Cannabis Health Search

ORAL CANCER  & cannabis studies completed


Oral cancer, a subtype of head and neck cancer, is any cancerous tissue growth located in the oral cavity.

It may arise as a primary
lesion originating in any of the oral tissues, by metastasis from a distant site of origin, or by extension from a neighbouring anatomic structure, such as the nasal cavity

Science and Research

Marijuana use and Risk of Oral Squamous Cell Carcinoma

Marijuana Use and Risk of Oral Squamous Cell Carcinoma
Karin A. Rosenblatt, Janet R. Daling, Chu Chen, Karen J. Sherman, and Stephen M. Schwartz
Department of Community Health, University of Illinois at Urbana-Champaign, Champaign, Illinois; Program in Epidemiology, Division of Public Health Sciences, Fred
Hutchinson Cancer Research Center, Seattle, Washington; 3Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle,
Washington; and 4Center for Health Studies, Group Health Cooperative, Seattle, Washington


Previous laboratory investigations, case reports, and a hospital-based case-control study have suggested that marijuana use may be a risk factor for squamous cell head and neck cancer. We conducted a populationbased case-control study to determine whether marijuana use is associated
with the development of oral squamous cell carcinoma (OSCC). Case subjects (n  407) were 18–65-year-old residents of three counties in western Washington State who were newly diagnosed with OSCC from 1985 through 1995. Control subjects (n  615), who were similar to the
cases with respect to age and sex, were selected from the general population using random-digit telephone dialing. Lifetime histories of marijuana use and exposure to known OSCC risk factors were ascertained using a structured questionnaire. Information on genetic polymorphisms in glutathione
S-transferase enzymes was obtained from assays on participant DNA.

Odds ratios for associations with features of marijuana use were adjusted for sex, education, birth year, alcohol consumption, and cigarette smoking. A similar proportion of case subjects (25.6%) and control subjects (24.4%) reported ever use of marijuana (adjusted odds ratio, 0.9; 95% confidence interval, 0.6 –1.3). There were no trends in risk observed with increasing duration or average frequency of use or time since first or last use. No subgroup defined by known or suspected OSCC risk factors (age, cigarette smoking, alcohol consumption, and genetic polymorphisms)
showed an increased risk. Marijuana use was not associated with OSCC risk in this large, population-based study.


Marijuana is the most commonly used illegal drug in the United States, and new users increased among minors during the 1990s. Marijuana smoke contains many known carcinogens, and experimental studies show that components of marijuana smoke are mutagenic in bacteria and cause molecular and cellular changes in bronchial tissue comparable with those seen among tobacco smokers and consistent with early steps in cancer development. Such findings raise the possibility that chronic marijuana use could cause premalignant changes in cells throughout the upper aerodigestive tract.

The possibility that marijuana use might be a risk factor for head and neck squamous cell carcinoma (HNSCC) was initially raised by several case reports. A small hospital-based study of HNSCC cases (n  173) and blood donor controls (n  176) found that ever users of marijuana were at 2-fold increased risk of HNSCC [odds ratio (OR), 2.6; 95% confidence interval (CI), 1.1– 6.6] and that the
risk increased with increasing frequency of marijuana use. Because hospital-based studies may be particularly susceptible to biases when lifestyle characteristics are the focus of investigation, we analyzed data from a population-based study to test the hypothesis that marijuana use is a risk factor for oral squamous cell carcinoma


Study Population. This report is based on participants, data, and biological specimens assembled during two population-based case-control studies originally designed to examine the association between human papillomavirus infection and OSCC risk. All participants were residents of King,
Pierce, or Snohomish counties, Washington State. Eligible cases in the first study were all 18–65-year-old men diagnosed with first, incident OSCC between January 1985 and December 1989. Eligible cases in the second study include all 18–65-year-old men and women diagnosed with first, incident
OSCC between January 1990 and June 1995. OSCC patients were ascertained through the population-based Cancer Surveillance System, a participant in the Surveillance, Epidemiology, and End Results program. In both studies, only individuals who could communicate in English were eligible. OSCC was defined as in situ and invasive tumors of the tongue, gums, floor of mouth, tonsils, oropharynx, and other intraoral sites. In the earliest of the two studies, the definition of OSCC also included cancers of the lip (exclusive of the vermilion border); these cases of lip cancer were excluded from the present report.

To be eligible, OSCC cases also were required to have residential telephones to ensure comparability with controls, who were identified for both previous studies through random-digit telephone dialing and frequencymatched to cases on age (18–19 years, 20–24 years, 25–29 years, . . . , 60–65
years) and sex.

The protocols for recruitment of cases and controls in both studies were approved by the institutional review board of the Fred Hutchinson Cancer Research Center. Combined across the two studies, 407 cases and 615 controls participated in an in-person interview (see below). These figures represented participation rates for cases in the first and second study of 54.4% and 63.3%, respectively,
and 59.7% overall.

Only eligible cases and controls were included in calculation of participation rates. Among the 275 nonparticipating cases in the previous two studies, 125 had died before they could be contacted for recruitment.

The control response rates for the two studies were 63% and 61%,respectively; these rates incorporate both the household screening phase of random-digit telephone dialing and the success at interviewing eligible controls from among screened households. Data and Biological Specimen Collection. Participating cases and controls were interviewed in-person by trained personnel using a structured questionnaire. The same questionnaire was used in both studies. To elicit histories of marijuana use, we initially asked each participant whether he or she had ever smoked marijuana or hashish (a stronger form of marijuana). If a participant had ever smoked these substances, he or she was asked about different episodes of marijuana or hashish use during his or her lifetime, with
each episode representing a different frequency of use (elicited in terms of times per day, week, month, or year). For each episode, each participant was asked about the frequency of use, the age (in whole years) he or she started and stopped using marijuana or hashish at that frequency, and whether marijuana, hashish, or both substances were used. The interview also elicited demographic
characteristics and extensive histories of tobacco use and alcohol consumption as described previously.

