Marijuana use and testicular germ cell tumors
Since the early 1970s, the incidence of testicular germ cell tumors (TGCTs) in the United States has been increasing; however, potential environmental exposures accounting for this increase have not been identified. A previous study reported a significant association between frequent and long-term current marijuana users and TGCT risk. The purpose of this study was to evaluate the relation between marijuana use and TGCTs in a hospital-based case-control study.
TGCT patients diagnosed between January 1990 and October 1996 (n = 187) and male friend controls (n = 148) were enrolled in the study. All participants were between the ages of 18 and 50 at the time of diagnosis and resided in Texas, Louisiana, Arkansas, or Oklahoma. Associations of marijuana use and TGCTs were estimated using unconditional logistic regression, adjusting for age, race, prior cryptorchidism, cigarette smoking, and alcohol intake.
Overall, patients with TGCTs were more likely to be frequent marijuana users (daily or greater) compared with controls (odds ratio [OR], 2.2; 95% confidence interval [CI], 1.0-5.1). Histological-specific analyses revealed that patients with nonseminoma were significantly more likely than controls to be frequent users (OR, 3.1; 95% CI, 1.2-8.2) and long-term (≥10 years) users (OR, 2.4; 95% CI, 1.0-6.1).
The finding of an association between frequent marijuana use and TGCTs, particularly among men with nonseminoma, was consistent with the findings of a previous report. Additional studies of marijuana use and TGCTs are warranted, especially studies evaluating the role of endocannabinoid signaling and cannabinoid receptors in TGCTs. Cancer 2011. © 2010 American Cancer Society.
Testicular germ cell tumors (TGCTs) are relatively rare malignancies, accounting for <2% of cancers in males; however, they are the most common malignant neoplasm occurring in males ages 15-44 in many countries, including the United States.1 Based on data from 2002-2006, TGCT incidence rates among young, white men in the United States were 6.3 per 100,000.2 In contrast, the incidence rates among black and Asian men range from 1 to 2 per 100,000.2 The highest incidence rates in the world occur in populations of northern European ancestry, regardless of their country of residence.3 In the United States, the incidence of TGCTs increased in white American men by 70.2% between 1975 and 2005.2 Similar increases in TGCT incidence have been observed among men of European heritage in Canada, many European countries, New Zealand, and Australia.3-7 Such an increase in incidence over a relatively short interval suggests the influence of environmental rather than genetic factors. TGCTs are hypothesized to develop as a result of the neoplastic transformation of germ cells into testicular carcinoma in situ. It is believed that these neoplasms arise early in fetal life and progress to invasive cancer under the influence of adult gonadotropic and androgenic hormones.8, 9
There are few established risk factors for TGCTs beyond age, race, history of cryptorchidism, and family history of TGCTs. Recently, marijuana use has been evaluated as a risk factor.10 Daling et al10 reported an increased risk of TGCTs with current marijuana use, specifically among frequent (≥1/wk) and long-term users (≥10 years). The exact mechanism by which heavy marijuana use might increase the risk of TGCTs is unknown; however, chronic marijuana exposure has multiple adverse effects on the endocrine and reproductive systems, including gynecomastia, impotence, reduced sperm counts, and suppressed testosterone.7, 11-13 We explored the relation between marijuana use and TGCT in an existing hospital-based case-control study conducted at The University of Texas M. D. Anderson Cancer Center. Specifically, we evaluated whether TGCT risk increased with increasing frequency and duration of marijuana use. We also assessed whether risk estimates for marijuana use varied by TGCT histology.
MATERIALS AND METHODS
TGCT patients and friend controls were enrolled into a hospital-based case-control study conducted at The University of Texas M. D. Anderson Cancer Center. Full details of the study design have been reported.14-16 Briefly, patients were men with incident primary TGCTs registered at The University of Texas M. D. Anderson Cancer Center between January 1990 and October 1996 through the Genitourinary Oncology clinic or who had been previously treated and were recruited from The University of Texas M. D. Anderson Tumor Registry (January 1990 to August 1994). To assemble a control population, patients were asked to provide the name of at least one adult male friend of similar age and race. All participants were between the ages of 18 and 50 at the time of diagnosis and resided in Texas, Louisiana, Arkansas, or Oklahoma.
