otite- tratamento - treatment and prevention om

7/27/2019 Otite- Tratamento - Treatment and Prevention OM

1/17

1452 The Annals of Pharmacotherapy 2000 December, Volume 34 www.theannals.com

Otitis media (OM) is second to the common cold as aninfectious cause of disease in infants and children.1Established risk factors include male gender, Native Amer-

ican or Eskimo descent, pacifier use, larger-group day caresettings, exposure to tobacco smoke, and lack of breast-feeding.2 The medical and surgical costs associated withtreatment of OM within the US have been estimated at be-tween $3 billion and $4 billion annually.3,4 This reviewsummarizes and highlights recent advances in the treat-

Treatment and Prevention of Otitis Media

John Erramouspe and Catherine A Heyneman

Pediatrics

OBJECTIVE: To review and summarize recent advances in the treatment and prevention of otitis media (OM).

DATA SOURCES: A MEDLINE search (1996March 2000) was performed to identify relevant primary and review articles. Referencesfrom these articles were also reviewed if deemed important.

STUDY SELECTION AND DATA EXTRACTION: English-language primary and review articles focusing on the treatment and prevention ofacute otitis media (AOM) were included. Studies focusing exclusively on OM with effusion or serous OM and chronic suppurative

OM were excluded. Information regarding prevention and drug therapy was reviewed, with an emphasis placed on advances made

in the last two years.

DATA SYNTHESIS: Recently, an expert panel of the Centers for Disease Control and Prevention recommended use of only three of 16systemic antibiotics approved by the Food and Drug Administration for treatment of AOM: amoxicillin, cefuroxime axetil, andceftriaxone. Controversy exists over the importance of key selection factors used by the expert panel in determining which antibiotics

to recommend in a two-step treatment algorithm, that is, in vitro data, pharmacodynamic profiles, and necessity for coverage of drug-

resistant Streptococcus pneumoniaeat all steps of empiric treatment. Additional antibiotic and patient selection factors useful for

individualizing therapy include clinical efficacy, adverse effects, frequency and duration of administration, taste, cost, comorbid

infections, and ramifications should bacterial resistance develop to the chosen antibiotic. Presumed or past patient/caregiver

adherence (especially when antibiotic failure has occurred) is also paramount in selecting antibiotic therapy. A three-step treatment

algorithm for refractory AOM that employs amoxicillin, trimethoprim/sulfamethoxazole (TMP/SMX), or high-dose amoxicillin/

clavulanate (depending on the prior dose of and adherence to amoxicillin therapy), and ceftriaxone or tympanocentesis at steps 1, 2,

and 3, respectively, appears rational and cost-effective. The recent upsurge in antimicrobial resistance is highlighted, and

recommendations are presented for the treatment of AOM and prevention of recurrent otitis media (rAOM).

CONCLUSIONS: Amoxicillin remains the antibiotic of choice for initial empiric treatment of AOM, although the traditional dosageshould be increased in patients at risk for drug-resistant S. pneumoniae. In cases refractory to high-dose amoxicillin, TMP/SMX

should be prescribed if adherence to prior therapy seemed good or complete, or high-dose amoxicillin/clavulanate if adherence wasincomplete or questionable. Ceftriaxone should be reserved as third-line treatment. The increasing prevalence of drug-resistant S.

pneumoniaeemphasizes the importance of alternative medical approaches for the prevention of OM, as well as judicious antibiotic

use in established cases. Removal of modifiable risk factors should be first-line therapy for prevention of rAOM. We support the use

of conjugate pneumococcal vaccine per guidelines for prevention of rAOM from the Advisory Committee on Immunization Practice

of the Centers for Disease Control and Prevention, with consideration given to influenza vaccine for cases of rAOM that historically

worsen during the flu season. Sulfisoxazole prophylaxis should be reserved for children who are immunocompromised, have

concurrent disease states exacerbated by AOM, or meet the criteria of rAOM despite conjugate pneumococcal and influenza

vaccination. Therapy should be intermittent, beginning at the first sign of an upper respiratory infection, and should continue for 10

days. The invasive nature and risks of anesthesia relegate myringotomy, tympanostomy tubes, and adenoidectomy to last-line

therapies for rAOM.

KEY WORDS: otitis media, antibiotics, prophylaxis, resistance, Streptococcus pneumoniae, oligosaccharides, xylitol, conjugatepneumococcal vaccine, influenza vaccine.

Ann Pharmacother2000;34:1452-68.

ACPE UNIVERSAL PROGRAM NUMBER: 407-000-00-028-H01

Author information provided at the end of the text.


2/17

ment and prevention of OM, with an emphasis placed oninformation reported over the past two years. More de-tailed background information regarding the etiology, epi-demiology, pathophysiology, and pharmacoepidemiologyof OM is beyond the scope of this article. Thorough re-views of these topics are available.2,5-7 The objectives ofthis article include updating the reader regarding the ratio-nale behind current acute OM (AOM) treatment guide-

lines, as well as reviewing and ranking alternative medicalapproaches for the prophylaxis of recurrent OM (rAOM)in a treatment algorithm.

Treatment of Acute Otitis Media

Streptococcus pneumoniae (pneumococcus) is the mostimportant bacterial cause of AOM, followed closely bynontypable Haemophilus influenzae and Moraxellacatarrhalis.8 In recent years, there has been a worldwideincrease in strains of pneumococcus that are resistant topenicillins as well as other antibiotics. The National Com-mittee for Clinical Laboratory Standards (NCCLS)9,10 de-

fines breakpoints for susceptible, intermediate-resistant,and resistant pneumococci according to minimum inhibito-ry concentrations (MICs) of 0.06, 0.121.0, and 2.0g/mL (for penicillin G) and 2.0g/mL (for amoxicillin), respectively. The increasing resis-tance of pneumococcus to many of the antibiotics typicallyused in treating AOM has affected recent treatment guide-lines in the US. In 1999, a panel of experts convened bythe Centers for Disease Control and Prevention (CDC) rec-ommended new guidelines for treating AOM with antibi-otics (Table 1).