All questions were directed toward the time period before each participant’s reference date. The reference date for a case was the month and year he or she was diagnosed. Reference dates for controls were assigned at random from among the possible case diagnosis dates that had
occurred before the selection of a particular control through random-digit
telephone dialing.

Genetic Polymorphism Analyses. Because the putative carcinogenicity of marijuana may derive, at least in part, from exposure to polycyclic aromatic hydrocarbons and other tobacco-related carcinogens, we assayed for polymorphisms in several glutathione S-transferase (GST) genes (GSTM1, GSTT1, and GSTP1), which are known to be involved in the biotransformation of these
compounds. Biological specimens from which genomic DNA could be extracted had been sought from each participating case and control during theoriginal studies. Across the two studies, such specimens were available for 365 of 407 interviewed cases (89.7%) and 576 of 615 interviewed controls (93.7%).

The null polymorphisms of GSTM1 and GSTT1 were assayed asdescribed previously (15, 16). The GSTP1 (I105V) polymorphism ( was assayed as follows:a 176-bp fragment was amplified using primers 5-ACC-CCA-GGG-CTCTAT- GGG-AA-3 and 5-TGA-GGG-CAC-AAG-AAG-CCC-CT-3. Each 30-l reaction contained 15 l of Qiagen Taq PCR Master Mix, 0.2 M each primer, and 100 ng of genomic DNA. Thermal cycling conditions were as
follows: 1 cycle at 94°C for 5 min; 40 cycles at 94°C for 2 min, 60°C for 1 min, and 72°C for 2 min; and 1 cycle at 72°C for 5 min. BsmA1 (New England Biolabs) restriction fragments were separated on a 4% nusieve gel.

Quality control samples included wells containing known genotype (positive controls), wells with PCR reagent only (negative controls), and paired replicate aliquots. Two reviewers independently read the gels and assigned genotypes without knowledge of the case-control status or other characteristics
of each participant.

Genetic analyses, which were restricted to whites because of variation in allele frequency across racial groups and because most of our study population was white (92.4% among controls), included the following numbers of subjects (percentage of total interviewed): GSTM1, 340 cases
(89.0%) and 548 controls (83.5%); GSTT1, 339 cases (88.8%) and 547 controls (83.3%); and GSTP1, 355 cases (91.5%) and 565 controls (87.2%).

Statistical Analysis. We created variables to characterize various features of marijuana use, including ever use, time since first use, time since last use, years of use, and average frequency of use. The average frequency of use was calculated by first determining, for each episode, the number of times the participant had used marijuana (based on the length of the episode and the
reported frequency of use).

We then summed the total number of uses over each participant’s lifetime and divided that number by the total weeks of use.

These quantitative measures of the extent of marijuana use were categorized for analyses. A separate exposure category, with respect to each of these measures, was established for participants whose who had 1 year of use because this exposure level was felt to be minimal and could not be distinguished from years since first use or years since last use (i.e., because their
reported age at first and last use of marijuana was the same).

The remaining categories used for marijuana use were established so that they represented
either 5- or 10-year intervals (years of use, years since first use, years since last use) or could be directly compared with the results of Zhang et al. (times used/week; Ref. 10). Participants who reported never using marijuana comprised the referent group for comparisons.

We used standard methods for statistical analysis of case-control studies. ORs and 95% CIs were calculated using unconditional logistic regression. Analyses were adjusted for sex, education, birth year (continuous), average drinks of alcohol/week (continuous), pack-years of cigarette smoking
(continuous), and whether the data came from the first (11) or second (12) of the previous studies.

We assessed whether the data were consistent with effect measure modification between marijuana use and other characteristics (age, sex, cigarette smoking, alcohol consumption, and genetic polymorphisms) by estimating stratum-specific ORs associated with marijuana use and by estimating
ORs jointly for marijuana use and each putative modifier relative to a common baseline consisting of individuals who never used marijuana and were in the a priori “low-risk” category of the modifier. We performed likelihood ratio tests of the fit of respective models with multiplicative and additive
interaction terms compared with the fit of models without such terms. For multiplicative models, all terms were log-linear, whereas for additive models, confounders were log-linear, and the interaction terms were linear.

Possible differences in the association with marijuana use according to tumor site (tongue, gum, floor of the mouth, tonsils and hypopharynx, and other sites) were assessed using polytomous logistic regression.

To evaluate the extent to which the reporting of marijuana use among our controls was consistent with other population-based studies, we analyzed publicly available data from the National Household Survey of Drug Abuse (NHSDA) conducted in 1988 and 1990–1994 [the NHSDA was not conducted
in 1989).

These years largely include the reference dates we used for eliciting risk factor histories. We compared the observed number of controls who reported ever use of marijuana with the expected number, based on the birth cohort (or in some analyses, age) and sex-specific prevalences of “ever
marijuana use” from the NHSDA and the birth cohort and sex-specific distribution
of our controls.

We calculated the observed:expected ratio of ever marijuana users and corresponding 95% CIs using the logarithmic transformation.

Cases had lower annual incomes and a lower educational level than controls (Table 1). The risk of OSCC was strongly related to cigarette smoking and alcohol consumption, as well as the combination of these characteristics (results not shown). Table 2 shows the characteristics of ever and never users of marijuana among controls. Ever users of marijuana were more likely than never users to have been born more recently and to be under 50 yearsof age, to be male, to have a low income, and to have attended graduate school. Marijuana users more often smoked tobacco, but they were less likely than never marijuana users to smoke tobacco at the higher levels (30 pack-years). Marijuana users drank alcohol more frequently than nonusers of marijuana.

