for reprints to Dr Lei-Meng Jiang, University oEndodontology, Louwesweg 1, 1066EA Amsterdam, the Netherland0099-2399/$0 - see front matter


1372 Jiang et al.

Because instruments used for root canal preparation can merely touch a small part ofthe canal, mechanical preparation by instruments obviously does not suffice for the

debridement of the complex root canal system (1, 2). Irrigation has therefore beengaining increasing attention to improve the cleanliness of root canal systems afterroot canal instrumentation. Passive ultrasonic irrigation (PUI) has been suggested tobe used to disinfect the areas beyond instruments by acoustically activating theirrigant (3–5). Acoustic streaming has been shown to be useful in cleaning the rootcanal system (6).

Ultrasonically powered handpieces are normally attached with an oscillatinginstrument and operated at a certain frequency domain of 20–40 kHz. Previous in vitroinvestigations have shown that oscillation of the file perpendicular to the dentin surfacehad a greater influence on dentin removal than an oscillation parallel to the surface(7, 8), indicating that the energy was distributed nonuniformly around theoscillating file. Lumley et al (9) have demonstrated a 3-dimensional streaming patternaround the ultrasonically activated files, while streaming occurred mainly in front of andbehind the file parallel to the handpiece. In another study, Lumley et al (10) found thata file oscillation directed toward oval recesses left less debris than a perpendicularoscillation. Despite these previous studies, a detailed description and understandingof the oscillation characteristics of ultrasonically driven files are still missing. Theaims of this study were therefore (1) to investigate whether the orientation of the ultra-sonically activated file had an influence on the increase of dentin debris removal fromartificially made grooves simulating uninstrumented canal extensions in standardizedroot canals and (2) to investigate the streaming pattern around an ultrasonicallyoscillating file by using visualization techniques.

Materials and MethodsDentin Debris Removal Model

Straight roots from 20 extracted human maxillary canines were decapitated toobtain uniform root sections of 15 mm. The roots were embedded in self-curing resin(GC Ostron 100; GC Europe, Leuven, Belgium) and then bisected longitudinally throughthe canal in mesiodistal direction with a saw microtome (Leica Microsystems SP1600,Wetzlar, Germany). The surfaces of both halves were ground successively with240-, P400-, and 600-grit sandpaper, resulting in smooth surfaces on which only littleof the original root canal lumen was left. Four holes were drilled in the resin part, andthe 2 halves were reassembled by 4 self-tapping bolts through the holes (Fig. 1A). Theroot canal space of the model was ensured as a closed system.

New root canals were prepared by K-files #15/.02 (Dentsply Maillefer, Ballaigues,Switzerland) and HERO 642 (MicroMega, Besancon, France) nickel-titanium rotary

f Dentistry Amsterdam (ACTA), University of Amsterdam and VU University, Amsterdam, Theechnology, University of Twente, and Research Institute for Biomedical Technology and Technical

f Amsterdam and VU University, Academic Centre of Dentistry Amsterdam (ACTA), Department ofs. E-mail address: [email protected].



Figure 1. (A) Schematic representations of the standardized root canal model, its groove (B1) and cross section (B2). (C1) Drawing of the optical setup, showingthe file inserted into the hole in the aluminum plate (details in C2), which is in turn inserted into a large water tank. The camera is positioned on the left, lookingtoward the hole. (This figure is available in color online at www.aae.org/joe/.)


instruments to a working length (WL) of 15 mm, ISO size 30 and taper0.06, resulting in standardized root canals. During preparation, thecanals were rinsed with 1 mL of 2% NaOCl after each file, deliveredby a 10-mL syringe (Terumo, Leuven, Belgium) and a 30-gauge needle(Navitip; Ultradent, South Jordan, UT).

A standard groove of 4 mm in length, 0.5 mm deep, and 0.2 mmwide, situated at 2–6 mm from WL (11), was cut in the wall of one half ofeach root canal with a customized ultrasonic tip (Fig.1B). A periodontalprobe with an adapted 0.2-mm-wide tip was used to verify the dimen-sion of each groove during and after preparation. The dimension of thegroove is comparable to an apical oval root canal (12). Each groove wasfilled with dentin debris, which was mixed with 2% NaOCl for 5 minutes,to simulate a situation in which dentin debris accumulates in uninstru-mented canal extensions (11). This model was introduced to stan-dardize the root canal space and the amount of dentin debris presentin the root canal before the irrigation procedure to increase the reli-ability of the dentin debris removal evaluation. The methodology issensitive, and the data are reproducible (13, 14). A pilot study hasshown that a single model could be reused up to at least 8 timeswithout any visible defect on the surface of the canal wall. Therefore,the 20 models were used repeatedly in the 3 experimental groups,which are shown in Table 1.

Irrigation ProcedureSpecimens in all the experimental groups were rinsed with 2 mL

irrigant (2% NaOCl) by using 10-mL syringes with 30-gauge needlesplaced 1 mm from WL. Then the irrigant was passively activated by anultrasonic file for 10 seconds, with the oscillation perpendicular tothe groove (group 1; Fig. 2C1) or toward the groove (group 2;Fig. 2C2). The ultrasonic activation was performed with a stainless steel

JOE — Volume 36, Number 8, August 2010 Influ

#20/.00 file (IrriSafe; Acteon, Merignac, France) driven by an ultra-sonic device (Suprasson PMax Newtron; Acteon) at power setting’’blue 4.’’ Every attempt was made to keep the file centered in the canalto minimize contact with the canal walls to do passive ultrasonic activa-tion. Group 3 acted as the control group, in which the ultrasonic file wasinserted but not activated. All the experimental specimens received 2 mLirrigant, which was delivered again by a syringe as final flush.

Image Evaluation and Statistical AnalysesBefore and after each irrigation procedure, the root halves were

separated, and the grooves were viewed through a stereomicroscope(Stemi SV6; Carl Zeiss, Gottingen, Germany) by using a cold light source(KL 2500 LCD; Carl Zeiss). Controls verified that no debris had fallen outof the groove during the assembly or disassembly process. Pictureswere taken with a digital camera (Axio Cam; Carl Zeiss).

The debris left in the groove after irrigation was scored indepen-dently by 3 calibrated dentists by using the following score system: 0, thegroove is empty; 1, less than half of the groove is filled with debris;2, more than half of the groove is filled with debris; and 3, the completegroove is filled with debris (11, 14). The percentage of interagreementshould be more than 95%; if this percentage was lower than 95%,a consensus had to be reached.

The differences in debris scores between the groups were analyzedby means of the Kruskal-Wallis test and the Mann-Whitney test. The levelof significance was set at a = 0.05.

High-speed Imaging ExperimentsAn optical setup was constructed to visualize the oscillation of the

same ultrasonically driven file used in the ex vivo study. To simulate the

ence of Oscillation Direction of Ultrasonic File on Cleaning Efficacy of PUI 1373


TABLE 1. Experimental Groups and the Number of Specimens at Each Score Rank after Irrigation Procedure

Score

Group(N = 20)

Orientation offile oscillation Irrigant Intensity

Totalduration (sec) 0 1 2 3

1 Perpendicular to the groove NaOCl Blue 4 10 10 (50%) 9 (45%) 1 (5%) 0 (0%)2 Toward the groove NaOCl Blue 4 10 19 (95%) 1 (5%) 0 (0%) 0 (0%)3 (control) N/A NaOCL 0 10 0 (0%) 0 (%) 0 (0%) 20 (100%)

Score 0, the groove is empty; score 1, less than half of the groove is filled with debris; score 2, more than half of the groove is filled with debris; score 3, the complete groove is filled with debris.


confinement of apical section of the root canal, a 1-mm-thick aluminumplate with a hole (F = 0.4 mm) and a 4-mm-thick plate with a hole(F = 0.4 mm) plus a groove with the same dimensions as the exvivo model were used. The plate was positioned in a water tank (dimen-sions, 75 � 64 � 60 mm), and the ultrasonic file was centered inthe hole (Fig. 1C). Tracer particles (hollow glass spheres, F = 11mm, r = 1.1$103 kg/m3; Sphericel; Potters Industries, South Yorkshire,UK) were added to the water for flow visualization.

The flow around the oscillating file was imaged through a micro-scope (BX-FM; Olympus, Tokyo, Japan) with a magnification of 1.25–20�. Illumination was performed in bright-field by a continuous wavelight source (ILP-1; Olympus, Tokyo, Japan). Recordings were madewith a high-speed camera (HPV-1; Shimadzu Corp, Kyoto, Japan) ata frame rate of 125,000 frames per second, starting 2 seconds after initi-ation of file oscillation to be able to avoid transient file motion at start-up. Recordings were analyzed by using a Particle Image Velocimetry(PIV) code developed in-house.

ResultsThe results of the ex vivo experiments are presented in Table 1.

There is a statistically significant difference between each of the exper-imental groups (P < .0001). When the irrigant was activated, signifi-cantly more dentin debris was removed than in the control group(no activation). Oscillation of the file toward the groove had a signifi-cantly greater influence on dentin debris removal than oscillationperpendicular to the groove (P = .002).

The time-averaged flow pattern caused by an oscillating file ina large water tank is shown in Fig. 2A. The steady part of the flowdepicted here consists of 2 ’’jets’’ in the direction of oscillation of thefile. There is an inflow toward the file from the direction perpendicularto the oscillation direction. We observe an unsteady flow that is locatedwithin a distance of approximately 1 diameter of the file. Fig. 2B1 showsa close-up of the instantaneous flow pattern, while the file is moving inthe direction indicated by the white arrow. In Fig. 2B2 we show theinstantaneous flow pattern when the confinement of the root canal isincluded; Fig. 2C1 and C2 show the average flow pattern whenthe groove is also included. Flow velocities in the groove when thefile oscillation is toward the groove are 3–5 times higher than whenthe file is oscillating perpendicular to the groove.

DiscussionThe results showed that debris was reduced significantly more by

PUI when the file oscillation was directed toward the groove than whenthe file oscillation was perpendicular to the groove, indicating that theoscillation direction of the ultrasonic file has a great influence.

The ultrasonically driven file oscillates mainly in the directionequal to the axis of the handpiece and a minor transverse vibration atright angles to the main one (9). Lumley et al (10) have shown moreeffective cleaning of an oval extension in the root canal when theoscillation of the file is directed toward the oval extension. They

1374 Jiang et al.

proposed 2 explanations: (1) the streaming forces are more intensetoward the oval recess, and (2) the file was less likely to be constrainedwhen it oscillated toward the recess. However, in the study by Lumley etal, the ultrasonic file was used for root canal instrumentation; in otherwords, the file was intentionally in contact with the root canal wall.Therefore, the file was unable to vibrate freely; acoustic microstreamingwould consequently be less intense, although it would not stopcompletely (15). It could be hypothesized that when oscillating towardthe groove, the file could vibrate somewhat more freely, despite inten-tional contact with the root canal wall, resulting in more intensestreaming forces toward the groove.

PUI was performed in the current study, and the experimentalsetup was such that contact of the file with the root canal wall wasprevented for both oscillation directions. Moreover, the oscillationamplitude of the file is approximately 28 mm according to the manufac-turer, which is smaller than the dimension of the root canal in thecurrent study; thus, the file could vibrate freely whether its orientationwas toward or perpendicular to the groove. Therefore, the only expla-nation for the different efficiency by the 2 ways of irrigation should bethe difference in streaming of the irrigant around the oscillating file,consisting of both the streaming orientation and strength.

This streaming has been visualized with high-speed imaging. Theresults showed a high-velocity jet from the file tip in one dimension anda slow inflow in the perpendicular direction, which could well explainthe consequences for the cleaning efficacy. Streaming has been heldresponsible for cleaning (6), in which the direction and the velocityof the flow might be the key factors.

The flow pattern as shown in Fig. 2A and B1 is qualitatively similarto the flow as described theoretically by Riley (16) and Stuart (17) andconfirmed experimentally by Bertelsen (18). The flow pattern withconfinement, as shown in Fig. 2B2, is qualitatively similar to the flowas described theoretically by Duck and Smith (19). All authors reporteda boundary layer close to the oscillating object, which consisted of anoscillatory and a steady component. Outside this boundary layer, onlythe steady component remains, visible as jets in Fig. 2A and C2. InFig. 2B1 and B2 it can be observed that the boundary layer is approx-imately 0.3 mm thick; therefore, in the apical area (when the groove isstill filled with debris) the flow is dominated by the oscillatory compo-nent, which causes a shear stress circumferentially. In addition to theshear stress, it is expected that there is a push-pull mechanism by whichremoval of debris in the oscillation direction will be enhanced. Thispush-pull effect is induced by the oscillation of the file.

Once (the entrance of) the groove is starting to be emptied byremoval of debris, there will be space available for the jet (steadystreaming component) to form when the file is oscillating toward thegroove. Shear stresses developed by this jet can enhance the removalof the debris in the groove. When the file oscillates perpendicular tothe groove, the jet has no space to develop; therefore, the flow is againdominated by the oscillatory component and its push-pull effect. Thefluid in the groove will contribute to the inflow toward the file; however,flow velocities inside the groove are smaller than when the file oscillates


Figure 2. Direction (solid arrows) and magnitude (colors) of the flow caused by an ultrasonically oscillating file. White arrows in the black circles indicate thedirection of oscillation. (A) Steady part of the flow when oscillating in a large water tank, averaged over 100 frames (0.8 ms). (B1) Unsteady part of the flow, singleframe only; (B2) unsteady part of the flow in the confinement of a root canal, single frame only. (C) Sketch of the cross section of the root canal indicating thedirection of oscillation with respect to the groove and the flow around the oscillating file; (C1) oscillation perpendicular to the groove, (C2) oscillation toward thegroove. (Black bar is 0.2 mm.) (This figure is available in color online at www.aae.org/joe/.)


toward the groove and are unlikely to contribute much to the removal ofdebris from the groove.

In the studies by Riley (16), Stuart (17), Bertelsen (18), andDuck and Smith (19), a cylindrical file was considered, whereasthe Irrisafe file used in this study has a square cross section, twistedalong the length of the file. This difference in cross section mightexplain the increased divergence of the measured jet comparedwith the theoretical solution by Duck and Smith. Experiments per-formed by Kim and Troesch (20) and Tatsuno (21) by using squarecylinders showed a flow pattern more similar to the flow patternobserved in this study.

The ex vivo dentin debris removal model used in this study isa closed system. The 2 halves of the root embedded in the resin matchedperfectly and were fixed by the 4 bolts well to prevent any irrigant flowapically or laterally (Fig. 1A). Because the apical fluid movement mech-anisms can be quite different between a closed and an open system(22), it is better to use a closed system like the model used in this study,which is more clinically relevant.

JOE — Volume 36, Number 8, August 2010 Influ

References1. Peters OA, Schonenberger K, Laib A. Effects of four Ni-Ti preparation techniques on

root canal geometry assessed by micro computed tomography. Int Endod J 2001;34:221–30.

2. Wu MK, van der Sluis LWM, Wesselink PR. The capability of two hand instrumen-tation techniques to remove the inner layer of dentin in oval canals. Int Endod J2003;36:218–24.


4. Burleson A, Nusstein J, Reader A, Beck M. The in vivo evaluation of hand/rotary/ultrasound instrumentation in necrotic, human mandibular molars. J Endod2007;33:782–7.

5. van der Sluis LW, Versluis M, Wu MK, Wesselink PR. Passive ultrasonic irrigation ofthe root canal: a review of the literature. Int Endod J 2007;40:415–26.

6. Ahmad M, Pitt Ford TJ, Crum LA. Ultrasonic debridement of root canals: acousticstreaming and its possible role. J Endod 1987;13:490–9.

7. Briggs PF, Gulabivala K, Stock CJ, Setchell DJ. Dentin-removing characteristics of anultrasonically energized K-file. Int Endod J 1989;22:259–68.

8. Gulabivala K, Briggs PF, Setchell DJ. A comparison of the dentin-removing charac-teristics of two endosonic units. Int Endod J 1993;26:26–36.

ence of Oscillation Direction of Ultrasonic File on Cleaning Efficacy of PUI 1375


9. Lumley PJ, Walmsley AD, Laird WR. Streaming patterns produced around endosonic
files. Int Endod J 1991;24:290–7.10. Lumley PJ, Walmsley AD, Walton RE, Rippin JW. Cleaning of oval canals using

ultrasonic or sonic instrumentation. J Endod 1993;19:453–7.11. Lee SJ, Wu MK, Wesselink PR. The effectiveness of syringe irrigation and ultrasonics

to remove debris from simulated irregularities within prepared root canal walls. IntEndod J 2004;37:672–8.

12. Wu MK, R’Oris A, Barkis D, Wesselink PR. Prevalence and extent of long oval canalsin the apical third. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;89:739–43.

13. van der Sluis LWM, Wu MK, Wesselink PR. The evaluation of removal of calciumhydroxide paste from an artificial standardized groove in the apical root canal usingdifferent irrigation methodologies. Int Endod J 2007;40:52–7.

14. Jiang L-M, Verhaagen B, Michel V, van der Sluis LWM. Evaluation of a sonicdevice designed to activate irrigant in the root canal. J Endod 2010;36:143–6.

1376 Jiang et al.

15. Lee SJ, Wu MK, Wesselink PR. The efficacy of ultrasonic irrigation to removeartificially placed dentin debris from different-sized simulated plastic root canals.Int Endod J 2004;37:607–12.

16. Riley N. Unsteady laminar boundary layers. SIAM Rev 1975;17:274–97.17. Stuart JT. Double boundary layers in oscillatory flow. J Fluid Mech 1966;24:673–87.18. Bertelsen AF. An experimental investigation of high Reynolds number steady

streaming generated by oscillating cylinders. J Fluid Mech 1974;64:589–97.19. Duck PW, Smith FT. Steady streaming induced between oscillating cylinders. J Fluid

Mech 1979;91:93–110.20. Kim SK, Troesch AW. Streaming flows generated by high frequency small amplitude

oscillations of arbitrarily shaped cylinders. Phys Fluids A 1989;1:975–85.21. Tatsuno M. Circulatory streaming in the vicinity of an oscillating square cylinder.

J Phys Soc Japan 2004;36:1185–91.22. Tay FR, Gu LS, Schoeffel GJ, et al. Effect of vapor lock on root canal debridement

by using a side-vented needle for positive-pressure irrigant delivery. J Endod2010;36:745–50.



Effect of Varying Water-to-Powder Ratios on the SettingExpansion of White and Gray Mineral Trioxide AggregateMichael Hawley, DDS, MS, Terry D. Webb, DDS, MS, and Gary G. Goodell, DDS, MS, MA

Abstract
Introduction: Clinicians commonly mix mineral trioxideaggregate (MTA) to a desired consistency rather thanuse the recommended amounts of powder and water.The purpose of this in vitro study was to evaluatehow varying the water-to-powder (WP) ratio affectsthe setting expansion of white MTA (WMTA) and grayMTA (GMTA). Methods: Eight combinations (n = 5)of WMTA and GMTA were mixed using varying WPratios. Randomized samples were placed in a linear vari-able displacement transformer and submerged underHank’s balanced salt solution for 25 hours. Expansionwas compared using a 2-way analysis of variance (a =0.05). Results: Mean percent expansions ranged from0.058%–0.093% for WMTA and 2.15%–2.56% forGMTA. GMTA expanded significantly more thanWMTA at p = .001. No differences in expansion werefound between WP ratios (p = .218). No significantinteraction was found between the WP ratio and mate-rial type (p = .228). Conclusions: GMTA expandedsignificantly more than WMTA; however, varying theWP ratio did not affect the setting expansion. (J Endod2010;36:1377–1379)
Key WordsExpansion, MTA, setting, water to powder ratio

From the Naval Postgraduate Dental School, Bethesda,Maryland.

Address requests for reprints to Dr Terry D. Webb, 4 Box-berry Court, Gaithersburg, MD 20879. E-mail address:[email protected]/$0 - see front matter

Published by Elsevier Inc. on behalf of the American Asso-ciation of Endodontists.doi:10.1016/j.joen.2010.03.010


Current rationale for endodontic therapy is based on the removal of tissue andbacteria from the pulp space and sealing of that cleaned space from the external

environment. In 1974, Schilder (1) listed the biologic objectives of endodontic therapy.They included cleaning and shaping procedures to remove pulp tissue, bacteria, andtheir endotoxins from the root canal system. To achieve predictable success, theroot canal system must be cleaned of organic remnants and shaped to receive a 3-dimensional hermetic filling of the entire root canal space.

Much research and development has been devoted to creating the ideal root canalobturating material. In 1998, Dentsply Tulsa Dental (now Tulsa Dental Specialties,Tulsa, OK) released a new endodontic material called ProRoot MTA. Since that time,a significant amount of research has been published regarding the biomechanicalproperties of this material.

Torabinejad and Chivian (2) initially outlined the uses for mineral trioxide aggregate(MTA) as a restorative material for pulp capping with reversible pulpitis, apexification,repair of root perforations, both nonsurgically and surgically, as well as its use asa root-end filling material. In 1993, Lee et al (3) demonstrated that MTA could seal exper-imentally induced root perforations. de Leimburg et al (4) demonstrated similar resistanceto leakage with obturations of varying thickness of MTA in pulpless teeth with open apices.

White ProRoot MTA (WMTA) (Tulsa Dental Specialties) was introduced in 2002for use in esthetic areas. The original formulation, now called gray MTA (GMTA), couldcause a gray line or show through tooth structure, which was noticeable in estheticareas. It was soon observed that the minor differences in chemical composition betweenthe two might result in differences in biomechanical properties. Matt et al (5) found thatthis original version of WMTA demonstrated significantly more leakage than GMTA inimmediate apical barriers. The authors hypothesized that perhaps slight volumetricshrinkage occurred with the WMTA that accounted for its increased leakage. As earlyas 1998, Fischer et al (6) suggested that the success of MTA was due to its expansionon setting, but this aspect was not investigated.

The manufacturer reformulated the product with reduced particle size in 2003 toaddress the differences in biomechanical properties. A more recent in vitro dye leakagestudy by Hamad et al (7) with the reformulation demonstrated no leakage differencesbetween WMTA and GMTA in furcation perforation repair.

Fridland and Rosado (8) investigated the solubility and porosity of GMTA usingdifferent water-to-powder (WP) ratios. They discovered that porosity and solubilityincreased with increasing WP ratio and that a WP ratio higher than 0.33 was not viscousenough for use, and a 0.26 WP ratio was the minimum that allowed a mix of putty consis-tency to be manipulated. Using a linear variable displacement transformer (LVDT) dila-tometer, Storm et al (9) compared the setting expansion of GMTA, WMTA, and Portlandcement. Using the manufacturer’s recommended WP ratio of 0.35 for mixing all mate-rials and allowing them to set in liquid medium, they found a significantly greater expan-sion for GMTA than for either WMTA or Portland cement.

With the exception of bismuth oxide and gypsum content, MTA has a compositionvery similar to Portland cement (10). Bentz (11) found that Portland cement shrinks(0.1% volumetrically) as it undergoes setting in an unsaturated (dry) environment asa result of chemical shrinkage after hydration. However, in a saturated environment,cement hydration is often accompanied by overall expansion because of crystal growthand possible swelling of gel hydration products (12). Thus in a saturated clinical envi-ronment, setting of GMTA, WMTA, and Portland cement would also be expected toexhibit slight linear expansion.

Effect of Varying WP Ratios on Setting Expansion of WMTA and GMTA 1377


TABLE 1. Percent Expansion of MTA

WP ratio WMTA GMTA

0.26 0.084% � 0.012% 2.42% � 0.324%0.28 0.058% � 0.044% 2.38% � 0.034%0.30 0.093% � 0.013% 2.56% � 0.393%0.35 .086% � 0.029% 2.15% � 0.337%

WP, water-to-powder ratio; WMTA, white mineral trioxide aggregate; GMTA, gray mineral trioxide

aggregate.

Figure 1. Percent expansion of MTA. (This figure is available in color onlineat www.aae.org/joe/.)

The manufacturer of MTA recommends a WP ratio of 0.35 g water
to 1 g powder supplied in one packet. This unit dose mixture results ina great deal of waste because very little material is actually used clini-cally. Because there is an ample amount of material in each packetfor several uses, clinicians commonly estimate the amounts of waterand powder chairside, thus deviating from the manufacturer’s guide-lines and using an unknown WP ratio. This variation from the recom-mended guidelines could have an effect on the expansion of settingMTA in a clinical setting. Therefore, the purpose of this in vitrostudy was to evaluate how varying the WP ratio affects the linear settingexpansion of WMTA and GMTA.

Materials and MethodsWMTA and GMTA were mixed using WP ratios of 0.26, 0.28, 0.30,

and 0.35. Each group (8 total) contained 5 samples (n = 5). Eachsample was assigned a number (1–40), and random numbers weregenerated to determine the testing sequence. Each sample contained1.00 g of MTA powder measured on an analytic balance. Using micro-pipettes, an appropriate amount of sterile water at room temperaturewas added to each sample to achieve the proper WP ratio. Each samplewas hand-mixed on a nonabsorbent pad in a standardized fashion.

As used by Storm et al (9), an LVDT dilatometer was used tomeasure setting expansion. After mixing, each MTA sample was gentlyvibrated into a cylindrical polyvinyl siloxane mold (10 mm height, 6mm diameter, sample mass ~0.60 g) to reduce the inclusion of airbubbles. The mold was constrained so that any displacement causedby linear change in the sample could only occur at one end of themold. Because MTA is often used with a moist cotton pellet or ina surgical environment in contact with body fluids, the samples andmold were covered with Hank’s balanced salt solution (HBSS). A nylonpiston attached to the LVDT was placed on the surface of the settingcement to record volumetric changes. Setting occurred at room temper-ature, and piston movement was recorded every 6 seconds for 25 hoursusing the American Dental Association Health Foundation Dilatometer(ADAHF) Serial Program for Windows version 2.10.0.1, whichmeasures the percent of sample volumetric change. WMTAand GMTA samples mixed with a WP ratio of 0.35 were consideredthe positive controls. The negative control was the testing apparatusrun without a test sample.

Data comparing WMTA and GMTA were analyzed using a two-wayanalysis of variance. To further determine whether there was any rela-tionship between WP ratio and expansion, regression analysis was alsoused. The level of significance was set at a = 0.05.

ResultsLinear expansion was noted in all WMTA and GMTA samples. The

average expansions for WMTA and GMTA with the 0.26, 0.28, 0.30, and0.35 WP ratios are listed in Table 1. When evaluated by two-way analysisof variance, there was no significant difference in linear expansionbetween the WP ratios within each group at p = .218. However, therewas a significant difference in the expansion between WMTA andGMTA at p = .001. No significant interaction was found between theWP ratio and material type at p = .228. Regression analysis furthershowed no correlation between WP ratios and expansion (Fig. 1).

DiscussionType I Portland cement and MTA have complex setting character-

istics. Most dental materials contract as a result of polymerization orchemical shrinkage. However, as previously mentioned, Portlandcements and MTA set by a different mechanism (12), expanding slightlywhen setting in a saturated environment (9, 13, 14). This study supports

1378 Hawley et al.

the findings of Storm et al (9), who reported that GMTA expands signif-icantly more than WMTA and suggested the effect might be due todifferences in the material composition. Although the 2 studies founda virtually identical expansion rate for WMTA, the present study founda nearly 4 times greater expansion rate for GMTA. The reason for thisdifference is unclear. The methodology and instrumentation usedwere identical. The only difference noted was that the studies testeddifferent lot numbers of GMTA. This suggests possible differencesbetween GMTA lot numbers in physical and chemical composition ormanufacture.

In an x-ray diffraction analysis, Song et al (15) found that thecrystalline structure and chemical composition of GMTA and WMTAwere similar except for the presence of iron in GMTA. Asgary et al(16) reported higher levels of aluminum oxide (+122%), magnesiumoxide (+130%), and iron oxide (+1000%) in GMTA than in WMTA.The differences in expansion noted between WMTA and GMTA mightbe explained by the setting chemistries of the 2 different powders.Camilleri (17) found a difference in the crystal forms of set WMTAwhen compared with ordinary Portland cement, noting low presenceof ettringite and monosulfate phases in the hydrated WMTA. Long,slender ettringite crystals are a common member in the setting Portlandcement crystal population. As the hydration mechanism proceeds, thesecrystals form needle-like structures and contribute to the majority ofexpansion seen in Portland cement. In addition, it is the aluminateand ferrite components of Portland cement that form the ettringitecrystals. Thus, if WMTA is lacking this component, less ettringite willbe produced, and less expansion should result. This expansionprobably contributes to MTA’s sealing ability. Most dental materialshave a tendency to shrink away from their interfacial margins, exposinga gap through which contaminating elements can transfer. MTA expands



against its confining margins, thus enhancing the seal and minimizingleakage.
In an article concerning the set of Portland cement in an unsatu-rated environment, Bentz and Aitcin (18) discussed the effects of WP(also called w/c) ratios on the physical properties of Portland cements.They stated that ‘‘w/c has a hidden meaning: it is directly linked to thespacing between cement particles in the cement paste (the smaller thew/c the tighter the spacing of particles and the shorter the setting time).Additionally, the smaller this spacing..the larger the stresses, gener-ating autogenous shrinkage.’’ The environment in the current studywas saturated with HBSS, so a slight chemical expansion rather thanshrinkage would be expected. This discussion highlights the importanceof clinical placement of a moist (wet) cotton pellet adjacent to thesetting MTA when applicable. A dry environment could contribute toautogenous shrinkage, resulting in an interfacial gap at the MTA-dentin margin.

In a clinical setting, MTA is usually mixed using estimated propor-tions of water and powder on the basis of a desired consistency. Underthe conditions of this in vitro study, the linear setting expansion ofGMTA was significantly greater than that of WMTA. However, no signif-icant differences were found between differing WP ratios and theirrespective expansions for both WMTA and GMTA, demonstrating thatdeviating from the manufacturer’s recommended WP mixing ratio ina saturated environment would likely not affect the expansion ofMTA. It is not known whether varying the WP ratio affects physical prop-erties such as compressive strength. Future studies of MTA expansioncould investigate possible introduction of latent stresses or microfrac-tures in dentin or the effect on expansion of any MTA modifications in-tended to improve its physical properties.

DisclaimerThe opinions or assertions expressed in this article are those of the

authors and are not to be construed as official policy or position of theDepartment of the Navy, Department of Defense or the U.S. Government.