FIRST STEP: HIGH-DOSE AMOXICILLIN

The CDCs expert panel8 advocated amoxicillin as first-line treatment of AOM but recommended increasing theinitial dosage used for empiric treatment from 40 to 45mg/kg/d (usual or standard dose) to 8090 mg/kg/d (highdose). The rationale behind this recommendation is that,

unlikeH. influenzae andM. catarrhalis, S. pneumoniaedoes not produce -lactamase.2 The mechanism of pneu-mococcal resistance is alteration in penicillin-binding pro-teins, a genetic ploy that could theoretically be overcomeby simply increasing penicillin/amoxicillin concentrationsin the middle ear fluid (MEF). What clinical evidence ex-ists to support such a dosage increase?

Lister et al.11 employed an in vitro pharmacodynamic

model to study the killing of three strains each of peni-cillin-susceptible (MIC 0.06 g/mL), penicillininterme-diate-resistant (MICs 1 strain, 0.25 g/mL; 2 strains 0.5g/mL), and penicillin-resistant (MIC 4 g/mL) pneumo-cocci. Three different peak amoxicillin concentrations, 3,6, and 9 g/mL (achieved every 12 h) and an eliminationhalf-life of 1.6 hours were simulated. Extrapolated from aconcentrationtime curve, it appeared that the percentagesof the dosing interval during which the amoxicillin con-centration exceeded the MIC for the susceptible, interme-diate-resistant, and resistant pneumococcal isolates were67%, 4755%, and 0% (3 g/mL peak profile); 100%,7084%, and 10% (6 g/mL peak profile); and 100%,7988%, and 22% (9 g/mL peak profile), respectively.These results suggested that peak amoxicillin concentra-tions of 6 and 9 g/mL might be sufficient for eliminationof penicillin-nonsusceptible pneumococcal strains, espe-cially those of intermediate resistance.

Canafax et al.12 administered a single oral dose of amox-icillin 25 mg/kg (following 23 d of standard oral dosingof amoxicillin 13.3 mg/kg/dose every 8 h) to 30 childrenwith AOM. The regular morning dose of 13.3 mg/kg hadbeen withheld prior to administering 25 mg/kg. At leastone MEF sample per patient (37 total samples) was col-lected between 0.5 and four hours after dosing. An MEFconcentration mean of approximately 9.5 g/mL (range,undetectable to 20.6 g/mL) occurred three hours after 25mg/kg was administered. The estimated time that the MEFamoxicillin concentrations were above 1.0 and 2.0 g/mLwas approximately four and 2.5 hours (50% and 31% of an8-h dosing interval), respectively. The MEF amoxicillinpenetration tended to be lower in patients with concurrent

The Annals of Pharmacotherapy 2000 December, Volume 34 1453www.theannals.com

Table 1. Recommended Antibiotics for Treating Acute Otitis Media in Children8

InitialClinical Treatment Failure

Antibiotics in Prior Month Alternative Choices Day 3 Days 1028

No HD amoxicillina HD amoxicillin/clavulanateb HD amoxicillin/clavulanateb

standard-dose amoxicillin cefuroxime axetil cefuroxime axetil

im ceftriaxonec im ceftriaxonec

Yes HD amoxicillin im ceftriaxonec HD amoxicillin/clavulanate

HD amoxicillin/clavulanate clindamycind or tympanocentesis cefuroxime axetil

cefuroxime axetil im ceftriaxonec or tympanocentesis

HD = high-dose.a8090 mg/kg/d.bRequires formulation not yet commercially available or two separate prescriptions: amoxicillin/clavulanate (from 6.4 to ~10 mg/kg/d of clavulanate)and amoxicillin to boost the cumulative dosage to 8090 mg/kg/d.c50 mg/kg im once daily for three days.dNot effective against Haemophilus influenzaeor Moraxella catarrhalis.


3/17

viral infection compared to those without. One weaknessof the study was the lack of measurement of baseline amox-icillin concentrations prior to the 25-mg/kg dose. Singledoses of amoxicillin 15 mg/kg have generated mean peakvalues of 36 g/mL.11 Using this information, the contri-bution from a dose of 13.3 mg/kg given at least eight hoursearlier is estimated at 1 g/mL according to the 3- and 6-g/mL peak profiles generated by Lister et al. Another

weakness of this study12 was the exclusion of 13 MEFsamples in which the amoxicillin concentration was unde-tectable. These samples were excluded because of low vol-umes (e.g., 8 of 13 were 2 y, noday care attendance).

We support the routine empiric first-line use of high-dose amoxicillin in children at high risk for DRSP-mediat-ed AOM who are not truly allergic to penicillins. Our supportis based on the following information. First, S. pneumoniaeis the most common cause of bacterial AOM (4050%8)and the increasing incidence of DRSP in the US.16

Second, there is a need to eradicate pathogenic bacteriafrom the middle ear rather than simply produce clinicalimprovement (clinical improvement alone may reflect thehigh natural spontaneous cure rate for AOM rather than anantibiotics effectiveness). A meta-analysis17 of 33 random-ized trials involving 5400 children with AOM reported aspontaneous cure rate of 81% (95% CI, 69% to 94%) aspart of the natural history of untreated AOM. Antibiotic

therapy statistically raised this cure rate by only 13.7% (95%CI, 8.2% to 19.2%). The high spontaneous cure rate forAOM is the reason practitioners in the Netherlands andScandinavian countries generally first treat AOM in chil-dren older than two years of age symptomatically withanalgesics alone, reserving antibiotic therapy for patientswhose AOM fails to resolve after three to four days.18-20

Conclusions about antibiotic efficacy for AOM based sole-ly on either clinical improvement or bacteriologic eradica-tion should be viewed with suspicion since they are not al-ways correlated.21,22 Both should be distinct goals of antibi-otic therapy for AOM and documented in clinical trials.Comparative trials using only clinical outcome criteria toassess efficacy between antibiotics have generally not dem-onstrated differences, quite possibly because they lackedadequate sample size or power. Larger sample sizes arenecessary to show true clinical differences between twoantibiotics, assuming they do exist. For example, assuminga power of 90% and an of 0.05, one would need to enrolla total of at least 2054 subjects to confirm the suspicionthat antibiotic A has 5% greater clinical efficacy than an-tibiotic B.23 The number of subjects required decreases, butonly to 473, if a difference in clinical efficacy of 10% is as-sumed. Adequate funding for larger, solely clinically basedoutcome studies is usually difficult to obtain. Clinical cou-

pled with bacteriologic outcome criteria may be more sen-sitive in detecting true differences between antibiotics andmore acceptable for the funding agency.

