EDITORIAL

Cannabis use and cancer

INTRODUCTION In many developed societies, cannabis is widely smoked in the same way as tobacco (Hall 1998). Cannabis smoking, like tobacco smoking, begins typically in adolescence (Bachman et al. 1997) but unlike tobacco, most cannabis users discontinue their use in their mid- to late 20s, with only a minority continuing to use cannabis into their 30s and beyond (Bachman et al. 1997). Given the cancer risks of tobacco smoking, is smoking cannabis likely to be a cause of cancer? An answer to this question requires a review of two research literatures. The first is research on whether substances in cannabis or its combustion products cause mutations in bodily cells that may lead to cancer. Cannabis could be a cause of cancer if its principal psychoactive ingredient, tetrahydrocannabinol (THC), or substances produced by its combustion, produced mutations in the genetic material of somatic cells (such as those of the lung) that were exposed to them (MacPhee et al. 1999). The second literature is clinical and epidemiological research on associations between cannabis use and the occurrence of specific cancers.

IS CANNABIS SMOKE MUTAGENIC AND CARCINOGENIC? There is more consistent evidence that cannabis smoke is mutagenic in vitro (MacPhee et al. 1999; Marselos & Karamanakos 1999; Leuchtenberger et al. 1983). Cannabis smoke produces chromosomal aberrations such as hypoploidy and mutagenicity in the Ames test (Bloch et al. 1983) and it causes cancers in the mouse skin test assay of carcinogenicity (MacPhee et al. 1999). The fact that it is cannabis smoke that is carcinogenic (Bloch et al. 1983) suggests that any cancers caused by cannabis smoking are most likely to develop in organs that receive maximum long-term exposure to cannabis smoke or its constituents, namely, the lung and possibly the upper aerodigestive tract (mouth, tongue, oesophagus) and bladder (MacPhee et al. 1999).

CANNABIS SMOKING AND CANCERS OF THE AERODIGESTIVE AND RESPIRATORY TRACT Reasons for concern

IS THC MUTAGENIC AND CARCINOGENIC? There is very little evidence that THC is mutagenic. Studies of animal cells suggest that THC can produce alterations to cell metabolism, DNA synthesis and cell division in the test tube (Nahas & Latour 1986). These changes, however, are more likely to delay or stop cell division rather than to produce cellular changes that may lead to cancer (MacPhee et al. 1999). THC and other cannabinoids are not mutagenic in standard microbial assays, such as the Ames test (MacPhee et al. 1999; Marselos & Karamanakos 1999). There is inconsistent evidence on whether the cannabinoids are clastogenic (i.e. produce breaks in chromosomes) (Marselos & Karamanakos 1999) However, even if they are clastogenic they are unlikely to be mutagenic (MacPhee et al. 1999), because chromosomal abnormalities are more likely to kill the affected cell than to produce malignant transformation and proliferation (MacPhee et al. 1999). A recent study in rats and mice has also found no evidence that THC was carcinogenic (Chan et al. 1996). © 2002 Society for the Study of Addiction to Alcohol and Other Drugs

There are good reasons for suspecting that cannabis smoking may contribute to the development of lung cancer and cancers of the aerodigestive tract (the oropharynx, nasal and sinus epithelium and the larynx). The first is the strong qualitative similarity between the carcinogens found in cannabis and tobacco smoke (Institute of Medicine 1999; Tashkin et al. 1999; Van Hoozen & Cross 1997). Tobacco smoke is a demonstrated cause of respiratory cancer (International Agency On Cancer 1990). Cannabis contains many of the same carcinogens and mutagens as tobacco but differs in containing cannabinoids rather than nicotine (Tashkin et al. 1999). There are quantitative differences in the amounts of some carcinogens, with cannabis smoke typically containing higher levels of some carcinogens than tobacco smoke (Institute of Medicine 1999; Tashkin et al. 1999) Secondly, chronic cannabis smokers show histopathological changes of a type that typically precede carcinoma development in tobacco smokers (Tashkin et al. 1999; Fligiel et al. 1988). Thirdly, there are case reports of cancers of the upper aerodigestive tract in young adults who have been Addiction, 97, 243–247