Twenty percent of cases and 16% of controls only used marijuana,
Table 1 Characteristics of OSCCa cases and controls
Case %
(N  407)
Control %
(N  615)
(95% CI)
Age at reference date (yrs)
18–39 6.9 10.1
40–49 20.4 23.2
50–59 37.6 35.0
60–65 35.1 31.7
Male 70.8 71.5
Female 29.2 28.5
Birth year
1919–1929 30.0 26.8
1930–1934 21.4 20.3
1935–1939 17.0 16.4
1940–1944 15.2 14.5
1945–1949 7.4 9.4
1950–1959 7.4 9.3
1960–1971 1.7 3.2
White 93.9 94.1 1.0
Black 3.7 2.8 1.2 (0.6–3.2)
Other 2.5 3.1 1.6 (0.5–2.6)
$15,000 23.2 5.8 1.0
$15,000 to $29,999 26.2 20.8 0.4 (0.2–0.7)
$30,000 to $44,999 24.7 25.6 0.3 (0.2–0.5)
$45,000 25.9 47.8 0.2 (0.1–0.4)
High school or less 44.7 29.0 1.0
Technical school 7.9 5.9 1.1 (0.6–1.9)
College 37.8 49.1 0.7 (0.5–1.0)
Graduate school 9.6 16.0 0.7 (0.5–1.2)
Cigarette smoking (pack-years)d
1 14.6 38.3 1.0
1–9 6.4 15.7 1.0 (0.6–1.8)
10–19 7.6 11.7 1.6 (0.9–2.7)
20–29 10.9 12.2 2.0 (1.2–3.3)
30 60.5 22.0 6.1 (4.1–9.3)
Alcohol consumption (drinks/week)e
1 13.5 26.7 1.0
1–7 26.8 43.1 1.1 (0.7–1.6)
8–14 15.5 14.5 1.6 (1.0–2.7)
15–28 17.2 9.8 2.2 (1.2–3.8)
29 27.0 6.0 4.4 (2.0–9.6)
a OSCC, oral squamous cell carcinoma; OR, odds ratio; CI, confidence interval.
b ORs and 95% CIs adjusted for sex, birth year (continuous), cigarette smoking
(continuous pack-years), alcohol consumption (continuous average drinks/week), and
study (first or second). Excludes two cases and two controls with missing data on
pack-years of cigarette smoking.
c Excludes 9 cases and 11 controls with missing data on income.
d ORs and 95% CIs adjusted for sex, birth year (continuous), alcohol consumption
(continuous average drinks/week), and study (first or second). Excludes two cases and two
controls with missing data on pack-years of cigarette smoking.
e ORs and 95% CIs adjusted for sex, birth year (continuous), cigarette smoking
(continuous pack-years), and study (first or second). Excludes two cases and two controls
with missing data on pack-years cigarette smoking.
5% of cases and 6% of controls used both marijuana and hashish, and
1% of cases and of controls only used hashish. Table 3 displays the
association between marijuana use and oral cancer, adjusted for sex,
education, birth year, average number of alcoholic drinks/week, and
pack-years of cigarette smoking. There was no association with ever
having used marijuana (OR, 0.9; 95% CI, 0.6–1.3), total years of
marijuana use, average frequency of marijuana use, years since first
use of marijuana, or years since last use of marijuana. There was no
discernable difference across oral tumor sites in the association with
ever having used marijuana, nor did the association vary by stage at
diagnosis (data not shown).
The magnitude of the association between ever marijuana use and
oral cancer varied little among subgroups defined by known or suspected
demographic, lifestyle, or genetic risk factors, whether such
assessments were made against the null hypothesis of multiplicative
or additive joint effects (Table 4). Individuals with at least one copy
of GSTM1 had a borderline statistically significant reduced risk of
OSCC associated with marijuana use (OR, 0.5; 95% CI, 0.3–1.0),
whereas those with the null GSTM1 genotype showed no association
with marijuana use (OR, 1.1; 95% CI, 0.7–1.9; P  0.05 for test of
heterogeneity between strata). Evaluated under an additive model of
combined effects, the association between marijuana use and OSCC
varied in a statistically significant manner across categories of cigarette
smoking status (current, former, and never), with an inverse
association among current and never smokers, but no association
among former smokers. ORs for marijuana use, combined with either
alcohol use or alcohol and pack-years of cigarette smoking together,
suggested stronger risks for these established OSCC risk factors
among never marijuana users.
We observed 150 ever users of marijuana among our controls, and
we would have expected 146.8 controls to have reported ever use of
marijuana based on the sex- and birth cohort-specific ever marijuana
use estimates from the NHSDA (observed:expected ratio, 1.0; 95%
CI, 0.8–1.3). A similar expected number of ever marijuana users
among controls (n  141.8) was obtained when the calculations were
based on the sex- and age-specific NHSDA data (observed:expected
ratio, 1.1; 95% CI, 0.8–1.3).
In this large, population-based study, we did not find any association
between marijuana use and OSCC risk. The absence of an
increased risk was largely consistent across analyses using different
measures of marijuana use (e.g., ever use, frequent versus infrequent
use, and long-term versus short-term use) and among subgroups
representing different levels of underlying risk.
The possibility that marijuana use might increase the risk of cancer
was initially raised more than 20 years ago when marijuana smoke
components yielded positive results in Ames salmonella/microsome
mutagenesis assays (4, 5). However, in other non-human model systems,
9-tetrahydrocannabinol, the psychoactive component of marijuana,
both induced expression, and inhibited the activity, of cytochrome