Third, amoxicillin displays a favorable pharmacody-namic profile (i.e., longest time above MIC90 for DRSP ofany commercially available oral antibiotic).24 The pharma-codynamic profile of an antibiotic used for AOM is con-sidered its ability to penetrate into and persist in MEFabove the MIC of typical AOM bacteria. Antibiotics thatexceed the MIC90 of typical AOM bacteria in MEF for atleast 4050% and 6070% of the dosing interval havebeen associated with bacteriologic eradication rates of8085% and approaching 100%, respectively. The time

that an antibiotic persists at the site of action in a concen-tration greater than the MIC of a pathogenic organism is asurrogate marker of microbiologic and clinical efficacythat appears appropriate to use in selecting antibiotics thatdemonstrate concentration-independent killing (e.g., amox-icillin, cephalosporins).

Finally, amoxicillin has traditionally been a well-tolerat-ed antibiotic, even at high dosages. No difference was de-tected in adverse effects among 274 children, aged three to10 years, randomly allocated to receive either amoxicillin125 mg three times daily for seven days or 750 mg twicedaily for two days.25 Although no weights were reported



4/17

for these patients, it is likely that some children receivingthe higher dosage may have been taking approximately75100 mg/kg/d, assuming an average weight range of1520 kg for children three to six years old. Only one caseof diarrhea was reported out of 43 infants and childrenwith OM who received amoxicillin 150 mg/kg/d.26

SECOND-STEP: WHEN AMOXICILLIN INITIALLY FAILS

What is the recommended second-line therapy for AOMwhen high-dose amoxicillin appears to have failed? Con-troversy exists regarding recommendations made by vari-ous authors for antibiotics to treat AOM refractory to high-dose amoxicillin.8,15,27-29 The main controversy dwells onthe need to provide coverage for DRSP if high-dose amox-icillin has already been prescribed. The traditional strategyhas been to prescribe an oral antibiotic that is effectiveagainst -lactamaseproducing bacteria (but not necessari-ly DRSP), is dosed infrequently (usually once to twice dai-ly), is low cost, and has acceptable palatability. However,some practitioners believe that this strategy may not be op-

timal in the present era of increasing prevalence of DRSP.8Two bacteriologic studies30,31 highlight this dilemma.

Gehanno et al.30 identified S. pneumoniae in 67 of 170bacterial isolates (39.4%) taken from 126 AOM patients inthe Paris region who had failed prior oral antibiotics. Themost frequently prescribed prior antibiotic was amoxi-cillin/clavulanate in 43% of cases. Reduced susceptibilityto penicillin (MIC 0.125 g/mL) was present in 52 of thepneumococcal isolates (77.6%).

Similarly, in Washington, DC, pneumococcus has beenfound to be responsible for approximately one-third of thebacterial isolates cultured from the MEF of children not re-

sponding to other oral antibiotics. Antibiotic resistanceamong these pneumococcal isolates was >60%.31 TheCDC panel of experts8 believed that empirically prescribedsecond-line antibiotics should be effective against not only-lactamaseproducingH. influenzae andM. catarrhalis,but also DRSP. One exception to this dual-coverage guide-line would be a culture of the MEF in which a singlepathogenic bacteria was identified. In this case, antibiotictherapy could be definitively narrowed (as opposed tobroad empiric therapy). For example, clindamycin is noteffective againstH. influenzae orM. catarrhalis, but is po-tentially very useful for culture-confirmed pneumococcalAOM, especially when caused by penicillin-resistant S.

pneumoniae (PRSP).The additional requirement for effectiveness against bothDRSP and -lactamaseproducing bacteria is extremelylimiting. Only three of the 16 systemic antibiotics presentlyapproved by the FDA for OM32 are acceptable according tothe CDCs expert panel.8 The other systemic antibiotics32

(i.e., amoxicillin, trimethoprim/sulfamethoxazole [TMP/SMX], erythromycin/sulfisoxazole, azithromycin, clarithro-mycin, cefixime, ceftibuten, cefdinir, loracarbef, cefprozil,cefpodoxime, cefaclor, cephalexin) were not endorsed be-cause of limited data showing these agents capable of eradi-cating DRSP from the middle ear and/or little -lactamase

coverage. The three antibiotics recommended by the expertpanel8 include high-dose amoxicillin/clavulanate (presentlyrequires additional prescription of amoxicillin to boost thedose to 8090 mg/kg/d while limiting the clavulanate to10 mg/kg/d), cefuroxime axetil, and intramuscular ceftri-axone (Table 1).