Certain commercial materials and equipment are identified in thispaper to specify the experimental procedure. In no instance does suchidentification imply recommendation or endorsement by the U.S. Navy,or that the material or the equipment identified is necessarily the bestavailable for the purpose.


AcknowledgmentsThe authors would like to thank Anthony Guisepetti, American

Dental Association Foundation Paffenbarger Research Center,National Institute of Standards and Technology, Gaithersburg,Maryland.

References1. Schilder H. Cleaning and shaping the root canal. Dent Clin North Am 1974;18. 69-9.2. Torabinejad M, Chivian N. Clinical applications of mineral trioxide aggregate for

repair of lateral root perforations. J Endod 1999;25:197–205.3. Lee SJ, Monsef M, Torabinejad M. Sealing ability of a mineral trioxide aggregate for

repair of lateral root perforations. J Endod 1993;19:541–4.4. de Leimburg ML, Angeretti A, Ceruti P, Lendini M, Pasqualini D, Berutti E. MTA obtu-

ration of pulpless teeth with open apices: bacterial leakage as detected by poly-merase chain reaction assay. J Endod 2004;30:883–6.

5. Matt GD, Thorpe JR, Strother JM, McClanahan SB. Comparative study of white andgray mineral trioxide aggregate (MTA) simulating a one- or two-step apical barriertechnique. J Endod 2004;30:876–9.

6. Fischer EJ, Arens DE, Miller CH. Bacterial leakage of mineral trioxide aggregate ascompared with zinc-free amalgam, intermediate restorative material, and super-EBAas a root-end filling material. J Endod 1998;24:176–9.

7. Hamad HA, Tordik PA, McClanahan SB. Furcation perforation repair comparinggray and white MTA: a dye extraction study. J Endod 2006;32:337–40.

8. Fridland M, Rosado R. Mineral trioxide aggregate (MTA) solubility and porosity withdifferent water-to-powder ratios. J Endod 2003;29:814–7.

9. Storm B, Eichmiller FC, Tordik PA, Goodell GG. Setting expansion of gray and whiteMTA and Portland cement. J Endod 2008;34:80–2.

10. Material safety data sheet (MSDS): ProRoot MTA (mineral trioxide aggregate) rootcanal repair material. Tulsa, OK: Dentsply Tulsa Dental; 2002.

11. Bentz D. Three dimensional computer simulation of Portland cement hydration andmicrostructure development. J Am Ceram Soc 1997;80:3–21.

12. Bentz DP, Jensen OM, Hansen KK, Olesen JF, Stang H, Haecker CJ. Influence ofcement particle-size distribution on early age autogenous strains and stresses incement-based materials. J Am Ceram Soc 2001;84:129–35.

13. Chng HK, Islam I, Yap AU. Properties of a new root-end filling material. J Endod2005;31:665–8.

14. Islam I, Chng HK, Yap AU. Comparison of the physical and mechanical properties ofMTA and Portland cement. J Endod 2006;32:193–7.

15. Song JS, Mante FK, Romanow WJ, Kim S. Chemical analysis of powder and set formsof Portland cement, gray ProRoot MTA, white ProRoot MTA, and gray MTA-Angelus.Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:809–15.

16. Asgary S, Parirokh M, Eghbal MJ, Brink F. Chemical differences between white andgray mineral trioxide aggregate. J Endod 2005;31:101–3.

17. Camilleri J. Hydration mechanisms of mineral trioxide aggregate. Int Endod J 2007;40:462–70.

18. Bentz DP, Aitcin PC. The hidden meaning of water-cement ratio. Concrete Interna-tional 2008. May:51–4.

Effect of Varying WP Ratios on Setting Expansion of WMTA and GMTA 1379


A Comparative Computational Analysis of the MechanicalBehavior of Two Nickel-Titanium Rotary EndodonticInstrumentsSilvia Necchi, MEng,* Lorenza Petrini, MEng, PhD,* Silvio Taschieri, MD,

†

and Francesco Migliavacca, MEng, PhD*

Abstract
Introduction: A finite element model of two nickel-titanium (Ni-Ti) rotary endodontic instruments(ProTaper and SystemGT; Dentsply-Maillefer, Bal-laigues, Switzerland) was developed to investigate themechanical behavior of these devices and to identifythe benefits/limitations of different geometries duringinstrumentation in various root canals. Methods:Instrument shape, curved root canal geometry, andNi-Ti alloy pseudo-elastic behavior were investigatedin this study using computational techniques. Twodifferent operating conditions were simulated: (1) thefile insertion-removal cycle which resembles the stan-dard working condition and (2) the file subjected to a tor-que in the counter-clockwise direction, which mimicsthe auto-reverse movement of the instrument whenthe tip is locked in the canal wall. Results: The simula-tions of standard and auto-reverse conditions producedbending and torsion loading conditions in the files,respectively. In the standard situation in which differentcanal shapes were considered, the strains in the Sys-temGT were generally lower than the strains in the Pro-Taper and always in the pseudo-elastic range; in only 1case did the ProTaper overcame the pseudo-elasticrange limit. In the auto-reverse situation, a betterbehavior of the ProTaper was detected. Conclusions:The two simulated conditions highlighted the differentmechanical properties of the files; the SystemGT showedslightly better performances under flexural solicitation,whereas the Protaper presented better behavior undertorsion solicitations. (J Endod 2010;36:1380–1384)
Key WordsFinite element analysis, nickel-titanium alloy, ProTaper,rotary endodontic files, SystemGT

From the *Laboratory of Biological Structure Mechanics, DepaTecnologies, Universita degli Studi di Milano and IRCCS Galeazzi, M

Supported by the MIUR within the project ‘‘Shape memory alloyAddress requests for reprints to Dr Lorenza Petrini, Laboratory of

Leonardo da Vinci, 32 20133 Milan, Italy. E-mail address: lorenza.p0099-2399/$0 - see front matter


1380 Necchi et al.

The introduction of rotary nickel-titanium (Ni-Ti) endodontic instruments (alsocalled files) into clinical practice has improved the effects of endodontic practice

in terms of procedural time, accuracy, and risk reduction (1–3) compared with thepreviously used, manual, stainless steel files. However, selecting the mostappropriate Ni-Ti file remains difficult for clinicians because of the large number ofNi-Ti rotary instruments available nowadays on the market. In the last 10 years,many publications were devoted to study the factors influencing the fracture of Ni-Tiinstruments (4–9) and to compare cyclic fatigue resistance (10–15), ease of use(16–18), and canal shaping/cleaning ability (19–21) of different Ni-Ti files using invitro experiments. Although these experiments are an essential tool to investigate fileperformance, computational studies may add important information and support theclinical decision process. Indeed, numeric analyses allow us to quantify parametersresponsible for the instruments’ failure as, for example, maximum strain and stress,which are difficult to measure in vivo and in vitro, and to easily compare differentworking conditions by simply changing the model boundaries and removing the oper-ator dependency factor.

In this study, a finite element analysis (FEA) of two files was performed with theaim of identifying potential benefits and/or limitations of various device geometriesduring instrumentation in different types of root canals. In particular, the analysesfocused on the evaluation of the instruments’ resistance to bending and torque asa result of standard and auto-reverse working conditions, respectively.

Materials and MethodsTwo commercially endodontic instruments, both manufactured by Dentsply-

Maillefer (Ballaigues, Switzerland) and usually used in the final step of the shapingsequence, were considered: the ProTaper finisher F1 (convex triangular cross section,variable taper from 7% to 5.5%, 1.2-mm shaft diameter) and the SystemGT series .06(U-shaped flutes cross section, 6% constant taper, 1-mm shaft diameter).

The instrument geometries, built using data retrieved by microscopyimages of the products (Fig. 1A) and 8 different geometries of curved root canals(Table 1), were modeled and implemented into the commercial computational codeSIMULIA (Dassault Systemes, Providence, RI) as previously described by Necchi et al(22), together with the characteristic pseudo-elastic behavior of the Ni-Ti alloy. Thealloy behavior was described using an ad hoc–developed constitutive model (23);average properties from the literature were considered to define the instrument mate-rial parameters (Fig. 1B).

rtment of Structural Engineering, Politecnico di Milano, Milan, Italy; and †Department of Healthilan, Italy.active microactuators and devices for biomedical applications.’’

Biological Structure Mechanics, Department of Structural Engineering, Politecnico di Milano, [email protected].



Figure 1. (A) Main geometric features of the studied files. (B) Stress-strain diagram of the Ni-Ti alloy and characteristic mechanical parameter values used in thenumeric analyses.


Two operating conditions, standard and auto-reverse, werestudied. In the former, the standard clinical procedure, which consistsof numerous cycles of insertion and removal of the rotating file in thecanal, was simulated for each instrument imposing ad hoc boundaryand constraint conditions (22). In the latter, the file was subjected toa torque in the counter-clockwise direction, which resembled theauto-reverse movement of the instrument once the tip is locked inthe canal wall. The instrument was inserted into the canal up to 1mm from the apex, and a counter-clockwise torque (Mz=2 N $ mm)was imposed to the shaft once the tip was blocked.

The following model simplifications were adopted: (1) the accu-mulation of plastic deformation consequent to cyclic loading in thepseudo-elastic range was neglected, (2) the shear strains induced inthe file during the standard procedure as a consequence of frictionand cutting actions of the instrument blade into the canal wall wereneglected, and (3) a reduced rotational speed (uz = 25 rpm) wasimposed.

The instruments’ performance was studied by analyzing the strainvalues reached during the phases of file insertion and removal into andout of the canals. The following variables were considered: the instan-taneous equivalent strain hoarded in the transformation plateau (e

eqtr )

and the logarithmic maximum principal strain (etot) experienced bythe instruments, calculated as the sum of the elastic and the transforma-tion component. The scalar values were compared with the selectedmaterial strain limits (Fig. 1b): eL ¼ 7%, which is the strain limit ofthe austenite to martensite transformation, and eM

y ¼ 7:7%, which isthe yielding strain of the martensite. The undeformed shape recoverycondition after the removal step was considered reached if e

eqtr \eL.

In the limit case of eeqtr ¼ eL (ie, completed austenite to martensite

transformation), the additional condition etot#lnð1þ eMy Þ ¼ 7:4%

was required to avoid plastic strain accumulation. Moreover, the straindistribution was evaluated.


ResultsTable 1 (upper part) shows the results when simulating both the

standard and auto-reverse conditions for the two instruments. The high-est strain levels were reached at the end of the insertion step in theformer condition and at the end of the sticking step in the latter.

In the simulation of the standard clinical procedure, the mostdemanding (in terms of strain) working condition was related to canaltype IV (2-mm radius, 30� angle, middle curvature) for both instru-ments. The ProTaper reached, only in this case, critical strain levels(e

eqtr ¼ eL ¼ 7% and etot ¼ 8:4%$7:4%), whereas the SystemGT

always remained slightly below the eL threshold. The less demandingworking condition was canal type I (5-mm radius, 30� angle, apicalcurvature) for SystemGT. ProTaper performed slightly better in canaltype V (5-mm radius, 45� angle, apical curvature) than in canal typeI. Higher levels of strains were recorded for the ProTaper, except forcanals III and VII (small radius of curvature in the apical position).

Considering the percentage of variation in the strain values (etot

and eeqtr ) as a function of the canal geometry parameters (Table 1),

in the same condition of angle and curvature position, a lowering ofthe radius produced a stronger increment of strains at the end of theinsertion step; in the same condition of radius and curvature position,an increased angle generally produced small increments in strains. Onlyfor the ProTaper operating in canals having a curvature in the apicalposition with a 5-mm radius, the change in the angle from 30� (typeI canal) to 45� (type V canal) produced a small decrement of strain;in the same condition of radius and angle, a repositioning of the curva-ture from apical to middle location produced significant increments ofstrain for canals with radius equal to 5 mm (II vs I and VI vs V) and smallincrements for canals with a 2-mm radius (IV vs III).

The behavior of the instruments in the auto-reverse condition wasstudied by simulating the tip locking at the end of the insertion step.Canal type I was selected for these analyses because it presented the

Mechanical Behavior of Two NiTi Rotary Instruments 1381

TABLE 1. Percentage Values Obtained for the Specific Variables (Logarithmic Maximum Principal Strain etot and Equivalent Transformation Strain eeqtr ) at the End of

the Insertion (standard condition) and Sticking (auto-reverse condition) Steps (upper part). Percentage of Variation of the Strain Specific Variables as a Functionof the Curvature Parameters (lower part). The Canal Parameters Refers to the Radius (R) (2 or 5 mm), Angle A (30� or 45�), and Position (P) (a = apical and m =medial) of the Curvature. PT, Protaper; GT = SystemGT.

Canal type Canal parameters PT GT step

R a P etot[%] eeqtr [%] etot[%] e

eqtr [%]

I 5 30� A 4.0 3.2 3.2 2.7

Insertion end

II 5 30� M 6.6 5.8 5.8 5.0III 2 30� A 7.2 6.4 7.5 6.5IV 2 30� M 8.4* 7.0* 8.2 6.9V 5 45� A 3.8 3.1 3.4 3.0VI 5 45� M 6.8 5.9 5.8 5.6VII 2 45� A 7.4 6.6 7.5 6.6

I 5 30� A 11.4 6.6 11.0 6.8 Sticking end

Radius:2 vs 5 mm

Canal p a etot eeqtr etot e

eqtr

III vs I Apical 30� +80% +100% +134% +141% Insertion endIV vs II Middle 30� +27% +21% +41% +38%VII vs V Apical 45� +95% +113% +121% +120%

Mean +67% +78% +99% +100%

Angle:45� vs 30�

Canal r P etot eeqtr etot e

eqtr

V vs I 5 Apical -3% -3% +6% +11% Insertion endVI vs II 5 Middle +2% +2% +0% +12%VII vs III 2 Apical +3% +3% +0% +2%

Mean +1% +1% +2% +8%

Position:middle vs apical

Canal r A etot eeqtr etot e

eqtr

II vs I 5 30� +65% +81% +81% +85% Insertion endIV vs III 2 30� +17% +9% +9% +6%VI vs V 5 45� +72% +90% +71% +87%

Mean +51% +60% +54% +59%

*Indicates that both strain limit of the austenite to martensite transformation and yielding strain of the martensite are reached.


lowest strains in the standard working condition simulation. This al-lowed us to highlight the effects of the block; at the end of the lockingstep, the SystemGT showed a lower total strain etot (11.0% vs 11.4%)than the ProTaper instrument but a higher equivalent transformationstrain e

eqtr (6.8% vs 6.6%). However, in both the cases the transforma-

tion strain threshold (eL) was not overcome.In Figure 2, the strain distributions during the standard (upper

part) and auto-reverse (lower part) working conditions are depictedfor the files inserted in canal type I.

DiscussionThe present study compared the mechanical behavior of two Ni-

Ti files during instrumentation in various root canals using FEA. FEAmay add useful information to in vivo and in vitro studies, allowingus to calculate quantities that are not measurable during experimentsand to easily compare many models with different geometries, mate-rials, boundary, and loading conditions. However, the accuracy andvalidity of the results are strictly dependent on the adopted compu-tational model, and it is therefore important to recognize the poten-tial limitations and simplifications of the described models. In thisarticle, instrument geometry (multiple tapers along the cuttingportion of the instruments), boundary conditions (curved canals),and material constitutive law were accurately described. However,approximations were adopted for the device material properties(no data from the company or from experimental tests were avail-able) and in the model restrictions discussed in the Material andMethods section. Accordingly, the results previously described shouldnot be considered in absolute terms, but they are important in thecomparison between different conditions. During the standardworking condition, normal stresses and strains arose because ofthe flexure induced by the canal shape. In all studied cases, the flex-ural deformation was lower and more uniformly distributed along

1382 Necchi et al.

a smaller length of the blade for SystemGT (Fig. 2, upper part)than for ProTaper. The highest values were found in the canalwith a sharp curvature in the medial portion. In this case, the Pro-Taper strain level overcame the transformation strain limit. There-fore, the ProTaper seemed to be more prone to accumulateplastic deformation and hence not to recover the original shape afterremoval from the canal.

For both instruments, a stronger increase in the strain levels wasfound as a consequence of a decrease of the curvature radius; a slightlylower increment was found by shifting the curvature position from themiddle to the apical zone. A raise in the curvature angle seemed not toaffect the strain levels (Table 1). Accordingly, the criticality of the oper-ation in standard conditions is dictated primarily by the canal curvatureradius, secondarily by the position, and only at last by the curvatureangle.

During the auto-reverse condition, shear stresses and strainsarose because of the torsion induced by the file rotation in thecounter-clockwise direction. The imposed contrarotational cycletended to ‘‘unroll’’ the files, straightening the tract immediately closeto the blocked tip. In this case, the strains increased radially outwardfrom the section center (Fig. 2, lower part) reaching high values ifcompared with the strains induced by cycles of insertion and removalin the same canal (shown in the upper part of Fig. 2) but without over-coming the strain limits. In the case of the SystemGT, high strain valuesdeveloped over a wide area, whereas in the ProTaper the deformationsquickly decreased moving away from the tip as a consequence of anincrease in the torsion resistance because of the files’ cross-sectionsenlargement. This is because of the higher torsion stiffness of theProTaper with respect to the SystemGT. It is interesting that inin vivo conditions shear strains (and hence stresses) are also presentduring the standard working condition. Indeed, friction caused by thefile cutting action (not taken into account in our simulations) produces


Figure 2. Logarithmic maximum principal strain levels reached in the ProTaper (PT) and SystemGT (GT) at the end of the insertion and sticking steps (canal typeI; section A and B at 3.7 and 2.4 mm from the apex in the z direction). (This figure is available in color online at www.aae.org/joe/.)


torsion solicitations along the blade. Accordingly, strains caused by thesum of torsion and flexural action should be limited in a file. Thissuggests that ProTaper should be preferred because of its geometry,which is more suitable for undergoing both shear and normalstrains/stresses.

These results partially disagree with those presented by Berutti et al(24) and Xu and Zheng (25). The first study compared the torsion andbending behavior of the ProTaper and ProFile instruments, whereas thesecond work investigated the influence of the cross-section profile on


the mechanical performance of different commercial Ni-Ti instruments,including the ProTaper and the ProFile. The ProFile is very similar to theSystemGT in the concave cross-section. Therefore, these two files can beconsidered similar in the evaluation of mechanical response. In botharticles, the results were presented in terms of Von Mises stresses,whereas in our work strain variables were preferred considering thatalong the pseudo-elastic plateau small stress variations correspondedto big strain changes. However, qualitative comparison between theoutputs can be assessed. The results of Berutti et al showed that under

Mechanical Behavior of Two NiTi Rotary Instruments 1383


both bending or torsion actions, ProTaper had Von Mises stresses lowerand more uniformly distributed over the surface than Profile. Moreover,ProTaper worked in the pseudo-elastic range in many load conditions,whereas ProFile operated frequently out of this range. These results agreewith the outcomes from Xu and Zheng that show how ProTaper (convexcross-section) exhibits the lowest Von Mises stress intensity, and modelswith convex and triple helix sections are more torque resistant.
The behavior under flexural deformation of the files with concaveor convex cross-section (ie, ProFile/SystemGT or ProTaper) describedin Berutti et al. (24) and Xu & Zheng (25) is clearly different from ourresults. As highlighted before, FEA accuracy is strictly related to thehypothesis of the mathematical model, and, in this case, there aresome important differences in the computational models adopted inthese three studies. In particular, Berutti et al (24) ignored the taperingof the files, blocked the model at one end, applied a concentratedtorsion moment (0-2.5 N/mm) or a bending moment (0-4.5 N/mm)at the free end, thus omitting file insertion-removal movement intoand from the canal. Xu and Zheng (25) considered a fixed taperingfor all the instruments and constrained the tip face while applyinga 2.5-N/mm torque to the free end.

In conclusion, our results suggested a slightly better performanceof the SystemGT compared with the ProTaper under bending solicita-tions. Conversely, according to literature, a better behavior of the Pro-Taper under torsion solicitations was highlighted.

AcknowledgmentsThe authors thank Daniele Aspesi, MEng, for his contribution

to the numerical analyses.

References1. Parashos P, Messer HH. The diffusion of innovation in dentistry: a review using

rotary nickel-titanium technology as an example. Oral Surg Oral Med Oral PatholOral Radiol Endod 2006;101:395–401.

2. Taschieri S, Necchi S, Rosano G, et al. Advantages and limits of nickel-titaniuminstruments for root canal preparation. A review of the current literature. SchweizMonatsschr Zahnmed 2005;115:1000–5.

3. Walia H, Brantley WA, Gerstein H. An initial investigation of the bending andtorsional properties of nitinol root canal files. J Endod 1988;14:346–51.

4. Martın B, Zelada G, Varela P, et al. Factors influencing the fracture of nickel-titaniumrotary instruments. Int Endod J 2003;36:262–6.

5. Parashos P, Gordon I, Messer HH. Factors influencing defects of rotary nickel-titanium endodontic instruments after clinical use. J Endod 2004;30:722–5.

6. Shen Y, Coil JM, McLean AG, et al. Defects in nickel-titanium instruments after clin-ical use. Part 5: single use from endodontic specialty practices. J Endod 2009;35:1363–7.

1384 Necchi et al.

7. Shen Y, Cheung GS, Bian Z, et al. Comparison of defects in ProFile and ProTapersystems after clinical use. J Endod 2006;32:61–5.

8. Spanaki-Voreadi AP, Kerezoudis NP, Zinelis S. Failure mechanism of ProTaper Ni-Tirotary instruments during clinical use: fractographic analysis. Int Endod J 2006;39:171–8.

9. Zinelis S, Darabara M, Takase T, et al. The effect of thermal treatment on the resis-tance of nickel-titanium rotary files in cyclic fatigue. Oral Surg Oral Med Oral PatholOral Radiol Endod 2007;103:843–7.

10. Bahia MGA, Melo MCC, Buono VTL. Influence of cyclic torsional loading on thefatigue resistance of K3 instruments. Int Endod J 2008;41:883–91.

11. Fife D, Gambarini G, Britto LR. Cyclic fatigue testing of ProTaper NiTi rotary instru-ments after clinical use. Oral Surg. Oral Med. Oral Pathol Oral Radiol Endod 2004;97:251–6.

12. Grande NM, Plotino G, Pecci R, et al. Cyclic fatigue resistance and three-dimensionalanalysis of instruments from two nickel-titanium rotary systems. Int Endod J 2006;39:755–63.

13. Kramkowski TR, Bahcall J. An in vitro comparison of torsional stress and cyclicfatigue resistance of ProFile GT and ProFile GT Series X rotary nickel-titanium files.J Endod 2009;35:404–7.

14. Lopes HP, Moreira EJ, Elias CN, et al. Cyclic fatigue of ProTaper instruments. J Endod2007;33:55–7.

15. Yared G. In vitro study of the torsional properties of new and used ProFile nickeltitanium rotary files. J Endod 2004;30:410–2.

16. Gekelman D, Ramamurthy R, Mirfarsi S, et al. Rotary nickel-titanium GT and Pro-Taper files for root canal shaping by novice operators: a radiographic andmicro-computed tomography evaluation. J Endod 2009;35:1584–8.

17. Peru M, Peru C, Mannocci F, et al. Hand and nickel-titanium root canal instrumen-tation performed by dental students: a micro-computed tomographic study. Euro JDent Edu 2006;10:52–9.

18. Pettiette MT, Delano EO, Trope M. Evaluation of success rate of endodontic treat-ment performed by students with stainless-steel K-files and nickel-titanium handfiles. J Endod 2001;27:124–7.

19. Ayar LR, Love RM. Shaping ability of ProFile and K3 rotary Ni-Ti instruments whenused in a variable tip sequence in simulated curved root canals. Int Endod J 2004;37:593–601.

20. Hulsmann M, Herbs U, Schafers F. Comparative study of root-canal preparationusing Lightspeed and Quantec SC rotary NiTi instruments. Int Endod J 2003;36:748–56.

21. Williamson AE, Sandor AJ, Justman BC. A comparison of three nickel titanium rotarysystems, EndoSequence, ProTaper universal, and profile GT, for canal-cleaningability. J Endod 2009;35:107–9.

22. Necchi S, Taschieri S, Petrini L, et al. Mechanical behavior of Ni-Ti rotary endodonticfiles in simulated clinical conditions: a computational study. Int Endod J 2008;41:939–49.

23. Auricchio F, Petrini L. A three-dimensional model describing stress-temperatureinduced solid phase transformations: solution algorithm and boundary value prob-lems. Int J Numer Methods Eng 2004;61:716–37.

24. Berutti E, Chiandussi G, Gaviglio I, et al. Comparative analysis of torsional andbending stresses in two mathematical models of nickel-titanium rotary instruments:ProTaper versus ProFile. J Endod 2003;29:15–9.

25. Xu X, Zheng Y. Comparative study of torsional and bending properties for six modelsof nickel-titanium root canal instruments with different cross-sections. J Endod2006;32:372–5.



Evaluation of the Effect of Maleic Acid andEthylenediaminetetraacetic Acid on the Microhardnessand Surface Roughness of Human Root Canal DentinNidambur Vasudev Ballal, MDS, Kundabala Mala, MDS, and Kadengodlu Seetharama Bhat, MDS

Abstract
Introduction: The aim of this in vitro study was toevaluate the effect of 7% maleic acid and 17% EDTAsolutions on the microhardness and the surface rough-ness of human root canal dentin. Methods: Forty-fiveextracted human maxillary central incisors weresectioned longitudinally into a total of 90 segments,were embedded in auto polymerizing acrylic resin, andwere grounded flat with silicon carbide abrasive papers.Based on the test solutions used, samples were dividedrandomly into three groups: (1) the EDTA group, 1 mL of17% EDTA for 1 minute (n = 30), (2) the maleic acidgroup, 1 mL of 7% maleic acid for 1 minute (n = 30),and (3) the control group, 1 mL of 0.9% saline for 1minute (n = 30). Every group was then divided intotwo subgroups of 15 specimens each. In group 1a, 2a,and 3a, specimens were used to determine the micro-hardness of the root canal dentine in the coronal,middle, and apical third using Vicker’s hardness tester.In groups 1b, 2b, and 3b, specimens were used for thedetermination of surface roughness of the root canaldentine using a roughness tester (Surtronic, Leicester,England). The data were statistically analyzed usingthe Kruskall Wallis and Mann Whitney U tests. Results:There was no significant difference between EDTA andmaleic acid in the reduction of microhardness. Theincrease in roughness was significantly greater withmaleic acid when compared with EDTA. Conclusion:Maleic acid reduced the microhardness of root dentinsimilar to EDTA but increased the surface roughnesssignificantly more than EDTA. (J Endod 2010;36:1385–1388)
Key WordsEDTA, maleic acid, microhardness, roughness

From the Manipal College of Dental Sciences, Manipal,Karnataka, India

Address requests for reprints to Dr Nidambur VasudevBallal, Manipal College of Dental Sciences, Manipal 576 104,Karnataka, India. E-mail address: [email protected]/$0 - see front matter



The success of root canal therapy depends on the technique and the quality of instru-mentation, irrigation, disinfection, and three-dimensional obturation of the root

canal system. Mechanical instrumentation of the root canal either using manual orrotary instruments produces a smear layer that covers the dentinal tubules (1). Thereis a controversy over whether to remove or maintain the smear layer, but a recentsystematic review and meta-analysis of leakage studies concluded that the removal ofthe smear layer improves the fluid tight seal of the root canal system (2). The smearlayer can be removed using various chelating agents like EDTA, citric acid, and a mixtureof tetracycline isomer (doxycycline), an acid (citric acid) and a detergent (Tween 80)(3–5), but the combination of EDTA and sodium hypochlorite is used most often forsmear layer removal (6). Chelation is a physicochemical process that prompts theuptake of multivalent positive ions by specific chemical substances. In radiculardentine, the agent reacts with the calcium ions of hydroxyapatite crystals. This processcan cause changes in the microstructure of the dentine and changes in the calcium tophosphorus ratio. Changes in the mineral content ratio may alter the original propor-tion of organic and inorganic components, which in turn reduces the microhardness,increases the permeability and solubility of the root canal dentin, and inhibits resistanceto bacterial ingress and permitting coronal leakage (7). It has been indicated that mi-crohardness determination can provide indirect evidence of mineral loss or gain indental hard tissues (8). Changes in the mineral content of root canal dentine mayalso adversely affect the sealing ability and adhesion of dental materials such asresin-based cements and root canal sealers (7). Several studies have shown thatEDTA is capable of decreasing the microhardness of root canal dentine (9–11).Since microhardness is sensitive to composition and surface changes of toothstructure (12), the following effects of several solutions on dentine hardness werepreviously evaluated: sodium perborate (13), EDTA and a combination of hydrogenperoxide and sodium hypochlorite (14), EDTAC, cyclohexane-1,2-diamine-tetra aceticacid and ethylene glycol-bis-(betamino-ethyl ether) N,N,N’,N’-tetra acetic acid (15),and the combination of sodium hypochlorite and EDTA (16, 17). Recently, 7%maleic acid in combination with 2.5% sodium hypochlorite has been found to besignificantly better than EDTA in the removal of the smear layer from the root canalsystem (18). To date, there are no studies evaluating the effect of maleic acid on themicrohardness of root canal dentine. The purpose of this in vitro study was to evaluatethe effect of 7% maleic acid and 17% EDTA solutions on the microhardness and thesurface roughness of human root canal dentine.

