Thursday, March 8, 2007
Neurology has a randomized, placebo-controlled, double-blind trial of smoked marijuana/cannabis for painful HIV-associated peripheral neuropathy. The study involved 50 people with HIV who had at least 3/10 pain from their neuropathy. They were randomized to a standardized marijuana cigarette or a placebo joint (marijuana with the THC extracted out): for those of you who follow these things the cigarettes were an average of 0.9oz and each contained 3.56% delta-9-tetrahydrocannabinol (delta-9-THC)--I think they were provided by the US government. Subjects smoked one cigarette three times daily for 5 days in an inpatient research setting; they were subject to the usual pain research tortures ("long thermal stimulation" etc.). Thankfully though the primary outcome was straightforward: subjects' rating of average pain for the preceding 24h on a 0-100mm visual analog scale. Intention to treat analysis was not used, but there were equal drop-outs in each group after randomization (2 each; 50 people completed the study).
Average age was ~50, most were male, ~35% were African-American, baseline pain scores were ~5/10, and almost all subjects were current cannabis users (they were requested to abstain prior to the research). The study's primary outcome favored marijuana: 52% of marijuana subjects vs. 24% of placebo subjects had at least a 30% reduction in mean pain scores (median reduction was 34% in the marijuana group vs. 17% in the placebo group). The number-needed-to-treat for a 30% pain reduction was 3.6. If 30% doesn't sound like much to you this is actually a really good, positive finding in the world of pharmacologic treatments for neuropathic pain: most trials that show any benefit show in at roughly this magnitude. Cannabis looked good in the experimental pain model part of the trial as well.
So this is some good, real-patient evidence furthering the finding that cannabis/THC is an effective analgesic. The trial was short unfortunately, but in order to control everything this was perhaps necessary (the researchers actually weighed the cigarettes after they were smoked to, I guess, ensure that the subject actually got the THC). The big problem with this is its generalizability. These patients were regular marijuana smokers--they were self-selected to 1) be able to tolerate the side effects, and 2) to perceive benefit--of whatever nature--from marijuana's CNS effects. Even with prescription forms of THC, the CNS/psychotomimetic side effects are often dose-limiting, particularly for older, frail patients. The next study should be using similar quantities of prescription THC for people over 50 years old with painful diabetic peripheral neuropathy who don't smoke marijuana at baseline--good results from that might be practice changing.