Amoxicillin/clavulanate 40 mg/kg/d in three divideddoses has been suggested24 as the best oral antibiotic avail-

able for maximizing the time in which the concentrationremains greater than the MIC (T > MIC) for the full spec-trum of AOM bacteria (i.e., S. pneumoniae, H. influenzae,

M. catarrhalis). The T > MIC for DRSP is increased bysupplementing this usual dose of amoxicillin/clavulanatewith a second prescription for additional amoxicillin. Inthe future, high-dose amoxicillin/clavulanate (90/6.4mg/kg/d, twice daily) will likely be commercially availableto facilitate adherence to and coverage of both -lacta-maseproducing bacteria and DRSP.33

Cohen et al.34 found no difference in efficacy between a10-day course of high-dose amoxicillin/clavulanate (80/10mg/kg/d divided in 3 doses daily) and a single dose of cef-triaxone (50 mg/kg im) in the treatment of newly diag-nosed patients with AOM in France (an area with a highrate of PRSP isolated from patients with AOM, >50% asof 1995). A weakness of the study is that tympanocentesiswas not done to identify MEF bacteria, although cultureand sensitivity tests using samples from nasopharyngealswabs were performed prior to treatment and on days1214 to study the effect of treatment on bacterial carriage.No difference existed in the carriage of S. pneumoniae, M.catarrhalis, andH. influenzae between the two groups pri-or to treatment. The reduction of S. pneumoniae andM.catarrhalis carriage at days 1214 was significantly

greater for the amoxicillin/clavulanate group comparedwith the ceftriaxone group (42.6% vs. 14.9% and 47% vs.12.9% reductions, respectively; p < 0.0001 for both). Amox-icillin/clavulanate has a high acquisition cost comparedwith older, low-cost antibiotics (i.e., amoxicillin, TMP/SMX,erythromycin/sulfisoxazole [Table 2]).35 Although diarrheawas a major problem with the original formulations ofamoxicillin/clavulanate (4:1 ratio), the newer formulations(7:1) have substantially lessened this problem.36 The nextgeneration of amoxicillin/clavulanate formulations (14:1)should maintain a similar reduced incidence of gastroin-testinal problems since they should contain a similaramount of clavulanate per dose as the 7:1 formulations.33

Cefuroxime axetil was included as a viable second-lineantibiotic by the CDCs expert panel8 due primarily to oneprospective study of children with pneumococcal AOM.Gehanno et al.37 treated 84 children with AOM with ce-furoxime axetil 30 mg/kg/d in two divided doses for eightdays. Treatment failure, defined as persistent signs of in-fection for >10 days (fever, otoscopic abnormalities, pain),occurred in three of 42 penicillin-susceptible (7%, MIC


5/17

and lack of a comparison/control group were methodolog-ic weaknesses of this study.

In a double-blind taste comparison of 22 antimicrobialsuspensions,38 cefuroxime axetil as well as oxacillin, di-cloxacillin, cefpodoxime, and erythromycin/sulfisoxazole,were judged as having the worst aftertastes, which the au-thors believed could potentially jeopardize compliance.Cefuroxime also has a high acquisition cost relative to old-

er, low-cost antibiotics (Table 2).35 When coverage of both-lactamasepositiveH. influenzae and DRSP is neces-sary, we recommend high-dose amoxicillin/clavulanaterather than cefuroxime (Table 3).38,39 Our selection is basedon greater palatability; similar cost, even with the addition-al amoxicillin prescription (Table 2); and higher pharma-codynamic profile of amoxicillin against DRSP.24,35,38 Themost recent commercially available amoxicillin/clavu-lanate formulations (7:1) should be prescribed at an amox-icillin/clavulanate dosage of 45/6.4 mg/kg/d; additionalamoxicillin should be prescribed to boost the cumulativedosage to 8090 mg/kg/d. Occasionally, some patients orcaregivers find it too difficult to administer two antibioticsat the same time. In these cases, cefuroxime axetil is an ap-propriate choice.

Controversy exists as to what place the oral antibioticsexcluded by the CDCs expert panel8 should hold in AOMtherapy. TMP/SMX has long been beneficial for treatingAOM due to its twice-daily dosing schedule, lack of ad-mixture requirement for the suspension, low cost (Table 2),

and broad effectiveness against most middle ear patho-gens.35 Recently, however, 11.9% of all isolates of S. pneu-moniae were found to be resistant to TMP/SMX (13.4%and 40.4% TMP/SMX resistance for penicillin intermedi-ate- and highly resistant pneumococcal isolates, respective-ly).40 Even isolates ofH. influenzae have been found to beresistant to TMP/SMX in up to 9% of cases.41 Despite theconcerns raised by results of in vitro studies, we are not

aware of any randomized, comparative trial that demon-strates TMP/SMX to have less clinical and bacteriologiceffectiveness than another antibiotic for initial treatment ofAOM. In addition, no randomized, comparative trials ofany antibiotics have been published addressing the treat-ment of AOM initially unresponsive to high-dose amoxi-cillin. Given the lack of comparison with other antibioticsand trends toward greater in vitro resistance, it is probablyappropriate to reserve the use of TMP/SMX to special AOMsituations, such as initial treatment of AOM in amoxicillin-allergic/intolerant patients who lack risk factors for DRSP(antibiotic use in the preceding month, age


6/17

tion is practically nonexistent and is probably more an ex-trapolation from the 10-day penicillin regimen used totreat streptococcal pharyngitis.43 A meta-analysis44 of 32randomized, controlled trials of the treatment of AOM inchildren with antibiotics of different durations suggestedthe efficacy of a five-day course of antibiotic. Dowell etal.43 recommended that uncomplicated AOM (i.e., absenceof perforated tympanic membrane, underlying medical

conditions, chronic/recurrent OM) may be treated with afive-day course of antibiotics in children two years of ageor older. Theoretical advantages to this shortened regimeninclude a reduction in selective pressure favoring resistantbacteria and greater adherence to therapy. We support theuse of the five-day oral antibiotic regimen (high-doseamoxicillin included) for the initial treatment of uncompli-cated AOM in children at least two years old. However, ifthe AOM fails to clear with the initial regimen, subsequentoral regimens should be 10 days long until resolution oc-curs (exception: azithromycin requires only 5 d of treat-ment). If the initial antibiotic chosen to treat AOM is stan-dard-dose amoxicillin or TMP/SMX and symptomatic im-provement (e.g., decreased ear pain, irritability, fever,sleeplessness, anorexia) together with a reduction in tym-panic membrane inflammation, redness, bulging, and/or

otorrhea have not occurred by day 3, resistant -lacta-maseproducingH. influenzae and/or pneumococcus shouldbe suspected. In this case, the second-step antibiotic cho-sen should cover both types of bacteria.