Materials and MethodsForty-five extracted human maxillary central incisors were selected. The selection

of teeth was based on their relative dimensions and similarity in morphology. Ethicalclearance was obtained from the Ethical Committee of Manipal University, Manipal,Karnataka, India. Superficial soft tissues were removed with a brush, and the teethwere stored in 0.2% sodium azide (Sigma Chemical Co, St Louis, MO) at 4�C. The teethwith caries, cracks, and dilacerations were excluded. The teeth were decoronated at thecementoenamel junction using a high-speed diamond point (Diatec, Coltene AG,Switzerland) under water cooling, after which the pulp tissue was removed by usinga barbed broach (Mani Inc-Tochigi Ken, Utsunomiya-shi, Japan). Each root wasthen sectioned longitudinally by starting from the cervical to the apical area with

Effect of Maleic Acid and EDTA on Human Root Canal Dentin 1385


a low-speed diamond disc (Horico, Berlin, Germany), separating eachroot into buccal and lingual segments and making a total of 90segments. The root segments were then horizontally embedded inauto polymerizing acrylic resin, leaving their dentine exposed to facili-tate manipulation and improve metallographic preparation. Then, thedentine surfaces of the mounted specimens were grounded flat andsmooth on a circular grinding machine with a series of ascendinggrades of Silicon carbide abrasive papers (500, 800, 1,000, and1,200 grit) under distilled water to remove any surface scratches andfinally polished with 0.1-mm alumina suspension (Ultra-Sol R; EminessTec Inc, Monroe, NC) on a rotary felt disk.
Ninety specimens were then divided into three groups (n = 30)and were prepared as follows: (1) the EDTA group, the specimenswere treated with 1 mL of 17% EDTA (Merck, Dermstadt, Germany)for 1 minute; (2) the maleic acid group, the specimens were treatedwith 1 mL of 7% maleic acid (KMC Pharmacy, Karnataka, India) for1 minute; and (3) the control group, the specimens were treatedwith 1 mL of 0.9% saline for 1 minute. Every group was then dividedinto two subgroups of 15 specimens each. In group 1a, 2a, and 3a spec-imens were used to determine the microhardness of the root canaldentine using a Vickers hardness tester (Matsuzawa Seiki Co Ltd, Tokyo,Japan). The indentations were made with a Vicker’s diamond indenterat three different locations. The locations were chosen at the 0.5-mmlevel to the root canal wall in the coronal, middle, and apical third ofthe roots. The indentations were made on each specimen usinga 200-g load and a 20-second dwell time. The diamond- shaped inden-tations were carefully observed in an optical microscope with a digitalcamera and image analysis software, allowing the accurate digitalmeasurement of their diagonals (Fig. 1). The average length of thetwo diagonals was used to calculate the microhardness value. Therepresentative hardness value for each specimen was obtained as theaverage of the results of the three indentations. These measurementswere then converted into Vicker’s numbers.

In groups 1b, 2b, 3b specimens were used for the determination ofsurface roughness (Ra, mm) of the root canal dentine using a roughnesstester (Surtronic, Leicester, England). Three tracings at different locationson each specimen were made. The mean Ra and standard deviations weredetermined. The Ra parameter describes the overall roughness of a surfaceand can be defined as the arithmetical value of all absolute distances of theroughness profile from the centerline within the measuring length. Themicrohardness and surface roughness values were statistically analyzedusing the Kruskall Wallis and Mann-Whitney U test.

Figure 1. Photomicrograph of the diamond-shaped indentation on the rootcanal surface. (This figure is available in color online at www.aae.org/joe/.)

1386 Ballal et al.

ResultsFigure 2 shows the mean microhardness among the different

experimental groups at the coronal, middle, and apical third of theroot canal system. The control group (saline) reduced the microhard-ness the least as compared with EDTA and maleic acid in the coronal,middle, and apical third, which was statistically significant (p < 0.001).A comparison between EDTA and maleic acid showed that there was nostatistical significant difference in the reduction of the microhardnessat the coronal, middle, and apical third of the root canal system(p = 0.520, p = 0.901, p = 0.089).

Figure 3 shows the mean surface roughness among the differentexperimental groups. When the results of the surface roughness wereanalyzed, the difference between all three experimental groups wasfound to be statistically significant (p < 0.001). When comparing thechange in the dentin surface roughness after using the test agents, theincrease in the dentin surface roughness was significantly greaterwith maleic acid when compared with EDTA (p = 0.014) and saline(p < 0.001). The control group (saline) showed the least roughness.

DiscussionCurrent concepts of chemomechanical preparation imply that

chelating solutions should be applied on instrumented root canalsurfaces in order to remove the smear layer (19). Such proceduresmay induce considerable changes in the surface morphology of dentin,which may also exert changes in its mechanical and physical properties(20, 21). In the present study, chelating solutions were applied on theroot canal dentine, and the surface microhardness and roughness testswere conducted to determine the changes on dentin surface.

Panighi and G’Sell (22) reported a positive corelationshipbetween microhardness and the mineral content of the tooth. Thus,the determination of microhardness can provide valuable evidence ofmineral loss or gain in the dental hard tissue (8). Results of the presentstudy indicated that there was no significant difference in the reductionof the microhardness between 17% EDTA and 7% maleic acid solutionsin the coronal, middle, and the apical third of the root canal system. Thisfinding corroborates with previous studies that have shown that EDTAsignificantly decreases the root canal dentin microhardness (9–11).The degree of mineral content and the amount of hydroxyapatite inthe intertubular substance are considerable factors in determiningthe intrinsic hardness profile of dentin structure (12). The chelatingaction of EDTA solution induces an adverse softening potential on thecalcified components of dentin, and, subsequently, a reduction in mi-crohardness was expected. Maleic acid is a mild organic acid used asan acid conditioner in adhesive dentistry (23). It has been found toposses the smear layer–removing quality when used as an acid etchantin restorative dentistry (24). Recently, 7% maleic acid proved to besignificantly better than EDTA in removing the smear layer from theapical third of the root canal system (18); 7% maleic acid is highlyacidic with a pH of 1.05. This acidic pH might have caused demineral-ization of the root canal dentin and subsequent reduction in the micro-hardness. There is no statistical significant difference between 7%maleic acid and 17% EDTA, which suggests that maleic acid can beused in place of EDTA. Saline, which was used as a control, was foundto have no effect on the microhardness of root canal dentin comparedwith that of EDTA and maleic acid. This may be because saline does nothave any chelating or dimenaralizing effect on the root dentin.

Previous investigations have shown that the suitability and practi-cality of the Vicker’s microhardness test for evaluating surface changesof dental hard tissues treated with chemical agents (25, 26). AlthoughKnoop hardness test was used for evaluating surface changes of dentalhard tissues in other studies (27, 28), the Vicker’s microhardness test



Figure 2. Mean microhardness scores observed among experimental groups.


was preferred in this study because of suitability of the method. In thepresent study, to measure the Vicker’s hardness values for dentin,indentations were made from the coronal, middle, and apical thirdsof the root canal and were done at the 0.5-mm level from the rootcanal walls for standardization, and their means for each samplewere calculated. Dentin hardness is related to location, and its valuedecreases as the indentation tested are made closer to the pulp (29).Pashly et al (29) reported that the microhardness of dentin declinedwhen dentin was tested from superficial to deep regions. They also re-ported an inverse correlation between dentin microhardness andtubular density. In addition to tubular density, the contact time of theirrigation solution needs to be considered as another determinant inthe posttreatment microhardness values of dentin.

Currently, there is no consensus on the optimal time a chelatingagent must be in contact within the root canal to adequately remove

Figure 3. Mean surface roughness scores observed among experimental groups.


the smear layer. However, we opted for a 1-minute time interval, whichis in accordance with various other studies (18, 30). The relativesoftening effect on the dentinal walls exerted by these chemicalsirrigants could be of clinical benefit because it permits rapidpreparation and facilitates the negotiation of tight small root canals(31). These alterations may also affect the sealing ability and the adhe-sion of sealers and adhesives to the root dentin (32).

For the surface roughness, results of the present study indicatedthat maleic acid produced maximum rough surface of the root canaldentin compared with that of EDTA. This could be because of the bettersmear layer removal capacity and the demineralizing ability of maleicacid compared with EDTA. In maleic acid samples, the dentinal tubulesbecame patent, and the surface roughness increased. This finding is inaccordance with the previous study (18), which has shown that maleicacid removes the smear layer significantly better than EDTA. An increase

Effect of Maleic Acid and EDTA on Human Root Canal Dentin 1387

in the surface roughness could be of clinical benefit in restorativedentistry/endodontics because of micromechanical bonding of adhe-sive restorative materials/root canal resin sealers that requires the pres-ence of irregularities on the surface of the adherent into which theadhesive can penetrate.
Further studies need to be conducted to check for the effect of 7%maleic acid on the microhardness and surface roughness of root dentinat different time intervals and also the fracture resistance of roots aftertreatment with 17% EDTA and 7% maleic acid solutions.

ConclusionBased on the results obtained and experimental conditions of the

present study, there was no significant difference between 17% EDTAand 7% maleic acid solutions in the reduction of microhardness.However, maleic acid produced maximum surface roughnesscompared with that of EDTA. Saline that was used as a control didnot alter the radicular dentin microhardness and surface roughness.

References1. Sen BH, Wesselink PR, Turkun M. The smear layer: a phenomenon in root canal

therapy. Int Endod J 1995;28:141–8.2. Shahravan A, Haghdoost A, Adl A, et al. Effect of smear layer on sealing ability of

canal obturation: a systematic review and meta-analysis. J Endod 2007;33:96–105.3. Torabinejad M, Khademi AA, Babagoli J, et al. A new solution for the removal of the

smear layer. J Endod 2003;29:170–5.4. Scelza MFZ, Antoniazzi JH, Scelza P. Efficacy of final irrigation- a scanning electron

microscopic evaluation. J Endod 2000;26:355–8.5. Calt S, Serper A. Smear layer removal by EGTA. J Endod 2000;26:459–61.6. Zehnder M. Root canal irrigants. J Endod 2006;32:389–98.7. Rotstein I, Dankner E, Goldman A, et al. Histochemical analysis of dental hard

tissues following bleaching. J Endod 1996;22:23–6.8. Arends J, ten Boscj JJ. Demineralization and remineralization evaluation techniques.

J Dent Res 1992;71:924–8.9. Ari H, Erdemir A, Belli S. Evaluation of the effect of endodontic irrigation solutions

on the microhardness and the roughness of root canal dentin. J Endod 2004;30:792–5.

10. De-Deus G, Paciornik S, Mauricio MHP. Evaluation of the effect of EDTA, EDTAC andcitric acid on the microhardness of root dentine. Int J Endod 2006;39:401–7.

11. Yu Qing, Akita Y, Kawano S, et al. Cleaning efficacy and dentin microhardness afterroot canal irrigation with a strong acid electrolytic water. J Endod 2006;32:1102–6.

12. Panighi M, G’Sell C. Influence of calcium concentration on the dentine wettability byan adhesive. J Biomed Mater Res 1992;26:1081–9.

1388 Ballal et al.

13. Chang HK, Palamara JE, Messer HH. Effect of hydrogen peroxide and sodiumperborate on biomechanical properties of human dentine. J Endod 2002;28:62–7.

14. Saleh A, Ettman W. Effect of endodontic irrigation solutions on microhardness ofroot canal dentine. J Dent 1999;27:43–6.

15. Cruz- Filho AM, Sousa-Neto MD, Saquy PC, et al. Evaluation of the effect of EDTAC,CTDA and EGTA on radicular dentine microhardness. J Endod 2001;27:183–4.

16. Zhang K, Kim YK, Cadenaro M, et al. Effects of different exposure times and concen-trations of sodium hypochlorite/ethylenediaminetetraacetic acid on the structuralintegrity of mineralized dentin. J Endod 2010;36:105–9.

17. Moreira DM, Almeida JFA, Ferraz CCR, Gomes BPFA, Line SRP, Zaia AA. Structuralanalysis of bovine root dentin after use of different endodontics auxiliary chemicalsubstances. J Endod 2009;35:1023–7.

18. Ballal NV, Kandian S, Mala K, et al. Comparison of the efficacy of maleic acid andethylenediaminetetraacetic acid in smear layer removal from instrumented humanroot canal: a scanning electron microscopic study. J Endod 2009;35:1573–6.

19. Torabinejad M, Handysides R, Khademi AA, et al. Clinical implications of the smearlayer in endodontics: a review. Oral Surg Oral Med Oral Pathol Oral Radiol Endod2002;94:658–66.

20. Dogan H, Calt S. Effect of chelating agents and sodium hypochlorite on mineralcontent of root dentin. J Endod 2001;27:578–80.

21. Ari H, Erdemir A. Effects of endodontic irrigation solutions on mineral content ofroot canal dentin using ICP-AES technique. J Endod 2005;31:187–9.

22. Panighi M, G’Sell C. Effect of the tooth microstrucutre on the shear bond strength ofa dental composite. J Biomed Mater Res 1993;27:975–81.

23. Wieczowski G, Davis EL, Joynt RB. Microleakage in various bonding agent compositeresin systems. Oper Dent 1992;5(suppl):62–7.

24. Erickson RL. Surface interactions of dentin adhesive materials. Oper Dent (supple-ment) 1992;5(suppl):81–94.

25. Tulga F, Ozok R, Gurbuz A, Ozkan P. Effect of different types of vital bleaching agentson microhardness of human enamel. Balkan J Stomatol 2000;4:164–6.

26. Kuramoto Junior M, Matson E, Turbino ML, et al. Microhardness of Nd:YAG laserirradiated enamel surfaces. Braz Dent J 2001;12:31–3.

27. Meredith N, Sherriff M, Setchell DJ, et al. Measurement of the microhardness andYoung’s modulus of human enamel and dentin using an indentation technique.Arch Oral Biol 1996;41:539–45.

28. Hosoya Y, Marshall SJ, Watanabe LG, et al. Microhardness of carious deciduousdentin. Oper Dent 2000;25:81–9.

29. Pashley D, Okabe A, Parham P. The relationship between dentin microhardness andtubule density. Endod Dent Traumatol 1985;1:176–9.

30. Yamada RS, Armas A, Goldman M, et al. A scanning electron microscopic compa-rison of a high volume final flush with several irrigating solutions: part 3. J Endod1983;9:137–42.

31. Cruz-Filho AM, Paula EA, Pecora JD, et al. Effect of different EGTA concentrations ondentin microhardness. Braz Dent J 2002;13:188–90.

32. Garcia-Godoy F, Loushine RJ, Itthagarun A, et al. Application of biologically orienteddentin bonding principles to the use of endodontic irrigants. Am J Dent 2005;18:281–90.



Antimicrobial Effects of Calcium Hydroxide andChlorhexidine on Enterococcus faecalisRonan J.R. Delgado, DDS, MSc,* Thaıs H. Gasparoto, DDS, MSc, PhD,* Carla R. Sipert, DDS, MSc,*

Claudia R. Pinheiro, DDS, MSc,* Ivaldo G. Moraes, DDS, MSc, PhD,†

Roberto B. Garcia, DDS, MSc,PhD,† Clovis M. Bramante, DDS, MSc, PhD,† Ana P. Campanelli, MSc, PhD,*

and Norberti Bernardineli, DDS, MSc, PhD†

Abstract
Introduction: Endodontic treatment is commonlybased on nonspecific elimination of intraradicular micro-organisms. Although some authors prefer single-visitroot canal operations for endodontic treatment, severalstudies have shown the importance of intracanal medi-cation between sessions to kill microorganisms thatbiomechanical preparations alone cannot achieve. Thepurpose of this study was to evaluate the efficacy ofcalcium hydroxide Ca(OH)2 and chlorhexidine gel onthe elimination of intratubular Enterococcus faeca-lis. Methods: Human uniradicular teeth contaminatedwith E. faecalis were treated with Ca(OH)2, 2% chlo-rhexidine gel, Ca(OH)2 plus 2% chlorhexidine gel, orsaline (0.9% NaCl) as a negative control. Samples ob-tained at a depth of 0 to 100 mm and 100 to 200 mmfrom these root canal preparations were analyzed forbacterial load by counting the number of colony-forming units (CFUs) and bacterial viability using fluo-rescence microscopy. Results: A significant decreasein the number of CFUs and the percentage of viable E.faecalis was observed after treatment with eitherCa(OH)2 or chlorhexidine when compared with thecontrol group. Additionally, chlorhexidine gel hada significantly higher antimicrobial efficacy as measuredby the number of CFUs and the percentage of viable cellsthan Ca(OH)2. No differences were observed betweenthe antimicrobial properties of chlorhexidine gel withand without the addition of Ca(OH)2. Conclusion:Both Ca(OH)2 and chlorhexidine have antimicrobialeffects on E. faecalis. Chlorhexidine had increasedantimicrobial activity when compared with Ca(OH)2.
Ca(OH)2 combined with chlorhexidine showed similarantimicrobial activity to chlorhexidine alone. (J Endod2010;36:1389–1393)

Key WordsCalcium hydroxide, chlorhexidine, endodontic infection,Enterococcus faecalis, intracanal dressings

From the Departments of *Biological Sciences and †EndodonticAddress requests for reprints to Dr Ronan J.R. Delgado, Departm

Pinheiro Brisolla, 9-75, Bauru, Sao Paulo 17012-901, Brazil. E-mail0099-2399/$0 - see front matter



Bacterial invasion of the root canal system is crucial for the onset and maintenance ofperiapical diseases; thus, one goal of endodontic treatment is to kill microorgan-

isms in the root canal system (1). Although biomechanical preparation and root canalshaping effectively reduce microbiota, these procedures do not completely eliminatebacteria in the lateral and accessory root canals, isthmi, and apical deltas (2). Thus,intracanal medication between appointments is recommended to further reducebacteria in the root canal system, and multiple visits may be required even in casesin which biological concerns are not an issue (3).

Calcium hydroxide (Ca[OH]2) is one of the most commonly used substances inendodontics, and its antibacterial property stems from its ability to increase a solutionspH (4). Chlorhexidine has emerged as an intracanal medication because of its wideantimicrobial spectrum, its ability to maintain its antibacterial action for a prolongedduration when adhered to anionic substrates, and its slow release as its concentrationdecreases (5, 6). The use of Ca(OH)2 and chlorhexidine combined has shown betterantimicrobial properties than Ca(OH)2 alone, presenting biocompatibility withoutaffecting the sealing ability of root canal obturation (7–9).

Enterococcus faecalis is a gram-positive bacterium often isolated in persistentroot canal infections. Furthermore, it can penetrate deeply into dentinal tubules andresist bactericidal substances commonly used in endodontic procedures (10, 11).E. faecalis was the first gram-positive bacterium referred to as able to enter intoa ‘‘viable but nonculturable’’ state (12, 13). Clinical procedures normallyconsidered to be bactericidal, as used during endodontic treatment, may allowbacteria to be in this metabolic state where they cease to grow on bacteriologicculture media but remain alive. When favorable conditions return, E. faecalis canthen return to a fully culturable state (12, 13). Therefore, the purpose of this studywas to test, through both culture growth method and viability count, whetherchlorhexidine alone or in combination with Ca(OH)2 could completely eliminateE. faecalis.

Materials and MethodsTeeth Preparation

Sixty extracted single-rooted teeth were cleaned and stored in 10% formaldehyde.Fifteen teeth served as a negative control, and 45 teeth were used in the experimentalgroups. Specimens were prepared according to methods previously described by Haa-pasalo and Orstavik (11). Briefly, crowns (2-3 mm from the cement-enamel junction)and 3 to 5 mm of the apical portion of the root were removed, and specimens (roots of 8mm) were stored in saline until experimental procedures were performed.

s, Bauru School of Dentistry, University of Sao Paulo, Bauru, SP, Brazil.ent of Biological Sciences, Bauru School of Dentistry, University of Sao Paulo, Alameda Dr Octavioaddress: [email protected].

Antimicrobial Effects of Ca(OH)2 and CHX on E. faecalis 1389



Bacterial GrowthE. faecalis (American Type Culture Collection 29212) was

cultured in brain heart infusion (BHI) broth. Bacterial morphologywas confirmed using the Gram method and a stereomicroscope(Wild Heerbrugg, Gais, Switzerland). Bacterial concentrations used inexperiments were determined by MacFarland Standards.

Experimental Root Canal InfectionRoot canals were infected by inoculating the upper chambers with

6.3� 108 colony-forming units (CFUs) per mL of E. faecalis diluted in4 mL of BHI, keeping the bacterial suspension in contact with the filledroots. Fresh BHI broth was supplemented in the upper chamber weeklyto ensure viability of bacteria. All specimens were incubated aerobicallyat 37�C. For 21 days, specimens were verified daily, and the turbidity ofthe E. faecalis was recorded for each specimen. After this period, eachsample had each root canal irrigated with sterile saline, dried withsterile paper points, and divided into four groups (n = 15).

Intracanal DressingsTeeth samples were submitted to the following intracanal dress-

ings: Ca(OH)2 (Calen, S.S. WHITE Artigos Dentarios Ltd, Rio de Janeiro,Brazil), 2% chlorhexidine gel, Ca(OH)2 combined with 2% chlorhexi-dine gel, and a negative control group with sterile saline. Each dressingwas applied under sterile conditions via a syringe and needle until thedressing was extravasated. After the excess material was removed,coronal and apical orifices were sealed with temporary restorativecement (Cimpat; Septodont, Saint-Maur-des-Fosses, France). Then,specimens were placed into Petri dishes and covered with humid sterilegauzes and incubated at 37�C for 14 days.

Dentinal Fragment SamplesAt the end of the incubation period, the restorative cement was

removed, and the root canal was washed with 5 mL of sterile saline.Subsequently, root canals were irrigated with neutralizing solution ac-cording to the following antimicrobial agents: 0.5% citric acid forCa(OH)2, 0.5% Tween 80 (Sigma-Aldrich; Saint Louis, MO) in 0.07%soy lecithin for 2% chlorhexidine gel, and 2% chlorhexidine gelcombined with Ca(OH)2. A last irrigation with sterile saline was thenperformed, and, lastly, root canals were dried with sterile paper points.

Dentinal fragment samples were collected using Gates Gliddenburs #5 for the depth of 0 to 100 mm and #6 for the depth of 100 to200 mm (500 rpm, 1 N, electric motor Endo Plus, Driller; Sao Paulo,Brazil) and maintained in a microcentrifuge tube containing 1 mL ofBHI broth. Each bur was used three times throughout the whole exten-sion of the root canal.

Assessment of Antibacterial ActivityImmediately after the collection, dentine samples were mixed for 1

minute; aliquots of 25 mL were seeded on BHI blood agar and incubatedat 37�C for 48 hours. After incubation, CFUs were counted usinga colony counter CP600 (Quimis; Diadema, Brazil). Bacterial puritywas confirmed by colony morphology and Gram staining.

For fluorescence microscopy analysis, dentin suspensions weremixed for 30 seconds, centrifuged at 600 g/5 min, and washed withphosphate buffered saline three times. The pellet was suspended in25 mL of phosphate buffered saline followed by vigorous agitation.Samples were stained with 2.5 mL of calcein-acetoxymethylester(calcein-AM) (viable bacteria stained green in color) and 1 mL of pro-pidium iodide (nonviable bacteria stained red in color) at 37�C for 15minutes and analyzed using confocal microscopy (TCS model SPE; Le-

1390 Delgado et al.

ica, Mannheim, Germany). Bacteria viability was expressed as the meanpercentage of viable E. faecalis over the total number of microorgan-isms by randomly counting of three fields.

Data AnalysisResults are expressed as the mean � one standard deviation.

Statistical analysis was performed using the nonparametric Kruskal-Wallis one-way analysis of variance test. The Miller test was used toestablish individual differences. Values of p < 0.05 were consideredstatistically significant.

ResultsAntimicrobial Activity of Chlorhexidine and Ca(OH)2Against E. faecalis

The effectiveness of removing the smear layer was analyzed(Fig. 1A). Twenty-one days after smear layer removal, bacterial contam-ination of all dentinal tubules was confirmed as shown by scanning elec-tron microscopy (Fig. 1B).

The antimicrobial activity of Ca(OH)2 and 2% chlorhexidine gelwas evaluated by counting the CFUs. After 14 days of intracanal medica-tion, our results showed that treatment with saline solution did not influ-ence the bacteria viability within the dentinal tubules (Fig. 1C). The useof Ca(OH)2 and 2% chlorhexidine reduced the number of CFUs ofE. faecalis recovered at depths of 0 to 100 mm and 100 to 200 mmwhen compared with control treatment (saline). The treatment withchlorhexidine presented the least number of CFUs. No significant differ-ences in the bacteria CFU (0- to 100-mm depth or 100- to 200-mmdepth) were observed between chlorhexidine with and withoutCa(OH)2 (Fig. 1C).

Bacterial viability was evaluated through confocal scanning elec-tron microscopy (Fig. 2). The intracanal dressings used significantlyreduced the number of E. faecalis when compared with the controls(p < 0.001) (Fig. 2A). Chlorhexidine significantly eliminated more E.faecalis (p < 0.001) when compared with Ca(OH)2 because viableE. faecalis was decreased (6.9% � 5.25%) compared with Ca(OH)2

(25%� 10.9%) (Fig. 2A, C, and D). There were no differences betweenthe antimicrobial activity of chlorhexidine with or without Ca(OH)2

(Fig. 2A and E).

DiscussionEndodontic therapy can be based on nonspecific elimination of in-

traradicular microorganisms. Furthermore, some authors prefera single-visit root canal treatment (14, 15) although many studieshave shown the importance of intracanal medication betweensessions in order to kill microorganisms that biomechanicalpreparations miss. Intracanal medication act beyond the root canallumen, inside dentinal tubules and apical resorptions (1–3).E. faecalis’ high resistance to antibacterial substances is widelydocumented, and this bacterium can enter in a viable butnonculturable state during environmental stress (12, 13). Therefore,this investigation analyzed the influence of chlorhexidine andCa(OH)2 on the survival of intratubular E. faecalis.

The methodology performed in this study has several advantagesover other methods (16). Moreover, the interaction between dentineand E. faecalis may also result in bacterium resistance (17). Themethod used in this study allowed the recovery of bacteria within bio-films inside dentinal tubules instead of planktonic microorganisms sus-pended at the lumen of the root canal (17, 18) and also resulted insimilar outcomes to those obtained in daily practice because themicroorganisms were grown in liquid medium containing openedteeth allowing the penetration of bacteria into dentinal tubules.


Figure 1. Antimicrobial activity of intracanal dressings against intratubular E. faecalis. Sixty extracted single-rooted teeth were infected with E. faecalis, and after21 days, bacterial contamination of dentinal tubules was confirmed. Ca(OH)2 and 2% chlorhexidine gel with and without Ca(OH)2 were applied to these teeth, andafter 14 days, dentine samples obtained at depths of 0 to 100 mm and 100 to 200 mm through the root canal preparation were analyzed by counting the CFUs. (A)Scanning electron microscopy showing smear layer removal. (B) Scanning electron microscopy showing E. faecalis in human dentinal tubules 21 days after infec-tion. (C) The bars show the number of CFUs of E. faecalis recovered after Ca(OH)2 and chlorhexidine treatment. Results represent the mean� standard deviationof two independent experiments. The data were analyzed by the Kruskal-Wallis and Miller post tests. )p < 0.05 compared with control group; #p < 0.05 comparedwith Ca(OH)2.


The present investigation evaluated the ability of E. faecalis tosurvive certain intracanal dressings while considering their metabolicstate and cultivable response. Typically, a bacterium’s physiologicalstate infecting the root canal system is close to a starvation condition.Despite the effectiveness of Ca(OH)2 in eliminating a wide range ofmicroorganisms, our results corroborate others findings showingE. faecalis’ resistance to Ca(OH)2 (19–21). Furthermore, E. faecalishas also been shown to be more susceptible to solutions containingchlorhexidine (7, 22–24) but not Ca(OH)2 (16, 25, 26). Becausestarvation may be one of the major factors that impacts the resistanceof E. faecalis (27), the metabolic state of this microorganism when sub-jected to intracanal dressings during the endodontic treatment must beconsidered.

The antibacterial properties of Ca(OH)2 are attributed to its alka-linity and its ability to destroy the cytoplasmic membrane, denaturebacterial proteins, and damage bacterial DNA (4). Chlorhexidine pene-trates into bacteria and exerts toxic effects through disturbance in theirmembrane charges (28). Both hydroxide ions and chlorhexidine exerttheir bactericidal activity by disintegrating membranes (29). Addition-ally, chlorhexidine may induce reactive oxygen species production inthe alkaline environment. The production of reactive oxygen speciesmay inhibit E. faecalis growth because of the destruction of the cellwall and the plasma membrane mediated by nitric oxide (30). On


the other hand, studies have reported that the antimicrobial activity ofchlorhexidine is reduced when it is associated with Ca(OH)2 (31).However, this finding was not observed in our study. Rather, weobserved a similar antimicrobial activity for chlorhexidine combinedwith Ca(OH)2 when compared with chlorhexidine alone against E. fae-calis when using an agar diffusion tests. This observation is corrobo-rated by Zerella et al (7). The combination of chlorhexidine withCa(OH)2 paste could remain in the root canal system as a barrier forlonger periods, eliminating large amounts of persistent microorganisms(32, 33).

This study evaluated the antimicrobial activity of Ca(OH)2 andchlorhexidine with and without Ca(OH)2 and found that all of thesetreatments were effective against E. faecalis. More specifically,Ca(OH)2 alone was significantly less effective than the other treatmentswith chlorhexidine. Therefore, chlorhexidine may be effective forendodontic therapy. However, viable E. faecalis was detected after 14days from the treatment indicating persistence of this pathogen toremain in a ‘‘viable but nonculturable’’ state.

AcknowledgmentsWe thank Dr Daniel Thomas Brozoski for his critical reading of

the manuscript and Marcia Graeff and Andre Luis da Silva from the


Figure 2. The viability of E. faecalis. After 14 days of intracanal medication, the antimicrobial activity of Ca(OH)2 and 2% chlorhexidine gel was analyzed byfluorescence microscopy. Dentin samples were obtained at 0 to 100 mm and 100 to 200 mm, and bacteria were stained with calcein-AM and propidium iodide.(A) The bars show the total number of viable bacteria. The data were analyzed by the Kruskal-Wallis and Miller post tests. )p < 0.05 compared with the controlgroup; #p < 0.05 compared with Ca(OH)2. (B) E. faecalis viability after treatment with saline solution, (C) calcium hydroxide, (D) 2% chlorhexidine gel and, (E)Ca(OH)2 combined with 2% chlorhexidine gel. (This figure is available in color online at www.aae.org/joe/.)


Department of Biological Sciences, Bauru School of Dentistry, fortheir technical assistance.

References1. Vianna ME, Horz HP, Conrads G, et al. Effect of root canal procedures on endotoxins

and endodontic pathogens. Oral Microbiol Immunol 2007;22:411–8.2. Vianna ME, Gomes BP. Efficacy of sodium hypochlorite combined with chlorhexi-

dine against Enterococcus faecalis in vitro. Oral Surg Oral Med Oral Pathol OralRadiol Endod 2009;107:585–9.

3. Sathorn C, Parashos P, Messer H. Australian endodontists’ perceptions of single andmultiple visit root canal treatment. Int Endod J 2009;42:811–8.

1392 Delgado et al.

4. Siqueira JF Jr, Lopes HP. Mechanisms of antimicrobial activity of calcium hydroxide:a critical review. Int Endod J 1999;32:361–9.