All patients diagnosed with TGCTs during the given period were eligible for inclusion regardless of ethnicity, tumor stage, or tumor histology. Because it is not clear whether gonadal and extragonadal germ cell tumors share common etiologies, we excluded patients diagnosed with extragonadal germ cell tumors to avoid introducing heterogeneity into the case group. Pathology reports were reviewed for all cases, and tumors were grouped as pure seminomas, nonseminomas (teratoma, embryonal carcinoma, and choriocarcinoma), and mixed germ cell tumors (both seminomatous and nonseminomatous elements).
Patients and controls completed a self-administered questionnaire ascertaining demographics, lifestyle habits, medical history, and diet. Information on drug use, including marijuana, was also collected. The questions pertaining to drug use appeared on a detachable throwaway coversheet to protect each participant's privacy. Participants were asked if they had ever used marijuana, and if so, to indicate the number of times used per week and the calendar years during which they used marijuana.
Analyses were conducted for all cases combined and for cases classified by histological type (seminomas, nonseminomas, and mixed germ cell tumors). Using the data collected on number of times per week and years of marijuana use, we created 3 analytic variables: ever used marijuana, frequency of marijuana use per week, and duration (years) of marijuana use. Study participants who self-reported never smoking marijuana served as the reference category for the marijuana use variables. Ever-use was defined as a self-report of smoking marijuana regardless of frequency or duration of use. The ever-users were then stratified by frequency and duration of marijuana use. Frequency of use was categorized as marijuana use less than once per day or at least once per day (daily or >1/d). Duration of use was categorized as <10 years of lifetime marijuana use or ≥10 years of lifetime marijuana use.
Several cases did not have a matched control because a name was not provided or the individual named was unwilling to participate. To avoid the loss of information by excluding unmatched cases, we evaluated the results using a matched analysis (conditional logistic regression) and an unmatched analysis (unconditional logistic regression adjusting for age and race). The point estimates for the 2 models were not substantially different (data not shown) and did not change the interpretation of the results; therefore, we have presented the results from the unmatched analysis. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated as estimates of relative risk using unconditional logistic regression. Polytomous logistic regression models were used to compare controls with each of the case groups defined by histological type. All analyses were adjusted for age at diagnosis for cases and time of interview for controls, race, and history of cryptorchidism. We also included 2 additional behaviors that may be correlated with marijuana use: cigarette smoking and alcohol consumption. Other covariates considered were education and income; however, because the addition of these 2 variables to a model controlling for age, history of cryptorchidism, alcohol consumption, and cigarette smoking did not substantially change the ORs (<10% change in the risk estimate), they were not included in the final analysis. To evaluate the robustness of our results, we conducted sensitivity analyses excluding the 60 study participants who reported very infrequent marijuana use (<1/wk). All analyses were performed with Stata/SE (Stata Statistical Software, version 10.1; StataCorp, College Station, TX).
The distribution of selected demographic and health characteristics for the patients with TGCTs and controls are provided in Table 1. The median age for controls and all cases combined was similar; the median age for nonseminoma cases was approximately 10 years younger than that of seminoma cases. Compared with controls, patients tended to report somewhat lower annual incomes and were less likely to have more than a high school education. Patients were more likely than controls to have a history of cryptorchidism (13.4% vs 2.0 %, respectively; age- and race-adjusted OR, 7.8; 95% CI, 2.3-26.5).
