Penicillin vs Cephalosporin for Strep Throat—Which Is Better?

Abstract & Commentary

Synopsis: Bacteriologic and clinical failures in adults with Streptococcal tonsillopharyngitis are twice as likely with oral penicillin therapy as with cephalosporin treatment. But, what is the significance of this finding?

Source: Casey JR, et al. Meta-analysis of Cephalosporins vs Penicillin for Treatment of Group A Streptococcal Tonsillopharyngitis in Adults. Clin Infect Dis. 2004;38:1526-1534.

In the 28 years since G. V. Glass described the meta-analysis concept, countless meta-analyses have been published in the medical literature. In the present study, penicillin is compared with cephalosporins for the treatment of adults with group A ß-hemolytic streptococcal (GABHS) tonsillopharyngitis.

Casey and colleagues critically reviewed all randomized, controlled therapy trials performed in patients 12 or more years old with streptococcal sore throat. Treatment consisted of an orally administered antibiotic for 10 days. Bacteriologic cure (defined as absence of GABHS in a post-treatment throat culture) and clinical cure (defined as "resolution or improvement" of signs and symptoms) were tabulated. Multiple statistical manipulations analyzed the effects of 1) lack of double-blinding, 2) inferior methodologic quality, 3) uncertainty concerning compliance, 4) failure to account for pharyngeal streptococcal carriage, and several other study design weaknesses.

Of 66 randomized clinical trials that were reviewed, 57 were excluded because of such analytic problems as predominance of children over adults in the patient population, failure to separate adult data from those of children, poorly defined cure rates, and inadequate duration of antibiotic treatment (< 10 days). The remaining studies included treatment with loracarbef (3 studies), cefadroxil and cefdinir (2 studies each), and cefpodoxime and cefetamet (1 each). Penicillin V was the comparator penicillin compound.

Analysis of all 2113 patients in 9 studies demonstrated an overall or summary odds ratio (OR) for bacterial cure of 1.83, thereby favoring cephalosporin treatment (an OR greater than 1 denotes a higher cure rate with cephalosporin therapy). For clinical cure, the OR was 2.29, again favoring cephalosporin over penicillin. Results were similar when sensitivity analysis parsed trials grouped by double-blinding, exclusion of carriers, trials deemed to be of high quality, and other methodologic measures. Stratified analysis, in which cephalosporins were analyzed individually and as generation groups (first-generation, second-generation, and so on), demonstrated both bacterial and clinical cure rate superiority in most cases, with the exception of clinical cure rate equivalence with loracarbef.

Although Casey and colleagues did not explicitly provide absolute overall cure rates, calculations from their data indicate that, in the 9 trials that were analyzed, bacteriologic and clinical cure rates of 92.4% and 95.3%, respectively, were found in cephalosporin-treated adult patients, compared with rates of 86.9% and 89.9% in penicillin-treated patients. Casey et al stated that 19 patients would need to be treated with a cephalosporin to result in 1 additional bacteriologic cure, were penicillin used instead. A similar benefit ratio results if one calculates a clinical cure estimate.

Comment by Jerry D. Smilack, MD

Readers of this article are left with the impression that an oral cephalosporin is the drug of choice for treatment of adults with GABHS tonsillopharyngitis. Before taking this statement at face value, Infectious Disease Alert readers would be well advised to peruse Dr. Alan Bisno’s editorial commentary (Clin Infect Dis. 2004;38:1535-1537) that accompanies Casey et al’s meta-analysis. Bisno takes issue with the latter’s conclusions by citing several important caveats. First, some of the reports included in the meta-analysis had "appreciable deficiencies" (such as issues of compliance, timing of follow-up cultures, and inclusion of GABHS carriers), flaws that "are not likely to be overcome by even the most rigorous analytic statistical techniques." Second, because GABHS tonsillopharyngitis is usually a self-limited disease for which antibiotic therapy may shorten the course by only 1 or 2 days, assessing the importance of small differences in clinical cure rates is problematic. Third, if penicillin therapy leads to significant failure rates, why have we not seen reports of increased disease-associated or post-streptococcal complications? Bisno concludes by asking whether a small difference in bacteriologic and clinical cure rates—even if such a difference truly exists—really makes a clinically significant difference. He answers the question by saying that it doesn’t, and that penicillin remains the drug of choice for treatment of adults with streptococcal tonsillopharyngitis.

Jerry D. Smilack, MD Infectious Disease Consultant Mayo Clinic Scottsdale Scottsdale, AZ, is Associate Editor for Infectious Disease Alert.