244

Editorial

chronic cannabis smokers. Donald (Donald 1991) reported 13 cases of advanced head and neck cancer occurring in young adults under 40 years of age. Eleven had been daily cannabis smokers but at least five also smoked tobacco and three were heavy alcohol consumers, both risk factors for these cancers (Vokes et al. 1993). Taylor (1988) reported a similar series of 10 upper respiratory tract cancers in adults under the age of 40 years. Five cases had a history of heavy cannabis smoking, two patients were ‘regular’ cannabis users, one was a ‘possible’ cannabis user and two did not use cannabis. Six were heavy alcohol consumers and six were cigarette smokers. Other investigators (Caplan & Brigham 1990; Nahsa & Latour 1992; Endicott et al. 1993) have reported cancers in young adults with histories of heavy cannabis use. Caplan & Brigham (1990) reported two cases of squamous cell carcinoma of the tongue in males aged 37 and 52 years. Neither smoked tobacco nor consumed alcohol; a history of long-term daily cannabis use was their only shared risk factor. These case reports raise a suspicion but provide limited support for the hypothesis that cannabis use can cause upper respiratory tract cancers. They do not compare rates of cannabis use in cases and controls; they assess cannabis exposure retrospectively, in the knowledge that the user has cancer; and they do not control for confounding factors such as alcohol and tobacco use.

Controlled studies Sidney et al. (1997) studied cancer incidence during an 8.6-year follow-up of 64 855 members of the Kaiser Permanente Medical Care Program (KPMCP). Study participants were asked about cannabis use during multiphasic screening between 1979 and 1985. Their average age at entry was 33 years and they were followed until death, a diagnosis of cancer or HIV/AIDS, exit from the KPMCP or 31 December 1993 (a mean of 8.6 years). At the time of entry to the cohort, 38% had never used cannabis, 20% were experimenters who had used less than six times, 20% were past users and 22% were current users. Data on cancer incidence were collected from a cancer registry and the California mortality data system. There was no overall excess of cancer when those who had ever used cannabis or who were current users were compared to those who were non-users at study entry (RR = 0.9, 95% confidence interval: 0.7, 1.2). Tobacco smokers had a higher rate of tobacco-related cancers than non-smokers (regardless of cannabis use) (RR = 2.7–3.8) but cannabis smokers did not. Males who had smoked cannabis had an increased risk of

© 2002 Society for the Study of Addiction to Alcohol and Other Drugs

prostate cancer (RR = 3.1, 95% confidence interval: 1.0, 9.5), as did males who were current cannabis smokers (RR = 4.7, 95% confidence interval: 1.4, 15.5) (Sidney et al. 1997). Zhang et al. (1999) reported a case–control study of cannabis use in head and neck cancers. They compared rates of cannabis use among 173 cases of primary squamous cell carcinoma of the head and neck with that in 176 controls. The cases were patients at the Memorial–Sloan Kettering Hospital in 1992–94 and the controls were age- and sex-matched cancer-free blood donors at the same hospital. The life-time use of cannabis was higher in cases than controls (14% and 10%, respectively) and the odds ratio for cannabis smoking was 2.6 (95% confidence interval: 1.1, 6.6) after adjustment for cigarette smoking, alcohol use and other known risk factors. There was a dose–response relationship between frequency and duration of cannabis use and the likelihood of having cancer. The relationship between cannabis smoking and these cancers was stronger among adults under the age of 55 years (OR = 3.1, 95% CI: 1.0, 9.7). There was a suggestion of an interaction between tobacco and cannabis smoking (Zhang et al. 1999). Zhang et al. considered the possibility that selection bias may have led them to overestimate the strength of the association between cannabis smoking and this cancer. They showed that the prevalence of cannabis use among their controls did not differ from that in general population surveys. This, and the fact that there was a dose–response relationship between risk and frequency and duration of cannabis use, makes selection bias an unlikely explanation of their findings. None the less, replication would be desirable. How do we reconcile the negative findings of the Sidney et al. study with the positive result from the Zhang et al. case–control study? The average age of Sidney et al. cohort was 43 years at follow-up, which is probably too young to see many excess cancers attributable to cannabis smoking. The chance of finding cancers attributable to cannabis smoking in the Sidney et al. cohort was also small because only 22% were cannabis users at study entry. As yet, there is no epidemiological evidence that regular cannabis smoking causes cancers of the lungs and lower respiratory tract of the type caused by cigarette smoking (Tashkin et al. 1999). Case–control studies of respiratory cancers would now be timely, given that the cannabis users among the postwar birth cohorts are reaching the age of 60 years, when the incidence of these cancers begins to increase steeply. A longer follow-up of the Sidney et al. cohort may also reveal whether cannabis smoking causes respiratory cancers.

