Precision of CNC-MilledTitanium Frameworks for Implant

Treatment in the Edentulous JawTorsten femt, LDS, Odont Dr*Tomas Back, MS&Anders Petersson, MS(f

Purpose: The purpose of this report was to describe a new technique to fabricate one-piece, implant-supported titanium frameworks by means of a computer numeric-controiied (CNO miiiing technique, and to compare the fit of these ramevvork withconventionai cast prostheses. Materials and Methods: The study comprised 20 patientswho were provided with 5 standard Brnemark implants each in the edentuiousmandible. The fit of the first 10 prostheses with CNC-milled frameworks was measured bymeans of a 3-dimensional photogrammetric technique. The distortion of the center pointof the framework cylinders was measured in relation to the master cast replicas. Thesemeasurements were compared with 10 randomly selected routine prostheses with castframeworks. Results: No significant differences were found between the 2 groups. The3-dimen5onal distortion of the cylinders in the compieted prostheses ranged from 3 to80 (jm: no passive framework was observed. More distortion was observed in thehorizontai plane (x and y axes) as compared to the distortion in tbe vertical direction2 axis). Conciusion: The precision of fit of the first CNC-milied prostbeses presented acomparable fit to the conventionai cast frameworks, Consequentiy, tbis new techniquecouid be a valid option for the routine fabrication of frameworks for implant prostheses.int] Prosthodont 1999,-12:209-215.

Edentulous patients have been treated with fixedprostheses supported by osseointegrated impiantsad modum Brnemark since 1965. The first patientswere provided with frameworks in Cr-Co alioy withresin teeth.' This prosthetic protocoi was modifiedovertime, and goid-aiioy castings were introduced forimpiant treatment in the eariy 1970s.' The reason for

'Associate Professor and Head, Prosthodontic Division, TheBrnemark Ciinic, Demat Health Care, Gteborg, Sweden.'Manager, tmpiant Prosthetic Production, Nobei Biocare,Gteborg, Sweden.^Manager, Electronics and Research Support, Nobel Biocare,Gteborg, Sweden.

Reprint requests: Dr Torsten Jemt. The Brnemark Clinic, PublicDental Health, Mediciriaregatan 12 C, S-413 90 Gteborg,Sweden. Fax: + 46(0131 82 73 29.e-maih torsten.iemt@odontologi.gu.se

using goid ailoy castings was to provide a more sta-bieocciusion in metai and to ai iow for porcelain ve-neering materials.'-- However, in many cases with se-vere bone rsorption, large amounts of goid aiioy hadto be cast when the occlusion was placed in metai.These heavy frameworks caused casting problemsand the risk of distortion of the frameworks increased.To compensate for the casting probiems associatedwith the iost-wax casting technique and to improveprecision of the frameworks, premachined goid-aiioycylinders were introduced on a reguiar basis in theBrnemark impiant system (Nobel Biocare) at thistime.' The cyiinders were first cast into the frame-work,^** but soldering and luting the cylinders to themetai frame were aiso tried.'-^ To simpiify the implantclinical and iaboratory protocol, further modifica-tions to use more standardized bar framewori

Precision of CNC-Milled Titanium Frameworks

Table 1 Clinical and Laboratory Protocol Used toFabricate Fixed Prostheses with CNC-Milled TitaniumFrameworks

Patient visit Clinic" Laboratory

123

45

Final impressionJaw recordingFunctional and

esthetic assessment

F rame werk try-inFinal insertion

iviaster cast trial baseArticulator tocth set upResin frame werk patternLaser scanningCNC milling in titaniumCompletion ot prosthesis

Tlie clinical protocol is the saim plant-sup ported prostlieses.

1 that used tor conventionally cast

Besides the conventional techniques for cast frame-works, other methods to manufacture prostheses forimplant rehabilitation have also been tested. Onesuch technique is to use prostheses based on carbon/graphite fiber-reinforced poly (methyl methacrylate)^;these prostheses have been shown to function wellin clinical foliow-up studies,'' Another noncastingapproach is to use laser-welded titanium frame-works,^'^ These frameworks are based on prefabri-cated titanium components, and different techniqueshave been described for manufacturing the metalframes.""^' Clinical experience with titanium frame-works has been encouraging, presenting results com-parable to those reported for conventional castingtechniques.'*''' The most recent laser-welded pros-theses (titanium type 3 frameworks) were manufac-tured by welding small, prefabricated titanium com-ponents fol lowed by manual contouringof the metalbeam by the technician before curing resin teeth tothe framework.'^'"'^ However, a major problemwith this technique is that the manual contouring ofthe metal is too time consuming in the laboratory, andfurther modifications of the titanium protocol there-fore had to be considered.

