August 4, 2000
After years of bureaucratic nightmare, a research team at the University of California has completed a safety study of smoked marijuana vs. oral dronabinol (Marinol(R), an FDA-approved oral drug with the same main active ingredient as marijuana) vs. placebo in HIV/AIDS patients being treated with an antiviral regimen including a protease inhibitor (either indinavir or nelfinavir). The main purpose of the study was to find out whether cannabinoids, the active ingredient in marijuana, could raise the viral load by interacting either with the antiretrovirals, or with the immune system. No such problem was found.(1)
All three study arms (marijuana, dronabinol, and placebo) gained weight -- not surprisingly as the study was conducted while volunteers remained 24 hours a day in a hospital research facility, and were probably less active than they would have been outside. But those in the marijuana and dronabinol arms gained more (3.51 kg and 3.18 kg, vs. 1.30 kg for the placebo arm). Again this is not surprising, as marijuana is often used as an appetite stimulant.
This study randomly assigned 67 patients to marijuana, dronabinol, or an oral placebo (the smoked marijuana arm was not blinded, probably because of the difficulty of providing a smoked placebo); interestingly, half of the placebo recipients thought they were getting dronabinol, and about a quarter of the dronabinol recipients thought they were getting placebo. Two volunteers in the marijuana arm and one in the dronabinol arm reported neuropsychiatric symptoms; two of these three (one in each arm) chose to leave the study as a result. (All of the volunteers were previously experienced with marijuana -- and were required not to have used it for 30 days before entry to the study.)
Note: In the poster printed quickly for distribution at the Durban conference, two minus signs were mistakenly omitted in reporting viral loads; viral load went down slightly in both the marijuana and the dronabinol group, and up slightly with the placebo. However, none of these changes was large enough to be statistically significant.
This trial had to be designed to look at safety instead of efficacy, because the Federal government must give special permission for any marijuana study, and provide the drug as well. It has clearly been Federal policy to only allow studies that look for possible problems with marijuana, so that the study will either find something negative, or find nothing; studies testing for medical benefit have not yet been allowed, so that no information favorable to marijuana could come into existence. In this case, the standard body weight measurement used in safety studies found a benefit. (Weight gain would be regarded as a benefit here, as marijuana is often used medically by patients who need to eat more and gain weight, not by those who are trying to lose weight.)
This study was presented as a late-breaker poster at the Durban conference; the summary noted that "future trials should investigate the effectiveness of marijuana in: appetite stimulation/weight gain nausea [and] pain."
If such a study is allowed to be done, we would suggest adding a fourth arm -- to compare placebo, vs. marijuana, vs. dronabinol, vs. *choice* (of either marijuana, dronabinol, or neither, allowing switching whenever the volunteers wanted). Clearly some patients do better with smoked marijuana, and others with oral dronabinol; so we suspect that for some individuals there could be large differences in effectiveness in different directions, which would tend to balance each other out, and could result in little difference for the group average. If this is the case, then a patients' choice among the drugs could well be more safe and/or effective (for many individuals, and also for the group average) than assignment to either drug alone.
ISSN # 1052-4207
Copyright 2000 by John S. James. Permission granted for noncommercial reproduction, provided that our address and phone number are included if more than short quotations are used.
Back to the AIDS Treatment News August 4, 2000 contents page.