Table 1. Selected Characteristics of Controls and Testicular Germ Cell Tumor Cases by Histology: The University of Texas M. D. Anderson Cancer Center, January 1990 to October 1996
|Age in 1996, median [range]||34 [18-63]||33 [18-57]||38.5 [24-57]||29 [18-51]||33.5 [20-51]|
|Race, no. (%)|
| White||130 (87.8)||151 (80.8)||41 (82.0)||75 (79.0)||35 (83.3)|
| Hispanic||14 (9.5)||28 (15.0)||7 (14.0)||13 (13.7)||7 (16.7)|
| Other||4 (2.7)||8 (4.2)||2 (4.0)||7 (7.4)||0 (0.0)|
|Cigarette smoking, no. (%)|
| Never||88 (59.5)||99 (52.9)||25 (50.0)||49 (51.6)||25 (59.5)|
| Former (quit >6 mo)||37 (25.0)||59 (31.6)||16 (32.0)||32 (33.7)||11 (26.2)|
| Current||23 (15.5)||29 (15.5)||9 (18.0)||14 (14.7)||6 (14.3)|
|Alcohol consumption, no. (%)|
| Never||14 (9.5)||23 (12.3)||6 (12.0)||10 (10.5)||7 (16.7)|
| Former (quit >6 mo)||37 (25.0)||53 (28.3)||15 (30.0)||29 (30.5)||9 (21.4)|
| Current||97 (65.5)||111 (59.4)||29 (58.0)||56 (59.0)||26 (61.9)|
|History of cryptorchidism, no. (%)a|
| Yes||3 (2.0)||25 (13.4)||7 (14.0)||12 (12.6)||6 (14.3)|
| No||145 (98.0)||162 (86.6)||43 (86.0)||83 (87.4)||36 (85.7)|
|Income, no. (%)a|
| <$25,000||20 (14.3)||55 (31.6)||11 (23.4)||30 (34.1)||14 (35.9)|
| $25,000-$54,999||43 (30.7)||52 (29.9)||12 (25.5)||34 (38.6)||6 (15.4)|
| $55,000-$74,999||36 (25.7)||29 (16.7)||8 (17.0)||13 (14.8)||8 (20.5)|
| >$75,000||41 (29.3)||38 (21.8)||16 (34.0)||11 (12.5)||11 (28.2)|
|Years of education, no. (%)a|
| ≤12||31 (21.0)||58 (31.0)||12 (24.0)||30 (31.6)||16 (38.1)|
| >12||117 (79.1)||129 (69.0)||38 (76.0)||65 (68.4)||26 (61.9)|
Overall, the frequency of ever having used marijuana was similar between cases and controls (OR, 0.7; 95% CI, 0.4-1.1) (Table 2). TGCT patients were more likely to be frequent marijuana users (daily or >1 per day) than controls (OR, 2.2; 95% CI, 1.0-5.1) and were less likely than controls to report infrequent (<1/d [OR, 0.5; 95% CI, 0.3-0.9]) and short-term (<10 years [OR, 0.6; 95% CI, 0.3-1.0]) marijuana use. Approximately three-quarters (75.5%) of the individuals who reported ever using marijuana initiated use at age 18 or younger (data not shown). Furthermore, we evaluated the intensity of marijuana use (frequency and duration) with TGCT risk and found that the results were consistent with a doubling of risk with very frequent and long-term use but were not statistically significant due to the small numbers within these categories (data not shown).
