Intended for healthcare professionals

Rapid response to:


Comparing cannabis with tobacco: Services are needed for acute and chronic effects of cannabis

BMJ 2003; 327 doi: (Published 17 July 2003) Cite this as: BMJ 2003;327:165

Rapid Response:

Carcinogenic Cannabis?

To date, there have been 10 studies on the effects of cannabinoids on
cancers in vivo. Eight of them demonstrated antineoplastic effects,
especially on highly aggressive malignancies liable to be caused by
polycyclic aromatic hydrocarbons (viz., lung adenocarcinoma, squamous
cell carcinoma, and leukemia). These 8 studies were conducted in Spain,
Italy, and the US. The remaining 2 studies were conducted at UCLA and
demonstrated that cannabinoids promote tumor growth via

What is interesting about the UCLA studies is that
they administered cannabinoids intravenously at inordinately high doses
-- a systemic route and dose irrelevant to cannabis smokers. If the UCLA
in vivo studies are relevant to anyone, it would be cancer patients who
have repeatedly overdosed on IV marinol. If the primary concern is that
the tissues exposed directly to cannabis smoke are vulnerable to
malignant transformation, then relevant effects would be most evident
from repeated applications of cannabinoids directly to pre-existing
tumors (since there has never been an in vivo study to demonstrate
cannabinoids or cannabis smoke induces cancer). As for evidence in
humans, there have been 8 epidemiological studies conducted to date. The
only prospective study showed an elevated risk in tobacco smokers but not
in cannabis smokers even though a significant number of the cannabis
smokers did not smoke tobacco. UCLA's much-vaunted proof of cannabis
causing cancer at an early age derives from two epidemiological studies
(one to be presented this summer), neither of which acknowledged the risk
entailed by chronic exposure to paraquat among cannabis smokers in the
southwest US, where, as opposed to other regions of the US, the available
evidence suggests that heavier cannabis smokers were significantly more
likely to be exposed to levels of paraquat too low to cause symptoms but
high enough to be carcinogenic.

That paraquat is carcinogenic in humans
has been suggested by some, but not all, in vivo studies; however, the
most convincing evidence that paraquat is a potent carcinogen in humans
derives from a recent case-report of an agricultural worker who
developed 100 squamous tumors on an area of his skin that was chronically
exposed to paraquat but not exposed to excessive sunlight.

Aside from the
UCLA studies, 3 of the studies were conducted in Europe: two reported a
significant risk, while one did not show a significant risk. The
remaining 2 epidemiological studies were conducted in the US -- in
Washington and in Maryland -- two regions where cannabis smokers would
not face an elevated risk for paraquat exposure. Both of these studies
were much larger than the UCLA study that has been widely reported in the
popular media. Also, these studies used random-digit selection
methodology and showed that tobacco -- not cannabis -- was a major risk

Competing interests:
None declared

Competing interests: No competing interests

26 February 2004
Cory Gordon