Nonclinical toxicology studies with zidovudine: genetic toxicity tests and carcinogenicity bioassays in mice and rats.

Zidovudine (ZDV), an antiviral drug active in the treatment of acquired immunodeficiency syndrome (recommended human dose, 100 mg every 4 hr while awake), was evaluated for mutagenic and carcinogenic potential in a battery of short-term in vitro and in vivo assays and in lifetime studies in mice and rats. In L5178Y mouse lymphoma cells (tk+/- locus), a weak positive result was obtained only at the highest concentrations tested (4000 to 5000 micrograms/ml) in the absence of metabolic activation. In the presence of metabolic activation, the drug was weakly mutagenic at concentrations of 1000 micrograms/ml and higher. Following 24 hr treatment in the absence of metabolic activation, ZDV was moderately mutagenic at concentrations up to 600 micrograms/ml; dose-related structural chromosomal alterations were seen at concentrations of 3 micrograms/ml and higher in cultured human lymphocytes. Such effects were not noted at the two lowest concentrations tested, 0.3 and 1 microgram/ml, and BALB/c-3T3 cells were transformed at concentrations of 0.5 microgram/ml and higher. No effects were seen in the Ames Salmonella plate incorporation and preincubation modification assays (possibly due to bacteriocidal activity of ZDV at low concentrations) at concentrations ranging from 0.01 to 10 micrograms/plate or in a single-dose intravenous bone marrow cytogenetic assay in CD rats. In multidose micronucleus studies, increases in micronucleated erythrocytes were seen in mice at doses of 100 to 1000 mg/kg/day. Similar results were seen in rats and mice after 4 or 7 days of dosing at 500 mg/kg/day. In carcinogenicity bioassays, adjusted doses of 20, 30, or 40 mg/kg/day and 80, 220, and 300 mg/kg/day were given to CD-1 mice and CD rats, respectively, for up to 22 months in mice and 24 months in rats. ZDV caused a macrocytic, normochromic anemia in both species. No evidence of carcinogenicity was seen in male mice or rats. In female mice, five malignant and two benign vaginal epithelial neoplasms occurred in animals given 40 mg/kg/day. A single benign vaginal epithelial tumor was seen in a mouse given 30 mg/kg/day. In rats, two malignant vaginal epithelial neoplasms were seen in animals given 300 mg/kg/day. In a 7-day study in mice, ZDV was shown to be devoid of estrogenic activity. In an oral pharmacokinetics study, the AUC was 17 and 140 micrograms/ml.hr in female mice and rats given 40 or 300 mg/kg of ZDV, respectively. In contrast, the average steady-state concentration in humans at the recommended daily dose is 0.62 microgram/ml. Twenty-four hour urine concentrations were 1245 and 4417 micrograms/ml in female mice and rats given 40 or 300 mg/kg of ZDV, respectively. These values were approximately 26- and 136-fold higher than the human urine concentration at the recommended daily dose. In a one- to three-day study with intravenously administered sodium fluoroscein in rats and mice, retrograde flow of urine into the vagina was demonstrated. In a subsequent lifetime carcinogenicity bioassay in mice in which ZDV was given intravaginally at concentrations of 5 or 20 mg ZDV/ml in saline, 13 vaginal squamous cell carcinomas were seen at the highest concentration tested. It was concluded that the vaginal tumors seen in the oral carcinogenicity studies were the result of chronic local exposure of the vaginal epithelium to high urine concentrations of ZDV.