Single Dose of Azithromycin or Three‐Day Course of Ciprofloxacin as Therapy for Epidemic Dysentery in Kenya

Epidemic dysentery caused by Shigella dysenteriae type 1 in developing nations is a difficult disease to treat because of increasing drug resistance. An epidemic caused predominately by S. dysenteriae type 1 swept western Kenya in 1995–1996 [1]. At that time, a study of chemoprophylaxis for malaria was testing weekly azithromycin (1 g) or daily doxycycline as a means of preventing malaria [2]. Analysis of intercurrent cases of dysentery during this study suggested that azithromycin might be effective treatment of acute dysentery caused by Shigella. A simultaneous hospital study in a nearby district showed that the predominant organism involved was S. dysenteriae type 1 highly resistant to both trimethoprim-sulfamethoxazole and amoxicillin [3]. Since these two drugs are the only antibiotics readily available to most rural health clinics, an alternative regimen was urgently needed. Quinolones such as nalidixic acid and ciprofloxacin have been successfully used for the treatment of epidemic dysentery, but an epidemic during the Rwandan refugee crisis of 1994 was due to a nalidixic acid–resistant organism [4]. Data from Bangladesh suggested that since 1.5 g of azithromycin over 5 days was adequate to treat acute S. dysenteriae type 1 infection [5], it might be possible to use single-dose azithromycin treatment for epidemic dysentery. On the basis of these findings, a randomized, double-blinded study comparing 1 g of azithromycin and a 3-day regimen of ciprofloxacin (500 mg twice a day) was conducted in 1996–1997 in western Kenya. Any adult patient with the primary diagnosis of acute dysentery at St. Mary’s Hospital, Mumias, Kenya, was contacted at admission to see if the study entry criteria were met. These criteria were as follows: visible blood in a recently passed unformed stool, not receiving antibiotics likely to be effective against Shigella species (i.e., nalidixic acid or norfloxacin), if female not pregnant as confirmed by a negative urine test, able to take oral medication, no study drug allergy, and no known alternative cause of dysentery. All study entrants were informed of the nature of the research and gave their informed consent. At admission, history was obtained, physical examination was performed, and a pretreatment stool specimen was collected. Volunteers were then immediately randomized to receive either a single dose of azithromycin or placebo (four tablets) and ciprofloxacin or placebo (two capsules) (Pfizer Central Research, Groton, CT, supplied all medications and placebos). Administration of five additional doses of ciprofloxacin or placebo was continued over 3 days under observation in the hospital. The placebos were indistinguishable from the active drugs, and neither the patient nor the investigators were aware of the drug assignments. Clinical courses of the volunteers in the hospital were followed up daily, and when possible, an outpatient visit on the 10th study day was conducted. Antibiotic susceptibility testing was completed by using the disk diffusion method. End points studied included the number of days until resolution of dysentery, fever, and therapeutic support with intravenous or oral fluids and electrolytes. One hundred thirty adults volunteered for inclusion into the study and were randomized to receive treatment. At study completion, 113 (87%) were evaluable; the most common reason for study withdrawal was leaving the hospital before completion of the study drug regimen. Sixty-eight men and 45 women finished the study, of whom 57 received a single dose of azithromycin and 56 received 3 days of ciprofloxacin (table 1). Dysentery in all study patients cleared (median time of clearance, 2 days) during observation. There were no significant differences between the drug assignment groups in terms of age, sex, the number of days until resolution of dysentery, fever, or therapeutic support with fluids on either an intent-to-treat or study completion basis. S dysenteriae type 1 isolates were resistant to all commonly prescribed drugs: 99% were resistant to tetracycline; 98%, to ampicillin; 100%, to trimethoprim-sulfamethoxazole; 69%, to amoxicillin/clavulanic acid; and 98%, to erythromycin. The same isolates were susceptible to nalidixic acid (83%), norfloxacin (100%), and gentamicin (100%). All 23 S. dysenteriae type 1 isolates tested were found to be susceptible to azithromycin (MIC ¶ 4 mg/mL). Cultures of follow-up stool specimens did not yield any Shigella species, and no serious adverse effects of medication were noted. Azithromycin in a 1-g oral dose is a potentially useful regimen when faced with a public health emergency caused by a severe dysentery epidemic; this treatment may be particularly helpful when dealing with pediatric patients, as a relative contraindication to the use of fluoroquinolones for children does exist. In areas where older antimicrobials are no longer reliably effective, a directly observed single-dose therapy might allow for cost containment by reducing the number of treatment failures due to noncompliance. A shorter duration of therapy with a quinolone has been studied [6]. Given this brief The Kenya National Ethical Review Committee and the U.S. Army Surgeon General’s Human Subjects Research Review Division approved this study. This article is published with the permission of the Director of the Kenya Medical Research Institute. The views expressed in this article are those of the authors and do not reflect the official policy of the U.S. Army or the U.S. Department of Defense. * Study group members are listed after the text. Reprints or correspondence: Dr. G. Dennis Shanks, U.S. Army Medical Research Unit–Kenya, Box 401, Unit 64109, APO AE 09831-4109, USA (dshanks@africaonline.co.ke).