Disadvantages of the older macrolide/sulfonamide com-bination product erythromycin/sulfisoxazole include its re-quirement for dosing three to four times per day, potentialfor gastrointestinal upset, and unpleasant taste.38 As with

TMP/SMX, the incidence of in vitro resistance of pneumo-coccus to erythromycin has increased in recent years.41 Ad-vantages of erythromycin/sulfisoxazole for AOM includelow cost and coverage of comorbid infections caused byatypical bacteria (e.g.,Mycoplasma or Chlamydia pneu-moniae). Neither erythromycin/sulfisoxazole, clarithromy-cin, nor azithromycin were recommended by the CDCsexpert panel8 due to concerns of resistance to pneumococ-cus (~10% of isolates), substantial cross-resistance with -lactams, and the inability to overcome pneumococcal re-sistance by increasing the dosage. Other authors15 have ex-pressed concern about potentially inadequate extracellularMEF concentrations of erythromycin, clarithromycin, andazithromycin to treatH. influenzae and have not recom-mended them as second-line agents for treatment of refrac-tory AOM. Although these antibiotics concentrate intracel-

Treatment and Prevention of Otitis Media

The Annals of Pharmacotherapy 2000 December, Volume 34 1457www.theannals.com

Table 3. Recommended Antibiotics for Treating AOM in Childrena

Step 1 Step 2a Step 3

Adherence to Step 1 HD Amoxicillin

Treatment Good/Completeb Poor/Incompletec

Preferred HD amoxicillin (8090 mg/kg/d) TMP/SMX HD amoxicillin/clavulanatee ceftriaxonef or tympanocentesis

or

SD amoxicillin (40 mg/kg/d)d

Alternative(s) TMP/SMXd SD amoxicillin/clavulanate cefdinirg

cefdinirg cefpodoxime proxetilh

cefixime cefprozili

cefpodoxime proxetilh cefuroxime axetilh

ceftibuten

cefuroxime axetilh

azithromycinj

clarithromycinj

ery-sulfh,j

AOM = acute otitis media; DRSP = drug-resistant Streptococcus pneumoniae; ery-sulf = erythromycinsulfisoxazole; HD = high dose; SD = standard

dose; TMP/SMX = trimethoprim/sulfamethoxazole.aSteps 2 and 3 indicated for clinical treatment failure after three days of antibiotic from prior step. Clinical treatment failure defined as lack of clinical

improvement in signs and symptoms (e.g., ear pain, irritability, fever, sleeplessness, anorexia) together with tympanic membrane findings of inflam-

mation, redness, bulging, or otorrhea.bCovers only -lactamasepositiveHaemophilus influenzae.cCovers -lactamasepositiveH. influenzaeand DRSP.dAcceptable if low risk of DRSP (see text); failure of this antibiotic as the initial treatment of AOM warrants treatment with a second-step antibiotic

that covers both -lactamasepositive H. influenzaeand DRSP.ePresently requires patient/caregiver adherence to the concomitant administration of amoxicillin/clavulanate (45/6.4 mg/kg/d) and amoxicillin to boost

the cumulative dose to 8090 mg/kg/d.fThree-dose regimen (50 mg/kg im once daily for 3 d).gNot clinically investigated for refractory AOM, although it possesses in vitro activity similar to that of cefpodoxime proxetil.hTaste may jeopardize compliance.38

iAlthough questionable in vitro activity against -lactamaseproducing H. influenzae, no difference in clinical efficacy was seen in one study with AOM

patients harboring -lactamasepositive and negative H. influenzae.39

jReserve for possible comorbid atypical bacterial infections (prescribe ery-sulf, clarithromycin, or azithromycin) or concerns of caregiver/patient adherence

(prescribe azithromycin).


7/17

lularly, the inflammatory cells in which they concentratemay be delivered in high amounts to the site of infection/inflammation (e.g., the inner ear).45

A recommendation has been made that macrolide/aza-lide antibiotics be considered for AOM refractory toamoxicillin.45 Clarithromycin and azithromycin have mod-erate patient acquisition cost; azithromycin is especiallycompetitively priced for children weighing 20 kg (Table

2).35 Azithromycin has other advantages over erythromy-cin/sulfisoxazole and clarithromycin, that is, an extendedhalf-life (3060 h), permitting once-daily dosing for fivedays, relative lack of association with significant drug in-teractions, and better taste.38 Due to concern for inducingresistantH. influenzae and, especially, pneumococcal iso-lates from macrolide/azalide antibiotic overuse, we recom-mend that erythromycin/sulfisoxazole, clarithromycin, andazithromycin be reserved for situations when, despite goodor complete adherence to high-dose amoxicillin, treatmentfailure has occurred, and either coverage of atypical bacte-ria for a comorbid infection is necessary or concerns of ad-herence by the patient/caregiver dictate that oral therapyfor a short duration is warranted (azithromycin is recom-mended in this instance).

Loracarbef, cefaclor, and cefprozil have excellent tastesand may be administered, conveniently, twice daily.38

However, these antibiotics have acquisition costs in themoderate (cefaclor) to high range (Table 2).35 Among thethree antibiotics, cefprozil appears to have the most in vitroactivity against DRSP, with one noncomparative, open-label trial suggesting clinical effectiveness for childrenwith persistent and recurrent AOM. Pichichero et al.39

treated AOM in 262 children aged six months to 12 yearswith cefprozil 30 mg/kg/d in two divided doses for 10

days. Eighteen percent of the children had persistent AOM(i.e., antibiotic used 7 d before study entry); 37% had re-current AOM (i.e., 4 episodes/12 mo or 3 episodes/6 mo immediately preceding study entry). All children un-derwent tympanocentesis and had pretreatment cultures ofthe MEF performed. Subsequently, the clinical signs andsymptoms of AOM were assessed by one telephone con-tact (days 3 6) and during two return visits (days 1115and 28 40).