p450 1A1 (3); the combination of tobacco extracts and 9-
tetrahydrocannabinol also led to reduced cytochrome p450 1A1
activity. In studies focusing directly on tumor development and
growth, cannabinoids have been shown to have both tumorigenic (20,
Table 3 Risk of OSCCa associated with use of marijuana
Marijuana use
Case %
(N  407)
Control %
(N  615)
(95% CI)
Ever use
Never 74.4 75.6 1.0b
Ever 25.6 24.4 0.9 (0.6–1.3)
Years of usec
1 yr 7.9 6.5 0.8 (0.4–1.2)
1 yr 1.0 3.1 0.2 (0.1–0.7)
2–5 yrs 4.7 3.9 1.3 (0.6–2.6)
6–15 yrs 5.9 6.5 0.7 (0.4–1.4)
15 yrs 6.1 4.4 1.2 (0.6–2.2)
Times used/weekc
1 year use 7.9 6.5 1.0 (0.6–1.8)
1 times/week 10.1 9.3 0.8 (0.5–1.4)
1–7 times/week 6.1 6.7 0.8 (0.4–1.6)
7 times/week 1.5 2.0 0.5 (0.2–1.6)
Years since first use
1 yr total use 7.9 6.5 1.0 (0.6–1.8)
15 yrs 3.7 3.1 0.7 (0.3–1.6)
16–20 yrs 4.2 5.4 0.7 (0.3–1.4)
21–25 yrs 5.6 6.3 0.9 (0.5–1.7)
25 yrs 4.2 3.1 0.9 (0.4–2.0)
Years since last usec
1 yr total use 7.9 6.5 1.0 (0.6–1.8)
Current use 8.1 5.4 1.1 (0.6–2.0)
10 yrs 3.0 3.6 0.7 (0.3–1.7)
11–20 yrs 4.9 7.3 0.7 (0.4–1.3)
20 yrs 1.7 1.6 0.7 (0.3–2.1)
a OSCC, oral squamous cell carcinoma; OR, odds ratio; CI, confidence interval.
b Reference group for calculation of ORs. All ORs are adjusted for sex, education, birth
year (continuous), alcohol consumption (continuous average drinks/week), cigarette
smoking (continuous pack-years), and study (first or second). Excludes two cases and two
controls with missing data on pack-years of cigarette smoking.
c Among individuals who used marijuana for at least 1 year and compared with persons
who had never used marijuana.
Table 2 Characteristics of ever users and never users of marijuana among controls
Ever user %
(N  150)
Never user %
(N  465)
Birth year
1919–1929 4.0 34.2
1930–1934 9.3 23.9
1935–1939 11.3 18.1
1940–1944 18.7 13.1
1945–1949 18.7 6.4
1950–1959 28.7 3.0
1960–1971 9.3 1.3
Age (yrs)
18–39 31.3 3.2
40–49 38.7 18.3
50–59 20.7 39.6
60–65 9.3 38.9
Male 78.7 69.2
Female 21.3 30.8
White 93.3 94.4
Black 4.0 2.4
Other 2.7 3.2
$15,000 11.5 3.9
$15,000 to $29,999 16.2 22.1
$30,000 to $44,999 24.3 26.1
$45,000 48.0 47.8
High school or less 22.7 31.2
Technical school 4.7 6.2
College 50.7 48.6
Graduate school 22.0 14.0
Cigarette smoking (pack-years)b
1 31.3 40.6
1–9 18.0 14.9
10–19 14.7 10.8
20–29 17.3 10.6
30 18.7 23.1
Alcohol consumption (drinks/week)
1 13.3 31.0
1–7 36.7 45.2
8–14 17.3 13.6
15–28 20.0 6.3
29 12.7 3.9
Cigarette smoking & alcohol consumptionb
20 pack-years/15 drinks/week 48.0 63.5
20 pack-years/15 drinks/week 16.0 2.8
20 pack-years/15 drinks/week 19.3 26.1
20 pack-years/15 drinks/week 16.7 7.6
a Excludes two users and seven never users who had missing information on income.
b Excludes two never users who had missing information on pack-years of smoking.
Table 4 Risk of OSCCa associated with use of marijuana in subgroups of the study population
Subgroup Marijuana use Case % (N  407) Control % (N  615) ORb (95% CI) ORc (95% CI)
55 No 26.0 28.6 1.0 1.0
Yes 18.4 20.6 0.8 (0.5–1.2) 0.8 (0.5–1.2)
55 No 48.4 47.0 1.0 0.6 (0.4–0.9)
Yes 7.1 3.7 1.5 (0.8–2.9) 1.0 (0.5–1.9)
LRTmult: P  0.11e
LRTadd: P  0.11f
Male No 50.4 52.4 1.0 1.0
Yes 20.4 19.2 0.9 (0.6–1.4) 0.9 (0.6–1.4)
Female No 24.1 23.2 1.0 2.3 (1.6–3.5)
Yes 5.2 5.2 0.7 (0.4–1.4) 1.7 (0.9–3.3)
LRTmult: P  0.516e
LRTadd: P  0.396f
Cigarette smoking status at reference dateh
Never smoker No 12.0 28.5 1.0 1.0
Yes 2.0 7.0 0.6 (0.2–1.4) 0.6 (0.2–1.4)
Former smoker No 17.0 32.4 1.0 1.2 (0.7–1.8)
Yes 7.4 8.8 1.3 (0.7–2.3) 1.5 (0.8–2.7)
Current smoker No 45.5 14.8 1.0 5.1 (3.3–7.8)
Yes 16.2 8.6 0.6 (0.3–0.9) 2.9 (1.7–5.0)
LRTmult: P  0.07e
LRTadd: P  0.029f
Py of cigarette smokingh
1 No 12.3 30.7 1.0 1.0
Yes 2.2 7.7 0.5 (0.2–1.2) 0.5 (0.2–1.2)
1–19 No 7.9 19.4 1.0 1.0 (0.5–1.9)
Yes 6.2 8.0 1.2 (0.6–2.4) 1.3 (0.7–2.4)
20–29 No 5.4 8.0 1.0 1.5 (0.8–2.7)
Yes 5.4 4.2 1.6 (0.7–3.6) 2.3 (1.2–4.6)
30 No 48.9 17.4 1.0 6.0 (3.8–9.3)
Yes 11.6 4.6 0.7 (0.4–1.2) 4.0 (2.2–7.5)
LRTmult: P  0.117e
LRTadd: P  0.112f
Alcohol consumption (dwk)i
1 No 12.0 23.4 1.0 1.0
Yes 1.5 3.2 0.7 (0.2–1.9) 0.7 (0.2–1.9)
1–14 No 31.9 44.3 1.0 1.0 (0.6–1.5)
Yes 10.3 13.2 1.2 (0.7–1.9) 1.3 (0.8–2.3)
15–28 No 13.0 4.9 1.0 3.1 (1.6–6.0)
Yes 4.7 5.0 0.5 (0.2–1.0) 1.4 (0.7–2.9)
29 No 17.4 2.9 1.0 6.4 (3.2–12.7)
Yes 9.1 2.9 0.6 (0.3–1.4) 3.9 (1.9–8.0)
LRTmult: P  0.73e
LRTadd: P  0.096f
Smoking and drinkingj
20 p-y & 15 d/wk No 18.0 48.0 1.0 1.0
Yes 5.4 11.7 1.1 (0.6–2.0) 1.1 (0.6–2.0)
20 p-y & 15 d/wk No 2.2 2.1 1.0 3.2 (1.3–8.1)
Yes 3.0 3.9 0.7 (0.2–2.1) 2.2 (1.0–4.7)
20 p-y & 15 d/wk No 25.9 19.7 1.0 3.7 (2.5–5.5)
Yes 6.2 4.7 1.0 (0.6–2.0) 3.9 (2.1–7.2)
20 p-y & 15 d/wk No 28.4 5.7 1.0 15.3 (9.3–25.1)
Yes 10.8 4.1 0.5 (0.3–1.0) 8.3 (4.6–14.8)
LRTmult: P  0.326e
LRTadd: P  0.254f
Non-null No 38.5 36.9 1.0 1.0
Yes 9.1 13.1 0.5 (0.3–1.0) 0.5 (0.3–1.0)
Null No 35.9 37.8 1.0 1.0 (0.7–1.4)
Yes 16.4 12.2 1.2 (0.7–2.0) 1.1 (0.7–1.9)
LRTmult: P  0.036e
LRTadd: P  0.053f
Non-null No 58.1 60.3 1.0 1.0
Yes 19.1 20.6 0.7 (0.5–1.2) 0.7 (0.5–1.2)
Null No 16.2 14.3 1.0 1.0 (0.7–1.7)
Yes 6.5 4.8 1.2 (0.6–2.7) 1.3 (0.6–2.6)
LRTmult: P  0.222e
LRTadd: P  0.349f