5. Tervit C, Paquette L, Torneck CD, et al. Proportion of healed teeth with apical pe-riodontitis medicated with two percent chlorhexidine gluconate liquid: a case-seriesstudy. J Endod 2009;35:1182–5.

6. Lee Y, Han SH, Hong SH, et al. Antimicrobial efficacy of a polymeric chlorhexidinerelease device using in vitro model of Enterococcus faecalis dentinal tubule infec-tion. J Endod 2008;34:855–8.

7. Zerella JA, Fouad AF, Spangberg LS. Effectiveness of a calcium hydroxide and chlo-rhexidine digluconate mixture as disinfectant during retreatment of failedendodontic cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;100:756–61.



8. Kontakiotis EG, Tsatsoulis IN, Papanakou SI, et al. Effect of 2% chlorhexidine gel
mixed with calcium hydroxide as an intracanal medication on sealing ability ofpermanent root canal filling: a 6-month follow-up. J Endod 2008;34:866–70.

9. da Silva RA, Leonardo MR, da Silva LA, et al. Effects of the association betweena calcium hydroxide paste and 0.4% chlorhexidine on the development of the oste-ogenic phenotype in vitro. J Endod 2008;34:1485–9.

10. Williams JM, Trope M, Caplan DJ, et al. Detection and quantitation of E. faecalis byreal-time PCR (qPCR), reverse transcription-PCR (RT-PCR), and cultivation duringendodontic treatment. J Endod 2006;32:715–21.

11. Haapasalo M, Orstavik D. In vitro infection and disinfection of dentinal tubules.J Dent Res 1987;66:1375–9.

12. Oliver JD. The viable but nonculturable state in bacteria. J Microbiol 2005;43:93–100.

13. Signoretto C, Lleo MM, Tafi MC, et al. Cell wall chemical composition of Entero-coccus faecalis in the viable but nonculturable state. Appl Environ Microbiol2000;66:1953–9.

14. Sathorn C, Parashos P, Messer HH. Effectiveness of single-versus multiple-visitendodontic treatment of teeth with apical periodontitis: a systematic review andmeta-analysis. Int Endod J 2005;38:347–55.

15. Malkhassian G, Manzur AJ, Legner M, et al. Antibacterial efficacy of MTAD final rinseand two percent chlorhexidine gel medication in teeth with apical periodontitis:a randomized double-blinded clinical trial. J Endod 2009;35:1483–90.

16. Siqueira JF Jr, de Uzeda M. Intracanal medicaments: evaluation of the antibacterialeffects of chlorhexidine, metronidazole, and calcium hydroxide associated withthree vehicles. J Endod 1997;23:167–9.

17. Kayaoglu G, Erten H, Bodrumlu E, et al. The resistance of collagen-associated,planktonic cells of Enterococcus faecalis to calcium hydroxide. J Endod 2009;35:46–9.

18. Ercan E, Dalli M, Duulgergil CT, et al. Effect of intracanal medication with calciumhydroxide and 1% chlorhexidine in endodontic retreatment cases with periapicallesion: an in vivo study. J Formos Med Assoc 2007;106:217–24.

19. Evans M, Davies JK, Sundqvist G, et al. Mechanisms involved in the resistance ofEnterococcus faecalis to calcium hydroxide. Int Endod J 2002;35:221–8.

20. Gomes BP, Ferraz CC, Vianna ME, et al. In vitro antimicrobial activity of severalconcentrations of sodium hypochlorite and chlorhexidine gluconate in the elimina-tion of Enterococcus faecalis. Int Endod J 2001;34:424–8.


21. Sukawat C, Srisuwan T. A comparison of the antimicrobial efficacy of three calciumhydroxide formulations on human dentin infected with Enterococcus faecalis. J En-dod 2002;28:102–4.

22. Basrani B, Tjaderhane L, Santos JM, et al. Efficacy of chlorhexidine-and calciumhydroxide-containing medicaments against Enterococcus faecalis in vitro. OralSurg Oral Med Oral Pathol Oral Radiol Endod 2003;96:618–24.

23. Gomes BP, Vianna ME, Sena NT, et al. In vitro evaluation of the antimicrobial activityof calcium hydroxide combined with chlorhexidine gel used as intracanal medica-ment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:544–50.

24. Turk BT, Sen BH, Ozturk T. In vitro antimicrobial activity of calcium hydroxidemixed with different vehicles against Enterococcus faecalis and Candida albicans.Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108:297–301.

25. Barbosa CA, Goncalves RB, Siqueira JF Jr, et al. Evaluation of the antibacterial activitiesof calcium hydroxide, chlorhexidine, and camphorated paramonochlorophenol as in-tracanal medicament. A clinical and laboratory study. J Endod 1997;23:297–300.

26. Komorowski R, Grad H, Wu XY, et al. Antimicrobial substantivity of chlorhexidine-treated bovine root dentin. J Endod 2000;26:315–7.

27. Portenier I, Waltimo T, Orstavik D, et al. The susceptibility of starved, stationaryphase, and growing cells of Enterococcus faecalis to endodontic medicaments. JEndod 2005;31:380–6.

28. Lindskog S, Pierce AM, Blomlof L. Chlorhexidine as a root canal medicament fortreating inflammatory lesions in the periodontal space. Endod Dent Traumatol1998;14:186–90.

29. McDonnell G, Russell AD. Antiseptics and disinfectants: activity, action, and resis-tance. Clin Microbiol Rev 1999;12:147–79.

30. Nishikawa T, Sato E, Choudhury T, et al. Effect of nitric oxide on the oxygen metab-olism and growth of E. faecalis. J Clin Biochem Nutr 2009;44:178–84.

31. de Souza-Filho FJ, Soares Ade J, Vianna ME, et al. Antimicrobial effect and pH ofchlorhexidine gel and calcium hydroxide alone and associated with other materials.Braz Dent J 2008;19:28–33.

32. Gomes BP, Souza SF, Ferraz CC, et al. Effectiveness of 2% chlorhexidine gel andcalcium hydroxide against Enterococcus faecalis in bovine root dentine in vitro.Int Endod J 2003;36:267–75.

33. Lin S, Kfir A, Laviv A, et al. The in vitro antibacterial effect of iodine-potassium iodideand calcium hydroxide in infected dentinal tubules at different time intervals. J Con-temp Dent Pract 2009;10:59–66.



Influence of Cross-sectional Design and Dimension onMechanical Behavior of Nickel-Titanium Instruments underTorsion and Bending: A Numerical AnalysisEn-Wei Zhang, PhD,* Gary S.P. Cheung, MSc, MDS, PhD,

†and Yu-Feng Zheng, PhD*

Abstract
Introduction: The aim of this study was to examine theinfluence of the cross-sectional configuration anddimensions (size and taper) on the torsional andbending behavior of nickel-titanium rotary instruments,taking into account the nonlinear mechanical propertiesof material. Methods: Ten cross-sectional configura-tions, square, triangular, U-type, S-type (large andsmall), convex-triangle, and 4 proprietary ones (ManiNRT and RT2, Quantec, and Mtwo), were analyzedunder torsion or bending by using a 3-dimensional finiteelement method. The von Mises stresses were corre-lated with the critical values for various phases of thenickel-titanium material. Results: Different loadingconditions led to unequal patterns of stress distribution.Increasing the applied torque or bending angle resultedin a rise in the corresponding stresses in the instrument.Favorable stress distribution without dangerous stressconcentration was observed if the material was under-going superelastic transformation at that applied load.The ultimate strength of the material was not exceededwhen the instrument was bent up to a 50-degree curva-ture. On the other hand, when a torsional moment ofgreater than 1.0 N�mm was applied, the maximumstresses developed in some designs would exceed theultimate strength of the material. Little variation in thevon Mises stresses was observed for instruments ofdifferent nominal sizes and tapers on bending to similarextent. Conclusions: The cross-sectional design hasa greater impact than taper or size of the instrumenton the stresses developed in the instrument under eithertorsion or bending. Certain cross-sectional configura-tions are prone to fracture by excess torsional stresses.(J Endod 2010;36:1394–1398)
Key WordsFinite element analysis, nickel-titanium, root canalinstrument, rotary files, superelastic, three-dimensionalmodel

From the *State Key Laboratory for Turbulence and Complex SyUniversity, Beijing, China; and †Area of Endodontics, Comprehensiv

Address requests for reprints to Prof. Yu-Feng Zheng, Departme100871, China. E-mail address: [email protected]/$0 - see front matter


1394 Zhang et al.

The need to cut dentin and to conform to the anatomical form (curvature) of the rootcanal dictates that nickel-titanium (NiTi) rotary instruments are subject to both

torsional and bending moments in use. Various studies have reported that the methodof use and its dimension relative to the canal have a direct impact on the (torsional)stress developed in the instrument (1, 2). On bending, the imposed strain (hencestresses) on the instrument is determined by the ratio of the radius of instrument tothat of the canal curvature (3, 4). Finite element analysis (FEA) has been used toanalyze the behavior of the instruments in the dental literature (5, 6), but it was notuntil the report of Berutti and Chiandussi (7) when the nonlinear behavior of NiTi mate-rial was considered for the first time. This latter report indicated a difference in thestresses between instruments of different cross sections (7). The result was supportedby Xu et al (8), who simulated the application of a 2.5 N�mm torque on 6 commercialbrands of NiTi rotary files in an FEA software. However, other torque values were notexamined, and neither were other parameters such as the size and taper of the instru-ment. The fact that stress-induced martensitic (SIM) transformation and its reversetransformation could alter the local stress value at different regions of the cross section,as well as along the length of the instrument, should be considered in any form of simu-lation of the superelastic material to allow a better understanding of the behavior of NiTirotary instruments subjected to various forms of applied stress. The purpose of thisstudy was to examine the stresses developed in NiTi files of various cross sections, takinginto account the nonlinear mechanical behavior of the material as well as the dimensionof the instrument.

Materials and MethodsFinite element models were constructed for instruments of 10 distinctly different

cross-sectional designs by expressing all boundary conditions and geometric configu-ration numerically (Fig. 1A). The cross-sectional configuration (for several brands)was obtained by serial grinding of an embedded instrument and capturing the shapewith an image measuring device (VCAD-1010; HLEO, Beijing, China). Then an idealizedshape was established in CAD software (SolidWorks; Dassault Systemes, Velizy-Villacou-blay, France). A smart sweep meshing method, which was optimized with sufficientnumber of elements with acceptable calculation time, was used to best-fit a mesh oneach instrument in FEA software (ANSYS, Canonsburg, PA); the number of nodes mightnot be the same for all designs of instrument. Data were entered into the softwarepackage (ANSYS), together with the nonlinear mechanical properties of the NiTi mate-rial. Critical values for the start and end point of SIM transformation and the ultimatestrength of the material were obtained from published data for SE508 alloy from whichNiTi rotary files are made (9, 10). Multilinear interpolation was made to estimate the

stem and Department of Advanced Materials and Nanotechnology, College of Engineering, Pekinge Dental Care, Faculty of Dentistry, the University of Hong Kong, HKSAR.nt of Advanced Materials and Nanotechnology, College of Engineering, Peking University, Beijing



Figure 1. (A) Example of an instrument of a small S-type cross section (apical 10-mm segment only, taper 0.04 and size 30) meshed for the 3-dimensionalanalysis; the finite element model had 44,423 nodes and 9,720 elements. (B) Contour map showing the von Mises stress distribution for this instrument of differentsizes and tapers loaded to a 30-degree curvature. (This figure is available in color online at www.aae.org/joe/.)


intermediate values for the different phases of NiTi material (Table 1). Astress value exceeding 1400 MPa (the ultimate tensile strength of NiTi)was deemed to result in breakage of the instrument.

Two modes of loading were simulated. Either a torsional orbending moment was applied to the tip of instrument, with the otherend (shank) fixed. For the torsional analysis, different torque values(0.25, 0.5, 1.0, 1.5, and 2.0 N�mm) were applied to the 10 cross-sectional configurations at their tip. The radius of the circle circum-scribing each cross section was fixed at 0.15 mm (ie, size 30). Forthe bending analysis, all groups (simulated size 30, with a 0.04 taper)were curved to a maximum of 50 degrees at 10-degree intervals, withthe length of the curved (arc) portion confined to 10 mm (Fig. 1B).The degree of curvature was determined by controlling the tip displace-ment, akin to the method described by Schneider (11). The bending testwas repeated with instruments of a different size or taper.

ResultsFig. 2 depicts the distribution of von Mises stresses in the torsion

test; the (pseudo)colors corresponded to the range of stress values ona nonlinear scale (Table 1), corresponding to the transformationbetween phases of the NiTi material. Generally, the maximum stressoccurred at the periphery (or border) of the cross section and oftenwas located near the base or ‘‘bottom’’ of the flute. When the appliedtorque was small (0.25 or 0.5 N�mm), most instruments were still intheir austenitic phase, ie, within the linear elastic range of the austeniticmaterial (blue and cyan regions in Fig. 2). As the torsional momentincreased to 1.0 N�mm, there was only a small rise in this maximum

TABLE 1. Properties of various phases of NiTi (SE 508) material

Phase Stress (MPa) Strain (%)

Austenite phase (A) 0–160 0–0.8160–320320–480 0.8–1.4

Transformation phase (A + M) 480–530 1.4–2.1530–580 2.1–6.3580–760 6.3–7.7

Martensitic phase (M) 760–1160 7.7–9.81160–1400 9.8–12.6

*Pseudocolors applied to the stress contour map in FEA models.

JOE — Volume 36, Number 8, August 2010 Influence of Cross-sec

reaction stress for the NRT, both dimensions of S-type, Mtwo,convex-triangular, and the square cross section, indicating that theseinstruments were within their superelastic range, that is, the materialwas transforming with both austenite and stress-induced martensitebeing present (green region in Fig. 2). For the U-type, Quantec, RT2,and the triangular cross section, the maximum stress developed wasgreater than 1200 MPa, which indicated that these instruments hadgone beyond the superelastic plateau stage. No obvious stress concen-tration (ie, spotty areas with high stresses) was observed for some butpresent in other configurations (Fig. 2). Increasing the torsionalmoment (to 1.5 N�mm or higher) led to increasing amount of stress-induced martensitic phase, followed by elastic and then plastic defor-mation of the martensite; high von Mises stresses were observed.Breakage was anticipated for some designs: U-type triangular (1.5N�mm), Quantec (1.5 N�mm), Mtwo (2.0 N�mm), and RT2 (2.0N�mm) (Fig. 2). Convex-triangle and the (large) S-type showed thehighest resistance (with a lower stress value and with a more even stressdistribution) to torsional failure among the 10 cross-sectional config-urations (Fig. 2). The U-type Quantec and triangular cross sectionshowed the greatest susceptibility to torsional failure, according tothe maximum stress values that would develop in the instrument(Fig. 3A).

In the case of bending, there existed a neutral plane where thestress was approximately zero (blue region in Fig. 1B, which wascolor-coded according to values in Table 1). The maximum stressesoccurred on the surface of the instrument at a site furthest away fromthis neutral plane. There was no significant stress concentration inthe flute for all cross sections examined. When the angle of curvature

Stage Color coding*

Linear elasticity BlueSky-blueCyan

Superelasticity (stress-induced martensitictransformation)

Green

End of SIM transformation, and elasticrange of the martensite

KellyYellow

Plastic defloration of martensite Red

tional Design and Dimension on Mechanical Behavior of NiTi Instruments 1395


Figure 2. Contour map showing the distribution of von Mises stresses for 10 instruments of different cross sections, with each (pseudo)color corresponding to thestress range described in Table 1. (This figure is available in color online at www.aae.org/joe/.)


1396 Zhang et al. JOE — Volume 36, Number 8, August 2010


Figure 3. (A and B) Maximum values of von Mises stress developed in a size 30, 0.04 tapered instrument under torsion and bending, respectively. Those valueswere also determined for the 10 instruments (C) of the same size of #30 but different tapers and (D) of the same 0.04 taper but different sizes, all bent to a 30-degree curvature.


was increased, the stress value rose rapidly at first. The rate of increasein stress then slowed down at curvatures between 20 and 40 degrees,before it rose again on further bending (Fig. 3B). At a curvature of50 degrees, several instruments (Quantec, NRT, Mtwo, RT2, and ProFilein that order) demonstrated a much higher stress than the others(Fig. 3B). When the taper or size of the instrument was altered, littleinfluence on the maximum stress developed for various amounts ofbending was observed. The maximum von Mises stress increasedslightly when the instrument size and taper increased (Fig. 3C, D),with the area undergoing SIM transformation increased in size (greenregion in Fig. 1B). The superelastic property of the material appeared tohave maintained the stress within a safe range on bending; no significantdifference in the von Mises stresses for the various cross-sectionalconfigurations could be found for instruments of the same size butdifferent tapers (Fig. 3C) or of the same taper but different sizes(Fig. 3D).

DiscussionThe force acting on the endodontic files in use might be resolved

into 2 components, torsion and bending. It has been suggested than theload-deflection behavior for combined torsion and bending might beresolved into these 2 pure modes for analysis (12). Results from thepresent numerical analysis indicated that the NiTi instrument is morelikely to succumb to (a single application of) excess torsion than excessbending. For instance, a curvature of 50 degrees would only give rise toa stress of about 700–800 MPa on the outermost fibers of the instru-ment, which value is just into the martensitic phase but much lowerthan the material’s ultimate tensile strength (1400 MPa). However,a torsional moment greater than 1.0 N�mm would cause some instru-ments of certain designs to develop local stresses to a rather high level.Torque of this value might easily be attained if the endodontic file is

JOE — Volume 36, Number 8, August 2010 Influence of Cross-se

‘‘locked’’ during canal preparation. In other words, the excess torsionis more ‘‘dangerous’’ than bending moment (although repeatedbending would undoubtedly lead to fatigue breakage). That justifiesthe use of torque-controlled motors to avoid shear fracture of the rotaryinstruments, especially for some vulnerable designs. The result alsoindicates that rotary instruments seldom fail as a result of a singleepisode of excessive bending (ie, in canal with a small radius of curva-ture) but rather as a result of fatigue failure. Even if the instrument mightbe loaded to beyond the SIM transformation, that bending stress isinsufficient to cause immediate fracture (Fig. 3B). Continuous rotation,however, will lead to failure due to low-cycle fatigue (3, 13), with theperiphery of the instrument strained well into and sometimes beyondthe superelastic range in the curved canal.

U-file, Quantec, and file of a triangular cross section showed theleast resistance to torsional breakage than the other instruments exam-ined. Indeed, 2 commercial products both of a triangular cross sectionhave been shown to be less resistant to repeated torsional loads,compared with those of a convex-triangular cross section (14). Thismight be related to the deep flutes cut into these cross sections, resultingin a small inner core diameter (15). Because the principle of mechanicsdictates that the stress at any point in a structure is inversely propor-tional to its radial distance to the centroid of the cross section (16),any instrument with a small core diameter would be prone to torsionaloverload, an observation that was noted for NiTi rotary instruments thatwere discarded after clinical use (16). It seems that instruments ofa cross-sectional design that distributes the torsional stress well wouldbe most suited for use in constricted canals, in which situation a highreaction stress in the material is likely (2). Those possessing high flex-ibility with relatively low reaction stresses on bending would be moresuitable for preparing the more severely curved canals, on the basisof mechanical considerations. In the clinical situation, the cutting effi-ciency of the design would also be a consideration.

ctional Design and Dimension on Mechanical Behavior of NiTi Instruments 1397

It is widely accepted that instruments of a smaller diameter are
able to withstand a higher number of cycles of flexural fatigue loading(rotational bending) than those larger instruments of the same design(17–19). However, the dimensions (tip size or taper) seem to have littleeffect on the maximum von Mises stress when instrument was bent toa similar angle (eg, 30 degrees) of curvature; only a small effect onthe maximum reaction stresses could be noticed. The materialapparently was deforming within its superelastic range. Thus, thestress value stayed at about the SIM transformation threshold, despitethe increased amount of deformation (strain) for those files ofsomewhat larger size or taper. Obviously, if one were to examine thefatigue behavior of NiTi rotary files, the surface strain should beconsidered rather than the stress value for the low-cycle fatiguebehavior (3, 13). When the maximum stress developed in thematerial is concerned, our results indicated that the cross-sectionalconfiguration is the main determinant of the stresses developed inthe instrument. In other FEA studies that simulated the use of NiTi instru-ments of various designs in a curved root canal, the stresses developedin the material were also found to differ significantly for different cross-sectional configurations (20, 21). This is probably due to the fact thatstress concentration is governed by the geometry of the part rather thanthe material property (16). It is alarming that some instruments thatdeveloped high von Mises stresses on bending (Quartec, NRT, andMTwo; Fig. 3B) were also susceptible to torsional failure. This under-lines the importance of design on durability and safety of an instrument.

In summary, from this numerical analysis, it might be concludedthat certain cross-sectional configurations are more prone to failure bytorsional overload than others. The cross-sectional design of the instru-ment has a significant impact on the bending stresses developed in theNiTi rotary instrument, more so than the size and taper.

References1. Blum JY, Cohen A, Machtou P, Micallef JP. Analysis of forces developed during

mechanical preparation of extracted teeth using ProFile NiTi rotary instruments.Int Endod J 1999;32:24–31.

2. Peters OA, Peters CI, Schoneberger, Barbakow F. ProTaper rotary root canalpreparation: assessment of torque and force in relation to canal anatomy.Int Endod J 2003;36:93–9.

1398 Zhang et al.

3. Cheung GSP, Darvell BW. Fatigue testing of a NiTi rotary instrument: part 1—strain-life relationship. Int Endod J 2007;40:612–8.

4. Huston RL. Principles of biomechanics. Boca Raton, FL: CRC Press; 2009:79–140.5. Turpin YL, Chagneau F, Vulcain JM. Impact of two theoretical cross-sections on

torsional and bending stresses of nickel-titanium root canal instrument models.J Endod 2000;26:414–7.

6. Turpin YL, Chagneau F, Bartier O, Cathelineau G, Vulcain JM. Impact of torsionaland bending inertia on root canal instruments. J Endod 2001;27:333–6.

7. Berutti E, Chiandussi G. Comparative analysis of torsional and bending stresses intwo mathematical model of nickel-titanium rotary instruments: Protaper versusProfile. J Endod 2003;29:15–9.

8. Xu X, Eng M, Zheng Y, Eng D. Comparative study of torsional and bending propertiesfor six models of nickel-titanium root canal instruments with different cross-sections. J Endod 2006;32:372–5.

9. Nitinol devices and components. Materials data sheet. Available at: http://www.nitinol.com/nitinol-university/material-properties. Accessed October 15, 2009.

10. Wang GZ. A finite element analysis of evolution of stress-strain and martensite trans-formation in front of a notch in shape memory alloy NiTi. Mater Sci Engng A2007;383–91.

11. Schneider SW. A comparison of canal preparations in straight and curved canals.Oral Surg 1971;32:271–5.

12. McNaney JM, Imbeni V, Jung U, Papadopoulos P, Ritchie RO. An experimental studyof the superelastic effect in a shape-memory Nitinol alloy under biaxial loading.Mech Mater 2003;35:969–86.

13. Cheung GSP, Darvell BW. Low-cycle fatigue of NiTi rotary instruments of variouscross-sectional shapes. Int Endod J 2007;40:626–32.

14. Park S-Y, Cheung GSP, Yum J, Hur B, Park J-K, Kim H-C. Dynamic torsional resis-tance of nickel-titanium rotary instruments [published online ahead of print March19, 2010]. J Endod doi:10.1016/j.joen.2010.02.016

15. Harty FJ, Pitt Ford TR. Harty’s endodontics in clinical practice. 5th ed. Edinburgh:Wright; 2004:62.

16. Gere JM. Mechanics of materials. 5th ed. Pacific Grove, CA: Brooks/Cole; 2001:815–39.

17. Shen Y, Coil JM. Defects in nickel-titanium instruments after clinical use: part 3—a 4-year retrospective study from an undergraduate clinic. J Endod 2009;35:193–6.

18. Pruett JP, Clement DJ, Carnes DL. Cyclic fatigue testing of nickel titanium endodonticinstruments. J Endod 1997;23:77–85.

19. Bergmans L, Van Cleynenbreugel J, Wevers M, Lambrechts P. Mechanical root canalpreparation with NiTi rotary instruments: rationale, performance and safety—statusreport for the American Journal of Dentistry. Am J Dent 2001;14:324–33.

20. Kim HC, Cheung GS, Lee CJ, Kim BM, Park JK, Kang SI. Comparison of forcesgenerated during root canal shaping and residual stresses of three nickel-titanium rotary files by using a three-dimensional finite-element analysis. J Endod2008;34:743–7.

21. Kim HC, Kim HJ, Lee CJ, Kim BM, Park JK, Versluis A. Mechanical response ofnickjel-titanium instruments with different cross-sectional designs during shapingof simulated curved canals. Int Endod J 2009;42:593–602.


http://www.nitinol.com/nitinol-university/material-properties

http://www.nitinol.com/nitinol-university/material-properties

http://dx.doi.org/doi:10.1016/j.joen.2010.02.016


Investigation of Apex Locators and Related MorphologicalFactorsJiangfeng Ding, DDS,* James L. Gutmann, DDS, PhD, FACD, FICD, FADI,† Bing Fan, DDS, MSc,PhD,* Yao Lu, DDS,* and Hao Chen, DDS*

Abstract
Introduction: The purpose of this study was to investi-gate the ability of three electronic apex locators (EALs)to detect the minor foramen and morphological influ-encing factors relative to working length determination.Methods: Three hundred fifty-six extracted teeth weredecoronated, and the coronal portion of the canal wasflared. The distance between the major foramen andthe file tips (DMFF) was determined by different EALs.The relationship between the DMFFs determined bythe EAL and the morphological features of the rootapex was analyzed by linear regression analysis.Results: The average DMFFs were 0.261mm, 0.376mm, and 0.383 mm for the Root ZX (J. Morita, Kyoto,Japan), Raypex 5 (VDW, Munich, Germany), andElements Apex Locator (SybronEndo, Anaheim, CA),respectively. The file tips determined by EALs weremuch closer to the major foramen in teeth with a ‘‘lateralmajor foramen’’ (p < 0.001). The area and diameters ofthe minor foramen were significantly related to the vari-ation of the DMFFs determined by EALs. Conclusion:When the ‘‘minor foramen’’ reading was given, the filetip connected to the Root ZX was much closer to themajor foramen than the other two EALs. The minor for-amen’s morphology and the major foramen’s locationwere both important influencing factors on the perfor-mance of EALs. (J Endod 2010;36:1399–1403)
Key WordsElectronic apex locator, Root ZX, Raypex 5, ElementsApex Locator, morphological factor

From the *State Key Laboratory Breeding Base of BasicScience of Stomatology (Hubei-MOST) and Key Laboratory ofOral Biomedicine Ministry of Education, School and Hospitalof Stomatology, Wuhan University, Wuhan, China; and†Department of Endodontics, Baylor College of Dentistry, TexasA&M University system Health Science Center, Dallas, TX, USA.

Supported by the National Natural Science Foundation ofChina (grant no. 30872881).

Address requests for reprints to Dr Bing Fan, The State KeyLaboratory Breeding Base of Basic Science of Stomatology (Hu-bei-MOST) and Key Laboratory of Oral Biomedicine Ministry ofEducation, School and Hospital of Stomatology, Wuhan Univer-sity, 237 Luoyu Road, Wuhan 430079, China. E-mail address:[email protected]/$0 - see front matter



Proper root canal therapy procedures exhibit the following features: completeremoval of infected pulp tissues, thorough canal cleaning, shaping, disinfection,

and three-dimensional obturation. These purposes can be achieved only when thetermination of root canal is determined accurately (1, 2). During this process, theminor foramen or physiological foramen is the desirable endpoint for theprocedures within the root canal (3, 4) . Traditionally, tactile and radiographicmethods have been used to determine the minor foramen; however, tactile sense isempirical and the radiograph can only provide a two-dimensional image of a three-dimensional object (1, 2).

Electronic apex locators (EALs) are now widely used for locating the minorforamen. The latest generation of EALs detects the minor foramen through calculatingthe subtle variation of impedance values between the file tip within the root canal gener-ated by electrical impedances with different frequencies. Because of the developmentand advances in electronic engineering technology in the past several decades, theprecision and stability of EALs have been greatly improved (5, 6). The accuracy ofvarious EALs has been reported to be from 31% to 97.37% (7–12). In previousstudies, accurate measurements were usually defined as measuring file tips within�0.50 or �1.00 mm around the preset endpoint of the root canal (1, 2). However,it remains questionable if the major foramen was set as the endpoint. In an ex vivostudy conducted by Ounsi and Naaman (13), part of file tips, which were consideredaccurately determined by the Root ZX to the apical constriction, had actually gonebeyond the root canal when the accuracy was defined as �0.5 mm around the majorforamen. Correspondingly, the measured working length was actually unacceptable forroot canal treatment. Therefore, it might be more objective to evaluate EALs by the distri-bution of measuring file tips in relation to the major foramen.

The file tip’s location determined by EALs has been observed using radiographic(14, 15), histologic (7, 8), and direct file length measuring methods (11, 12, 16).Radiographic interpretation can be compromised by image distortions, whereas thehistologic examination is destructive to the specimens. In both in and ex vivostudies, the conventional method is to represent the file tip’s location by calculatingthe difference between the measuring file and the actual root canal length (1, 2, 11,12, 16). However, the unexpected movement of the rubber stop and the lack ofparallelism between the measuring file and gauge might result in procedural errorsthat would influence the study results (17). ElAyouti and Lost developed a mountingand measuring unit (MMU) to assist locating file tips. They found that the MMU wassuperior to the conventional method in repeatability and produced less measurementerrors (17). Preferably, it would be more appropriate to use MMU instead of othermethods to study the performance of EALs in laboratory studies.

The diameter of apical foramen has been thought to be a major factor that influ-ences the performance of EALs (6, 18). Stein et al (19) reported that the accuracy ofEAL was associated with the major foramen’s diameter but was not affected by the minorforamen’s diameter, whereas other researchers found that the measurements of EALsvaried with the minor foramen’s diameter (10, 20, 21). Moreover, Pagavino et al (10)reported that the measuring error of EAL was significantly different in teeth with differentmajor foramen locations. The purpose of this study was to investigate the ability of threeEALs, Root ZX (J. Morita, Kyoto, Japan), Raypex 5 (VDW, Munich, Germany), andElements Apex Locator (SybronEndo, Anaheim, CA), to detect the minor foramenand related morphological influencing factors during working length determination.