Single and multiple bacterial pathogens were identifiedin 57% and 11% of the MEF samples, respectively. Satis-factory clinical responses in the children occurred in 13 of14 withM. catarrhalis (93%), 56 of 75 withH. influenzae

(75%), and 70 of 93 with S. pneumoniae (75%). There wasno difference in the clinical success of patients harboring-lactamasepositive (30/42, 71%) and -lactamasenega-tive (26/33, 79%)H. influenzae. The clinical response ratesfor patients infected with penicillin-susceptible (MIC0.06 g/mL; 39/50, 78%), intermediate-resistant (MIC0.11 g/mL; 11/12, 92%), and -resistant (MIC 2 g/mL;21/31, 68%) S. pneumoniae were similar to those reportedfrom France by Gehanno et al.37 who used cefuroxime ax-etil to treat DRSP AOM. Both studies failed to include acomparison/control group and repeat tympanocentesis af-ter completion of therapy.

A major concern when using loracarbef, cefaclor, andcefprozil for AOM refractory to high-dose amoxicillin istheir in vitro coverage ofH. influenzae. A recent US sur-veillance study46 of the activity of oral antibiotics againstpneumococcus andH. influenzae used pharmacodynami-cally derived breakpoints to determine susceptibilities. Forthe -lactams tested, the breakpoints for susceptibilitieswere based on drug concentrations in serum that main-

tained serum concentrations greater than the bacteriasMIC for 4050% of the dosing interval. It has not been de-termined whether this new susceptibility testing methodshould someday replace the method presently used by theNCCLS.9,10 Out of 1676 clinical isolates of untypableH.influenzae (41.6% of the isolates produced -lactamase),the percentages of isolates reported susceptible to cefaclor,loracarbef, and cefprozil were 2%, 9%, and 14%, respec-tively, compared with 57%, 78%, 98%, and 100% foramoxicillin, cefuroxime, amoxicillin/clavulanate, and ce-fixime, respectively. Among 1476 strains of S. pneumoni-ae, over half (50.4%) were resistant, with roughly two-thirds of these (64.5%) displaying high-level resistance.The overall susceptibilities of all S. pneumoniae isolateswere 11% (loracarbef), 22% (cefaclor), 52% (cefixime),63% (cefprozil and cefuroxime), 69% (azithromycin andclarithromycin), and 94% (amoxicillin and amoxicillin/clavulanate).

This study46 highlights a disturbing trend the increasedresistance of key pathogens of AOM to antibiotics exten-sively used in the past for treating this infection. At pre-sent, the use of cefaclor and loracarbef for AOM refractoryto high-dose amoxicillin seems difficult to support, unlesssensitive bacteria have been identified. Cefprozil might beused in refractory patients whose adherence to high-dose

amoxicillin was poor or incomplete, since the clinical re-sponse rate for DRSP appears similar to that of cefuroximeaxetil.37,39 In addition, despite questionable in vitro suscep-tibility, the clinical response rate for -lactamaseproduc-ingH. influenzae to cefprozil has not been proven to differfrom -lactamasenegativeH. influenzae.39,46

The CDCs expert panel report8 has been criticized fornot giving greater consideration to advocating a second-step approach to AOM initially refractory to amoxicillin,which concentrates on -lactamaseproducing bacteriaalone.28 Use of this approach assumes that good adherenceto an adequate regimen of amoxicillin has occurred as theinitial AOM step and, therefore, -lactamaseproducingH.

influenzae orM. catarrhalis are the most likely bacteria tohave persisted. TMP/SMX seems the most cost-effectivechoice with this approach unless palatability and/or ad-verse effects preclude its use, or the community in whichthe patient lives has a high incidence of resistance ofH. in-

fluenzae to TMP/SMX. In vitro resistance rates to TMP/SMX of only 9.0% for H. influenzae and 6.5% for M.catarrhalis have been reported.47,48 Oral antibiotics consid-ered as acceptable alternatives to TMP/SMX, albeit withgreater acquisition costs, include standard-dose amoxi-cillin/clavulanate (45 mg/kg/d), cefuroxime, azithromycin,clarithromycin, erythromycin/sulfisoxazole, cefdinir, cef-



8/17

ixime, cefpodoxime, and ceftibuten (Table 2).35 The lastfour antibiotics (third-generation cephalosporins) may begiven once daily, have high acquisition costs (Table 2),and, with one exception (cefpodoxime), good taste.35,38

Cefpodoxime and cefdinir have similar in vitro antibacteri-al activity to organisms including penicillin-susceptibleand intermediate-resistant S. pneumoniae, although nei-ther antibiotic has been extensively evaluated in children

with refractory AOM due to DRSP.49 Cefixime and cefti-buten are less active against pneumococci, especially DRSP,than cefpodoxime and cefdinir.

THIRD-STEP: CEFTRIAXONE AND TYMPANOCENTESIS

Presently, ceftriaxone is FDA-approved as only a single-dose regimen for the treatment of AOM. Studies34,50-53 oftreatment with single-dose ceftriaxone are limited to pa-tients with new AOM. The CDC panel of experts8 believedthat this duration may not be sufficient to treat refractoryAOM due to PRSP and advocated a three-day regimen ofceftriaxone (50 mg/kg/dose im once daily for 3 d). Three

studies21,54,55 have investigated this regimen in patients whoinitially failed oral antibiotic therapy; however, compari-son with the single-dose regimen occurred in only one ofthese studies.55