Addiction, 97, 243–247

Editorial

THE PUBLIC HEALTH IMPACT OF CANCERS CAUSED BY CANNABIS SMOKING On current patterns of use, cannabis smoking will make, at most, a small contribution to the occurrence of respiratory cancers. This is so even if we assume that the risks of daily cannabis smoking are comparable to those of daily tobacco smoking (Hall 1998). The reason is that in most western societies there are many more daily tobacco (25–30%) than daily cannabis smokers (1–3%) (Hall 1995); most cannabis smokers stop in their mid- to late 20s (Bachman et al. 1997); and the 1% or fewer people who smoke cannabis daily over decades typically smoke 1–3 cannabis cigarettes per day rather than 10–30 tobacco cigarettes a day (Didcott et al. 1997). Among this minority, however, cannabis smoking into the fourth and later decades may increase the risk of respiratory cancer in tobacco smokers who concurrently smoke cannabis. It remains uncertain what effect a change in the legal status of cannabis would have on patterns of cannabis use (Hall & Babor 2000). It seems likely, given our experience with alcohol and tobacco, that an increase in the availability of cannabis and increased promotion of its use would increase the proportion of the population who were regular users and the duration of their use may possibly increase (Kleiman 1992). If these changes occurred, then the contribution that cannabis smoking made to respiratory cancer could increase. Better assessments of the magnitude of such cancer risk will indicate how big any increase may be.

MATERNAL CANNABIS SMOKING DURING PREGNANCY AND CHILDHOOD CANCERS Cannabis smoking has also been linked with cancers among children born to mothers who used cannabis during pregnancy. The evidence for the risks of childhood cancers derives from three case–control studies that examined cannabis use among a range of other risk factors. There was no a priori reason to expect a relationship between cannabis use and these cancers, as there was in the case of respiratory cancers. An association between maternal cannabis use and childhood cancer was reported in a case–control study of acute non-lymphoblastic leukemia (ANLL), a rare form of childhood cancer (Robinson et al. 1989). The study was not designed to assess any relationship between cannabis use and ANLL; it was designed instead to assess the relationship between this childhood cancer and maternal and paternal environmental exposures to petrochemi-