Investigation of Apex Locators 1399


Figure 1. Photograph of the modified mounting and measuring unit. (A) Baseof the unit, (B) micrometer and measuring file, (C) container for sodiumchloride solution, and (D) electrodes of EALs.


Materials and MethodsA total of 356 permanent teeth with a single, straight root canal that

were extracted for periodontal, orthodontic, or prosthetic reasons wereselected for this study. All the extracted teeth were numbered and storedin 10% formalin solution. The teeth were decoronated at the cementoe-namel junction, and Gate-Glidden drills (Dentsply-Maillefer, Ballaigues,Switzerland) numbers 1 through 3 were used to flare the coronal twothirds of each root canal. A physiologic sodium chloride solution(0.9%) was used for irrigation during the process, and the patencyof the apical foramen was maintained with a no. 10 K-file.

A modified MMU was built up for electronic measurements(Fig. 1). A digital micrometer (Tricle, Shanghai, China) connectedwith a no. 10 K-file (Dentsply-Maillefer) was used as the measuringfile. By rotating the screw a full circle, the no. 10 K-file moved 0.5mm vertically up or down. Before the electronic measurements, themeasuring file was introduced into the root canal until the tipbecame just visible at the most coronal border of the major foramen(�20) as viewed under a surgical operating microscope (Pico, CarlZeiss, Jena, Germany), and the reading of digital micrometer was re-corded as L0. Subsequently, the root apices were immersed intoa plastic container filled with 0.9% sodium chloride, and the RootZX, Raypex 5, and Elements Apex Locator following the manufactur-er’s instructions were used to detect the minor foramen. After fillingthe root canal with 0.9% sodium chloride, the lip clip was attachedto the container, and the electrode was connected to the K-file. Forthe Root ZX, the digital reading was recorded as the file was with-drawn from the reading ‘‘Apex’’ to the ‘‘0.5’’ flashing bar; for the Ray-pex 5, the reading was recorded when all three green bars werereached, and the symbol represented the position of minorforamen for the Elements Apex Locator. All electronic measurementswere made in triplicate, and the mean value of readings was re-corded as Le. The value was calculated by Le minus L0 and indicated

1400 Ding et al.

the distance between the major foramen and the file tips (DMFF).Positive values represented the file tip short of the major foramen,whereas negative values represented the file tip beyond the majorforamen.

Subsequently, the apical anatomic features of the tooth, includingthe minor foramen’s morphology and the major foramen’s location,were identified and measured with the aid of a stereomicroscope (StemiSv11-Apo; Carl Zeiss, Jena, Germany). This procedure was similar to theprocess described by Cheung et al (22). The major foramen wasadjusted parallel to the objective lens (�20). The teeth with tip majorforamen were labeled as ‘‘0’’ (Fig. 2A), and those with a lateral majorforamen were labeled as ‘‘1’’ (Fig. 2C). Subsequently, the image ofminor foramen was captured by adjusting the focus onto the smallestinner canal diameter through the major foramen (�40) (Fig. 2Band D). The image software (Image-Pro Plus 4.5; Media Cybernetics,Silver Spring, MD) was used to measure the minor foramen’s morpho-logical features, such as area, diameters, perimeter, and roundness. Theroundness was used to describe the shape of the minor foramen; thevalue of ‘‘1’’ represented a round object.

The difference between the DMFFs determined by three EALswas analyzed using the Friedman test followed by the Wilcoxonsigned rank test. The DMFFs determined by each EAL in the teethwith different major foramen’s location were compared with theMann-Whitney U test. The statistical difference was considered at p0.05. The influence of morphological factors on the DMFFs wasanalyzed by linear regression using the stepwise method, and theassociation between them was expressed in the linear regressionequation as follows: Y = b+bnXn, where Y is the response variableof DMFF, X is the influencing factors, b is the coefficient of regres-sion indicating that the DMFF increases or decreases with the varia-tion of the influencing factors, and b is a constant that represents themeasurements of the control.

ResultsThe average of DMFFs was 0.261 mm for the Root ZX, 0.376 mm

for the Raypex 5, and 0.383 mm for the Elements Apex Locator. Statis-tical analysis showed that there was a significant difference between theRoot ZX and the Raypex 5 (p < 0.001) as well as the Root ZX and theElements Apex Locator (p < 0.001), but no significant difference wasfound between the Raypex 5 and the Elements Apex Locator (p =0.507). Of the 356 teeth, 180 had the major foramen at the root tip,whereas 176 had a lateral major foramen. For each EAL, the DMFFsshowed a significant difference in teeth with different major foramenlocations (p < 0.001) (Table 1).

The distribution of DMFFs determined by each EAL was presentedin Table 2. File tips were located less than 0.5 mm and within the rangeof 0.5 to 1.0 mm coronal to the major foramen in 82.87% and 11.24%of cases for the Root ZX, 67.70% and 26.41% for the Raypex 5, and64.05% and 22.47% for the Elements Apex Locator, respectively. Therewere six file tips identified beyond the major foramen with the Root ZX,with two and four tips beyond when using the Raypex 5 and ElementsApex Locator, respectively. There were two file tips that were over 1.5mm short of the major foramen when using both the Root ZX and Ray-pex 5, and 24 tips were identified short when using the Elements ApexLocator.

The morphological features of the minor foramen are given inTable 3. The DMFFs determined by EALs were significantly associatedwith the apical anatomic features of the teeth (p = 0.001-0.011)(Table 4). The relationship between them was expressed by the linearregression equation as follows: for the Root ZX: Y = 0.187+1.615X1�0.08X2; for the Raypex 5: Y = 0.196 + 0.799X1� 0.061X2 + 0.714X3;


Figure 2. The major foramen (outlined in red) and the minor foramen (outlined in green). (A) Tip major foramen, (B) the corresponding minor foramen, (C)lateral major foramen, and (D) the corresponding minor foramen. (This figure is available in color online at www.aae.org/joe/.)


and for the Elements Apex Locator, Y =�0.007 + 3.106X1� 0.128X2

+ 0.879X4. In these equations, X1 is the area of minor foramen, X2 is thelocation of major foramen, X3 is the narrow diameter of minor foramen,and X4 is the wide diameter of minor foramen.

TABLE 1. The Median, Q1 (25%), and Q3 (75%) Values for the Distance fromthe File Tip to the Major Foramen (mm)

Median Q1 (25%) Q3 (75%) N

Root ZX†

T 0.359* 0.183 0.434 180L 0.209 0.080 0.337 176Total 0.261 0.128 0.394 356

Raypex 5T 0.483* 0.277 0.606 180L 0.337 0.202 0.512 176Total 0.376 0.247 0.565 356

Elements Apex LocatorT 0.667* 0.311 0.683 180L 0.335 0.211 0.520 176Total 0.383 0.248 0.599 356

Positive values indicate file tip short of major foramen.

T, the extracted teeth with tip major foramen; L, the extracted teeth with major foramen deviated from

the root main axis.

*A significant difference between the two groups of extracted teeth with different major foramen’s

location for each EAL.†A significant difference between Root ZX and the other two EALs at p < 0.001.

DiscussionIn the current study, a modified MMU was constructed to measure

the DMFFs determined by EALs. The digital micrometer was mounted toa fixed base that ensured the stability of measuring file during the oper-ation. The different values recorded by the micrometer indicated thedistance that the measuring file traveled. With the aid of the MMU, itwas unnecessary for the operators to adjust the rubber stop andmeasure the file length manually. Therefore, the procedural errorswere reduced significantly (17).

The findings of the present study, which showed that the averagedistance from the EAL-indicated position of the minor foramen to themajor foramen was 0.261 mm for the Root ZX and 0.376 mm for theRaypex 5, are supported by previous studies. By exposing the apicalpart of 20 permanent teeth, Wrbas et al (8) reported that the file tipwas identified 0.12 and 0.15 mm short of the major foramen when usingthe Root ZX and Raypex 5, respectively. Vajrabhaya et al (23) found thatthe file tips were determined on average to be 0.2 mm coronal to theapical foramen when using the Root ZX on permanent teeth. However,in an ex vivo study of Pascon et al (12) on sixty extracted teeth, the file


tip was found 0.69 mm and 1.10 mm coronal to the major foramen forthe Raypex 5 and Elements Apex Locator, respectively, which weregreater than the results in this study (0.376 mm for the Raypex 5 and0.383 mm for the Elements Apex Locator). The disagreement might



TABLE 2. Location of File Tips in Relation to the Major Foramen

Root ZX Raypex 5 Elements Apex Locator

DMFF (mm)* N % N % N %

�0.5 < DMFF # 0.0 6 1.68 2 0.56 4 1.120.0 DMFF # 0.5 295 82.87 241 67.70 228 64.050.5 < DMFF # 1.0 40 11.24 94 26.41 80 22.471. 0 < DMFF # 1.5 13 3.65 17 4.77 20 5.62DMFF > 1.5 2 0.56 2 0.56 24 6.74

*Positive values indicate file tip short of major foramen. Negative values indicate file tip beyond major foramen; DMFF, the distance of the major foramen to the file tips.

DMFF, distance between the major foramen and the file tips.

TABLE 3. Morphological Features of the Minor Foramen

Area (mm2) Mean diameter (mm) Wide diameter (mm)Narrow

diameter (mm) Perimeter (mm) Roundness

Minimum 0.015 0.134 0.150 0.112 0.439 1.00Maximum 0.943 1.086 1.342 0.934 3.611 1.91Median 0.063 0.276 0.316 0.238 0.906 1.03Q1 (25%) 0.044 0.232 0.269 0.197 0.766 1.01Q3 (75%) 0.090 0.332 0.407 0.289 1.105 1.07

TABLE 4. Results of Linear Regression of Morphological Features as the Influencing Factors

Root ZX Raypex 5 Elements Apex Locator

b 95% CI p B 95% CI p b 95% CI p

Area 1.615 1.454-1.775 0.003 0.799 0.185-1.413 0.011 3.106 2.218-3.994 0.002Wide

diameter— — — — — — 0.879 0.323-1.435 0.001

Narrowdiameter

— — — 0.714 0.185- 1.243 0.008 — — —

Location �0.080 �0.118 to �0.042 0.004 �0.061 �0.107 to �0.016 0.009 �0.128 �0.202to �0.054

0.002

If the linear relationship is not statistically significant, it is represented as —.

CI, 95% confidence interval; b, coefficient of regression.


be attributed to the number and category of teeth and different researchmethods used in different studies (1, 2, 17).

The DMFFs determined by the same EAL varied among the teethalthough all measurements were conducted under the same condition.Huang (18) stated that the moisture content in root canals and thediameter of the apical foramen are two main factors influencing theperformance of EALs. In the present study, the DMFF determined bythe Root ZX was found to increase as the minor foramen’s areaincreased, which was in accordance with previous studies. In a labora-tory study with glass tubules, Fan et al (24) found that the increase of thetubule diameter decreased the accuracy of the Root ZX when the tubuleswere filled with electrolytes. Ebrahim et al (25) relocated the smallestdiameter of the canal to the ‘‘major foramen’’ and reported that themeasured length by the Root ZX became shorter as the average diameterof the root canal increased. In an ex vivo study, Herrera et al (21)enlarged the apical constriction to three different sizes and reportedthat the precision of Root ZX decreased as the average diameter ofthe apical constriction increased. Using histologic assessment, Steinet al (19) found the file tips determined by Neosono-D (AmadentMedical and Dental Corp., Cherry Hill, NJ) deviated from the majorforamen as its diameter increased but were not affected by the diameterof minor foramen. The disparity might be explained as the results ofdifferent methods used to measure the parameters of the minorforamen as well as the different EALs used in these studies. The DMFFsdetermined by the Raypex 5 were found to increase as the area and

1402 Ding et al.

narrow diameter of the minor foramen increased, and the DMFFs deter-mined by the Elements Apex Locator varied along with the area and widediameter of the minor foramen. The difference among the three EALsmight be caused by their design concepts for processing the impedancefrom electrical currents. Root ZX calculates the ratio of impedances gener-ated from alternating currents with 8-kHz and 0.4-kHz frequencies. Raypex5 makes use of the same frequencies of alternating currents but bases themeasurement on the mean square root values of the electrical signals.Elements Apex Locator is designed to compare the resistance and capac-itance elements of the impedance separately with a built-in database todetermine the file tip’s location (1, 2, 6).

As for the major foramen’s location, the finding of this study wassupported by the research of Pagavino et al (10). The file tips weredetermined much closer to the major foramen in teeth with a lateralmajor foramen when three EALs gave the ‘‘minor foramen’’ reading.They also stressed that the DMFFs would be slightly over calculated ifthe most coronal border of major foramen was considered as the apicalreference in the case of lateral foramina. According to the study of Mer-edith and Gulabivala (26), the impedance in a root canal was a complexelectrical network comprising resistive and capacitive series andparallel elements. Nekoofar et al. (6) stated that the impedance withinroot canal varied with the shape of the canal. Therefore, it is worth inves-tigating in the further study, whether the morphological features of theapical foramen would exert influence on the intra-canal impedance,and then on the performance of EALs.


Under the conditions of this ex vivo study, the ability of three EALs
to detect the minor foramen was found to be significantly different.When the ‘‘minor foramen’’ reading was given, the file tips determinedby the Root ZX were much closer to the major foramen than when theRaypex 5 and Elements Apex Locator were used. The minor foramen’smorphology and the major foramen’s location were both importantinfluencing factors on the performance of EALs.

AcknowledgmentsThe authors thank Professor Dong-E Chen for her assistance in

the statistical part of this research.

References1. Kim E, Lee SJ. Electronic apex locator. Dent Clin North Am 2004;48:35–54.2. Gordon MP, Chandler NP. Electronic apex locators. Int Endod J 2004;37:425–37.3. Ricucci D, Langeland K. Apical limit of root canal instrumentation and obturation,

part 2. A histological study. Int Endod J 1998;31:394–409.4. Wu MK, Wesselink PR, Walton RE. Apical terminus location of root canal treatment

procedures. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;89:99–103.5. Kobayashi C. Electronic canal length measurement. Oral Surg Oral Med Oral Pathol

Oral Radiol Endod 1995;79:226–31.6. Nekoofar MH, Ghandi MM, Hayes SJ, et al. The fundamental operating principles of

electronic root canal length measurement devices. Int Endod J 2006;39:595–609.7. Tselnik M, Baumgartner JC, Marshall JG. An evaluation of root ZX and elements

diagnostic apex locators. J Endod 2005;31:507–9.8. Wrbas KT, Ziegler AA, Altenburger MJ, et al. In vivo comparison of working length

determination with two electronic apex locators. Int Endod J 2007;40:133–8.9. Briseno-Marroquin B, Frajlich S, Goldberg F, et al. Influence of instrument size on

the accuracy of different apex locators: an in vitro study. J Endod 2008;34:698–702.10. Pagavino G, Pace R, Baccetti T. A SEM study of in vivo accuracy of the Root ZX elec-

tronic apex locator. J Endod 1998;24:438–41.11. Bernardes RA, Duarte MA, Vasconcelos BC, et al. Evaluation of precision of length

determination with 3 electronic apex locators: Root ZX, Elements Diagnostic Unit


and Apex Locator, and RomiAPEX D-30. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2007;104:e91–4.

12. Pascon EA, Marrelli M, Congi O, et al. An ex vivo comparison of working lengthdetermination by 3 electronic apex locators. Oral Surg Oral Med Oral Pathol OralRadiol Endod 2009;108:e147–51.

13. Ounsi HF, Naaman A. In vitro evaluation of the reliability of the Root ZX electronicapex locator. Int Endod J 1999;32:120–3.

14. Akisue E, Gavini G, de Figueiredo JA. Influence of pulp vitality on length determina-tion by using the Elements Diagnostic Unit and Apex Locator. Oral Surg Oral MedOral Pathol Oral Radiol Endod 2007;104:e129–32.

15. Martinez-Lozano MA, Forner-Navarro L, Sanchez-Cortes JL, et al. Methodologicalconsiderations in the determination of working length. Int Endod J 2001;34:371–6.

16. Plotino G, Grande NM, Brigante L, et al. Ex vivo accuracy of three electronic apexlocators: Root ZX, Elements Diagnostic Unit and Apex Locator and ProPex. Int En-dod J 2006;39:408–14.

17. ElAyouti A, Lost C. A simple mounting model for consistent determination of theaccuracy and repeatability of apex locators. Int Endod J 2006;39:108–12.

18. Huang L. An experimental study of the principle of electronic root canal measure-ment. J Endod 1987;13:60–4.

19. Stein TJ, Corcoran JF, Zillich RM. Influence of the major and minor foramen diam-eters on apical electronic probe measurements. J Endod 1990;16:520–2.

20. Fouad AF, Rivera EM, Krell KV. Accuracy of the Endex with variations in canal irri-gants and foramen size. J Endod 1993;19:63–7.

21. Herrera M, Abalos C, Planas AJ, et al. Influence of apical constriction diameter onRoot ZX apex locator precision. J Endod 2007;33:995–8.

22. Cheung GS, Yang J, Fan B. Morphometric study of the apical anatomy of C-shapedroot canal systems in mandibular second molars. Int Endod J 2007;40:239–46.

23. Vajrabhaya L, Tepmongkol P. Accuracy of apex locator. Endod Dent Traumatol1997;13:180–2.

24. Fan W, Fan B, Gutmann JL, et al. Evaluation of the accuracy of three electronic apexlocators using glass tubules. Int Endod J 2006;39:127–35.

25. Ebrahim AK, Yoshioka T, Kobayashi C, et al. The effects of file size, sodium hypo-chlorite and blood on the accuracy of Root ZX apex locator in enlarged root canals:an in vitro study. Aust Dent J 2006;51:153–7.

26. Meredith N, Gulabivala K. Electrical impedance measurements of root canal length.Endod Dent Traumatol 1997;13:126–31.



Development of Virtual Simulation Platform for Investigationof the Radiographic Features of Periapical Bone LesionYuan Gao, DDS, PhD,* Markus Haapasalo, DDS, PhD,† Ya Shen, DDS, PhD,† Hongkun Wu, DDS,PhD,* Huiyong Jiang, MD,

‡and Xuedong Zhou, DDS, PhD*

Abstract
Introduction: The periapical radiograph is used as animportant tool in the assessment of periapical bonelesions in endodontic therapy. The purpose of this studywas to develop a virtual simulation platform for radio-graphic research of periapical bone lesions based ona digitally reconstructed radiograph and to investigatethe radiographic features of different simulated periap-ical bone lesions. Methods: A cadaver mandible wasscanned by microcomputed tomography. The applica-tion framework for the creation of a digitally recon-structed radiograph with virtual periapical lesions wasconstructed. Subsequently, different size and shapeperiapical lesions were created virtually in an incisor,a premolar, and a molar, and the digitally reconstructedradiographs were produced. Result: The detection ofperiapical lesions based on digitally reconstructed radio-graphs was depended on lesion size, position, shape,and tooth position. Virtual periapical lesions could notbe visualized with lesions smaller than 1 mm in theincisor, 2 mm in the premolar, and 3 mm in the molar,and these virtual lesions were confined within thecancellous bone. A 4-mm lesion in the molar was stillnot visualized even if it encroached on the corticalbone. If the lesions encroached on the junctional trabec-ulae and cortical bone or the lesion was created with themaximal buccal-lingual dimension in ellipsoid shape andconfined within the cancellous bone giving it anabnormal shape, it could be seen, except for the ‘‘thin-nest’’ 1-mm lesion in incisor region. Conclusions: Thevirtual simulation platform described here providesa reproducible assessment of periapical lesions andaids in a better understanding of the characteristics ofperiapical lesions. (J Endod 2010;36:1404–1409)
Key WordsDigitally reconstructed radiograph, periapical bonelesions, radiography, virtual simulation

From the *State Key Laboratory of Oral Diseases, West China CoDepartment of Oral Biological and Medical Sciences, University of BrJiujiang, China.

Supported by the Open Research Fund Program of the State KeyUniversity (grant No. 2008072).

Address requests for reprints to Dr Xuedong Zhou, State Key Lab3rd section of RenMin Nan Road, Chengdu, China 610041. E-mail a0099-2399/$0 - see front matter


1404 Gao et al.

The periapical radiograph is used as an important tool in the assessment of periapicalbone pathology (1, 2). Several studies have reported that routine periapical

radiography does not always reliably reflect the presence of a lesion and does notshow the real size of a lesion and its spatial relationship with anatomic structures(3–5). Thus, more advanced radiographic techniques for the detection of periapicallesions have been used in dentistry, including digital radiography, densitometrymethods, cone beam computed tomography (CBCT), magnetic resonance imaging,ultrasound, and nuclear techniques (6–10). Despite the advantage of thesetechniques and equipments, the apical radiograph may nonetheless be the primarydiagnostic tool available to most clinicians. The recognition and understanding ofthe radiographic features of periapical lesions will help in the diagnosis andtreatment planning of endodontic infections and in control and evaluation of healing.

Most studies of experimentally created periapical lesions have frequently led to theconclusion that the lesions are visible in the radiographs only if the junction of thecortex and cancellous bone is eroded (4, 11). In contrast to these studies, Shoha etal (12) reported that experimental lesions associated with human mandibular premo-lars and confined to cancellous bone only could be detected radiographically. Theseclassic experiments in the past several decades used dry or wet maxillas or mandiblesand created artificial periapical lesions by drills or burs and then investigated the radio-graphic features of the lesions. However, the size of artificial periapical lesions withinthe bone by drill or bur cannot be controlled easily and precisely in three-dimensionallevels, and the boundary or transitional zone between cortical and cancellous bone canseldom be accurately defined (5). Therefore, it is difficult to compare the conclusionsby various authors, in particular regarding lesions located in the transitional (or junc-tion) area between cortical and cancellous bone (6).

In recent years, the digitally reconstructed radiograph has become an importanttool in radiotherapy treatment planning, treatment verification, and computer-aidedsurgery (13). The digitally reconstructed radiograph, also called a simulated x-rayimage, is a direct ray casting technique that consists of simulating x-rays passing throughthe reconstructed computed tomography (CT) volume based on the optical absorptionmodel, thus generating an x-ray like image. According to the principle of digitally re-constructed radiographs, it is possible to get a simulated dental x-ray image withouttaking a real radiograph (13–15).

If a methodology can develop an accurate and reproducible simulated periapicallesion and produce a radiograph of it, then a correlation between the radiographicfeatures and three-dimensional morphological characteristics of periapical lesionscan be established to achieve an accurate, simulated evaluation of the lesions, which

llege and Hospital of Stomatology, Sichuan University, Chengdu, China; †Division of Endodontics,itish Columbia, Vancouver, Canada; and ‡Anatomy Laboratory, Medical College, Jiujiang University,

Laboratory of Oral Diseases of China (grant no. SKLOD008) and the Youth Fund Program of Sichuan

oratory of Oral Diseases, West China College and Hospital of Stomatology, Sichuan University, 14,ddress: [email protected].



will improve feature recognition leading to better diagnostic accuracy ofperiapical lesions. Therefore, the purpose of this study was to developa virtual simulation platform for radiographic research of periapicallesions based on a digitally reconstructed radiograph and investigatethe radiographic demonstration of simulated the lesions with variationin lesion size and location.
Materials and MethodsSelection of Specimen and Micro-CT Scanning

One male adult’s (35 years old) mandibular cadaver specimen wasobtained from the anatomy laboratory at the Medical School of JiujiangUniversity. Ethical permission was obtained from the ResearchCommittee of the West China Stomatological Hospital. The specimenwas selected on the basis that it was structurally sound and without apicalpathology and signs of previous dental treatment in the area. The spec-imen was stored in 10% formalin before use. The mandible was scannedby using a micro-CT system (m-CT-80; Scanco Medical, Bassersdorf,Switzerland) with an isotropic voxel size of 37 mm at 70 kV and 114mA. A total of 2,330 cross-sectional slice images in TIFF format wereacquired in 2,048� 2,048 pixels, and the volume image of the mandib-ular specimen as cropped into three regions of volume image for thecreation of digitally reconstructed radiographs with virtual periapicallesions, including a mandibular incisor, premolar, and molar.

Application Framework for the Creation of a DigitallyReconstructed Radiograph with a Virtual PeriapicalLesion

The MeVisLab package (available from www.mevislab.de/download/?no_cache=1) provides a visual data-flow program environ-ment on its graphic user interface (16). The graphic user interfacecontains modules connected by data pipes. Each module encapsulatesa specific function; it has a parameter panel providing a control to itsfunctions, whereas the data pipes carry input and output data betweenthem. The modules and data pipes comprise a data-flow framework forcreation of digitally reconstructed radiograph with virtual periapicallesions. The main three parts of the creation of digitally reconstructedradiograph with virtual periapical lesions are as follows:

Part 1: navigating the position of a bone lesionThe image data from the three regions of a mandibular incisor,

premolar, and molar were transferred to an image-load module, andthe region data were displayed semitransparent with volume rendering.The virtual periapical bone lesion with a given shape was navigatedunder the apex of interest, and the coordinates of the virtual lesionwere recorded for further use.

Part 2: creating the virtual periapical bone lesionsVirtual lesions of different sizes and coordinates created in part 1

were imported, and the shape of the lesion was set. The virtual lesioneliminated the bony trabeculae and lamina dura around the rootapex, but the root was not affected, simulating the ‘‘halo’’ appearanceof naturally occurring lesions according to previous literature (17).The grayscale value of the virtual lesions was obtained from values rep-resenting the bone marrow space in the area to mimic the destruction oftrabeculae and lamina dura around the root apex.

Part 3: creating a digitally reconstructed radiograph ofthe virtual periapical bone lesions

The edited image volumes with virtual periapical lesions weretransferred into the digitally reconstructed radiograph projection


module and the simulated x-ray image of the virtual periapicallesions was calculated by using a ray casting algorithm (13). Thematrix size of the simulated x-ray image was 1,081 � 811 withthe pixel size of 37 mm, which corresponds to a size of a simulatedx-ray image of 4 � 3 cm, same as with a real apical radiograph. Thesimulated x-ray image was generated to correspond to an image ob-tained by the parallel techniques. The process can be described asthe attenuation of virtual x-ray beams emitted from a source;a parallel ray x-ray source was assumed, passing through themandible with a virtual periapical lesion and ending at a ‘‘detector.’’The simulated x-ray images were calculated and displayed in a sepa-rate window and saved.

Digitally Reconstructed Radiograph of Different VirtualPeriapical Lesions and Radiographic Classification

According to the previously described application framework andprocedure, virtual periapical lesions were created in two ways: (1) thenormal way, the lesion was created around the tooth root and the centerof the lesion under the root apex with spherical or ellipsoid shape; and(2) the abnormal way, the lesion was created in the cancellous bonewith oval shape under the tooth root, and the buccolingual dimensionof the lesion was expanded as large as possible without encroaching thejunctional and cortical bone.

In the apical region of the first right incisor, five spherical virtualperiapical lesions were created with the diameter of 1 to 5 mm. In theapical region of the second left premolar, six spherical shape virtualperiapical lesions were created with a diameter of 1 to 6 mm. In addi-tion, two ellipsoid virtual periapical lesions were created with a bucco-lingual long axis of 7 and 8 mm and a 6-mm mesiodistal diameter(round shape). In the first left mandibular molar apical region, sixspherical virtual periapical lesions were created with a diameter of 1to 6 mm. In addition, four ellipsoid virtual periapical lesions werecreated with a buccolingual long axis of 7 to 10 mm, all with a mesio-distal diameter of 6 mm. All of these virtual periapical lesions created innormal shape.

In addition, different-sized ellipsoid virtual periapical bone lesionswere created in the cancellous bone with an oval shape under the toothroot in an abnormal shape; the buccal-lingual dimension of the lesionwas expanded as large as possible without encroaching the junctionaland cortical bone, with a mesiodistal diameter (round shape) of 1,2, and 3 mm in the incisor and 1, 2, 3, and 4 mm in the premolarand molar. Finally, the digitally reconstructed radiographs of eachvirtual periapical lesions with different sizes and positions were calcu-lated, and 34 sets of digitally reconstructed radiograph were obtained.Radiolucent areas were classified into three categories: detection/visu-alization with a definite radiolucency (V), slight or questionable visual-ization of radiolucency (SQV), and no visualization of any radiolucentareas (NV). The digital radiographic images were evaluated by threeendodontists, and the classification was decided by consensus whenthere was a disagreement initially. All data were processed on a DellT7400 workstation (Dell Computer Corporation, Round Rock, TX)running on 64-bit Windows XP (Microsoft Corp, Redmond, WA).

ResultsA virtual simulation platform was developed for simulation radiog-

raphy. Virtual periapical lesions of different sizes can be created withaccurate dimensions and position, and the corresponding digitally re-constructed radiograph can be generated accordingly as shown inFigures 1 through 4.

The detection of the different bone lesions is shown in Table 1.Virtual lesions could not be visualized by the radiographs if the virtual

Virtual Simulation Platform 1405

http://www.mevislab.de/download/?no_cache=1

http://www.mevislab.de/download/?no_cache=1

Figure 1. (A-E) The virtual periapcial bone lesion with a diameter of 1 to 5 mm. The 3D volume-rendering images and digitally reconstructed radiographs ofdifferent-sized virtual periapical lesions in the incisor. The left image pair shows the dimension and the relation of periapical lesion with bone, and the right imageshows the digitally reconstructed radiograph in each image from A to E. (This figure is available in color online at www.aae.org/joe/.)


periapical lesions were confined within the cancellous structure (Figs.1A, 2A and, and 3A-D) and created in a normal shape. However, virtualperiapical lesions could be seen (Fig. 1B, 2C, and 3E) when ‘‘normal’’periapical lesions were enlarged, and they encroached on the junctionaltrabeculae and cortical bone or when the periapical lesions werecreated within cancellous structure in an abnormal shape, with theexception of the smallest 1-mm lesion in the incisor region.

Figure 2. (A-H) The virtual periapical bone lesion with a diameter of 1 to 6 mm anddigitally reconstructed radiographs of different-sized virtual periapical lesions in the premwith bone, and the bottom image shows the digitally reconstructed radiograph in each

1406 Gao et al.

Virtual Periapical Bone Lesions with Normal ShapeIn the incisor region, the 1-mm lesion was confined to cancellous

bone, and no radiolucency was detected (Fig. 1A). The outer edges ofthe 2-mm lesion touched the junctional trabeculae but without obviouscortical involvement, and a slight radiolucency can be observed(Fig. 1B). The 3-mm lesion penetrated into the buccal cortical plateand included half of its thickness (Fig. 1C), whereas the 4- and 5-

7 to 8 mm in the buccolingual dimension. The 3D volume-rendering images andolar. The top image pair shows the dimension and the relation of periapical lesion

image from A to H. (This figure is available in color online at www.aae.org/joe/.)