105Ile/Ile No 32.1 35.0 1.0 1.0
Yes 10.1 11.5 0.8 (0.4–1.4) 0.8 (0.4–1.4)
105Val/Ile or 105Val/Val No 42.0 39.8 1.0 1.1 (0.8–1.6)
Yes 15.8 13.6 1.0 (0.6–1.7) 1.1 (0.6–1.8)
LRTmult: P  0.424e
LRTadd: P  0.419f
a OSCC, oral squamous cell carcinoma; OR, odds ratio; CI, confidence interval; LRT, likelihood ratio test; p-y, pack-years; d/wk, average drinks/week; mult, multiplicative; add, additive.
b OR for ever marijuana use versus never marijuana use, within indicated subgroups. Excludes two cases and two controls with missing data on pack-years cigarette smoking.
c Odds ratio for joint association of ever marijuana use and each characteristic, relative to common reference group. Excludes two cases and two controls with missing data on
pack-years cigarette smoking.
d Adjusted for sex, education, alcohol use (continuous average drinks/week), cigarette smoking (continuous pack-years), and study (first or second).
e P values from likelihood ratio test of hypothesis that the joint association does not depart from multiplicative model.
f P values from likelihood ratio test of hypothesis that the joint association does not depart from additive model.
g Adjusted for age, education, birth year (continuous), alcohol use (continuous average drinks/week), cigarette smoking (continuous pack-years), and study (first or second).
h Adjusted for sex, education, birth year (continuous), alcohol use (continuous average drinks/week), and study (first or second).
i Adjusted for sex, birth year (continuous), cigarette smoking (continuous pack-years), and study (first or second).
j Adjusted for sex, education, birth year (continuous), and study.
k Adjusted for sex, education, birth year (continuous), alcohol use (continuous average drinks/week), cigarette smoking (continuous pack-years), and study (first or second), restricted
to white participants. Analyses stratified by GSTM1, GSTT1, and GSTP1 polymorphisms are based on 340 cases and 548 controls, 339 cases and 547 controls, and 355 cases and 565
controls, respectively. GSTP1 genotypes were in Hardy-Weinberg equilibrium among controls (P  0.06).
21) and antitumor (22, 23) properties. These findings suggest that the
ultimate effect of marijuana use on OSCC development, if any, results
from opposing physiological pathways. Nonetheless, human in vivo
studies of chronic marijuana smokers have found increased premalignant
changes in bronchial tissue (8, 24) akin to those observed in lung
tissue from tobacco smokers.
Reports of young patients with OSCC and other respiratory tract
cancers raised the question of whether marijuana use contributed to
these malignancies (9). These reports lacked comparison groups and
control for established risk factors and thus provide little evidence for
or against the existence of an association. A cohort study did not
provide results specific to OSCC or HNSCC, but for lung cancer and
other smoking-related and/or alcohol-related cancers in general, no
association was found (25). The only epidemiological study of marijuana
use and HSNCC found a 2–3-fold increased risk associated
with ever marijuana use, a dose-response relationship with frequency
and duration of use, and evidence of particularly strong associations
with ever use among several subgroups [e.g., younger individuals,
smokers, and those exhibiting mutagen sensitivity (10)]. Although our
study had a greater proportion of participants who had used marijuana
for 5 years [10.9% versus 3.5% in Zhang et al. (10)], we did not
confirm their findings.
Differences in the extent to which control groups represent the
population from which the cases arise potentially explain the discrepancy
between our null findings and the increased risks observed by
Zhang et al. (10). Blood donors comprise a highly self-selected
population likely to be depleted of individuals with high-risk lifestyle
behaviors; thus, the prevalence of marijuana use by blood donor
controls in Zhang et al. (10) may have been spuriously low in
comparison with what would have been observed in a group of
controls that more closely reflected the source population for their
cases. Alternatively, as Zhang et al. (10) hypothesized, controls may
in general underreport marijuana use to a greater extent than cases.
Zhang et al. (10) used published NHSDA estimates (26) to show that
the observed prevalence of ever marijuana use among their controls
was similar to the prevalence expected in the general population (10).
The published data from which Zhang et al. (10) calculated expected
numbers, however, excluded persons who had initiated marijuana use
after 20 years of age (Table 5 in Ref. 10; Table 3 in Ref. 26), whereas
the observed number of users among their controls included individuals
who had used marijuana regardless of the age at initiation. We
recalculated the sex- and birth cohort-adjusted expected prevalence of
ever marijuana use among controls in Zhang et al. (10) without the
exclusion based on age at initiation (using publicly available NHSDA
data for individuals who were 18 years old in 1992–1994). The
expected number of ever marijuana users was 40.6, whereas only 17