© 2002 Society for the Study of Addiction to Alcohol and Other Drugs

245

cals, pesticides and radiation. Maternal cannabis use during pregnancy was assessed so that it could be statistically controlled when assessing the relationship between ANLL and maternal and paternal environmental exposures. An unexpectedly strong association was found between maternal cannabis use and ANLL. The mothers of cases were 11 times more likely to have used cannabis before and during their pregnancy than were the mothers of controls. The relationship persisted after statistical adjustment for other risk factors. Reporting bias on the part of the mothers of cases was an alternative explanation, because the reports of cannabis use were obtained after diagnosis of the ANLL. The authors compared rates of cannabis use reported in this study with the rates in case–control studies of other childhood cancers and found that the rate of cannabis use was lower among controls in the ANLL study. When the rate of cannabis use among controls was adjusted the odds ratio was greater than 3 and statistically significant. Two other case–control studies have reported an increased risk of rhabdomyosarcoma (Grufferman et al. 1993) and astrocytomas (Kuitjen et al. 1992) in children born to women who reported using cannabis during their pregnancies. Neither study was a planned investigation of the association between childhood cancer and maternal cannabis use. In each case, cannabis use was one of a large number of possible confounding variables that were measured to control for in statistical analyses of the relationship between the exposure of principal interest and the childhood cancer. There is no suggestion that the incidence of any of these cancers has increased over the period 1979–95 in a way that could be accounted for by maternal cannabis use during pregnancy (Reis et al. 2000). The incidence of ANLL, for example, remained steady between 1979 and 1995 (Smith et al. 2000). The same was true for softtissue sarcomas (which include rhabdomyosarcomas) (Gurney et al. 2000). This null finding is most convincing in the case of ANLL because of the very high relative risk reported for this cancer. Central nervous system (CNS) malignancies (about 52% of which are astrocytomas) did increase in incidence between 1979 and 1995 (Gurney et al. 2000) but in a way that was more likely to reflect improvements in detection than changes in maternal cannabis use. The incidence of these cancers was steady between 1979 and 1985 when it abruptly increased; it remained steady thereafter (Gurney et al. 2000). Magnetic resonance imaging became widely available in the United States in 1985, which suggests that the increase was the result of improved diagnosis rather than an increase in incidence (Gurney et al. 2000).

Addiction, 97, 243–247

246

Editorial

IMPLICATIONS The experimental and epidemiological evidence on the cancer risks of cannabis use is still too meagre to warrant strong conclusions, but it raises concerns that for reasons of prudence should be communicated to cannabis users on public health grounds. The strongest reason for concern is that cannabis smoke contains many of the same carcinogens as tobacco smoke and it is mutagenic in microbial assays and carcinogenic in some animal tests. It is therefore a potential cause of cancer in body cells that are regularly and chronically exposed to it such as those of the aerodigestive and respiratory tracts. The epidemiological evidence is conflicting. There are case reports of aerodigestive tract cancers among relatively young adults who have been daily cannabis users and one case–control study has found an association between cannabis smoking and head and neck cancer. The only prospective study to date did not find an increased incidence of head and neck or respiratory cancers. It did find unexpectedly an increased rate of prostate cancer, a finding that needs to be confirmed. The relative youth and the low prevalence of regular cannabis use among participants in this cohort study may have reduced its ability to detect an increase in respiratory cancers. Further follow-ups of this cohort and case–control and prospective epidemiological studies are required to assess more effectively the cancer risks of cannabis use. Such studies may explain why cancers of the upper respiratory tract have been found among cannabis smokers before cancers of the respiratory tract. They will also enable a judgement to be made about the significance of the finding on prostate cancer. Given the high rates of tobacco use among regular cannabis users, it will be difficult to obtain definitive evidence that cancers observed among cannabis users are due to cannabis rather than to tobacco smoking. The best way of deciding the issue may be to examine the age of onset of these cancers among tobacco-only smokers and people who smoke both cannabis and tobacco. While awaiting the results of these studies, the similarities between cannabis and tobacco smoke are sufficiently strong to warrant advising regular cannabis smokers, especially those who also smoke tobacco, that they may be at increased risk of developing cancers of the upper respiratory tract (Marselos & Karamanakos 1999). The evidence that cannabis smoking during pregnancy increases the risk of childhood cancer is much weaker. Three case–control studies of three different types of cancer have reported an association with maternal cannabis use. None was a planned study of the role of cannabis use in these cancers and none of these © 2002 Society for the Study of Addiction to Alcohol and Other Drugs

cancers has shown an increase of the type that may be expected if maternal cannabis use was a cause of any of them. Because of the adverse effects of maternal tobacco smoking on fetal growth, it would none the less be prudent to advise women against smoking cannabis during pregnancy for the same reasons that we advise them against smoking tobacco during pregnancy. WAYNE HALL Professor and Executive Director National Drug and Alcohol Research Centre University of New South Wales Sydney Australia DONALD MACPHEE Chief of Radiobiology Radiation Effects Research Foundation Hiroshima Japan