Leibovitz et al.21 prospectively studied 92 pediatric pa-tients with AOM who did not respond to treatment with ei-ther amoxicillin, amoxicillin/clavulanate, or cefaclor. MEFwas collected prior to beginning the three-day ceftriaxoneregimen and again on days 4 to 10. Among bacteria iso-lates recovered from the MEF, eradication occurred in 54(100%)H. influenzae, 13 (100%) penicillin-susceptiblepneumococcus, 28 of 34 (82%) penicillinintermediate-re-

sistant pneumococcus (MIC 0.11.0 g/mL), one of two(50%)M. catarrhalis, and two (100%) S. pyogenes withtreatment. Although all isolates of pneumococcus weresusceptible to ceftriaxone (MIC 48 h during the 14 d preceding tympanocentesis) toreceive one day (group A, n = 29) or three days (group B,n = 31) of intramuscular ceftriaxone 50 mg/kg/d. Amoxi-cillin and amoxicillin/clavulanate were the main antibioticsadministered before randomization (45% and 21% forgroup A; 32% and 35% for group B, respectively). Bacte-ria recovered included pneumococcus (n = 37),H. influen-

zae (n = 36), andBranhamella catarrhalis (n = 2). Pneu-mococcal isolates in group A were intermediately resistantto penicillin (MIC 0.11.0 g/mL) and ceftriaxone (MIC>0.5, but 1.0 g/mL) in 13 of 17 (76%) and 2 of 17(12%) cases, respectively; two of 17 isolates (12%) werehighly resistant (MIC >1.0 g/mL) to penicillin. In groupB, pneumococcal isolates were resistant to penicillin andceftriaxone (no MIC data specified) in 17 of 20 (85%) andone of 20 (5%) cases, respectively. Bacteriologic eradica-tion occurred in allH. influenzae, B. catarrhalis, andpenicillin-susceptible pneumococcal cases. Bacteriologicfailure (positive culture on days 45) was seen more fre-quently in group A, the single-day ceftriaxone regimen(7/13, 54%), than in group B, the three-day (1/17, 6%) cef-triaxone regimen (p = 0.009). Both group A and B patientswere followed until day 30 2 with no apparent differenceevident in clinical (8 and 7 cases) and bacteriologic (6 and

4 cases, all with new pathogens) relapses, respectively. Inconclusion, the three-day regimen was significantly supe-rior to the one-day regimen in eradicating DRSP from theMEF of patients with nonresponsive AOM.

The CDCs expert panel8 suggested that ceftriaxonemight be used as second-line treatment for AOM initiallyrefractory to high-dose amoxicillin. We believe that ceftri-axone regimens are probably better reserved as third-linetreatment when both first- and second-line oral antibioticshave failed unless the risk of nonadherence to a five- to10-day course of oral antibiotic appears very likely.Widespread use of ceftriaxone for mild infections such asAOM has the potential to select for resistant isolates,

which could compromise ceftriaxones utility for moderateto severe infections. The three-day intramuscular ceftriax-one regimen has been shown to have superior bacteriolog-ic eradication compared with the single-dose regimen.55

Ceftriaxone regimens that employ less-frequent dosing(e.g., every other day or every third day) or cumulativedosages (i.e.,


9/17

axone MEF concentration (35 g/mL) from a single 50-mg/kg intramuscular dose given to children with chronicmiddle-ear effusion occurred 24 hours later and was35580 times higher than the MIC90 for the three majorbacteria causing AOM, including PRSP. The estimatedhalf-life in the MEF was 25 hours, and the T > MIC 90 var-ied between 100 and >200 hours depending on the bacte-ria. Therefore, it might be possible to achieve the same lev-

el of bacterial eradication with fewer doses administeredover a more prolonged period. The effect on adherence ofceftriaxone regimens employing fewer doses administeredless frequently also needs to be investigated. Until addi-tional studies are done, the three-day ceftriaxone regimenshould generally be considered the preferred empiric third-line treatment if tympanocentesis is not done.

The CDCs expert panel8 suggested a diagnostic tympa-nocentesis be done when treatment failures occur in pa-tients who have recently received multiple courses of an-tibiotics. Tympanocentesis is not a new or unique proce-dure.57 It is our opinion that, presently, few primary carephysicians have been adequately trained to perform tym-panocentesis. This lack of skilled training is probably theresult of decades when the standard practice to counteractmost AOM failures has been to treat empirically with adifferent antibiotic rather than perform tympanocentesis.Even manufacturers of antibiotics for AOM have voiceddifficulties in finding physicians willing to perform tympa-nocentesis on children participating in studies.58 Undoubt-edly, some physicians will remain reluctant to view tympa-nocentesis as necessary. Physician concerns regardingtympanocentesis may include the time and costs requiredto perform it and the attendant bacteriologic studies, care-giver/patient acceptance, liability in the event of a bad out-

come, and lost clientele should they choose to refer pa-tients for tympanocentesis. Despite these concerns, we rec-ommend tympanocentesis as a viable third-line option andencourage physicians to improve their skills in this area.

In summary, we recommend high-dose amoxicillin forinitial empiric treatment of AOM in patients at risk ofDRSP (Table 3).38,39 If patient risk factors for DRSP appearlow, standard-dose amoxicillin or TMP/SMX are also ap-propriate for initial therapy, although symptomatic failurewith either of these two regimens should warrant coverageof both -lactamasepositiveH. influenzae and DRSP withthe next antibiotic prescribed. If failure occurs after threedays of good or complete adherence to high-dose amoxi-

cillin, -lactamasepositiveH. influenzae seems the mostlikely organism, and we recommend TMP/SMX for treat-ment. If there is a reason to avoid TMP/SMX in this situa-tion, appropriate alternatives include standard-dose amoxi-cillin/clavulanate, oral third-generation cephalosporins, ce-furoxime axetil, erythromycin/sulfisoxazole, azithromycin,or clarithromycin. Poor or incomplete adherence to andfailure of high-dose amoxicillin as a first-step antibioticshould prompt coverage for both -lactamasepositiveH.influenzae and DRSP. We prefer high-dose amoxicillin/clavulanate, provided the patient/caregiver can adhere tothe administration of two antibiotics at the same time. If

adherence seems to be a potential problem, cefdinir, cefpo-doxime proxetil, cefprozil, or cefuroxime axetil can beconsidered. We do not recommend loracarbef or cefaclorfor empiric treatment of AOM due to the array of antibiot-ic alternatives available that have superior in vitro activityand pharmacodynamic profiles. The three-day intramuscu-lar ceftriaxone regimen is the preferred empiric third-stepchoice. However, tympanocentesis-directed antibiotic ther-

apy is also appropriate at this point, and we advocate itsuse although we suspect that, at present, few physiciansfeel comfortable performing it. Given the present risks anddifficulties associated with treating AOM, renewed effortsat prevention appear warranted.