Figure 3. (A-J) The virtual periapical lesion with a diameter of 1 to 6 mm and 7 to 10 mm in the buccolingual dimension. The 3D volume-rendering images and digitallyreconstructed radiographs of different-sized virtual periapical lesions in the molar. The top image pair shows the dimension and the relation of periapical bone lesionwith bone, and the bottom image shows the digitally reconstructed radiograph in each image from A to J. (This figure is available in color online at www.aae.org/joe/.)


mm lesions involved the full thickness of the buccal cortical plate andperforated it (Fig. 1D and E). In the premolar region, the lesions witha diameter of 1 and 2 mm were confined to the cancellous bone (Fig. 2Aand B). The outer edge of the 3-mm lesion was in contact with the junc-tional bone with little cortical involvement, and radiolucency was barelyobserved (Fig. 2C). The 4- to 6-mm lesions (round shape) (Fig. 2D-F)and ellipsoid lesions with a 7-mm long axis (buccolingual) (Fig. 2G)involved the buccal cortical plate from one fourth (0.5 mm) to fullthickness (2 mm) of the buccal cortical bone, and 8-mm ellipsoidlesions perforated buccal cortical bone (Fig. 2H). In the molar region,the 1- to 3-mm lesions (Fig. 3A-C) were confined to the cancellousbone, and the 4-mm lesion touched the junctional bone, but no radio-lucency was observed (Fig. 3D). The 5-mm lesion (Fig. 3E) penetratedthe cortical bone for about 0.5 mm, and radiolucency was observed.The round 6-mm lesion and ellipsoid lesions with a 7- to 9-mm longaxis (buccolingual) (Fig. 3F-I) involved the buccal cortical platefrom 1 mm to full thickness (2.5 mm). Ellipsoid lesions with a 10-mm long axis perforated the buccal cortical bone, and clear radiolu-cencies could be observed (Fig. 3J).

Virtual Periapical Bone Lesions with Abnormal ShapeIn the incisor region, the entire virtual lesion was confined to

cancellous bone (Fig. 4A). No radiolucency was observed when thelesion diameter was only 1 mm (Fig. 4A1). However, lesions witha diameter of 2 mm or more were detected in the digitally reconstructedradiograph although the lesions did not affect cortex (Fig. 4A2-3). In thepremolar and molar regions, all lesions with abnormal shape


(maximum buccolingual length without cortical involvement) were de-tected in the digitally reconstructed radiograph (Fig.4B1-4C1-4), evenwith the smallest mesiodistal diameter of 1 mm.

DiscussionAn understanding of how periapical bone lesions are represented

on the radiographic image is fundamental to the use of these images asa diagnostic aid or in designing research involving radiographic images.In this article, a platform for intraoral radiographic simulation of peri-apical lesions is presented. A digitally reconstructed radiograph is theartificial version of an x-ray image, and the digitally reconstructed radio-graph module in MeVisLab package was used to create a two-dimensional digitally reconstructed radiograph projection of a three-dimensional (3D) CT dataset. In 3D CT, the data provide informationof the x-ray attenuation of the object (18). Although CT providesa 3D representation of the object’s attenuation coefficients and enablesthe computation of virtual radiographs in any imaging geometry,modern spiral CT, or CBCT devices still provide insufficient spatial reso-lution for realistic virtual intraoral imaging. Hence, we used micro-CT togenerate the digitally reconstructed radiograph, with a resolutionsimilar to conventional periapical radiography. Based on the high-resolution micro-CT image, 3D image manipulation techniques byapplication framework were applied virtually to periapical bone lesionmodel to facilitate realistic simulation of lesions or other changes thatmay have taken place between the virtual evaluations. The imagesshowed that the results after these manipulations are similar to thoseafter the removal of local bone or tissue by means of a drill. More



Figure 4. The 3D volume-rendering images and digitally reconstructed radiographs of different-sized virtual periapical lesions in the incisor, premolar, and molarregions created into an abnormal shape. (A-C) The 3D volume-rendering images with a different-sized periapical bone lesions in the incisor, the premolar, and themolar. The digitally reconstructed radiographs of different-sized virtual periapical lesions: incisor, lesion diameter 1 to 3 mm (A1-A3); premolar, lesion diameter 1to 4 mm (B1-B4); molar, lesion diameter 1 to 4 mm (C1-C4). (This figure is available in color online at www.aae.org/joe/.)


sophisticated models of tissue modification for dental hard structurecan also be performed in this virtual simulation platform. In endodonticradiology education, it would be beneficial to simulate radiographicprojections to facilitate and in part replace traditional teachingmethods. The simulation technique can also be useful when analyzingprojections (eg, by observing relevant bone lesion inside the models).This would probably enhance understanding of the characteristics ofperiapical bone lesions and thus improve the diagnostic accuracy ofperiapical bone lesions.

The contribution of radiography to the diagnosis of periapicalbone lesions has been the subject of extensive research. It is commonto find clinical signs of bone disease despite negative radiographic find-ings. In these instances, there may be pronounced tissue destruction,but the destroyed tissue is entirely in cancellous bone. Inflammatorycells are substituted for marrow cells, a pathologic condition thatdoes not allow detection in a radiograph. Previous investigations ofnaturally occurring and experimentally created periapical bone lesionshad shown that lesions can be distinguished on roentgenograms if theyerode the junction area of the cortex and cancellous bone or perforatethe cortex because cortical bone contains much more calcium per unitof volume than cancellous bone (3). In the present study, we created thevirtual periapical bone lesions in two different shapes. First, the lesionwas created around the tooth root and the center of the lesion under theroot apex with a spherical or ellipsoid shape, simulating normal shapeand position of a periapical lesion. The findings favor the concept thatcortical plate or at the least junctional trabeculae involvement isa prerequisite for the radiographic delectability of periapical lesionsin incisor, premolar, and molar region. However, a 4-mm lesion in

1408 Gao et al.

the molar region was barely seen despite the involvement of junctionaltrabeculae. It is possible that a periapical lesion of a certain size can bedetected in a region covered by a thin cortex, whereas a similar lesionwill not be detected in a region covered by a thicker cortex (19). Inaddition to ‘‘normal’’ lesions, we also created a number of untypicallesions entirely confined in cancellous bone to further analyze possi-bility of detecting such lesions. This may be explained by the fact thatthe junctional and transitional area between the cortical and cancellousbone as defined in our experiment may not be the same as in earlierstudies. We found that the junction area may not exist as an anatomicallydemonstrable structure; it is the transitional area between cortical andcancellous bone where the marrow spaces become smaller than in thecancellous bone but are still visible in the micro-CT image. However, thedistinct boundary cannot be determined because it was reported by vander Stelt (5). In addition, the shape of the artificial lesions in the presentstudy was well defined, but with clinical lesions this is often not the case.Some previous studies have also shown that the removal of a largeamount of cancellous bone can make a lesion noticeable in certaincircumstances (5, 12).

In view of the experimental evidence presented, early stages ofperiapical bone inflammation cannot be detected by means of conven-tional roentgenograms, and the radiographic appearance of differentlesions is affected by many factors, including the lesion shape, location,and size. Thus, although early detection of a change is an important stepin the assessment of treatment efficacy and patient management, it maybe difficult because the changes are subtle. Hopefully, more sensitivetechnique such as CBCT and DSR will provide more reliable method-ology for early detection of periapical pathology (20, 21).



TABLE 1. Radiographic Detection of Different Virtual Periapical Bone Lesions

Periapical bone lesion of normal shape Periapical bone lesion of abnormal shape

Bone location Bone lesion size (mm) Radiolucency Bone lesion size (mm) Radiolucency

Incisor 1 NV 1 NV2 SQV 2 SQV3 V 3 SQV4 V5 V

Premolar 1 NV 1 V2 NV 2 V3 SQV 3 V4 V 4 V5 V6 V

7 (B-L long axis) V8 (B-L long axis) V

Molar 1 NV 1 V2 NV 2 V3 NV 3 V4 NV 4 V5 SQV6 V

7 (B-L long axis) V8 (B-L long axis) V9 (B-L long axis) V10 (B-L long axis) V

V, lesion visible; SQV, slight or questionable visualization; NV, lesion not visible; B-L, buccolingual.


The results of the present study have clinical, educational, andresearch implications. It is important that clinicians and dental studentsrealize the limits of conventional radiographic techniques. Although theresults from virtual simulation models cannot always completely extrap-olate to the in vivo/patient situation, they can provide valuable insightinto the early detection of periapical lesions.

References1. Stheeman SE, Mileman PA, van ’t Hof MA, et al. Diagnostic confidence and the accu-

racy of treatment decisions for radiopaque periapical lesions. Int Endod J 1995;28:121–8.

2. Barbat J, Messer HH. Detectability of artificial periapical lesions using direct digitaland conventional radiography. J Endod 1998;24:837–42.

3. Cotti E, Campisi G, Garau V, et al. A new technique for the study of periapical bonelesions: ultrasound real time imaging. Int Endod J 2002;35:148–52.

4. Bender IB, Seltzer S. Roentgenographic and direct observation of experimentallesions in bone: 2. J Am Dent Assoc 1961;62:708–16.

5. van der Stelt PF. Experimentally produced bone lesions. Oral Surg Oral Med OralPathol 1985;59:306–12.

6. Huumonen S, Ørstavik D. Radiological aspects of apical periodontitis. EndodonticTopics 2002;1:3–25.

7. Cotti E, Campisi G. Advanced radiographic techniques for the detection of lesions inbone. Endodontic Topics 2004;7:52–72.

8. Nielsen RB, Alyassin AM, Peters DD, Carnes DL, Lancaster J. Microcomputed tomog-raphy: an advanced system for detailed endodontic research. J Endod 1995;21:561–8.

9. Cotton TP, Geisler TM, Holden DT, Schwartz SA, Schindler WG. Endodontic appli-cations of cone-beam volumetric tomography. J Endod 2007;33:1121–32.


10. Nair MK, Nair UP. Digital and advanced imaging in endodontics: a review. J Endod2007;33:1–6.

11. Bender IB, Seltzer S. Roentgenographic and direct observation of experimentallesions in bone: 1. J Am Dent Assoc 1961;62:152–60.

12. Shoha RR, Dowson J, Richards AG. Radiographic interpretation of experimen-tally produced bony lesions. Oral Surg Oral Med Oral Pathol 1974;38:294–303.

13. Li X, Yang J, Zhu Y. Digitally reconstructed radiograph generation by an adaptiveMonte Carlo method. Phys Med Biol 2006;51:2745–52.

14. Pehlivan B, Pichenot C, Castaing M, et al. Interfractional set-up errors evaluation bydaily electronic portal imaging of IMRT in head and neck cancer patients. Acta On-col 2009;48:440–5.

15. Thompson CM, Hamilton CS, Vaarkamp J. Thorax set-up verification with multipleoblique treatment portal images. Br J Radiol 2009;82:950–5.

16. Gao Y, Peters OA, Wu H, et al. An application framework of three-dimensionalreconstruction and measurement for endodontic research. J Endod 2009;35:269–74.

17. Radiographic appearance of artificially prepared periapical lesions confined tocancellous bone. Int Endod J 1986;19:64–72.

18. Nilsson T, Ahlqvist J, Johansson M, et al. Virtual reality for simulation of radio-graphic projections: validation of projection geometry. Dentomaxillofac Radiol2004;33:44–50.

19. Estrela C, Bueno MR, Leles CR, et al. Accuracy of cone beam computed tomographyand panoramic and periapical radiography for detection of apical periodontitis. JEndod 2008;34:273–9.

20. Cotton TP, Geisler TM, Holden DT, et al. Endodontic applications of cone-beamvolumetric tomography. J Endod 2007;33:1121–32.

21. Miguens SA Jr, Veeck EB, Fontanella VR, et al. A comparison between panoramicdigital and digitized images to detect simulated periapical lesions using radiographicsubtraction. J Endod 2008;34:1500–3.



Comparison of the Vibringe System with Syringe and PassiveUltrasonic Irrigation in Removing Debris from SimulatedRoot Canal IrregularitiesTina Rodig, Dr. med. dent,* Meral Bozkurt,* Frank Konietschke, Dr,

†

and Michael Hulsmann, Prof. Dr. med. dent*

Abstract
Introduction: The aim of this study was to compare theefficiency of a sonic device (Vibringe), syringe irrigation,and passive ultrasonic irrigation in the removal of debrisfrom simulated root canal irregularities. Methods: Rootcanals with 2 standardized grooves in the apical andcoronal parts were filled with dentin debris. Threedifferent irrigation procedures were performed withNaOCl (1%) and (1) syringe irrigation, (2) Vibringe,and (3) passive ultrasonic irrigation. The amount of re-maining debris was evaluated by using a 4-gradescoring system. Results: Ultrasonic irrigation removeddebris significantly better from the artificial canal irreg-ularities than the Vibringe System and syringe irrigation(P < .0001). The Vibringe System demonstrated signifi-cantly better results than syringe irrigation in the apicalpart of the root canal (P = .011). Conclusions: Passiveultrasonic irrigation is more effective than the VibringeSystem or syringe irrigation in removing debris. Thesonic device demonstrated significantly better resultsthan syringe irrigation in the apical root canal third.(J Endod 2010;36:1410–1413)
Key WordsDebris, irrigation, root canal, sonic, ultrasonic, VibringeSystem

From the )Department of Preventive Dentistry, Periodon-tology and Cariology, and †Centre for Statistics, University ofGottingen, Gottingen, Germany.

Address requests for reprints to Dr T. Rodig, Department ofPreventive Dentistry, Periodontology and Cariology, Universityof Gottingen, Robert-Koch-Str. 40, 37075 Gottingen, Germany.E-mail address: [email protected]/$0 - see front matter


1410 Rodig et al.

Disinfection of the root canal system by using antimicrobial and tissue-dissolving ir-rigants is considered an essential part of chemomechanical debridement (1).

Residual pulp tissue, bacteria, and dentin debris remain in areas that are routinelyleft uninstrumented after root canal preparation (2, 3). Reports on the efficacy ofirrigation at different coronal-apical levels have been contradictory (4, 5).Therefore, a coronal groove was added to an existing research model (6) to evaluatedebris removal not only from apical but also from coronal thirds of the root canal.Conventional syringe irrigation is still widely accepted (3), although its flushing actionis not sufficient in removing debris from root canal irregularities (2, 7). Enhancementof the flushing action of irrigants by ultrasound is well-documented (8, 9) and has thepotential to remove dentin debris and organic tissue from inaccessible root canal areas(10, 11). Conflicting results regarding the effectiveness of sonic or ultrasonic activationof the irrigant to remove smear layer, debris, and bacteria (12–14) have beenpublished. Recently, the Vibringe System (Vibringe B. V. Corp, Amsterdam,Netherlands), an irrigation device that combines manual delivery and sonicactivation of the solution, has been introduced. The Vibringe is a cordless handpiecethat fits in a special disposable 10-mL Luer-Lock syringe that is compatible with everyirrigation needle. No data on the effectiveness of this system are available at present.

The aim of the present study was to compare the efficiency of conventional syringeirrigation, the Vibringe System, and passive ultrasonic irrigation (PUI) in the removal ofdentin debris.

Material and MethodsSpecimen Preparation

Ten extracted maxillary lateral incisors with straight roots were decoronated toobtain a standardized root length of 17 mm. The root canals were prepared with Flex-Master (VDW, Munich, Germany) nickel-titanium rotary instruments to a workinglength (WL) of 16 mm and an apical size of #35/02. Between the instruments, irrigationwas performed with 2 mL NaOCl (1%) by using a syringe and a 30-gauge needle (Na-viTip; Ultradent, South Jordan, UT). The roots were split longitudinally into 2 halves,allowing subsequent reassembling. A modified finger spreader was inserted into anultrasonic handpiece (Piezon Master 400; EMS, Nyon, Switzerland) to cut 1 longitudinalgroove of 4.0-mm length, 0.2-mm width, and 0.5-mm depth into root canal dentin ofeach half. The locations of the grooves were 2–6 mm from WL in one root half (apicalsection) and 10–14 mm from WL in the opposite half (coronal section). This experi-mental design is based on a previous study (6) and has been used in several investiga-tions concerning the removal of debris (15–17). Subsequently, digital photographs ofthe root halves were taken before and after irrigation from identical angles by usinga microscope (MOTIC Ergonomic Trinokular Zoom Stereo Mikroskop; Motic,Wetzlar, Germany) with 30� magnification. Dentin debris was produced by mixing100 mg dentin chips with 0.175 mL NaOCl (1%) in a standardized ratio to achievea wet sand-like consistency. Each groove was filled with debris to simulate a clinicalsituation when dentin debris accumulates in uninstrumented root canal extensions.Subsequently, the root halves were reassembled and fixed by using a clamp.




Irrigation ProceduresPreliminary experiments had shown that a single specimen could

be reused up to at least 5 times without visible damage to the root canalsurface. Therefore, the 10 teeth were used repeatedly in the 3 experi-mental groups. The irrigation procedures were performed consecu-tively with a random sequence of the specimens. In group 1,irrigation was accomplished with a 10-mL syringe and a 30-gauge nee-dle (NaviTip). In group 2, the irrigant was delivered and sonically acti-vated via the Vibringe System by using a 30-gauge needle (NaviTip)according to manufacturer’s instructions. In group 3, irrigation wasperformed with an ultrasonic device (Piezon Master 400) and a stainlesssteel K-type file size 15 (Endosonore; Maillefer, Ballaigues,Switzerland), with its power set at the 1⁄4 of the scale. In all groups a totalvolume of 20 mL NaOCl (1%) was delivered. The flow rate in groups 1and 2 was approximately 5 mL/min. In group 3, the delivery rate duringPUI with a continuous flush of the irrigant was 10 mL/min. Insertiondepth of the needle and the ultrasonic file was 1 mm short of WL inall groups. After irrigation the root halves were separated to take digitalphotographs of the canal walls. The remaining debris was removedfrom the grooves, followed by a complete refilling of the root canalextensions before the next irrigation procedure. All measures werecarried out under a microscope at 30� magnification.

Scoring Procedure and Statistical AnalysisThe amount of remaining debris in the grooves was scored under

the microscope with 30� magnification by 2 calibrated dentists witha scoring system described previously (18): 0, the groove is empty;1, less than half of the groove is filled with debris; 2, more than halfof the groove is filled with debris; 3, the complete groove is filledwith debris (Fig. 1). Intraobserver reproducibility and interobserveragreement were calculated. In cases of differences, both scores wereincluded in the statistical analysis that was performed with a nonpara-

Figure 1. Standardized debris score for grooves according to van der Sluis et al (1filled with debris; (C) score 2: more than half of the groove is filled with debris; (D)


metric analysis of variance for factorial longitudinal data and the closedtesting principle (P = .05).

ResultsInterobserver agreement was 90% (k = 0.9057, confidence

interval = 0.8310–0.9804), and intraobserver reproducibility was98% (k = 0.9843, confidence interval = 0.9626–1), with no significantinfluence of the observer (P = .9825). The results of the scoring proce-dure are presented in Fig. 2 and Table 1. There were statistically signif-icant differences between the experimental groups (P < .0001) and thelocation of the groove (P < .0001). A significant interaction between theirrigation protocol and the location of the groove (P = .018) was found.For syringe irrigation and Vibringe System, pairwise comparisonsdemonstrated significantly better cleanliness of the apical groove(P = .005; P = .002, respectively). No difference between the grooveswas detected for ultrasonics (P = .160), which removed debris signif-icantly better than Vibringe System and syringe irrigation (P < .0001),irrespective of the location of the groove. For the coronal groove, thedifference between syringe irrigation and the use of the Vibringe Systemwas not statistically significant (P = 1). In contrast, the results for theapical groove demonstrated a significantly better performance for Vi-bringe System than for syringe irrigation (P = .011). Overall, the clean-liness of the apical groove was significantly superior in comparison tothe coronal groove (P < .0001). None of the specimens irrigated witha syringe showed completely clean artificial root canal irregularitieswithout any remaining debris (score 0). Debris was completelyremoved after irrigation with the Vibringe System in 5% of the speci-mens and after PUI in 92.5% of the specimens.

DiscussionThe design of the present study is comparable to that described

by Lee et al (19) and has been used in several other investigations

8). (A) Score 0: the groove is empty; (B) score 1: less than half of the groove isscore 3: the complete groove is filled with debris. Original magnification,�30.

Efficiency of Sonic Device, Syringe Irrigation, and PUI in Removal of Debris 1411

Figure 2. Frequencies of pooled evaluations for investigator and root half.Readings were generated from 20 root halves per group scored by 2 observers,resulting in 40 readings per irrigation procedure (40 grooves).


(15, 17, 18, 20). The advantage of the groove model is the standardizedsize and location of the grooves, allowing a consistent evaluation withhigh intraobserver reproducibility and good interobserver agreement.Because the needle tip and the ultrasonic file do not have a physicaleffect on the debris in the groove, the flushing action of the irrigant isthe main factor for debris removal. The major disadvantage of thismodel is that the standardized grooves do not represent thecomplexity of a natural root canal system. Therefore, it might beeasier to remove dentin debris from artificial grooves than fromisthmuses or oval extensions in vivo, resulting in an overestimationof the removal efficiency of different irrigation techniques.

The results indicated that PUI removed significantly more debrisfrom root canal irregularities than the sonically activated VibringeSystem and syringe irrigation. A more effective removal of debris withultrasonic irrigation compared with sonic activation has been demon-strated (12, 14, 16), which could be due to the higher driving frequencyof ultrasound (30 kHz) in comparison to the sonic device (150 Hz).Therefore, the flow velocity and the cleaning efficiency are lower forsonically activated irrigation (14, 21), resulting in less effectivedelivery of irrigant to the root canal extensions. The generalconsensus that ultrasonic irrigation is more effective than syringeirrigation in removing remnants of debris (11, 19, 22) is confirmedby the results of the present study.

The percentages of complete debris removal (score 0) for sonicand ultrasonic irrigation were 5% and 92.5%, respectively. Theseresults are in agreement with a recent study that reported on 5.5%–6.7% completely clean root canals after sonic irrigation and 89% afterultrasonic irrigation (16).

In this study, the flow rate was approximately 5 mL/min forconventional manual or sonically activated irrigation and 10 mL/minfor PUI. During PUI with a continuous flush, the volume and flow

TABLE 1. Frequency Distribution of Debris Score by Experimental Groups andLocation of the Groove

Score

Group Location of the groove 0 1 2 3

Syringe Coronal 0 3 13 4Apical 0 10 10 0

Vibringe Coronal 0 5 9 6Apical 2 15 3 0

Ultrasonics Coronal 17 3 0 0Apical 20 0 0 0

Readings were pooled for both observers.

Score 0: groove is empty; score 1: less than half of groove is filled with debris; score 2: more than half

of groove is filled with debris; score 3: complete groove is filled with debris.

1412 Rodig et al.

rate of the irrigant that enters the apical part of the root canal cannotbe standardized (23). Although irrigant flow rate is considered a highlysignificant factor determining flow pattern in fluid dynamics (24), itis unknown whether it influenced the performance of ultrasonicirrigation.

Controversy exists regarding the removal of debris from differentparts of the root canal. Debris removal from the coronal part is consid-ered to be easier than from the apical part (14), whereas other authorsfound no differences among root canal thirds (25). In this study, a previ-ously described tooth model (6) was modified to determine root canalcleanliness at apical as well as coronal levels. The overall evaluation re-vealed that cleanliness of the apical groove was superior to cleanlinessof the coronal groove. All irrigation devices were placed 1 mm short ofWL in close proximity to the location of the apical groove. Therefore, theintroduction depth of the needle tip and the distance to the groovesseem to play an important role in the removal of debris, reinforcingthe benefit of the physical flushing action (4, 26). Although a finalirrigation protocol with apical cleaning as the main goal was tested,both grooves were filled with standardized amounts of debris. Itmight be speculated that the coronal third is flushed more often withNaOCl during the clinical procedure, resulting in better debridementcoronally.

The Vibringe System performed similarly to conventional syringeirrigation in the coronal part but removed significantly more debrisin the apical part. A possible explanation is that the oscillation amplitudeof the sonically activated irrigation needle is higher at the tip than at theattached end (16, 27), resulting in an increased fluid velocity. In thecoronal part the larger distance of the needle or file tips to thegroove seems to reduce the efficacy of the agitation of the irrigant.

In conclusion, the present study showed that none of the testedirrigation devices were able to completely remove debris from artificialextensions in straight root canals. PUI removed significantly moredebris than syringe irrigation or a sonically activated device (Vibringe).The Vibringe System performed significantly better than conventionalsyringe irrigation in the apical part of the root canal.

References1. Haapasalo M, Endal U, Zandi H, Coil J. Eradication of endodontic infection by instru-

mentation and irrigation solutions. Endodontic Topics 2005;10:77–102.2. Wu MK, Wesselink PR. A primary observation on the preparation and obturation of

oval canals. Int Endod J 2001;34:137–41.3. Peters OA. Current challenges and concepts in the preparation of root canal systems:

a review. J Endod 2004;30:559–67.4. Huang TY, Gulabivala K, Ng YL. A bio-molecular film ex-vivo model to evaluate the

influence of canal dimensions and irrigation variables on the efficacy of irrigation.Int Endod J 2008;41:60–71.

5. Wu MK, Wesselink PR. Efficacy of three techniques in cleaning the apical portion ofcurved root canals. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1995;79:492–6.

6. Lee SJ, Wu MK, Wesselink PR. The efficacy of ultrasonic irrigation to remove arti-ficially placed dentine debris from different-sized simulated plastic root canals.Int Endod J 2004;37:607–12.

7. Cunningham WT, Martin H, Forrest WR. Evaluation of root canal debridement by theendosonic ultrasonic synergistic system. Oral Surg Oral Med Oral Pathol 1982;53:401–4.

8. Stock CJ. Current status of the use of ultrasound in endodontics. Int Dent J 1991;41:175–82.

9. van der Sluis LW, Versluis M, Wu MK, Wesselink PR. Passive ultrasonic irrigation ofthe root canal: a review of the literature. Int Endod J 2007;40:415–26.

10. Gutarts R, Nusstein J, Reader A, Beck M. In vivo debridement efficacy of ultrasonicirrigation following hand-rotary instrumentation in human mandibular molars. J En-dod 2005;31:166–70.

11. Passarinho-Neto JG, Marchesan MA, Ferreira RB, Silva RG, Silva-Sousa YT, Sousa-Neto MD. In vitro evaluation of endodontic debris removal as obtained by rotaryinstrumentation coupled with ultrasonic irrigation. Aust Endod J 2006;32:123–8.


12. Stamos DE, Sadeghi EM, Haasch GC, Gerstein H. An in vitro comparison study to
quantitate the debridement ability of hand, sonic, and ultrasonic instrumentation.J Endod 1987;13:434–40.

13. Jensen SA, Walker TL, Hutter JW, Nicoll BK. Comparison of the cleaning efficacy ofpassive sonic activation and passive ultrasonic activation after hand instrumentationin molar root canals. J Endod 1999;25:735–8.


15. van der Sluis LW, Wu MK, Wesselink PR. The efficacy of ultrasonic irrigation to re-move artificially placed dentine debris from human root canals prepared usinginstruments of varying taper. Int Endod J 2005;38:764–8.

16. Jiang LM, Verhaagen B, Versluis M, van der Sluis LW. Evaluation of a sonic devicedesigned to activate irrigant in the root canal. J Endod 2010;36:143–6.

17. van der Sluis LW, Wu MK, Wesselink PR. A comparison between a smooth wire anda K-file in removing artificially placed dentine debris from root canals in resinblocks during ultrasonic irrigation. Int Endod J 2005;38:593–6.

18. van der Sluis LW, Wu MK, Wesselink PR. The evaluation of removal ofcalcium hydroxide paste from an artificial standardized groove in the apicalroot canal using different irrigation methodologies. Int Endod J 2007;40:52–7.

JOE — Volume 36, Number 8, August 2010 E

19. Lee SJ, Wu MK, Wesselink PR. The effectiveness of syringe irrigation and ultrasonicsto remove debris from simulated irregularities within prepared root canal walls. IntEndod J 2004;37:672–8.

20. van der Sluis LW, Wu MK, Wesselink P. Comparison of 2 flushing methods usedduring passive ultrasonic irrigation of the root canal. Quintessence Int 2009;40:875–9.

21. Ahmad M, Pitt Ford TR, Crum LA, Walton AJ. Ultrasonic debridement of root canals:acoustic cavitation and its relevance. J Endod 1988;14:486–93.

22. Cheung GS, Stock CJ. In vitro cleaning ability of root canal irrigants with and withoutendosonics. Int Endod J 1993;26:334–43.

23. van der Sluis LW, Gambarini G, Wu MK, Wesselink PR. The influence of volume, type ofirrigant and flushing method on removing artificially placed dentine debris from theapical root canal during passive ultrasonic irrigation. Int Endod J 2006;39:472–6.

24. Tilton JN. Fluid and particle dynamics. In: Perry RH, Green DW, Maloney JO, eds.Perry’s chemical engineer’s handbook. 7th ed. New York: McGraw-Hill; 1999:1–50.

25. Munley PJ, Goodell GG. Comparison of passive ultrasonic debridement betweenfluted and nonfluted instruments in root canals. J Endod 2007;33:578–80.

26. Abou-Rass M, Piccinino MV. The effectiveness of four clinical irrigation methods onthe removal of root canal debris. Oral Surg Oral Med Oral Pathol 1982;54:323–8.