REFERENCES Bachman, J. G., Wadsworth, K. N., O’Malley, P., Johnston, L. & Sschulenberg, J. (1997) Smoking, Drinking and Drug Use in Young Adulthood: the Impacts of New Freedoms and Responsibilities. Mahwah, NJ: Lawrence Erlbaum. Bloch, E. (1983) Effects of marijuana and cannabinoids on reproduction, endocrine function, development and chromosomes. In: Fehr, K. & Kalant, H., eds. Cannabis and Health Hazards, pp. 355–432. Toronto: Addiction Research Foundation. Caplan, G. A. & Brigham, B. A. (1990) Marijuana smoking and carcinoma of the tongue. Is there an association? Cancer, 66, 1005–1006. Chan, P., Sills, R., Braun, A., Haseman, J. & Bucher, J. (1996) Toxicity and carcinogenicity of delta-9-tetrahydrocannabinol in Fischer rats and B6C2F1 mice. Fundamental and Applied Toxicology, 30, 109–117. Didcott, P., Reilly, D., Swift, W. & Hall, W. (1997) Long Term Cannabis Users on the New South Wales North Coast. NDARC Monograph no. 30. Sydney: National Drug and Alcohol Research Centre, UNSW. Donald, P. (1991) Marijuana and upper aerodigestive tract malignancy in young patients. In: Nhas, G. & Latour, C., eds. Physiopathology of Illicit Drugs: Cannabis, Cocaine, Opiates, pp. 39–54. Oxford: Pergamon Press. Endicott, J. N., Skipper, P. & Hernandez, L. (1993) Marijuana and head and neck cancer. Advances in Experimental Medicine and Biology, 335, 107–113. Fligiel, S., Beals, T., Venkat, H., Stuth, S., Gong, H. & Tashkin, D. (1988) Pulmonary pathology in marijuana smokers. In: Chesher, G., Consroe, P. & Musty, R., eds. Marijuana: an International Research Report. NCADA Monograph no. 7, pp. 43–45. Canberra: Australian Government Publishing Service. Grufferman, S., Schwartz, A. G., Ruymann, F. B. & Maurer, H. M. (1993) Parents’ use of cocaine and marijuana and increased risk of rhabdomyosarcoma in their children. Cancer Causes and Control, 4, 217–224. Addiction, 97, 243–247