A completely new protocol for fabricating titaniumframeworks has been developed, in which no laserwelding or prefabricated components are used. Theaims of the present study were to describe this newtechnique for fabrication of fixed implant-supportedprostheses in theedentulous aw, to measure the pre-cision of tbe fabricated prostheses to the master cast,and to compare these measurements with the preci-sion of routine conventional cast frameworks.

Materials and Methods

The present study covered 2 different groups of fixedprostheses tbat were manufactured and clinically ac-cepted for final connection to implants in tbe edentu-lous mandible. Each group comprised 10 prostbeses.

and all prostheses were made to fit 5 standardBrnemark system abutment cylinders eacb. Onegroup of prostheses was made with titanium frame-works according to the new computer numeric-controlled (CNC] milling technique described below.The second group of prostheses was a control thatused conventional cast frameworks with resin teeth.

Reference Group

Ten edentulous patients who were treated witb stan-dard fixed prostbeses with conventional cast frame-works"-'^ were randomly selected to form the controlgroup. These patients and prostheses have been de-scribed in more detail by Jemt and Lie.^^ Fabricationof the frameworks was performed by two dental lab-oratories with experience of more tban 1,000 implantcases eacb during more tban 10 years. All patients re-ceived a type III gold-ailoy casting witb standardgold-alloy cylinders (DCA 072, Nobel Biocarej.Theframeworks were cast in one piece, and after try-inof the metal frames resin teeth were cured to thecastings.'' Try-in of tbe prostheses was performed asdescribed below.

Test Graup

The test group comprised the first 10 consecutiveprostheses that were milled from one piece of tita-nium in a CNC milling machine. These prostheseswere randomly selected from patients that weretreated with fixed implant-supported restorations inthe edentulous mandible at one c l in ic (TheBrnemark Clinic, GOteborg, Sweden) from Januaryto June 1997. The frameworks were fabricated by onemanufacturer (Nobel Biocare).

Clinical Procedures

The clinical restorative protocol was the same for the2 groups (Table 1), basically following a routine 5-appointment clinical procedure.^^ After completion,all prostheses were clinically tested for acceptable fit.This technique of clinical assessment has been de-scribed in more detail elsewhere.^^ In brief, tbe try-in was performed by following a protocol in whichthe bridge locking-gold alloy screw was initiallytightened to the first resistance felt with the screw-driver.^^ From this position of the screw not more tbanhalf a turn of further tightening was accepted untilcomplete seating of tbe screw was achieved, using atorque moment of 10 to 15 Ncm. When they wereclinically accepted, a 3-dimensional (3-d) photo-graph was taken of the master cast and the fit surfacesof the prostbeses in the 2 groups (Fig 1).

i oi PfOBthodontit 210 Volume T2, Number?, ' " " "

leml el al Precision of CNC-Milled Titanium Frameworks

Fig 1 Photograph of a complete CNC-milled prosthesis frame-work taken with the oamera equipment described. At least 3 im-ages of the object are given in the same exposure.

Fig 2 (Rtghl) Trial base with final denture tooth arrangement(top), framework resin pattern (middle), and final CNC-milled ti-tanium framework (bottom). The titanium framework is a milledcopy of the scanned resin pattern in one piece.

Fig 3 Shape of the framework resin pattern is scanned. Fig 4 Milling machine with the titanium framewori

Precision of CNC-Milled Titanium Frameworks

Fig 5 fLeft; Completed CNC-milled ti-tanium framework. Note that the oylin-ders are integral parts of tbe framework,and tbers are no welding joints.

Fig 6 (Right) Completed prosthesis withCNC-milied titanium framework and resinteeth.

Table 2 Mean and Standard Deviation (SD) of Distance Between 2 Terminal Implants (Lateral) and DistanceBetween Most Distal and Most Anterior Implants (Sagittal)

Group

ReferenceCPJC-mitled titanium

Lateral arch distance in mm(x axis)

Mean

30.728.9

SD

2.13.2

Range

27.2-33.924.6-33.1

Sagittal arch distance in mm(y axis)

Mean

8.47.3

SD Range

1.0 7.1-1D.41.7 3.7-10.0

Mean

0.30.3

y/x ratio

SD

0.030.05

(mm)Range

0,2-0.30.2-0.3

The photographs were placed in an analytic plot-ter under stereoscopic vision (Leica, Kern) for 3-dmeasurements"''' of the prosthesis and master castreplica cylinder center points (Atlas 2000, Metimur).Each pair of measurements (master cast-framework)was superimposed in the computer and analyzed foroptimal fit by means of the least squares method ,^ ''"'^ ''Distortion of the frameworks was measured and re-ported in relation to the center point of the cylinders,comparing the difference of x, y, and z positions aswell as the 3-d distortion for each individual cylinder.

