Study: Smoking marijuana does not cause lung cancer

[Julie Gardener]

STUDY: SMOKING MARIJUANA DOES NOT CAUSE LUNG CANCER​

July 2005

Marijuana smoking -"even heavy longterm use"- does not cause cancer of the lung, upper airwaves, or esophagus, Donald Tashkin reported at this year's meeting of the International Cannabinoid Research Society. Coming from Tashkin, this conclusion had extra significance for the assembled drug-company and university-based scientists ( most of whom get funding from the U.S. National Institute on Drug Abuse ). Over the years, Tashkin's lab at UCLA has produced irrefutable evidence of the damage that marijuana smoke wreaks on bronchial tissue. With NIDA's support, Tashkin and colleagues have identified the potent carcinogens in marijuana smoke, biopsied and made photomicrographs of pre-malignant cells, and studied the molecular changes occurring within them. It is Tashkin's research that the Drug Czar's office cites in ads linking marijuana to lung cancer. Tashkin himself has long believed in a causal relationship, despite a study in which Stephen Sidney examined the files of 64,000 Kaiser patients and found that marijuana users didn't develop lung cancer at a higher rate or die earlier than non-users. Of five smaller studies on the question, only two -involving a total of about 300 patients- concluded that marijuana smoking causes lung cancer. Tashkin decided to settle the question by conducting a large, prospectively designed, population-based, case-controlled study. "Our major hypothesis," he told the ICRS, "was that heavy, longterm use of marijuana will increase the risk of lung and upper-airwaves cancers."

The Los Angeles County Cancer Surveillance program provided Tashkin's team with the names of 1,209 L.A. residents aged 59 or younger with cancer ( 611 lung, 403 oral/pharyngeal, 90 laryngeal, 108 esophageal ). Interviewers collected extensive lifetime histories of marijuana, tobacco, alcohol and other drug use, and data on diet, occupational exposures, family history of cancer, and various "socio-demographic factors." Exposure to marijuana was measured in joint years ( joints per day x 365 ). Controls were found based on age, gender and neighborhood. Among them, 46% had never used marijuana, 31% had used less than one joint year, 12% had used 10-30 j-yrs, 2% had used 30-60 j-yrs, and 3% had used for more than 60 j-yrs. Tashkin controlled for tobacco use and calculated the relative risk of marijuana use resulting in lung and upper airwaves cancers. All the odds ratios turned out to be less than one ( one being equal to the control group's chances )! Compared with subjects who had used less than one joint year, the estimated odds ratios for lung cancer were .78; for 1-10 j-yrs, .74; for 10-30 j-yrs, .85 for 30-60 j-yrs; and 0.81 for more than 60 j-yrs. The estimated odds ratios for oral/pharyngeal cancers were 0.92 for 1-10 j-yrs; 0.89 for 10-30 j-yrs; 0.81 for 30-60 j-yrs; and 1.0 for more than 60 j-yrs. "Similar, though less precise results were obtained for the other cancer sites," Tashkin reported. "We found absolutely no suggestion of a dose response." The data on tobacco use, as expected, revealed "a very potent effect and a clear dose-response relationship -a 21-fold greater risk of developing lung cancer if you smoke more than two packs a day." Similarly high odds obtained for oral/pharyngeal cancer, laryngeal cancer and esophageal cancer. "So, in summary" Tashkin concluded, "we failed to observe a positive association of marijuana use and other potential confounders."

There was time for only one question, said the moderator, and San Francisco oncologist Donald Abrams, M.D., was already at the microphone: "You don't see any positive correlation, but in at least one category [marijuana-only smokers and lung cancer], it almost looked like there was a negative correlation, i.e., a protective effect. Could you comment on that?"

"Yes," said Tashkin. "The odds ratios are less than one almost consistently, and in one category that relationship was significant, but I think that it would be difficult to extract from these data the conclusion that marijuana is protective against lung cancer. But that is not an unreasonable hypothesis."

Abrams had results of his own to report at the ICRS meeting. He and his colleagues at San Francisco General Hospital had conducted a randomized, placebo-controlled study involving 50 patients with HIV-related peripheral neuropathy. Over the course of five days, patients recorded their pain levels in a diary after smoking either NIDA-supplied marijuana cigarettes or cigarettes from which the THC had been extracted. About 25% didn't know or guessed wrong as to whether they were smoking the placebos, which suggests that the blinding worked. Abrams requested that his results not be described in detail prior to publication in a peer-reviewed medical journal, but we can generalize: they exceeded expectations, and show marijuana providing pain relief comparable to Gabapentin, the most widely used treatment for a condition that afflicts some 30% of patients with HIV.

To a questioner who bemoaned the difficulty of "separating the high from the clinical benefits," Abrams replied: "I'm an oncologist as well as an AIDS doctor and I don't think that a drug that creates euphoria in patients with terminal diseases is having an adverse effect." His study was funded by the University of California's Center for Medicinal Cannabis Research.

Source: MapInc