users were observed. If a similar deficit was not present among the
HNSCC cases in Zhang et al. (10), some or all of the 2.6-fold
association with marijuana use they observed would be due to a
spuriously low exposure prevalence among their controls. The results
of similar calculations performed for our control group showed no
difference in the observed and expected number of ever marijuana
Marijuana cigarettes do not contain filters, as tobacco cigarettes do,
and are typically smoked well into the proximal end. Furthermore,
marijuana smokers may inhale more deeply and hold the smoke in
their lungs longer than tobacco smokers (27, 28). These latter characteristics
of marijuana smoking may explain why lung tar levels are
higher for marijuana cigarettes compared with tobacco cigarettes (29).
If marijuana is an oral cavity carcinogen, we nonetheless may not
have observed an association because the amount of marijuana consumed
by a typical user is substantially less than the amount of
tobacco consumed by a typical tobacco smoker (1), and a substantial
proportion of our population were not chronic, long-term users of this
drug. We also did not ask about the number of marijuana cigarettes or
bowls of hashish smoked or the depth of inhalation.
Our study included only OSCC, whereas laryngeal carcinomas
comprised a large proportion (27%) of the cases in Zhang et al. (10).
Marijuana use was reported similarly by their laryngeal cancer cases
(22%) and tongue carcinoma cases (approximately 19%), and we
found no evidence that the marijuana use association varied by OSCC
site. Potential etiologic heterogeneity among HNSCC of different
organs thus does not seem to explain the differences between our
results and those of Zhang et al. (10).
Our study was not without important limitations. We had relatively
low participation, and any association between marijuana use and
participation status that differed between cases and controls could
have biased our results. If an association between marijuana use and
OSCC truly exists, but the prevalence of marijuana use among our
controls is close to that expected (as discussed above), then underrecruitment
of cases who had used marijuana or under-reporting of
marijuana use by cases due to the drug’s illegality could have led to
our null associations. Our study observed well-established associations
with tobacco smoking and alcohol drinking, however, providing
some reassurance that low participation among our cases has not
noticeably affected our study. Our study also did not have data on
mutagen sensitivity status, which Zhang et al. (10) found to be a
strong modifier of the risk associated with marijuana use. Because
Zhang et al. (10) observed an association between HNSCC and
marijuana use even without considering mutagen sensitivity status, we
also should have observed an association if the prevalence of mutagen
sensitivity among our participants was similar to that in Zhang et al.
Studies of HNSCC have not observed consistent associations with
the GSTM1-null, GSTT1-null, or GSTP1 I105V polymorphisms (30–
34), but few have examined whether these polymorphisms modify the
risk associated with exposure to sources of carcinogens that the
enzymes metabolize. We did not find that carriers of the “high-risk”
genotypes of GSTM1, GSTT1, or GSTP1 who also smoked marijuana
were at greater risk than predicted on a multiplicative or additive
scale. Marijuana use was associated with a borderline statistically
significant 50% reduced risk of OSCC among individuals carrying at
least one copy of the GSTM1 gene, statistically distinguishable from
the absence of an association among GSTM1-null homozygotes. The
reduced risk of OSCC among carriers of GSTM1 could represent a
chance finding or could be consistent with induction of detoxifying
GSTM1 activity by marijuana constituents. Although we also observed
statistically distinguishable heterogeneity in the association
with marijuana use among current, former, and never cigarette smokers,
the pattern of ORs did not fit any obvious biologically based
model, nor was similar effect modification seen with pack-years of
cigarette smoking. In general, our study was not sufficiently large to
allow us to reliably assess the potential modifying effects of known or
suspected OSCC risk factors on risks associated with marijuana use.
Although the evidence from nonepidemiological investigations
suggests that marijuana smoking could cause upper respiratory tract
cancer, we did not observe an association with OSCC in this study.
Nonetheless, because our data included relatively few individuals who
had used marijuana for many years, we cannot discount the possibility
that long-term use of this drug is related to OSCC risk. As individuals
born since the 1940s age into their sixth decade of life (when the
baseline rates of OSCC start to rise dramatically), the prevalence of
long-term use marijuana use in the population will increase. This
demographic change will permit future studies to assess more definitively
the role of marijuana in OSCC development.x-Received 11/2/03; revised 3/11/04; accepted 3/23/04.