The error of this measuring technique has been re-potted in more detail elsewhere.^' Measurements fthe individual center points were calculated to a mean3-d precision of 1 2 |jm in repeated measurements.Arch Form

All prostheses in the 2 groups were supported by 5implants each, which were provided with standardabutments. The width of the implant arch'^ was as-sessed by measuring the distance between the cen-ters of the 2 most distal implants (x axis). The curva-ture of the arch was determined by measuring thesagittal distance from a line between the 2 most dis-tal implantstothecenterofthemost anteriorly placedabutment (y axis). The measurements of the archforms of the 2 groups are given in Table 2.

Statistics

Conventional descriptive statistics were used to presentthe distortion of the frarTieworks. The measurements

were also presented in absolute figures to show the de-gree of distortion without considering the direction ofthe displaced cylinder, ie, left-right (x axis), anterior-posterior (y axis), or up-down (z axis). Wilcoxon'sranked sum test^ ^ was used for statistical comparisonsbetween the 2 prosthesis designs. A P value of > 0.05was not considered significantly different.

Results

Reference Group

Distortion of individual cylinders showed a widerrange in the horizontal plane (x and y axes) as com-pared to the vertical direction (z axis) (Table 3). Therange of mean 3-d distortion measured was from 16to 80 |jm (Tables 3 and 4), with an overall mean 3-ddistortion of 42 pm (standard deviation 15 |jm).Accordingly, no prosthesis showed a perfect fit.

When considering the direction of displacement,the mean distortion in the 3 individual axes was onaverage insignificant, but the standard deviations ofthe means indicated more variation of misfit in thehorizontal plane (x and y axes; Table 4). When an-alyzing the same measurements as the mean of thedisplacement in absolute figures (disregarding di-rection of distortion), the misfit in the horizontalplane became more obvious as compared with thesagittal displacement (2 axis] of the cylinder (Table5), No pattern of distortion in relation to position ofthe implants could be observed (Tables 4 and S).

Tiie International louind of Prostiiodontics 212 Voli impl2, Number3

Jemt et al Precision of CNC-Milled Titanium Frameworks

Table 3 Maximal Range of Individual Center-Point Distortions of Frameworks (pm)Reference group CNC-milled titanium group

xaxisy axiszaxis3-d

Min

- 6 1- 5 3-a4

16

Max

63654080

Range

124118

6464

Min

- 7 5- 6 8- 4 2

3

Max

72443476

Range

147 ^11276 73

Table 4 Mean and Standard Deviation (SD) of Distortion of Center-PointOrientation of Frameworks (pm)'

Position

1MeanSD

2MeanSD

3MeanSD

4MeanSD

5MeanSD

TotalMeanSD

X

529

626

- 1 632

929

736

030

Reference

y

135

435

- 1 118

334

927

Q30

1 group

z

47

- 612

116

720

59

015

3-d

4215

4117

3919

477

4219

4215

X

922

729

1420

- 2 631

- 2 634

030

CNC-milled titanium

y

615

-1127

1123

524

-923

030

z

28

20

918

- 613

28

217

group

3-d

2612

4020

-am

389

3519

4421

3718

Mean error vector ot center peint trom right (position 1) to iett (position 5).

Table 5 Mean and Standard Deviation (SD) of Distortion of Center-PointOrientation of Frameworks in Absolute Figures (Mm)'

Position

1MeanSD

MeanSD

MeanSD

MeanSD

MeanSD

TotalMeanSD

X

2416

1917

3018

2515

2823

2517

Reference group

y

2819

3114

1516

2817

2315

2517

z

64

611

1210

1513

87

109

3-d

4215

4117

3919

477

4219

4215

K

1815

2514

1813

2220

3622

2418

CNC-miiled titanium group

V

1210

2121

1914

2012

1815

1915

z

65

1514

1412

128

84

1110

3-d

2612

4020

389

3519

4421

3718

Mean absolute error vector ot center poinl from nght (position 1 to left (position 5].

: Number 3,1999 2 1 3 The [ntemationaI lournai of ProthodenticB

Precision of CNC-Milled Titanium Frirreworks

Fig 7 Conventional prosthesis with cast gold-alioy trameworkand resin teeth (top replaced by CNC-milled titanium framewort^with porcelain veneers [bottom).

