744   Short-Term Safety of Cannabinoids in HIV Patients.

D. Abrams*, R. Leiser, S. Shade, F. Aweeka, B. Bredt, T. Elbeik, J. Hilton, and M. Schambelan.
Univ. of California, San Francisco.

Background:Widespread use of smoked marijuana prompted this study of its safety in patients with HIV. Cannabinoids could impact on HIV RNA levels by interacting with either the immune system or the cytochrome P450 enzyme system, thereby altering protease inhibitor metabolism.

Methods:A prospective randomized placebo-controlled 21-day in-patient trial was conducted in the General Clinical Research Center. HIV⁺patients with a stable viral load on a stable antiviral regimen that contained either nelfinavir (NFV) or indinavir (IDV) were randomized to one of three treatment interventions—3.95% tetrahydrocannabinol marijuana cigarette (M), dronabinol 2.5 mg (D)or placebo (P)—each tid before meals. Change in HIV RNA at 21 days was the primary endpoint, with CD4⁺cell counts, plasma levels of protease inhibitors and testosterone evaluated as secondary safety parameters. Data for day 0 through day 21 were analyzed using repeated measures.

Results:Of 67 subjects enrolled, 21 were randomized to M, 25 to D and 21 to P. The majority were male (96%), people of color (52%) and over 40 years of age (67%). The HIV RNA level was <50 copies/mL for 37 (55%) at baseline, and the median CD4⁺cell count was 300/mm³. Thirty subjects were on IDV 800 mg q8h; 37 were on NFV 750 mg tid. There were no statistically significant differences among the treatment groups by any of these baseline characteristics. Log₁₀HIV RNA levels declined in the M (-0.24 log₁₀; p < 0.02) and D (-0.20 log₁₀; p < 0.05) arms compared to those on P. Change in CD4⁺T lymphocyte counts did not differ in the M (+21.4 cells/ mm³; p = 0.47) and D (+21.1 cells/mm³; p = 0.46) arms compared to those on P. Although there were no statistically significant difference across arms, there was a 17% reduction in NFV AUC₈(p < 0.04) and a 24% reduction in the IDV AUC₈(p < 0.07) in the M group at day 14 compared to baseline (using paired t-tests). A non-significant decline in serum testosterone levels compared to P was observed in both the M (-73.8 ng/dl) and D (-40.9 ng/dl) groups. There were 3 grade 2-3 on-study adverse experiences in the M and D subjects, with none in the P group.

Conclusion:Smoked marijuana has no significant short-term detrimental effects on standard safety parameters in HIV⁺patients on a stable PI-containing regimen.