The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Stephen M. Schwartz, Program in Epidemiology, Fred
Hutchinson Cancer Research Center, M4-C308, P. O. Box 19024, Seattle, WA 98109-
1024. E-mail:

1. Rhodes W, Layne M, Bruen A-M, Johnston P, Becchetti L. (Abt Associates,
Cambridge, MA). What America’s users spend on illegal drugs, 1988–2000. Office
of National Drug Control Policy (US); 2001 Dec. Contact No.: 282-98-0006.
2. Office of Applied Statistics, Substance Abuse and Mental Health Services Administration.
Marijuana use and drug dependence. The NHSDA Report [fact sheet on the
Internet]. 2002 Aug [cited 2004 Apr 20]. Available from: http://www.oas.samhsa.
3. Roth MD, Marques-Magallanes JA, Yuan M, et al. Induction and regulation of the
carcinogen-metabolizing enzyme CYP1A1 by marijuana smoke and delta (9)-tetrahydrocannabinol.
Am J Respir Cell Mol Biol 2001;24:339–44.
4. Wehner FC, van Rensburg SJ, Thiel PG. Mutagenicity of marijuana and Transkei
tobacco smoke condensates in the Salmonella/microsome assay. Mutat Res 1980;77:
5. Busch FW, Seid DA, Wei ET. Mutagenic activity of marijuana smoke condensates.
Cancer Lett 1979;6:319–24.
6. Leuchtenberger C, Leuchtenberger R, Ritter U, Inui N. Effects of marijuana and
tobacco smoke on DNA and chromosomal complement in human lung explants.
Nature (Lond) 1973;242:403–4.
7. Sherman MP, Aeberhard EE, Wong VZ, et al. Effects of smoking marijuana, tobacco
or cocaine alone or in combination on DNA damage in human alveolar macrophages.
Life Sci 1995;56:2201–7.
8. Barsky SH, Roth MD, Kleerup EC, Simmons M, Tashkin DP. Histopathologic and
molecular alterations in bronchial epithelium in habitual smokers of marijuana,
cocaine, and/or tobacco. J Natl Cancer Inst (Bethesda) 1998;90:1198–205.
9. Firth NA. Marijuana use and oral cancer: a review. Oral Oncol 1997;33:398–401.
10. Zhang ZF, Morgenstern H, Spitz MR, et al. Marijuana use and increased risk of
squamous cell carcinoma of the head and neck. Cancer Epidemiol Biomark Prev
11. Maden C, Beckmann AM, Thomas DB, et al. Human papillomaviruses, herpes
simplex viruses, and the risk of oral cancer in men. Am J Epidemiol 1992;135:1093–
12. Schwartz SM, Daling JR, Doody DR, et al. Oral cancer risk in relation to sexual
history and evidence of human papillomavirus infection. J Natl Cancer Inst (Bethesda)
13. Schwartz SM, Doody DR, Fitzgibbons ED, Ricks S, Porter PL, Chen C. Oral
squamous cell cancer risk in relation to alcohol consumption and alcohol dehydrogenase-
3 genotypes. Cancer Epidemiol Biomark Prev 2001;10:1137–44.
14. Hankey BF, Ries LA, Edwards BK. The Surveillance, Epidemiology, and End Results
program: a national resource. Cancer Epidemiol Biomark Prev 1999;8:1117–21.
15. Crump C, Chen C, Appelbaum FR, et al. Glutathione S-transferase theta 1 gene
deletion and risk of acute myeloid leukemia. Cancer Epidemiol Biomark Prev
16. Chen C, Madeleine MM, Weiss NS, Daling JR. Glutathione S-transferase M1
genotypes and the risk of vulvar cancer: a population-based case-control study. Am J
Epidemiol 1999;150:437–42.
17. Thompson WD. Statistical analysis of case-control studies. Epidemiol Rev 1994;16:
18. Dubin N, Pasternack BS. Risk assessment for case-control subgroups by polychotomous
logistic regression. Am J Epidemiol 1986;123:1101–17.
19. Substance Abuse and Mental Hygiene Data Archive [database on the Internet].
Rockville (MD): Substance Abuse and Mental Health Services Administration
(US) [cited 2004 Apr 20]. Available from:
20. Zhu LX, Sharma S, Stolina M, et al. Delta-9-tetrahydrocannabinol inhibits antitumor
immunity by a CB2 receptor-mediated, cytokine-dependent pathway. J Immunol
21. Sarafian TA, Tashkin DP, Roth MD. Marijuana smoke and delta(9)-tetrahydrocannabinol
promote necrotic cell death but inhibit Fas-mediated apoptosis. Toxicol Appl
Pharmacol 2001;174:264–72.
22. Guzman M. Cannabinoids: potential anticancer agents. Nat Rev Cancer 2003;3:
23. Blazquez C, Casanova ML, Planas A, et al. Inhibition of tumor angiogenesis by
cannabinoids. FASEB J 2003;17:529–31.
24. Fligiel SE, Venkat H, Gong H Jr, Tashkin DP. Bronchial pathology in chronic
marijuana smokers: a light and electron microscopic study. J Psychoactive Drugs
25. Sidney S, Quesenberry CP Jr, Friedman GD, Tekawa IS. Marijuana use and cancer
incidence (California, United States). Cancer Causes Control 1997;8:722–8.
26. Johnson RA, Gerstein DR. Initiation of use of alcohol, cigarettes, marijuana, cocaine,
and other substances in US birth cohorts since 1919. Am J Public Health 1998;88:
27. Tashkin DP, Gliederer F, Rose J, et al. Delta 9THC delivery from the 1st and 2nd
halves of a marijuana cigarette. Pharmacol Biochem Behav 1991;40:657–61.
28. Tashkin DP, Gliederer F, Rose J, et al. Effects of varying marijuana smoking profile
on deposition of tar and absorption of CO and delta-9-THC. Pharmacol Biochem
Behav 1991;40:651–6.
29. Wu TC, Tashkin DP, Djahed B, Rose JE. Pulmonary hazards of smoking marijuana
as compared with tobacco. N Engl J Med 1988;318:347–51.
30. Geisler SA, Olshan AF. GSTM1, GSTT1, and the risk of squamous cell carcinoma of
the head and neck: a mini-HuGE review. Am J Epidemiol 2001;154:96–105.
31. Sreelekha TT, Ramadas K, Pandey M, et al. Genetic polymorphism of CYP1A1,
GSTM1 and GSTT1 genes in Indian oral cancer. Oral Oncol 2001;37:593–8.
32. Buch SC, Notani PN, Bhisey RA. Polymorphism at GSTM1, GSTM3 and GSTT1
gene loci and susceptibility to oral cancer in an Indian population. Carcinogenesis
(Lond) 2002;23:803–7.
33. Hahn M, Hagedorn G, Kuhlisch E, Schackert HK, Eckelt U. Genetic polymorphisms
of drug-metabolizing enzymes and susceptibility to oral cavity cancer. Oral Oncol
34. Lazarus P, Park JY. Metabolizing enzyme genotype and risk for upper aerodigestive
tract cancer. Oral Oncol 2000;36:421–31.
35. Hayes JD, Pulford DJ. The glutathione S-transferase supergene family: regulation of
GST and the contribution of the isoenzymes to cancer chemoprotection and drug
resistance. Crit Rev Biochem Mol Biol 1995;30:445–600.



