To improve the treatment of the acquired immunodeficiency syndrome, clinical trials are under way in which the nucleoside analogs 2',3'-dideoxycytidine (ddC) and 3'-azido-2'3'-dideoxythymidine (zidovudine, AZT) are being tested simultaneously, intermittently, or in an alternating regimen [1, 2]. Combination of these two antiretroviral agents might allow the use of lower or less frequent doses, thus minimizing the individual or selective drug toxicities or both [1]. None of the preliminary reports published have identified acute pancreatitis as a toxic effect of ddC [3]. Preliminary reports of clinical trials using another nucleoside analog, 2',3'-dideoxyinosine (ddI), mention sporadic cases of acute pancreatitis as a side effect [3].

In August 1991, a 45-year-old woman was admitted to our emergency room for evaluation of nausea, vomiting, and severe abdominal pain lasting less than 72 hours. The clinical impression was acute pancreatitis, possibly drug-induced. The patient denied drug or alcohol abuse. Past medical history was remarkable for a positive human immunodeficiency virus test and Pneumocystis carinii pneumonia. She began taking zidovudine on 22 September 1990 and one tablet of trimethoprim-sulfamethoxazole daily (800 mg and 160 mg per tablet, respectively) for P. carinii pneumonia prophylaxis in November 1990. She tolerated both drugs well for 1 year. Approximately 6 weeks before admission, the patient was enrolled in a triple-arm, double-blind, national clinical trial study (started on 7 July 1991) to evaluate the use of ddC and zidovudine. When the patient's protocol code was broken shortly before she died, she was found to be receiving both drugs.

Autopsy findings showed extensive fat and parenchymal hemorrhagic necrosis of the pancreas. Light microscopy showed coagulation necrosis, hemorrhage, and accumulation of neutrophils. These changes are characteristic of acute pancreatitis.

We could not confirm a direct causal relation between ddC and acute pancreatitis in this case; however, after microbiologic studies and the use of light and electron microscopy, we ruled out other possible causes of pancreatitis [4]. A good temporal relation exists between the administration of ddC and the appearance of symptoms; therefore, we accepted the association. We speculate that ddC or its metabolites, or both, interfere with the pancreatic process of the exocytosis of zymogen granules, resulting in granular accumulation and subsequent intracellular lysosomal fusion [5]. Future studies will better define the use of the nucleoside analogs; in the meantime, physicians need to be aware of the possible association between ddC and pancreatitis.