27. Ahmad M, Pitt Ford TJ, Crum LA. Ultrasonic debridement of root canals: acousticstreaming and its possible role. J Endod 1987;13:490–9.

fficiency of Sonic Device, Syringe Irrigation, and PUI in Removal of Debris 1413


Effects of Storage Temperature on Surface Hardness,Microstructure, and Phase Formation of White MineralTrioxide AggregateMohammad Ali Saghiri, BSc, MSc,* Mehrdad Lotfi, DMD, MSc,

†Morteza Daliri Joupari, PhD,

‡

Mohammad Aeinehchi, DMD, MS,§

and Ali Mohammad Saghiri, BSc, MSck

Abstract
Introduction: Storage temperature influences the prop-erties of Portland cement during mixing. Because of simi-larities between Portland cement and mineral trioxideaggregate, the aim of the present study was to evaluatesurface microhardness, topography, and phase structureof white mineral trioxide aggregate (WMTA) after storagein a range of temperatures. Methods: Thirty WMTAsachets were divided into 3 groups of 10. The 3 groupswere stored at 4�C, 25�C, and 40�C for 48 hours withaccompanying ampules. Sachets were immediatelymixed after removal from storage according to manufac-turer’s instructions and mixed and packed into cylindricalglass tubes at room temperature. Surface microhardnessof each specimen was measured after 3 days. Four spec-imens from each group were prepared and observedunder scanning electron microscope and x-ray diffrac-tion. Data were subjected to one-way analysis of vari-ance and a post hoc Tukey test at P <.05. Results:Mean surface hardness � standard deviation afterstorage at 4�C, 25�C, and 40�C were 25.23 � 5.99,53.56 � 3.28, and 62.89 � 1.76, respectively. Statisti-cally significant differences were observed among thegroups (P < .001). More voids and a disorganized,flake-like topography were observed in specimens storedat 4�C in comparison with those stored at 25�C and40�C. X-ray diffraction meter generated similar peaksat 40�C and 25�C, but slight differences were observedat 4�C. Conclusions: This study indicated that storagetemperature might influence surface hardness andmicrostructure of WMTA. (J Endod 2010;36:1414–1418)
Key WordsSurface hardness, temperature, white mineral trioxideaggregate

From the *Department of Biomedical Engineering, Science and Rtechnology and Department of Endodontics, Dental Faculty, Tabriz Unology, Karaj, Iran; §Department of Endodontics, Faculty of DentistrAmirkabir University of Technology, Tehran, Iran.

Address requests for reprints to Mohammad Ali Saghiri, PhD StudTehran, Iran. E-mail address: [email protected]/$0 - see front matter


1414 Saghiri et al.

Mineral trioxide aggregate (MTA) is currently used for multiple purposes, includingpulp capping (1), apexification (2, 3), perforation repair, root-end filling mate-

rial (4, 5), and for pulpotomy (6) and partial pulpotomy (7). MTA powder contains finehydrophilic particles that set in the presence of moisture (8) and is a mechanicalmixture of Portland cement (75%), bismuth oxide (20%), and gypsum (5%) (9).In addition, a recent study has shown that MTA contains several oxides, includingCaO, SiO2, and Al2O3, similar to Portland cement (10).

The principal setting process of Portland cement is initiated on contact with waterwhen a chemical reaction between water and cement begins (11). The particle size(12), powder-to-liquid ratio (13), environmental temperature (14, 15), andpresence of air in the mixture (16) might all affect the physical properties, progressof hydration, and kinetics of Portland cement.

Hydration of MTA produces calcium hydroxide and calcium silicate hydrate gel(C-S-H). C-S-H gel contributes to the strength of cement, and it also contributes tothe durability of hydrated cement (17). Some research studies have investigated theeffect of environmental conditions on mechanical and physical properties of WMTA,including evaluation of the morphology of WMTA stored under various conditions(18), sealing ability at different pH values (19), surface hardness (20, 21), andsetting time (22, 23), to improve the properties of WMTA.

Studies have demonstrated that the hydration of Portland cement is sensitive totemperature, and the reaction is exothermic (24, 25). Portland cement and MTAhave almost a similar chemical composition. This similarity leads to the speculationthat WMTA might possess the same physical properties of Portland cement duringhydration. In addition, studies have depicted that storing dental restorativecomposite in a cool place might increase shelf life (26, 27). Temperature hasa great influence on the physical properties of dental gutta-percha cones (28, 29).

WMTA is used worldwide under different climatic conditions, and some dentalpractitioners store it in the refrigerator alongside dental restorative composites andgutta-percha cones. A search of the literature indicated that few articles have been pub-lished on the potential effect of weather conditions on the properties of Portland cement(30, 31). Therefore, the purpose of the present study was to investigate the effect ofdifferent storage temperatures on microhardness, surface topography, and phasestructure of WMTA.

esearch Branch, Islamic Azad University, Tehran, Iran; †Research Center for Pharmaceutical Nano-niversity (Medical Sciences), Tabriz, Iran; ‡National Institute for Genetic Engineering and Biotech-y, Azad University of Medical Sciences, Tehran, Iran; and kDepartment of Computer Engineering,

ent, Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University,



Figure 1. Box plots of hardness values in each group stored at 4�C (I), 25�C(II), and 40�C (III), which illustrate the mean� standard deviation, minimumand maximum amount of hardness values, as well as the variance in each exper-imental group. (This figure is available in color online at www.aae.org/joe/.)


Materials and MethodsThis study was divided into 2 parts. In part I, thirty sealed sachets

each containing 1 g of WMTA (Tooth-colored formula; Dentsply, TulsaDental, OK) along with 30 ampules of sterile water (each weighing 0.3 g)were selected. Sachets and ampules were divided into 3 groups of 10each accompanied with ampules. All the specimens of the first groupwere stored in a refrigerator (Absal Co, Tehran, Iran) at 4�C. Thetemperature fluctuations of the refrigerator were less than �0.5�C in24 hours. The specimens of the second and third groups were storedat 25�C and 40�C, respectively, in an incubator (ATOI-050, ShanghaiAll-Time Commercial, Shanghai, China) for 48 hours. The temperaturefluctuations of the air thermostat were less than �0.1�C in 24 hours.Immediately after being retrieved from the storage environment, eachsachet was mixed with its ampule according to manufacturer’s instruc-tions. Each mixed sachet was packed into a cylindrical glass tube with aninternal diameter of 8 mm and a height of 10 mm by using a nonsurgicalmanual MTA carrier (Dentsply, Tulsa Dental) after the hand condenser(Hu-Friedy, Chicago, IL) was checked to make sure the tip would fit thelength. One operator condensed WMTA by using hand pressure accord-ing to previous studies (32, 33). The filled cylindrical glass tubes werestored at 37�C and 100% relative humidity for 72 hours to ensure thatWMTA hardened completely.

In part II, the experimental protocol was similar to the one used ina study carried out by Saghiri et al (21). In brief, surfaces containing setWMTA were polished with 600-grit and 1200-grit silicon carbide papersunder water cooling to create flat surfaces. The polished platforms wererinsed with deionized water for 1 minute and dried with oil-free air for 5seconds. The Vickers microhardness test was performed for each spec-imen by using a Clemex CMT microhardness tester (Clemex Technolo-gies Inc, Longueuil, Canada).

Three indentations were made on the polished surface. The diag-onal lengths of the resulting indentations were measured under a micro-scope, and the Vickers microhardness value was calculated. The meanvalue of the hardness was used as the hardness value for each specimen.After 6 indentations were made on surfaces, standard reference material(SRM) blocks were used to calibrate the testing machines. The SRMblock was cleaned with ethylene alcohol and a soft wipe material,and 2 indentations were made on it. Then the diagonal lengths weremeasured, and the machine was adjusted. Differences between themeans were analyzed by one-way analysis of variance and a post hocTukey test at a significance level of P <.05.

In addition, 4 specimens from each group were selected andprepared for scanning electron microscopy and x-ray diffraction (XRD)for the evaluation of microstructure morphologies and phase structure.

The surfaces of 2 samples from each group were sputter-coatedand observed under scanning electron microscope (SEM) (Leo 440i;Oxford Microscopy, Oxford, UK) by using secondary electron (SE)and back-scattered electron (BSE) detectors. Two specimens fromeach group were prepared for XRD. After removing the 2 specimensof WMTA out of the glass tubes, they were milled into the powder bymortar and pestle. XRD patterns were recorded with the XRD device(Seifert XRD 3000; Seifert Co, Ahrensburg, Germany; Co-Ka radiation).Samples were scanned at a range of 0–90, and data were collected ina continuous scan mode at a scanning rate of 4�/min.

Subsequently in each of the 3 groups, crystalline phases of WMTAafter hydration were determined by XRD analysis.

ResultsMicrohardness

Mean surface hardness� standard deviation after storage at 4�C,25�C, and 40�C were 25.23� 5.99, 53.56� 3.28, and 62.89� 1.76,

JOE — Volume 36, Number 8, August 2010 Effects of Storage Temper

respectively (P < .001). Post hoc Tukey test revealed significant differ-ences among all the groups (P < .001) (Fig. 1).

SEM AnalysisSEM micrographs are displayed in Fig. 2. SE detector revealed that

uniform needle-shaped crystals predominantly covered the surfaces ofthe 25�C and 40�C groups; however, needle-shaped crystals were notfound on the surface of the specimens stored at 4�C. BSE detector illus-trated that the uniformity had disappeared, and more voids and bubbleswere observed at 4�C samples compared with other groups. In addition,agglomeration of fine particles of C-S-H gel was observed.

The surfaces of group I specimens (4�C) showed that anhydrateparticles were bare, with irregular particles lying on top (Fig. 2a, b).BSE and SE detectors revealed that the surfaces of the cement particleswere still bare, but lumps and platelets had formed in addition to thefragments already present on the anhydrate particles. Fig. 2c, d showsthat group II specimens (25�C) consisted of needle-like protrudingstructures and regular structures on top of homogeneity. Fig. 2e, fshows that group III specimens (40�C) consisted of uniform hexagonalstructures and interlocking solids.

Phase CompositionXRD results of WMTA samples stored at 4�C, 25�C, and 40�C are

presented in Fig. 3. The same main constituent phases were observed inWMTA diffractograms in the 3 groups. Patterns of groups II and III(25�C and 40�C) were similar; however, for group I (4�C) the intensityof the peak decreased at 2q = 29.3� and increased at 2q = 47.2�.

DiscussionIn this study, the 40�C conditions were selected to simulate trop-

ical conditions; 4�C was selected to consider cool conditions such asrefrigerator, and the room temperature (25�C) was selected as control.SE and BSE detectors have been successfully used in several studies toevaluate surface topography and some deeper parts of cements (21,34). In the current study, applying these 2 SEM methods provideda better opportunity to understand structural differences of WMTA atdifferent temperatures. This ability is attributed to the fact that higher

ature on Surface Hardness, Microstructure, and Phase Formation of WMTA 1415


Figure 2. SEM images of specimens stored at 4�C (I), 25�C (II), 40�C (III), BSE and SE, respectively. More voids (<) can be seen on the surface of WMTA storedat 4�C (group I) in addition to uniform hexagonal (/)structures and interlocked solids stored at 40�C (original magnification,�1000). (This figure is availablein color online at www.aae.org/joe/.)


atomic elements appear brighter in BSE, so that differences in the graylevel observed in hardened cement paste allow the distinction, indescending order, of brightness between anhydrous phases, calciumhydroxide, C-S-H gel, and porosity (34–36).

XRD system has been widely used as a nondestructive analysis toidentify the phase composition of materials. WMTA is a complex mate-rial, and its hydration possibly provides additional complexity. In fact,no single method exists that completely determines all the chemicalreactions taking place in a WMTA structure from the mixing andonward. Therefore, several complementary techniques must be used.

Studies have indicated that storage of Portland cement in coolconditions leads to the extension of initial setting time, potential ofplastic shrinkage, and crack formation (37).

Low powder-to-liquid ratios, short setting time, and the presenceof defects all lead to lower strengths (30, 38).

A previous study (24) has speculated on the possible mechanismsmediating the effect of temperature and has reported that at low temper-atures hydration starts very slowly, allowing the dissolved ions moretime for diffusion before the hydrates precipitate, resulting in a less

1416 Saghiri et al.

dense C–S–H. In addition, this study (31) reported that in the caseof increased temperature, fast hydration might occur, leading toa greater interlocking solid state of hydration products and buildingup of a dense mass.

With knowledge of the initial setting time and primary mechanicalstrength, dental practitioners would be able to plan treatment comple-tion. By storing WMTA below 4�C, for example in a refrigerator withother dental materials such as dental composite and gutta-perchacones, adverse effects might ensue. Up to now some research studies(22, 39, 40) have tried to improve WMTA properties by adding someingredients to it; however, the current study shows that surfacehardness and surface topography of WMTA could be amendedwithout any additives. It also shows that lowering storage temperaturemight decrease microhardness of WMTA as well. Some previousstudies (17, 41, 42) have shown the suitability and reliability of XRDtest for evaluating hydration of MTA. XRD structural analysis resultsrevealed that specimens in both 25�C and 40�C groups werecompletely crystalline and showed similar patterns. The results ofXRD analysis of WMTA cement are consistent with previous results



Figure 3. XRD patterns of WMTA stored at 4�C (I), 25�C (II), and 40�C (III), indicating presence of the same main constituents (same phases in the material’scomposition after hydration; however, the intensity of the peak at 2q = 29.3� for 4�C group decreased). (This figure is available in color online at www.aae.org/joe/.)


reported by Islam et al (43) and Camilleri et al (18). XRD showeda diffraction peak at 2q = 29.3�, which is ascribed to C3S and C-S-Hphases, the main binding phases in the Portland cement–based system(44). In group I (4�C) the intensity of the peak at 2q = 29.3�

decreased, which might be attributed to lower C-S-H content in the finalhydrated product leading to harmful effects on hydration reaction suchas higher setting time (45).

Despite lack of significant differences in XRD patterns, the micro-hardness values of the 3 groups revealed significant differences. Thesefindings probably can be attributed to the influence of storage temper-ature on the exothermic reaction of the cement during hydration, whichmight affect physical properties of WMTA to some extent. The authorsattributed this phenomenon to the exothermic nature of hydration reac-tions versus the storage temperature, leading to decreases or increasesin hexagonal (needle-like) shapes and porosity in the bulk of WMTA.

In summary, at increased temperatures a regular shape wasformed, and the total porosity of the cement decreased, which mightproduce a mechanical resistance against the indenter penetration andconsequently altered microhardness.

ConclusionsThe main conclusions extracted from this investigation can be

summarized as follows. Surface microhardness and total porosity ofcement paste are drastically affected by the storage temperature. Inthe case of storage at higher temperatures, the hydrates are more homo-geneously distributed, which results in not only smaller pores but alsobetter interlocking of the different phases. In addition, needle-likestructures would form, leading to better interlocking of the solids.Lower storage temperatures were found to increase the hardness ofthe coarse porous surface, which could also be lowered by a weakerinterlocking between the hydrate products caused by their more hetero-geneous distribution and the presence of very short needles. In addi-tion, porosity generally correlates negatively with measured surfacehardness, ie, surface hardness increases with decreasing porosity.The lower surface hardness observed at decreased storage tempera-tures is probably mainly caused by the observed increase in porosity.

Furthermore, the room temperature or higher temperatures up to40�C seem to provide appropriate thermal conditions for WMTAstorage because at these temperatures the cement’s microhardness issuitable for clinical use with desirable hydration value.

Only a handful of studies support ambient temperatures for storingWMTA; the results of this study cannot be extrapolated to clinicalsuccess of WMTA as a root-end filling material because it merely repre-sents how microhardness, topographic image, and phase structure of

JOE — Volume 36, Number 8, August 2010 Effects of Storage Temper

WMTA are affected by storage temperature. Observations after thestorage of WMTA at high temperatures suggest that its leakage mightbe less than that of WMTA stored at room temperature. Because thereis no published report on these issues, further investigation is recom-mended.

AcknowledgmentsThis article was based on the thesis presented by the first

author to the Faculty of Biomaterials at Azad University Scienceand Research Branch of Tehran, Iran, in partial fulfillment to therequirements for PhD in Biomaterial Engineering. We are indebtedto Professor Kamal Asgar for the provision of laboratory facilities inthe Department of Dental and Biological Materials at the Universityof Michigan. Also special thanks to Drs Alireza Vosoughhosseini,Houtan Aghili, Mohammad Saghiri, HajarAfsar Ladjvardi, SaharDadvand, Kasra Karamifar, and Majid Abdolrahimi for all of theircontributions to this research.

References1. Aeinehchi M, Eslami B, Ghanbariha M, Saffar AS. Mineral trioxide aggregate (MTA)

and calcium hydroxide as pulp-capping agents in human teeth: a preliminary report.Int Endod J 2003;36:225–31.

2. Steinig TH, Regan JD, Gutmann JL. The use and predictable placement of mineraltrioxide aggregate in one-visit apexification cases. Aust Endod J 2003;29:34–42.

3. Levenstein H. Obturating teeth with wide open apices using mineral trioxide aggre-gate: a case report. SADJ 2002;57:270–3.

4. Torabinejad M, Watson T, Pitt Ford T. Sealing ability of a mineral trioxide aggregatewhen used as a root end filling material. J Endod 1993;19:591–5.

5. Torabinejad M, Smith P, Kettering J, Pitt Ford T. Comparative investigation ofmarginal adaptation of mineral trioxide aggregate and other commonly usedroot-end filling materials. J Endod 1995;21:295–9.

6. Witherspoon DE, Small JC, Harris GZ. Mineral trioxide aggregate pulpotomies:a case series outcomes assessment. J Am Dent Assoc 2006;137:610–8.

7. Barrieshi-Nusair K, Qudeimat M. A prospective clinical study of mineral trioxideaggregate for partial pulpotomy in cariously exposed permanent teeth. J Endod2006;32:731–5.

8. Parirokh M, Torabinejad M. Mineral trioxide aggregate: a comprehensive literaturereview—part I: chemical, physical, and antibacterial properties. J Endod 2010;36:16–27.

9. ProRootMTA safety data sheet. Available at: http://store.maillefer.com/lit2/pdfs/MTA-MSDS-W_01-02C.pdf. Accessed March 10, 2010.

10. Asgary S, Shahabi S, Jafarzadeh T, Amini S, Kheirieh S. The properties of a newendodontic material. J Endod 2008;34:990–3.

11. Taylor HFW. Cement chemistry. 2nd ed. London: Thomas Telford; 1997.12. Osbaeck B, Johansen V. Particle size distribution and rate of strength development

of Portland cement. J Am Ceram Soc 1989;72:197–201.13. Bentz DP. Influence of water-to-cement ratio on hydration kinetics: simple models

based on spatial considerations. Cem Concr Res 2006;36:238–44.

ature on Surface Hardness, Microstructure, and Phase Formation of WMTA 1417

http://store.maillefer.com/lit2/pdfs/MTA-MSDS-W_01-02C.pdf

http://store.maillefer.com/lit2/pdfs/MTA-MSDS-W_01-02C.pdf


14. Klieger P. Effect of mixing and curing temperature on concrete strength. ACI J Proc
1958;54:1063–81.15. Mouret M, Bascoul A, Escadeillas G. Strength impairment of concrete mixed in hot

weather: relation to porosity of bulk fresh concrete paste and maturity. MagConcrete Res 2003;55:215–23.

16. Buenfeld N, Okundi E. Release of air from unsaturated aggregate during setting ofconcrete. J Construction and Building Materials 1999;13:143–7.

17. Camilleri J. Hydration mechanisms of mineral trioxide aggregate. Int Endod J 2007;40:462–70.

18. Camilleri J, Montesin F, Brady K, Sweeney R, Curtis R, Ford T. The constitution ofmineral trioxide aggregate. Dent Mater 2005;21:297–303.

19. Saghiri MA, Lotfi M, Saghiri AM, et al. Effect of pH on sealing ability of white mineraltrioxide aggregate as a root-end filling material. J Endod 2008;34:1226–9.

20. Namazikhah MS, Nekoofar MH, Sheykhrezae MS, et al. The effect of pH on surface hard-ness and microstructure of mineral trioxide aggregate. Int Endod J 2008;41:108–16.

21. Saghiri M, Lotfi M, Saghiri A, Vosoughhosseini S, Aeinehchi M, Ranjkesh B. Scanningelectron micrograph and surface hardness of mineral trioxide aggregate in the pres-ence of alkaline pH. J Endod 2009;35:706–10.

22. Kogan P, He J, Glickman G, Watanabe I. The effects of various additives on settingproperties of MTA. J Endod 2006;32:569–72.

23. Huang T, Shie M, Kao C, Ding S. The effect of setting accelerator on properties ofmineral trioxide aggregate. J Endod 2008;34:590–3.

24. Escalante-Garcıa JI, Sharp JH. The microstructure and mechanical properties ofblended cements hydrated at various temperatures. Cem Concr Res 2001;31:695–702.

25. Price WH. Factors influencing concrete strength. J Aci J Proc 1951;47:417–32.26. de la Torre-Moreno FJ, Rosales-Leal JI, Bravo M. Effect of cooled composite inserts

in the sealing ability of resin composite restorations placed at intraoral tempera-tures: an in vitro study. Oper Dent 2003;28:297–302.

27. Bausch JR, Delange C, Davidson CL. The influence of temperature on some physicalproperties of dental composites. J Oral Rehabil 2007;8:309–17.

28. Friedman C, Sandrik J, Heuer M, Rapp G. Composition and physical properties ofgutta-percha endodontic filling materials. J Endod 1977;3:304–8.

29. Heuer MA. Instruments and materials. In: Cohen S, ed. Pathways of the pulp. 3rd ed.St Louis, MO: CV Mosby; 1984.

30. Lothenbach B, Winnefeld F, Alder C, Wieland E, Lunk P. Effect of temperature on thepore solution, microstructure and hydration products of Portland cement pastes.J Cement and Concrete Research 2007;37:483–91.

1418 Saghiri et al.

31. Escalante-Garcia JI, Sharp JH. Effect of temperature on the hydration of the mainclinker phases in Portland cements: part I—neat cements. Cem Concr Res 1998;28:1245–57.

32. Watts J, Holt D, Beeson T, Kirkpatrick T, Rutledge R. Effects of pH and mixing agentson the temporal setting of tooth-colored and gray mineral trioxide aggregate. J En-dod 2007;33:970–3.

33. Aminoshariae A, Hartwell G, Moon P. Placement of mineral trioxide aggregate usingtwo different techniques. J Endod 2003;29:679–82.

34. Maso JC. Pore structure and construction materials properties. London, New York:Chapman and Hall; 1987.

35. Scrivener KL, Patel HH, Pratt PL, Parrott LJ. Analysis of phases in cement paste usingbackscattered electron images, methanol adsorption and thermogravimetric anal-ysis. Mater Res Soc Symp Proc 1987;85:66–7.

36. Zhao H, Darwin D. Quantitative backscattered electron analysis of cement paste.Cem Concr Res 1992;22:695–706.

37. Hua C, Acker P, Ehrlacher A. Analyses and models of the autogenous shrinkage ofhardening cement paste: I—modelling at macroscopic scale. Cem Concr Res 1995;25:1457–68.

38. Porter M, Berto A, Primus C, Watanabe I. Physical and chemical properties of new-generation endodontic materials. J Endod 2010;36:524–8.

39. Lotfi M, Vosoughhosseini S, Saghiri MA, Mesgariabbasi M, Ranjkesh B. Effect ofwhite mineral trioxide aggregate mixed with disodium hydrogen phosphate oninflammatory cells. J Endod 2009;35:703–5.

40. Wiltbank K, Schwartz S, Schindler W. Effect of selected accelerants on the physicalproperties of mineral trioxide aggregate and Portland cement. J Endod 2007;33:1235–8.

41. Belıo-Reyes I, Bucio L, Cruz-Chavez E. Phase composition of ProRoot mineraltrioxide aggregate by x-ray powder diffraction. J Endod 2009;35:875–8.

42. Lee YL, Lee BS, Lin FH, Lin AY, Lan WH, Lin CP. Effects of physiological environmentson the hydration behavior of mineral trioxide aggregate. Biomaterials 2004;25:787–93.

43. Islam I, Kheng Chng H, Jin Yap A. Comparison of the physical and mechanical prop-erties of MTA and Portland cement. J Endod 2006;32:193–7.

44. Older I. Hydration. In: Hewlett PC, ed. Lea’s chemistry of cement and concrete. 4thed. Oxford, UK: Butterworth-Heinemann; 2004.

45. Lin FH, Wang WH, Lin CP. Transition element contained partial stabilized cement(PSC) as a dental retrograde-filling material. Biomaterials 2003;24:219–33.


Case Report/Clinical Techniques

Inferior Alveolar Nerve Paresthesia after Overfilling ofEndodontic Sealer into the Mandibular CanalMaribel Gonzalez-Martın, PhD, DDS, Daniel Torres-Lagares, PhD, DDS, Jose LuisGutierrez-Perez, PhD, MD, DDS, and Juan Jose Segura-Egea, PhD, MD, DDS

Abstract
The present study describes a case of endodontic sealer(AH Plus) penetration within and along the mandibularcanal from the periapical zone of a lower second molarafter endodontic treatment. The clinical manifestationscomprised anesthesia of the left side of the lower lip,paresthesia and anesthesia of the gums in the thirdquadrant, and paresthesia and anesthesia of the leftmental nerve, appearing immediately after endodontictreatment. The paresthesia and anesthesia of the lipand gums were seen to decrease, but the mental nerveparesthesia and anesthesia persisted after 3.5 years.This case illustrates the need to expend great carewith all endodontic techniques when performingnonsurgical root canal therapy, especially when theroot apices are in close proximity to vital anatomic struc-tures such as the inferior alveolar canal. (J Endod2010;36:1419–1421)
Key WordsEndodontic complications, paresthesias of the inferiordental nerve

From the Department of Stomatology, School of Dentistry,University of Seville, Seville, Spain.

Address requests for reprints to Dr Juan Jose Segura-Egea,Professor of Endodontics, Department of Stomatology, Schoolof Dentistry, University of Seville, C/Avicena s/n, 41009 Seville,Spain. E-mail address: [email protected]/$0 - see front matter



The elimination of all diseased pulp and dentin, adequate cleaning and shaping of theroot canal system, and its 3-dimensional obturation and sealing constitute the basic

principles of endodontic treatment. Ideally, the filling material should be limited to theroot canal without extending to periapical tissues or other neighboring structures.However, overinstrumentation of the root canal with hand or mechanically driven filescan perforate the mandibular canal, allowing the extrusion of sealers, dressing agents,and irrigation solutions and the passage of microorganisms into the canal duringendodontic treatment (1).

Totally biocompatible materials are not available. Consequently, their spreadbeyond the apical foramen can give rise to clinical manifestations in relation to thetoxicity of the product, although minor material extrusions are generally well-tolerated by the periradicular tissues (2). Sealers and filling materials differ chemicallyand include, among others, AH Plus, AH 26, Hydron, Diaket, Iodoform, Calasept, Endo-seal, and chloropercha. Some of them can cause serious neurotoxic complicationswhen extruded into the mandibular canal (3–5). Undesirable complications such asanesthesia, paresthesia, hypoesthesia, hyperesthesia, and dysesthesia can follow theextrusion of an endodontic sealer into the mandibular canal (6–12).

The first symptom of the overfilling into the mandibular canal is sudden pain ex-pressed by the patient during obturation of the root canal, which persists after the disap-pearance of the local anesthetic effects (13). The pain can be accompanied by localinflammatory signs, with the endodontically treated tooth being painful to percussion,painful on palpation of the vestibular alveolar process, or a combination of signs ofmechanical lesions and inferior dental nerve inflammation with pain or numbness ofthe lower lip or otalgia (6). Some patients experience the persistence of the localanesthesia (12, 14).

The present study describes a case in which endodontic sealer AH Plus spread tothe mandibular canal, causing paresthesia and anesthesia in the area of innervation ofthe inferior alveolar nerve.

Case ReportA 32-year-old woman was referred for root canal treatment in the left second

molar tooth because of an apical periodontitis subsequent to caries. The initial diag-nosis for the tooth was acute apical abscess (15). She neither smoked nor consumedalcohol and had no personal or family disease antecedents of interest. Two years before,her first mandibular left molar had been extracted, and an endosseous root-formedimplant had been placed, supporting a metal-ceramic crown. After adequate anesthesiaand isolation with rubber dam, an endodontic access cavity was established. Threecanal orifices were defined. After apical patency, the root length was estimated by usingan apex detector (AFA Apex Finder; Analytic Technology, Orange County, CA) and thenconfirmed with a periapical radiograph. The root canal treatment was carried out byinvolving canal shaping with hand files by using the step-back technique and saline irri-gation. After cleaning and shaping, the canal was dried and obturated with AH 26(Dentsply DeTrey GmbH, Konstanz, Germany) and gutta-percha by using the lateralcondensation technique. A small quantity of sealer was introduced into the root canalby using a manual instrument, then the main cone was placed and covered witha minimal quantity of sealer, and cold condensation was performed. Moreover, eachadditional cone was covered with a minimal quantity of sealer. There were no compli-cations during treatment.

Inferior Alveolar Nerve Paresthesia after Overfilling of Endodontic Sealer 1419


Figure 2. Panoramic radiograph taken the day after showing endodonticpaste in the periapical zone of tooth #37 and the inferior alveolar canal.

Just after endodontic treatment, the postoperative periapical
radiograph revealed the presence of radiopaque canal sealer in themandibular canal (Fig. 1). Nevertheless, the patient was still underthe effect of the anesthetic, and she reported no pain or other discom-fort after the root canal treatment.

The day after, no swelling, redness, or other signs of inflammationwere observed on intraoral exploration. However, the patient reportednumbness on the left side of the lower lip and a tingling sensation in thevestibular gingival and in the lower left premolar and anterior teeth. Ex-traoral examination likewise failed to identify swelling, alterations inskin color, or adenopathies. The anesthetized zone was delimited bytactile exploration, and anesthesia in the region served by the left infe-rior alveolar and mental nerve was observed. The buccal gingival tissuesover the left mandibular molar and premolar teeth felt no sensation;there was no sensation to thermal or mechanical stimuli in either theleft lower lip or buccal gingivae. The lingual gingival tissues respondedwithin normal limits to stimulation with an explorer. A panoramic radio-graph was taken (Fig. 2), revealing the presence of radiopaque material(AH Plus) in the periapical area of tooth #18 (universal) and spreadingdistally along the mandibular canal.

After discussing treatment options with the patient, it was decidedto monitor the progress with periodic follow-up visits. The patientnoticed a very rapid improvement during the first months after the inci-dent. Seven months later, she showed significantly less paresthesia, andthe anesthesia in the region of the lower left lip was decreased comparedwith the initial situation. However, there have been no significantimprovements during the subsequent 3 years. The skin anesthesia onthe left side of the lower lip persists (Fig. 3), and the radiopaque mate-rial in the periapical area of tooth #37 is still radiographically evident(Fig. 4).

Figure 1. Post-treatment periapical radiograph. Presence of the extrudedroot canal sealer in the mandibular canal is evident.

