Reply to Genton and D'Acremont

To the Editor—Genton and D'Acremont [1] question the conclusions of our recent article comparing single-dose and 3-day azithromycin-based regimens with a 3-day levofloxacin regimen for the treatment of traveler's diarrhea in Thailand [2]. They comment on concerns relating to study design, the significance of the findings, the generalizability of the results, and broader concerns related to antibiotic treatment of traveler's diarrhea. They state that a placebo comparator should have been included. Antibiotic treatment has been repeatedly demonstrated to be effective in reducing the duration of illness by 1 or several days [3, 4]. A Cochrane review evaluated 12 placebo-controlled trials and concluded that treatment with antibiotics led to a statistically significant reduction in the duration of diarrhea for subjects with traveler's diarrhea [3] and was also beneficial in early treatment of campylobacteriosis (within 72 h of onset) [5]. Two separate Infectious Diseases Society of America guidelines committees have recommended short-course antibiotic use for treatment of moderate-to-severe traveler's diarrhea (Infectious Diseases Society of America—US Public Health Service rating, A-I) [6, 7]. Given the known prevalence of Campylobacter infection in Thailand, the increased severity and prolonged duration of campylobacteriosis [8], and the body of evidence for antibiotic efficacy, we do not believe that a placebo-controlled trial would have been appropriate.

The difference in treatment effect between azithromycin-based regimens and levofloxacin-based regimens is not a “small improvement in the cure rate at 72 h,” as stated by Genton and D'Acremont [1, p. 1520]. Single-dose azithromycin led to complete resolution of symptoms in 96% of individuals by 72 h, whereas 29% of the patients treated with levofloxacin remained ill. The comment that clinical outcomes are the same despite pathogen susceptibility does not agree with the findings of our study. Stratification by levofloxacin susceptibility among the Campylobacter-infected patients in the levofloxacin treatment group (shown in figure 2 of our article [2]) demonstrated a mean time to last unformed stool of 41.2 h for patients with susceptible isolates, compared with 76.4 h for patients with resistant isolates. A probable explanation for the discordant results between pathogen susceptibility and clinical outcome observed in other studies may be susceptibility cutoff determinations linked to achievable serum levels without regard to mucosal infection and fecal antibiotic levels.

The study population consisted primarily of young adult men without chronic illness, which is similar to the populations in studies involving students or Peace Corps workers. Application of these results to individuals at the extremes of age or with comorbid illnesses was not proposed. The disease severity is typical for an area in which Campylobacter infection is not the predominant cause of traveler's diarrhea. The pathogen distribution is consistent with previous reports involving military populations in Thailand [9, 10], as well as with high rates of Campylobacter infection in civilian traveler populations returning from this region with diarrhea [11]. As discussed by Dr. DuPont in the editorial [12] that accompanied our article [2], there is a paucity of recent studies of traveler's diarrhea in civilian visitors to Thailand. The progressive increase in fluoroquinolone resistance in Campylobacter species in Thailand has accelerated beyond that in other regions (the rate of Campylobacter infection has increased exceeding 85% since 1998), providing the rationale for our study [2]; however, increasing levels of resistance are being observed in many regions [13]. The concomitant use of doxycycline is acknowledged to be a complicating factor for the interpretation of results and is discussed in detail in our article [2].

On the basis of our study findings, we recommended the use of single-dose (1-g) azithromycin for empirical therapy for traveler's diarrhea acquired in Thailand, with appropriate restrictions for judicious application. Our recommendation of azithromycin as a reasonable first-line option was based on a synthesis of the clinical studies and not solely on our study results. Azithromycin has demonstrated efficacy in other studies of traveler's diarrhea [10, 14], including a study conducted in a region in which enterotoxigenic Escherichia coli is the predominant pathogen (which is more typical of traveler's diarrhea globally), as well as efficacy in the treatment of shigellosis [15]. DuPont [12] states that, compared with other drugs, azithromycin has the broadest activity against the common bacterial enteropathogens, although it is likely that azithromycin is most useful in the treatment of febrile dysentery [12]. Studies to further define effective regimens that assess the potential for adverse effects, unintended consequences (such as acquisition of drug resistance), and cost-effectiveness continue to be important.

Acknowledgments

Financial support. Pfizer Pharmaceuticals Clinical Research Division supplied study medications and placebo formulations for the study at no cost.

Potential conflicts of interest. D.R.T.: no conflicts.