Prevention of Otitis Media

ANTIBIOTICS

rAOM has been defined as three or more distinctepisodes of AOM in a six-month period, or four occur-rences within a single year.59 Prevention of rAOM is unar-

guably superior to treatment after the fact, particularlywhen the possible sequelae of hearing loss and potentiallong-range learning disabilities are considered.60,61 Contin-uous antimicrobial prophylaxis for rAOM has generallybeen considered efficacious. The most commonly recom-mended regimens have included oral amoxicillin 20 mg/kg/dor sulfisoxazole 3575 mg/kg/d given either once daily atbedtime or in two divided doses.5,61-63

Several reports64-67 assessing prophylactic antibiotic effi-cacy in children with rAOM have documented a two- tothreefold reduction in AOM recurrences compared withplacebo. One study68 reported no statistical difference inrecurrences between treatment and placebo groups, despiteuse of amoxicillin 20 mg/kg/d. A 1993 meta-analysis ofnine studies69 found an overall reduction of 0.11 AOMepisodes per patient per month (95% CI 0.03 to 0.19) at-tributable to prophylactic antibiotic use. Considering theouter ranges of the 95% CI reported, this suggests that theprophylactic use of antibiotics could, on average, reduceyearly occurrences of AOM by approximately 0.42.3episodes per child.

These unimpressive results must be considered in lightof the documented emergence of DRSP. Baquero et al.70

reported a clear and direct correlation between annualaminopenicillin use and the incidence of S. pneumoniae re-

sistance. A recent Canadian study71

noted an increasingprevalence of pneumococci with reduced susceptibility tofluoroquinolones and postulated that this increase has oc-curred secondary to selective pressure from increased useof this antibiotic class. The use of prophylactic antibioticshas been documented as a risk factor for nasopharyngealcarriage of DRSP (p < 0.001).72 A meta-analysis73 reportedthat antibiotic use increases the risk of nasopharyngeal car-riage of DRSP two- to fivefold. The importance of increas-ing pneumococcal resistance patterns is highlighted by thefact that penicillin-resistant species are, more often thannot, resistant to multiple antibiotics and -lactams (includ-



10/17

ing cephalosporins) as well as non--lactam agents (e.g.,macrolides, TMP/SMX).74

The importance of judicious antibiotic use is empha-sized by the French study conducted by Gehanno et al.30

These authors evaluated the bacteriologic epidemiology oftreatment failures for AOM in infants. While the percent-age distribution of causative organisms isolated from MEFremained similar to that seen overall in AOM (40% S.

pneumoniae, 36%H. influenzae, 5%M. catarrhalis), morethan 75% of the S. pneumoniae isolates had reduced sus-ceptibility to penicillin. In contrast, estimates of penicillinresistance in S. pneumoniae cultures obtained from USchildren are approximately 31%.74

Compelling evidence of the dangers of prophylactic an-tibiotic overuse has been reported recently.75 In this study,nasopharyngeal cultures were obtained from children tak-ing daily or every-other-day oral TMP/SMX prophylaxisfor at least six weeks and were compared with those fromchildren taking no antibiotics. While the overall S. pneu-moniae colonization rate was similar in the two groups, in-

creased resistance to penicillin and multiple antibiotics wasidentified more frequently in the prophylaxis group (82%vs. 7% for penicillin resistance, p value NR; 82% vs. 0%for multiple-drug resistance, p = 0.00001, respectively).

Overuse of antibiotics for AOM, rAOM, and OM witheffusion is postulated to act as a natural selective pressuretoward resistance.42,74 The use of strict criteria for patientselection has been advocated before prophylactic antibi-otics are administered.30,42

Given that >90% of AOM cases are preceded by a res-piratory tract infection, intermittent dosing of prophylacticantibiotics seems rational.76 The evidence supporting inter-mittent versus continuous prophylaxis is equivocal, how-

ever. One study77 comparing oral administration of azithro-mycin 5 or 10 mg/kg/wk with amoxicillin 20 mg/kg/d overa six-month period reported that the recurrence rate withazithromycin 10 mg/kg/wk was significantly lower (p 38 C with signs of otalgia, age less than twoyears, and prior antibiotic therapy with erythromycin/sul-fisoxazole have been documented as independently predic-tive risk factors for PRSP.30

SURGICAL OPTIONS

Myringotomy and tympanostomy tube placement havebeen proposed as reasonable alternatives to repeated cours-es of antibiotics.88 Data comparing the efficacy of myrin-gotomy or tympanostomy tube placement with chemopro-phylaxis are sparse and equivocal, however; myringotomyhas no demonstrable influence on the course of AOM.89-91

An increased incidence of retraction or atrophy of the tym-panic membrane has been noted in patients who receivedmyringotomy when compared with placebo controls.92

Gonzalez et al.93 compared tympanostomy tube insertionwith sulfisoxazole 5001000 mg given orally twice daily

or placebo in 65 children with rAOM. These authors re-ported that tympanostomy tube placement significantly re-duced treatment failures when compared with placebo(23% vs. 60%, respectively; p < 0.02). The sulfisoxazoleprophylaxis group reported 38% treatment failures, but thiswas not significantly different from either the tympanosto-my tube group or the placebo group. It is important to notethat these results were confounded by inclusion of 18 pa-tients (27.7%) who had OM with effusion. In a larger tri-al,94 tympanostomy tube insertion was compared withamoxicillin 20 mg/kg/d or placebo in 264 children agedseven to 35 months. This study used stricter rAOM inclu-sion criteria, excluding all patients with OM with effusion.

Both tube placement and chemoprophylaxis proved supe-rior to placebo for the prevention of AOM recurrences (p

otite- tratamento - treatment and prevention om

Documents