Editorial

Gurney, J., Young, J., Roffers, S., Smith, M. A. & Bunin, C. (2000) CNS and miscellaneous intracranial and intraspinal neoplasms. In: Reis, L., Eisner, M., Kosary, C., Hankey, B. F., Miller, B. A., Clegg, L. & Edwards, B. K., eds. SEER Cancer Statistics Review, 1973–97, pp. 51–63. Bethseda, MD: National Cancer Institute. Gurney, J., Young, J., Roffers, S., Smith, M. A. & Bunin, C. (2000) Soft-tissue sarcomas. In: Reis, L., Eisner, M., Kosary, C., Hankey, B. F., Miller, B. A., Clegg, L. & Edwards, B. K., eds. SEER Cancer Statistics Review, 1973–97, pp. 111–123. Bethseda, MD: National Cancer Institute. Hall, W. (1995) The public-health significance of cannabis use in Australia. Australian Journal of Public Health, 19, 235–242. Hall, W. (1998) The respiratory risks of cannabis smoking. Addiction, 93, 1461–1463. Hall, W. & Babor, T. F. (2000) Cannabis use and public health: assessing the burden. Addiction, 95, 485–490. Institute of Medicine (1999) Marijuana and Medicine: Assessing the Science Base. Washington, DC: National Academy Press. International Agency on Cancer (1990) Cancer: Causes, Occurrence and Control. Lyon: International Agency on Cancer. Kleiman, M. (1992) Against Excess: Drug Policy for Results. New York: Basic Books. Kuitjen, R. R., Bunin, G. R., Nass, C. C. & Meadows, A. T. (1992) Parental occupation and childhood astrocytoma. Cancer Research, 52, 782–786. Leuchtenberger, C. (1983) Effects of marihuana (cannabis) smoke on cellular biochemistry of in vitro test systems. In: Fehr, K. & Kalant, H., eds. Cannabis and Health Hazards, pp. 177–224. Toronto: Addiction Research Foundation. MacPhee, D. (1999) Effects of marijuana on cell nuclei: A review of the literature relating to the genotoxicity of cannabis. In: Kalant, H., Corrigal, W., Hall, W. & Smart, R., eds. The Health Effects of Cannabis, pp. 293–309. Toronto: Centre for Addiction and Mental Health. Marselos, M. & Karamanakos, P. (1999) Mutagenicity, developmental toxicity and carcinogenicity of cannabis. Addiction Biology, 4, 5–12.

© 2002 Society for the Study of Addiction to Alcohol and Other Drugs

247

Nahas, G. & Latour, C. (1986) Cannabis: toxicological properties and epidemiological aspects. Medical Journal of Australia, 145, 82–87. Nahsa, G. & Latour, C. (1992) The human toxicity of marjiuana. Medical Journal of Australia, 156, 495–497. Reis, L., Eisner, M., Kosary, C., Hankey, B. F., Miller, B. A., Clegg, L. & Edwards, B. K. (2000) SEER Cancer Statistics Review, 1973–97. Bethesda, MD: National Cancer Institute. Robinson, L., Buckley, J., Daigle, A., Wells, R., Benjamin, D., Arthur, D. C. & Hammond, G. D. (1989) Maternal drug use and the risk of childhood nonlymphoblastic leukemia among offspring: an epidemiologic investigation implicating marijuana. Cancer, 63, 1904–1911. Sidney, S., Beck, J. E., Tekawa, I. S., Quesenberry, C. P. Jr & Friedman, G. D. (1997) Marijuana use and mortality. American Journal of Public Health, 87, 585–590. Smith, M. A., Gloekler, Reis, J., Gurney, J. & Ross, J. (2000) Leukemia. In: Reis, L., Eisner, M., Kosary, C., Hankey, B. F., Miller, B. A., Clegg, L. & Edwards, B. K., eds. SEER Cancer Statistics Review, 1973–97, pp. 17–34. Bethseda, MD: National Cancer Institute. Tashkin, D. (1999) Effects of cannabis on the respiratory system. In: Kalant, H., Corrigal, W., Hall, W. & Smart, R., eds. The Health Effects of Cannabis, pp. 311–346. Toronto: Addiction Research Foundation. Taylor, I. F. (1988) Marijuana as a potential respiratory tract carcinogen: a retrospective analysis of a community hospital population. Southern Medical Journal, 81, 1213–1216. Van Hoozen, B. E. & Cross, C. E. (1997) Marijuana. Respiratory tract effects. Clinical Reviews in Allergy and Immunology, 15, 243–269. Vokes, E., Weischelsbaum, R., Lippman, S. & Hong, W. (1993) Head and neck cancer. New England Journal of Medicine, 328, 183–194. Zhang, Z.-F., Morgenstern, H., Spitz, M., Tashkin, D. P., Yu, G. P., Marshall, J. R., Hsu, T. C. & Schantz, S. P. (1999) Marijuana use and increased risk of squamous cell carcinoma of the head and neck. Cancer Epidemiology, Biomarkers and Prevention, 8, 1071–1078.

Addiction, 97, 243–247