LETTER

Top Letter References

A recent paper by Patel et al. (8) comparing the pharmacokinetics of clarithromycin and azithromycin in plasma, epithelial lining fluid (ELF), and alveolar macrophage (AM) cells concluded that absolute concentrations of clarithromycin were greater than those of azithromycin in these compartments at the end of a typical course of therapy. While these data are interesting and support the findings of Conte et al. (1), we feel that certain issues need clarification.

Patel et al. did not follow antibiotic concentrations past a 24-h period. Although it has been previously established that this sampling scheme adequately represents the pharmacokinetics of clarithromycin, it clearly does not do so for azithromycin (9). Had the authors included a 48-h bronchoalveolar lavage sample, it would have been evident, based on past investigators' data, that the concentrations in AMs and ELF of both clarithromycin and its 14-hydroxymetabolite would have been negligible at best (1). However, previous investigators have demonstrated that as a result of prolonged elimination from the tissue compartment, azithromycin concentrations in AMs remain constant beyond 120 h after a single 500-mg dose (1) and continue to be detectable beyond 21 days after the start of a standard 5-day dosage regimen (7).

Although Patel et al. were unable to consistently detect microbiologically active azithromycin concentrations in ELF, other investigators have established the existence of such concentrations up to 24 h after the final dose of a standard 5-day regimen consisting of 500 mg on day 1, followed by 250 mg for 4 days (5).

The authors also stated that serum azithromycin concentrations were below the typical MICs for certain bacteria such as Streptococcus pneumoniae and Haemophilus influenzae and that traditional pharmacodynamic variables predict unsatisfactory eradication rates; thus they concluded that the reasons for this antibiotic's success are still "hypothetical." However, they failed to mention previous studies that demonstrated that the addition of 50% human serum to an in vitro model decreases the MIC at which 90% of the isolates are inhibited (MIC₉₀) for these organisms two- to sixfold (2, 6). It has also been demonstrated by two of these same authors that azithromycin accumulates to a greater degree in inflammatory blister fluid (an infection model) than in noninflammatory blister fluid (an interstitial fluid model) and remains elevated for a prolonged period of time (3). Also, as neutrophils and macrophages accumulate azithromycin in concentrations at least 100-fold higher than those in the surrounding serum (4), the drug is able to be delivered to the site of infection via common, and long delineated, chemotactic mechanisms. These data better explain the clinical efficacy of azithromycin.

Finally, it is curious that these authors maintained that advanced-generation macrolide leukocyte delivery is currently a speculative mechanism. It is particularly puzzling considering it was these authors that convincingly stated in a previous publication that this mechanism is most likely the major source of azithromycin delivery to the site of infection (3).

We agree that both leukocyte and serum deliveries are integral to the concentration of clarithromycin and azithromycin at the infection site. However, it is the inherent differences in chemical structure and pharmacokinetic properties between these two agents that result in their widely divergent penetrations and retentions in tissue.