CNC-Milled Titanium Group

The range of distortion of the center points of theCNC-miiled titanium frameworks was on a similarievei to that observed for tbe reference group (Tabie3). Misfit with regard to direction of dispiacement(Table 4) as weli as in absoiute figures (Tabie 5) wasalso in accordance with the cast goid-aiioy frame-works. No significant differences of distortion of frame-works couid be found between the groups (P>0.05|.

Discussion

This paper presents a compieteiy new technique tofabricate frameworks for impiant-supported prosthe-ses with precision comparabie to that found in con-ventionai cast frameworks made by very experiencediaboratories. Since these first 10 prostheses weremade, another 600 CNC-miiied titanium frameworkshave been fabricated duringthe first year, indicatingthat this new laboratory procedure (AI i-in-One, NolielBiocare) is a vaiid alternative to conventional tech-niques for routine impiant rehabiiitation.

One of the reasons for deveioping a laboratoryprocedure for titanium frameworks was to try to im-prove the fit of the superstructures to the impiants byavoiding some of the inherent probiems associatedwith the iost-wax casting technique.'^''' By using anindustriai manufacturing protocoi for the frameworks,many factors reiated to manuai handiing of the con-ventionai castings can also be controiied andavoided. As shown in the present study, even the firstciinicai cases reached a ievel of fit that compares fa-vorably with cast prostheses from weii-establishediaboratory teams. Even though there is no data avaii-abie to indicate that a better fit than that presentedhere is needed,'*-'^''"'^^ it is reasonable to assume thatthe precision of the CNC-milied frameworks may befurther improved by additional refinement of the laserscanning and CNC miiiing procedures.

The distortion of the cast impiant framework has ear-iier been shown to increase as the curvature of the Im-piant arch and the weight of the casting olioy in-crease."'These inherent variabies for then i.iventionaiiost-wax casting technique are not present lur the CNCmiiiing procedure. This aiiows for better fii of the tita-nium frameworks in difficult ciinicai cases with manyimplants in large, curved impiant arches. Some ciini-cians have refrained from using porceiain-fused-to-metai impiant prostheses in the severeiy resorbed eden-tuious jaw because of the cost and fit probiemsassociated with conventionai goid aiioy castings. TheCNC miiiing procedure provides a solution to this fitprobiem, as the macbining of the fit surfaces of theframework is not affected by the totai amount of metalin the superstructure. Accordingly, CNC-miiied tita-nium frameworks can also be considered for rehabii-itation with prostheses with porcelain veneers in thesevereiy resorbed edentuious jaw (Fig 7).

The ciinicai performance of fixed impiant-supported prostheses with titanium frameworks hasshown resuits comparabie to those found for con-ventionai restorations,^'""'"' Few severe probiemsbave been reported by these patients, and none ofthese problems has been positively associated withthe metai in tbe superstructure,'^ iHowever, eventhough some studies have reported few, if any, frac-tures of the titanium metal frames,"''^ others have ob-served a siightiy higher incidence of fractured frame-works for the titanium prostheses,'"* These fractureshave been in reiation to the welding joints. With apenetration depth of about 0.6 mm in theweid,^''thisarea must be considered to be the weak part of theframework and shouid not be adjusted by grindingafter completion,'^"''' it is reasonabie to assume thatthe risi< of fractures of the metai framework wouid besignificantiy reduced by fabricating the metai super-structure from one soiid biock of titanium.

As discussed above, the CNC miiiing procedureeliminates probiems with distortion caused by con-ventionai iost-wax casting techniques, tbereby ai-iowing for more predictabie manufacturing of theframeworks to the master casts, ' ' iHowever, the iimi-tation of this new technique is that the intraoral fit ofthe framework wiii never be better than the accuracyof the master cast."'^'' Since the master cast never pre-sents an exact orientation of the intraorai impiants-^^it shouid be of interest not only to optimize the fit ofthe framework to the master cast, but aiso to furtherimprove the accuracy and the precision of the regis-tration of intraorai impiant position. Since this CNCmiiiing technique was introduced, it has become pos-sibie to use new digital means for recording impiantorientation based on numeric impiant position datafrom the master cast. Recent exprimentai ciinical

The Iniernational Journal of Prosthodontics 2 1 4 Volume 12, Number 3,1999

leint et a\ Precisicn of CNC-Milied Titanium Frameworks

studies have measured bone and framework defor-mations caused by misfit,-^ soft tissue volume adjacentto implants,-' and oral implant positions^^ by meansof the photogram metric technique used in the presentstudy. These recent studies indicate a new option forrecording implant positions without using conven-tional impression procedures.

Acknowledgments

The CNC tramework prostheses used in thi study were fabricatedby Mr Tond |alevi