Top      Home


A Population-Based Case-Control Study of Marijuana Use and Head and Neck Squamous Cell

Cancer Prev Res (Phila). 2009 Aug;2(8):759-68. Epub 2009 Jul 28

Liang C, McClean MD, Marsit C, Christensen B, Peters E, Nelson HH, Kelsey KT

Department of Community Health, Department of Pathology and Laboratory Medicine, Division of Biology and Medicine, Brown University, Providence, RI, USA.


Cannabinoids, constituents of marijuana smoke, have been recognized to have potential antitumor properties. However, the epidemiologic evidence addressing the relationship between marijuana use and the induction of head and neck squamous cell carcinoma (HNSCC) is inconsistent and conflicting. Cases (n = 434) were patients with incident HNSCC disease from nine medical facilities in the Greater Boston, MA area between December 1999 and December 2003. Controls (n = 547) were frequency matched to cases on age (+/-3 years), gender, and town of residence, randomly selected from Massachusetts town books.

A questionnaire was adopted to collect information on lifetime marijuana use (decade-specific exposures) and associations evaluated using unconditional logistic regression. After adjusting for potential confounders (including smoking and alcohol drinking), 10 to 20 years of marijuana use was associated with a significantly reduced risk of HNSCC [odds ratio (OR)(10-<20 years versus never users), 0.38; 95% confidence interval (CI), 0.22-0.67].

Among marijuana users moderate weekly use was associated with reduced risk (OR(0.5-<1.5 times versus <0.5 time), 0.52; 95% CI, 0.32-0.85). The magnitude of reduced risk was more pronounced for those who started use at an older age (OR(15-<20 years versus never users), 0.53; 95% CI, 0.30-0.95; OR(> or =20 years versus never users), 0.39; 95% CI, 0.17-0.90; P(trend) < 0.001). These inverse associations did not depend on human papillomavirus 16 antibody status. However, for the subjects who have the same level of smoking or alcohol drinking, we observed attenuated risk of HNSCC among those who use marijuana compared with those who do not. Our study suggests that moderate marijuana use is associated with reduced risk of HNSCC.



Top      Home

Smoking of cannabis does not increase risk for oral cancer


Contrary to previous research findings cannabis smoking does not appear to be associated with an increased risk of developing oral cancer, according to a large study from the USA. The results appeared in the June issue of Cancer Research. The study of Dr. Karin A. Rosenblatt and colleagues found no association between cannabis use and increased oral-cancer risk, regardless of how long, how much or how often a person has used cannabis.

"When asking whether any marijuana use puts you at increased risk of oral cancer, our study is pretty solid in saying there's nothing going on there," said Dr. Stephen M. Schwartz, one of the authors of the study. But because the incidence of extensive, long-term cannabis use was low among the study population it is unclear whether extremely heavy use over many years is related to the risk for oral cancer, Schwartz said. The study involved 407 oral-cancer cases and 615 healthy control subjects from the state of Washington who had been interviewed in detail about their history of marijuana use, among other lifestyle factors. Participants ranged in age from 18 to 65.

The new study counters findings from a smaller investigation with 173 oral cancer patients, published in 1999, which suggested that ever-users of marijuana were at more than twice the risk of getting head-and-neck cancer as compared to non-users. The earlier study had a number of limitations, most importantly the fact that its control, or comparison, group was comprised of individuals who had donated blood at the same hospital where the oral-cancer cases had been treated. Blood donors tend to have fewer high-risk habits than the general population.

(Sources: Rosenblatt KA, et al. Marijuana use and risk of oral squamous cell carcinoma. Cancer Res 2004;64:4049-54. Press release of the Fred Hutchinson Cancer Research Center of 1 June 2004)


Top      Home