1420 Gonzalez-Martın et al.

DiscussionSensory loss or alteration in the territory of the inferior alveolar

nerve, the chin region, and lower homolateral half of the lip is a relativelyinfrequent complication in daily dental practice and is normally theresult of an inadequate dental treatment (12). One of the potential iatro-genic causes of this problem is the incorrect treatment of the root canalsof a lower molar or premolar (overextension and/or overfilling). Theproximity of the mandibular canal to the apices of the premolar andmolar teeth requires a careful radiographic diagnosis when endodontictreatment of these teeth is planned. An initial pretreatment radiograph ofthe mandibular teeth will reveal the proximity of the canal to the apices(2). Preventive measures such as the use of an electronic apex detector,the application of a good apical stop, or moderate condensation willhelp avoid overfilling or overextending the endodontic material (11).

During endodontic treatment, it is extremely important to addressthe cleaning and shaping of the apical third accurately, knowing cor-rected length and width. The use of an electronic apex detector togetherwith a radiograph taken with the files in position will not only ensure thecorrect working length but also prevent perforation of the canal andpossible subsequent damage to the inferior alveolar nerve resultingfrom the endodontic treatment.

In the case reported here, preventive measures were taken. Theroot length was estimated by using an apex detector and confirmedwith a periapical radiograph. Poor length control does not seem tobe the obvious cause of the overextension of sealer in this case. More-over, only a small quantity of sealer was introduced into the root canalby using a manual instrument. The main cone was placed covered with

Figure 3. Area of mental nerve anesthesia after 3.5 years is outlined on theskin. (This figure is available in color online at www.aae.org/joe/.)



Figure 4. Panoramic radiograph taken 3.5 years after the accident showingthe persistence of the endodontic paste in the periapical zone of tooth #37 andthe inferior alveolar canal.


a minimal quantity of sealer as well as each additional cone. Likewise,root does not seem to be significantly overfilled. However, the cliniciandid not use a master gutta-percha cone film before obturation.

Experimental studies have shown that eugenol and paraformalde-hyde are the main materials causing neurotoxic reactions (3). Theirrigation solutions, such as sodium hypochlorite and ethylenediamine-tetraacetic acid (EDTA), might leak into the canal and damage the nervechemically. Cytotoxic effects of EDTA (16), eugenol (17), andhypochlorite (18) have been described. Escoda-Francoli et al (19) sug-gested that some of the materials used in certain sealer cements, such ascalcium tungstate (scheelite [CaWO4]) and zirconium oxide (badde-leyite [ZrO2]), are not totally reabsorbable or innocuous and do notseem to be well-tolerated when overextended beyond the apicalforamen.

AH Plus is one of the epoxy resin-based root canal sealers mostcommonly used. The monomer 2,2-bis[4-(2-hydroxy-3-methacrylyl-oxypropoxy)phenyl]-propane (BisGMA), prepared from bisphenol Aand glycidyl methacrylate, is the major ingredient of the epoxy resin-based root canal sealers AH 26 and AH Plus (20). AH 26 cures withthe generation of formaldehyde as a by-product, but AH Plus onlyreleases small amounts of formaldehyde (21). However, AH Plus cancause cytotoxic effects (22) when extruded into the mandibular canal(4). Moreover, it has been shown that its component bisphenol Acan cause cytotoxic effects (23).

Serper et al (24) investigated the neurotoxic effects of the rootcanal filling materials Endomethasone, N2 Universal, TraitementSPAD, Sealapex, and Calciobiotic Root Canal Sealer on isolated ratsciatic nerves after local application, demonstrating the neurotoxiceffects of root canal filling materials. They observed that recoveryfrom chemical insults to nerve structures was relatively slow and wasincomplete in the in vitro conditions. A larger and more rapid andappreciable recovery was found with sealers without paraformaldehydeand eugenol, which seems to be less toxic to nervous structures whencompared with other compounds containing these substances.

In the present case, the buccal gingival tissues over the left mandib-ular molar and premolar teeth felt no sensation. This could be explainedbecause although the buccal nerve normally supplies these tissues,accessory branches of the inferior alveolar nerve have been described(25, 26).

A literature review of paresthesia and anesthesia cases attributableto the extrusion of a root canal sealer indicated that the surgical removalof the sealer from the mandibular canal is an effective treatment andmight restore normal sensation in the affected region (9, 19).However, in the present case, the patient did not want surgical


treatment, even though she exhibited complete lip skin anesthesia 3.5years after the endodontic mishap. This case illustrates the need toexpend great care with all endodontic techniques when performingnonsurgical root canal therapy, especially when the root apices are inclose proximity to vital anatomic structures such as the inferioralveolar canal.

References1. Koseoglu BG, Tanrikulu S, Subay RK, Sencer S. Anesthesia following overfilling of

a root canal sealer into the mandibular canal: a case report. Oral Surg Oral MedOral Pathol Oral Radiol Endod 2006;101:803–6.

2. Poveda R, Bagan JV, Diaz Fernandez JM, Sanchis JM. Mental nerve paresthesia asso-ciated with endodontic paste within the mandibular canal: report of a case. OralSurg Oral Med Oral Pathol Oral Radiol Endod 2006;102:e46–9.

3. Morse DR. Infection-related mental and inferior alveolar nerve paresthesia: litera-ture review and presentation of two cases. J Endod 1997;23:457–60.

4. Tamse A, Kaffe I, Littner MM. Paraesthesia following over extension of AH26: reportof two cases and review of the literature. J Endod 1982;8:88–90.

5. Rowe AHR. Damage to the inferior dental nerve during or following endodontictreatment. Br Dent J 1983;153:306–7.

6. Brodin P, Røed A, Aars H, Ørstavik D. Neurotoxic effects of root filling materials onrat phrenic nerve in vitro. J Dent Res 1982;6:1020–3.

7. Gatot A, Tovi F. Prednisone treatment for injury and compression of inferior alveolarnerve: report of a case of anesthesia following endodontic treatment. Oral Surg OralMed Oral Pathol Oral Radiol Endod 1986;62:704–6.

8. Kothary P, Cannell H. Bilateral mandibular nerve damage following root canaltherapy. Br Dent J 1996;180:189–90.

9. Scolozzi P, Lombardi T, Jaques B. Successful inferior alveolar nerve decompressionfor dysesthesia following endodontic treatment: report of 4 cases treated by mandib-ular sagittal osteotomy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004;97:625–31.

10. Kaufman AY, Rosenberg L. Paraesthesia caused by Endomethasone. J Endod 1980;6:529–31.

11. Neaverth EJ. Disabling complications following inadvertent overextension of a rootcanal filling material. J Endod 1989;15:135–9.

12. Gallas-Torreira MM, Reboiras-Lopez MD, Garcıa-Garcıa A, Gandara-Rey J. Mandib-ular nerve paresthesia caused by endodontic treatment. Med Oral 2003;8:299–303.

13. LaBlanc JP, Epker BN. Serious inferior alveolar nerve dyesthesia alter endodonticprocedure: report of three cases. J Am Dent Assoc 1984;108:605–7.

14. Grotz KA, Al-Nawas B, de Aguiar EG, Schulz A, Wagner W. Treatment of injuries to theinferior alveolar nerve after endodontic procedures. Clin Oral Invest 1998;2:73–6.

15. ABE (American Board of Endodontics). Pulpal & periapical diagnostic terminology.Available at: http://www.aae.org/NR/rdonlyres/0A9E773B-506D-4B63-884E-DA68381CEAB0/0/ABETerminologyMay2007.doc. Accessed June 2, 2009.

16. Segura JJ, Calvo JR, Guerrero JM, Jimenez-Planas A, Sampedro C. EDTA inhibits invitro substrate adherence capacity of macrophages: endodontic implications. J En-dod 1997;23:205–8.

17. Segura JJ, Jimenez-Rubio A. Effect of eugenol on macrophage adhesion in vitro toplastic surfaces. Endod Dent Traumatol 1998;14:72–4.

18. Jimenez Rubio A, Segura JJ, Jimenez A, Guerrero JM, Calvo JR. ‘‘In vitro’’ study of theeffect of sodium hypochlorite and glutaraldehyde on substrate adherence capacity ofmacrophages. J Endod 1997;23:562–4.

19. Escoda-Francoli J, Canalda-Sahli C, Soler A, Figueiredo R, Gay-Escoda C. Inferioralveolar nerve damage because of overextended endodontic material: a problemof sealer cement biocompatibility? J Endod 2007;33:1484–9.

20. Peutzfeldt A. Resin composites in dentistry: the monomer systems. Eur J Oral Sci1997;105:97–116.

21. Leonardo MR, Bezerra da Silva LA, Filho MT, Santana da Silva R. Release of form-aldehyde by four endodontic sealers. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 1999;88:221–5.

22. Pulgar R, Segura-Egea JJ, Fernandez MF, Serna A, Olea N. The effect of AH 26 and AHPlus on MCF-7 breast cancer cells proliferation in vitro. Int Endod J 2002;35:551–6.

23. Segura-Egea JJ, Jimenez-Rubio A, Pulgar R, Olea N, Guerrero JM, Calvo JR. In vitroeffect of the resin component bisphenol A on macrophage adhesion to plasticsurfaces. J Endod 1999;25:341–4.

24. Serper A, Ucer O, Onur R, Etikan I. Comparative neurototxic effects of root canalmaterials on rat sciatic nerve. J Endod 1998;24:592–4.

25. Carter RB, Keen EN. The intramandibular course of the inferior alveolar nerve.J Anat 1971;108:433–40.

26. Loizeaux AD, Devos BJ. Inferior alveolar nerve anomaly. J Hawaii Dent Assoc 1981;12:10–1.

Inferior Alveolar Nerve Paresthesia after Overfilling of Endodontic Sealer 1421

http://www.aae.org/NR/rdonlyres/0A9E773B-506D-4B63-884E-DA68381CEAB0/0/ABETerminologyMay2007.doc

http://www.aae.org/NR/rdonlyres/0A9E773B-506D-4B63-884E-DA68381CEAB0/0/ABETerminologyMay2007.doc


Autotransplantation of Teeth with Complete RootFormation: A Case SeriesJi-Hyun Bae, DDS, PhD,* Yong-Hoon Choi, DDS, MSD,* Byeong-Hoon Cho, DDS, PhD,*

Young-Kyun Kim, DDS, PhD,†

and Su-Gwan Kim, DDS, PhD‡

Abstract
Introduction: Autotransplantation is a viable option fortreating missing teeth when a donor tooth is available.This retrospective study reports the success rate for theautotransplantation of 19 molars with complete rootformation. Methods: The study enrolled 19 patients(11 men and 8 women) in whom 19 molars were trans-planted. The mean age was 38.5 years (range, 19-67).The transplanted third molars were stabilized witha silk suture or wire splint for 2 to 3 weeks. Root canaltreatment of the transplanted teeth was performedbefore surgery in six patients and 1 to 2 weeks aftertransplantation in 13 patients. Postoperatively, themarginal and periapical conditions were examined clin-ically and radiographically. Results: In 16 of the 19cases, the outcome met the success criteria, for an84% success rate. Conclusions: In autogenous toothtransplantation, even if the donor tooth has completeroot formation, a high success rate can be achieved ifthe cases are selected and treated properly. (J Endod2010;36:1422–1426)
Key WordsAutotransplantation, complete root formation, successrate

From the Departments of *Conservative Dentistry and †Oraland Maxillofacial Surgery, Section of Dentistry, Seoul NationalUniversity Bundang Hospital, Seongnam, Republic of Korea;and ‡Department of Oral and Maxillofacial Surgery, School ofDentistry, Chosun University, Gwangju, Republic of Korea.

Address requests for reprints to Dr Su-Gwan Kim, Depart-ment of Oral and Maxillofacial Surgery, School of Dentistry,Chosun University, 375, SeoSukDong, DongGu, GwangJuCity, South Korea. E-mail address: [email protected]/$0 - see front matter


1422 Bae et al.

Autogenous tooth transplantation refers to the repositioning of autogenous teeth inanother tooth extraction site or a surgically formed recipient site to replace teeth

that are, for example, missing congenitally or involve ectopic eruption, severe caries,periodontal disease, trauma, or endodontic failure when a suitable donor tooth is avail-able (1, 2).

The success rate of autogenous tooth transplantation in the 1950s was approxi-mately 50% because of the difficulty in predicting root development after transplanta-tion and dental root resorption (3, 4). Because too little was known of the causes andprevention of root resorption of transplanted autogenous teeth, the procedure was usedinfrequently. Since the 1990s, many studies have examined the healing of periodontaltissues and periodontal membrane and dental root resorption, and the transplantsuccess rate increased rapidly, drawing new clinical interest (5–7). Tsukiboshi (5)reported a 90% survival rate and an 82% success rate for 250 cases observed for 6years. Lundberg and Isaksson (6) reported a 94% success rate in cases with incom-pletely formed roots and 84% in cases with completely formed roots and a highersuccess rate in cases with immature teeth, whereas Majare et al (7) reported a highsuccess rate for cases with mature teeth.

Complications of autogenous tooth transplant include root resorption and attach-ment loss, and its success rate is lower than for implants. Nonetheless, autogenous teethresult in good utilization, the maintenance and regeneration of alveolar bone, and themaintenance of attached gingiva with a natural shape. Hence, the esthetic results arebetter, the cost is low, one-stage surgery can be used, orthodontic movement ispossible, and the procedure can be performed in growing patients (2, 8).Therefore, this retrospective study examined the autotransplantation of mature teethclinically and radiographically; a case series is reported with a review of the literature.

Material and MethodsThis study analyzed the medical records and radiographs of 19 patients (11 men

and 8 women) in whom 19 molars with complete root formation were autotransplantedin the Departments of Conservative Dentistry and Oral and Maxillofacial Surgery of SeoulNational University Bundang Hospital. The ages of the patients at the time of surgeryranged from 19 to 67 years with a mean age of 38.5 years (Table 1). All of the patientswere in good health, and a routine examination found no systemic or local contraindi-cations to surgical treatment.

The reasons for the transplantation were dental caries in seven patients; peri-odontal disease in three patients; cracked teeth and a request for autogenous toothtransplant after tooth extraction in a dental private office in two patients each; a verticalroot fracture, congenitally missing tooth, and endodontic treatment failure in onepatient each; and unknown reasons for extraction in two patients. The donor teethwere all third molars, except one patient in which an upper second molar was used.In all cases, the recipient sites were the first or second molar areas (Table 2).

In this study, the criteria for successful transplantation were as follows (1, 2):

1. The transplanted tooth functioned normally without excessive mobility; however,physiologic mobility was allowed. Tooth mobility was classified as grade I: slightlymore than normal, grade II: moderately more than normal, and grade III: severemobility faciolingually and mesiodistally combined with vertical displacement (9).

2. Clinically, no discomfort and a normal periodontal probing.



TABLE 1. Patient Age at the Time of Surgery

Age Sex Number

10-19 M 1F 1

20-29 M 1F 3

30-39 M 3F 1

40-49 M 4F 2

60-69 M 3Total 19

M, male; F, female.


3. Radiographically, no root resorption and a normal periodontal liga-ment space and lamina dura are present.

Surgical ProcedureTo prepare the recipient site, when a three-dimensional model of

the donor tooth was available, a computer-aided rapid prototyping(CARP) model was used (Fig. 1); otherwise, the transplant site wasbased on the size measured on radiographs in advance. The recipientsocket was prepared with a surgical round bur with copious saline irri-gation. Minimizing trauma, the tooth to be transplanted was extractedcarefully. To reduce the injury to the periodontal ligament, the toothwas wrapped with gauze wet with saline, an apicoectomy was performedwith a diamond point, and the cavity for retrograde filling was formedand filled with mineral trioxide aggregate (MTA) (Fig. 2). To reduce theextraoral time, an endodontic treatment was performed before extract-ing the tooth to be transplanted if possible. When endodontic treatmentcould not be performed in advance, it was performed 1 to 2 weeks aftertransplantation.

The donor tooth was fitted in the prepared transplant site. If addi-tional preparation was required, the additional area was formed usingthe same method. After tooth transplantation, autogenous bone frag-ments were placed in the adjacent defect or allogenic bone was grafted.When the transplant site was close to the maxillary sinus, the septal bonewas elevated from the maxillary sinus floor using an osteotome and thenthe tooth was transplanted. When the septal bone was lightly tapped andlifted using a Summers’ Osteotome (Implant Innovations, Inc, WestPalm Beach, FL), the sinus membrane was elevated, and the space inthe recipient site where the donor tooth was grafted was secured(Fig. 3). For early stabilization, the donor tooth was rotated andtransplanted in some cases.

The transplanted teeth were fixed with a wire or fiber splint, andany occlusal interference was corrected. When the root was long andearly stabilization was good, an over-crown suture was tied over theocclusal surface. The transplanted teeth were splinted for 2 to 3 weeks.The extraoral exposure of the transplanted teeth was from 3 to 16minutes. Six patients underwent endodontic treatment before transplan-tation and 13 patients after transplantation. The follow-up ranged from

TABLE 2. The Number of Autotransplanted Teeth Distributed According to the Re

Transplanted tooth #3 #2 #14

#1 1 5#15#16#17#32Sum 1 5 0


3 to 21 months and averaged 15 months. Prosthetic treatment was per-formed 3 to 6 months after transplant.

ResultsOn radiographs taken immediately after the transplantation, the

transplanted tooth was seen in a wide tooth extraction socket. Twoweeks after the transplantation, the pain and tenderness had decreasedalthough the tooth mobility was grade 3. One month after the transplan-tation, the morphology of the transplanted tooth and surroundinggingiva were similar to that of the adjacent teeth. Six months after thetransplantation, the mobility of the transplanted tooth had stabilizedat grade 1, and the periodontal condition was good. On radiographs,no pathologic radiolucency or tooth resorption was observed. Themarginal bone support appeared similar to that of the neighboringteeth. A continuous periodontal space was present radiographicallyaround the transplanted teeth.

Good healing took place when the septal bone was elevated fromthe maxillary sinus floor, replacement of autogenous bone fragments ortransplanted allogenic bone occurred, or the tooth was rotated andgrafted. Comparable results were seen in the cases without bone graft-ing. Prosthetic restoration was performed after 3 to 6 months when theprognosis of the tooth could be predicted. In 16 of the 19 transplantedteeth, no inflammation occurred during the healing period. In addition,no pain, discomfort, or other side effects were noted, and the toothbecame stable over time. In the other three cases, severe inflammationand tooth mobility were seen, and the teeth were not fixed within thetransplant site; these were considered failures and extracted. In allfailed cases, the periodontal condition was poor, and the primarystability of the transplanted tooth was poor. The transplanted teethmet the success criteria in 16 cases for an 84% success rate.

DiscussionTo increase the success rate of autogenous tooth transplantation,

a healthy periodontal membrane should be present on the transplantedtooth and the root morphology of the tooth to be transplanted should besimple. In addition, infection should be absent in the recipient site, andduring surgery, the extraoral period should be short and trauma shouldbe minimized (1, 10, 11). The most important factor for the success ofautogenous tooth transplantation is the vitality of the periodontalligament attached to the transplanted tooth (12). The periodontal liga-ment is sensitive to pH and osmotic potential, and its viability is reducedif extraoral dry time is long (13). Previous studies showed that theviability of periodontal ligament exposed to the extraoral spacedecreased rapidly after 18 minutes (12, 14). In our series, the teethto be transplanted were wrapped with gauze wet with sterile salineduring preparation of the recipient site. In all cases, the transplantwas performed within 3 to 16 minutes.

If the tooth to be transplanted had reduced root development (ie,an immature root), it increased the probability of pulp healing. In addi-tion, if the tooth was covered with a thick follicle and periodontal

cipient Site

Recipient sites

#15 #19 #18 #30 #31

11

2 11 4

1 22 3 5 0 3

Autotransplantation of Teeth with Complete Root Formation 1423

Figure 1. Photographs of a left mandibular third molar extracted fortransplantation (left) and a computer-aided rapid prototyping model madeof resin (right). (This figure is available in color online at www.aae.org/joe/.)


membrane and was extracted with a weak force, the injury to the rootsurface was minimized, and replacement resorption rarely occurred. Inmost cases, the pulp healed and no endodontic treatment was required,shortening the treatment time and the future possibility of developingpulp disease; the risk of root fracture was almost zero. In the casesshowing pulp necrosis, the root development ceased and inflammatoryroot resorption occurred (6, 10).

The pulp of a completely mature tooth cannot regenerate. There-fore, if the tooth to be transplanted is accessible, endodontic treatmentshould be completed before transplantation. Otherwise, the endodontictreatment should be initiated 1 to 2 weeks after autogenous tooth trans-plantation. The 1- to 2-week interval is very important because ifendodontic treatment is performed too early after autogenous toothtransplantation, additional injury to the periodontal ligament mayoccur, whereas after 2 weeks, inflammatory resorption may developfrom the infected root canal (5).

During autogenous tooth transplantation, extraoral endodontictreatment prolongs the extraoral time, and during manipulation ofthe instruments, Hertwig’s epithelial root sheath of the root cementalsurface is injured, increasing the possibility of root resorption (11).

Figure 2. A photograph of MTA retrograde filling. To reduce the injury to theperiodontal membrane, the tooth was wrapped in gauze wet with saline. Usinga diamond point, an apicoectomy was performed and an MTA retrograde fillingperformed. (This figure is available in color online at www.aae.org/joe/.)

1424 Bae et al.

In this study, in the six patients for whom endodontic treatment was per-formed before surgery, the transplanted teeth were grabbed with gauzewet with saline, retrograde filling was performed rapidly using MTA, andthe tooth was then transplanted; in none of these cases was the extraoraltime longer than 18 minutes.

A long time is required to form the bone socket at the recipient siteafter extracting the tooth to be transplanted by referring to the shape ofthe extracted tooth. Skilled surgeons could form a recipient site ina short time using techniques similar to implant drilling. However, itmay require more than 30 minutes in most cases. When a longertime is required, the period of time that the donor tooth is exposedto the extraoral cavity becomes longer. In addition, while fitting the ex-tracted tooth to the bone socket, the root surface may be injured. InCARP, software is used to produce a shape identical to the real tooth.This was first applied clinically in the 1980s (15). To prepare a modelof the donor tooth, three-dimensional data on the tooth to be extractedare obtained (Fig. 1A) and converted into a Digital Imaging andCommunications in Medicine format file. Then, a resin or starch modeltooth is prepared using computer prototyping (Fig. 1B) (2, 16). Whenthe data are sent to a company specializing in preparing computer-aided rapid prototypes, a model can be prepared in 3 or 4 days. Ifa CARP model identical to the tooth to be transplanted is preparedbefore surgery, the bone preparation time is shortened, and injury tothe root surface is reduced.

Fong (17) stated that maxillary transplants should not be donebecause of the extreme variation in the size and shape of the maxillarythird molars and because of the proximity of the maxillary antrum to themolar sockets. In our series, however, for the sites adjacent to themaxillary sinus, the septal bone was elevated from the maxillary sinusfloor using an osteotome, and for the cases in which the tooth to betransplanted did not fit the recipient site, it was rotated or autogenousbone fragments or allogenic bone was used and good results were ob-tained. The most important factor in bone formation is the cervicalapproximation of the transplanted tooth and bone in the recipientarea. If the cervical approximation is good, because the bone tissuebelow the cervical portion is a closed wound and there is a lower chanceof infection, there is a tendency to heal well without problems (2). Whena maxillary tooth is moved to the mandible because the buccolingualwidth of the maxillary tooth is wider than the recipient area in themandible in most cases excessive bone must sometimes be removed.In such cases, if the maxillary tooth is rotated before it is placed, itcan be positioned in an anatomically appropriate manner withoutremoving excessive alveolar bone (2).

In autogenous tooth transplantation, long-term firm fixation mayhave negative effects on healing, whereas nonrigid fixation for 7 to 10days stimulates the activation of alveolar ligament cells and bone healing(18, 19). Tsukiboshi et al (1) reported that the tooth should be fixed forbetween 2 weeks and 2 months depending on whether the mobility isreduced. In our series, the fixation was removed after 2 to 3 weekswhen any vertical mobility had disappeared.

Following Chamberlin and Goerig (20), tooth transplantation wasjudged successful if the tooth was fixed in its socket without residualinflammation; masticatory function was satisfactory and withoutdiscomfort; the tooth was not mobile; no pathological condition wasapparent radiographically; the lamina dura appeared normal radio-graphically; the tooth showed radiographic evidence of root growth;and the depth of the pocket, gingival contour, and gingival colorwere all normal. The prognosis of autogenous tooth transplantationdepends on the level of root development, the formation of the rootapex, the condition of the periodontal ligament of the transplantedtooth, the method of tooth fixation, the match between the transplantedtooth and recipient socket, and the time of endodontic treatment (1, 8,




Figure 3. When the transplant site is close to the maxillary sinus, the septal bone in the extraction socket was elevated. (A) A panoramic radiograph of a 25-year-oldwoman with severe caries of the right maxillary first molar (#3). The tooth could not be restored, and it was decided to transplant the third molar (#1). (B) The septalbone was elevated into the sinus cavity using a Summers osteotome. (C) A panoramic radiograph immediately after transplantation. Because the recipient site wasclose to the maxillary sinus, the septal bone was elevated from the maxillary sinus floor and the tooth transplanted. (D) A photograph taken immediately after trans-plant. The #1 tooth was transplanted to the #3 position. (E) An intraoral radiograph 1 week after transplantation. The septal bone is elevated in the recipient extractionsocket. (F) An intraoral radiograph immediately after endodontic treatment (1 month after transplantation). The features of the root apex of the transplanted tooth andthe distal bone regeneration are shown. (G) An intraoral radiograph 6 months after transplantation. (H) A photograph after prosthesis treatment. (I) An intraoralradiograph 19 months after transplantation. No inflammation or resorption was detected. The marginal bone support appeared similar to that of the neighboring teeth.A continuous periodontal space was present radiographically around the transplanted teeth. (This figure is available in color online at www.aae.org/joe/.)


10). Transplanted teeth have a poor outcome because of the failure ofperiodontal reattachment or the occurrence of root resorption in theengrafted cementum-root surface (2, 21). The failure of cementumreattachment may be induced by periodontal inflammation,inflammation in the alveolar socket, or in cases with insufficient earlyfixation after transplantation. Therefore, transplantation iscontraindicated in cases with infection in the root apex. During toothextraction, chronic inflammatory tissues should be removedcompletely. In this study, the follow-up period averaged 15 months,which is shorter than in other studies. Nevertheless, 16 of the 19 casesmet the success criteria for an 84% success rate. No inflammation orreplacement root resorption developed during the follow-up period.This result is comparable to the success rates of autogenous tooth trans-plantation of teeth with a mature root apex reported by Lundberg andIsaksson (6) and Majare et al (7). In our series, three cases failed.The possible cause of the failure in all cases was the poor periodontalcondition caused by the incomplete removal of chronic inflammatorytissues. In one case, early stabilization could not be achieved, and splintfixation was required for 3 months; however, the level of tooth mobilitydid not decrease, and it was considered a failure and extracted.


Autogenous tooth transplantation is a procedure used in caseswhen restoration is impossible because of, for example, severe dentalcaries, root fracture, alveolar problems, or the failure of endodontictreatment. It involves transplanting and fixing another of the patient’steeth. If the cases are selected properly and appropriate surgery andmaintenance are performed, the success rate is relatively high, and itcontributes greatly to prolonging the function of the natural teeth.

References1. Tsukiboshi M, Andreasen J, Asai Y. Autotransplantation of teeth. Chicago; Quintes-

sence, 2001:10–4, 97, 152–67.2. Lee S. Transplantation and replantation of teeth. Seoul: Shinhung; 2008. 8–15, 92–116.3. Apfel H. Autoplasty of enucleated prefunctional third molars. J Oral Surg (Chic)

1950;8:289–96.4. Miller HM. Transplantation; a case report. J Am Dent Assoc 1950;40:237.5. Tsukiboshi M. Autotransplantation of teeth: requirements for predictable success.

Dent Traumatol 2002;18:157–80.6. Lundberg T, Isaksson S. A clinical follow-up study of 278 autotransplanted teeth. Br

J Oral Maxillofac Surg 1996;34:181–5.7. Mejare B, Wannfors K, Jansson L. A prospective study on transplantation of third

molars with complete root formation. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2004;97:231–8.

Autotransplantation of Teeth with Complete Root Formation 1425


8. Schwartz-Arad D, Herzberg R, Levin L. Evaluation of long-term implant success. J
Periodontol 2005;76:1623–8.9. Newman M, Takei H, Carranza F, et al. Carranza’s Clinical Periodontology. 10th ed.

Philadelphia, PA: WB Saunders Co; 2006:546.10. Schwartz O, Bergmann P, Klausen B. Autotransplantation of human teeth. A life-table

analysis of prognostic factors. Int J Oral Surg 1985;14:245–58.11. Smith JJ, Wayman BE. Successful autotransplantation. J Endod 1987;13:77–80.12. Andreasen JO. Periodontal healing after replantation and autotransplantation of

incisors in monkeys. Int J Oral Surg 1981;10:54–61.13. Lindskog S, Blomlof L. Influence of osmolality and composition of some storage

media on human periodontal ligament cells. Acta Odontol Scand 1982;40:435–41.

14. Andreasen JO. Effect of extra-alveolar period and storage media upon periodontaland pulpal healing after replantation of mature permanent incisors in monkeys. Int JOral Surg 1981;10:43–53.

1426 Bae et al.

15. Herman GT, Coin CG. The use of three-dimensional computer display in the study ofdisk disease. J Comput Assist Tomogr 1980;4:564–7.

16. Lee SJ, Jung IY, Lee CY, et al. Clinical application of computer-aided rapid prototyp-ing for tooth transplantation. Dent Traumatol 2001;17:114–9.

17. Fong CC. Transplantation of the third molar. Oral Surg Oral Med Oral Pathol 1953;6:917–26.

18. Pogrel MA. Evaluation of over 400 autogenous tooth transplants. J Oral MaxillofacSurg 1987;45:205–11.

19. Sange S, Thilander B. Transalveolar transplantation of maxillary canines. A follow-up study. Eur J Orthod 1990;12:140–7.

20. Chamberlin JH, Goerig AC. Rationale for treatment and management of avulsedteeth. J Am Dent Assoc 1980;101:471–5.

21. Andreasen JO. Relationship between surface and inflammatory resorption andchanges in the pulp after replantation of permanent incisors in monkeys. J Endod1981;7:294–301.


journal of tics joe 2010 issue august

Documents