Table 2. Associations of Testicular Germ Cell Tumors and Marijuana Use According to Histology: The University of Texas M. D. Anderson Cancer Center, January 1990 to October 1996
|Ever used marijuana|
| No||66 (45.2)||96 (51.3)||1.0 (Referent)||30 (60.0)||1.0 (Referent)||44 (46.3)||1.0 (Referent)||22 (52.4)||1.0 (Referent)|
| Yes||80 (54.8)||91 (48.7)||0.7 (0.4-1.1)||20 (40.0)||0.5 (0.3-1.1)||51 (53.7)||0.8 (0.4-1.4)||20 (47.6)||0.8 (0.3-1.6)|
|Frequency of marijuana use|
| Never||66 (46.8)||96 (52.2)||1.0 (Referent)||30 (61.2)||1.0 (Referent)||44 (46.8)||1.0 (Referent)||22 (53.7)||1.0 (Referent)|
| <1/d||65 (46.1)||54 (29.4)||0.5 (0.3-0.9)||12 (24.5)||0.4 (0.2-0.9)||29 (30.9)||0.6 (0.3-1.1)||13 (31.7)||0.6 (0.3-1.4)|
| Daily or >1/d||10 (7.1)||34 (18.5)||2.2 (1.0-5.1)||7 (14.3)||1.4 (0.4-4.6)||21 (22.3)||3.1 (1.2-8.2)||6 (14.6)||2.0 (0.6-7.0)|
|Duration of marijuana use|
| Never||66 (46.8)||96 (52.8)||1.0 (Referent)||30 (61.2)||1.0 (Referent)||44 (47.3)||1.0 (Referent)||22 (55.0)||1.0 (Referent)|
| <10 y||59 (41.8)||57 (31.3)||0.6 (0.3-1.0)||14 (28.6)||0.5 (0.2-1.2)||29 (31.2)||0.6 (0.3-1.1)||14 (35.0)||0.7 (0.3-1.6)|
| ≥10 y||16 (11.4)||29 (15.9)||1.2 (0.6-2.8)||5 (10.2)||0.6 (0.2-1.9)||20 (21.5)||2.4 (1.0-6.1)||4 (10.0)||0.9 (0.2-3.5)|
In the analyses by histological type, patients with nonseminoma were significantly more likely to be frequent users than controls (daily or >1/d; OR, 3.1; 95% CI, 1.2-8.2) and long-term users (≥10 years; OR, 2.4; 95% CI, 1.0-6.1). In contrast, in the analysis limited to seminomas, cases were significantly less likely than controls to be infrequent users (OR, 0.4; 95% CI, 0.2-0.9), and there was no significant association with duration of use.
The results of sensitivity analyses, excluding study participants reporting very infrequent marijuana use, support the relation between frequent marijuana use and all TGCTs (data not shown). Sensitivity analyses also supported the association between frequent and long-term use among patients with nonseminomas (data not shown).
We observed a 2-fold increased risk of TGCTs associated with frequent marijuana use. In histological-specific analyses, the associations were limited to nonseminomas. These associations were independent of known risk factors for TGCTs: age, race, and prior cryptorchidism, as well as cigarette smoking and alcohol consumption. We also observed reduced risk of TGCTs associated with infrequent and short-term use; however, these findings were not robust to sensitivity analyses and require further evaluation.
There is limited evidence that marijuana use may modulate TGCT risk. The only other published study to date examined the association of marijuana use and TGCTs in a population-based case-control study in western Washington.10 Daling and colleagues reported that men with TGCTs were significantly more likely than controls to be current marijuana users. The association was particularly strong among men who were frequent current users (at least weekly) or current users of long duration (≥10 years). In addition, the associations appeared to be limited to cases with nonseminoma or mixed germ cell tumors. Unlike the study by Daling et al, the results of our study are not supportive of an association between weekly marijuana use and TGCTs; instead, our results are supportive of a substantially increased risk of TGCTs with very frequent marijuana use (daily or greater). In addition, although we were not able to evaluate current use with our data (<10% of our study population reported current marijuana use), our findings are supportive of the association between frequent and long-term marijuana use and nonseminoma.
Puberty may be a period of development during which environmental factors increase the risk of TGCTs.17 Exposures during this time may be more relevant to nonseminoma risk, given that the peak occurrence of nonseminoma occurs 10 years earlier than seminoma, and within a biologically relevant period after puberty. It is plausible that the use of marijuana during puberty perturbs the hypothalamic-pituitary-gonadal axis leading to altered levels of pituitary gonadotropins (follicle-stimulating hormone and luteinizing hormone) and sex-steroid hormones and potentially increased risk of nonseminoma.
Studies have demonstrated that acute or chronic treatment of cannabis extract in male mice and tetrahydrocannabinol, the active ingredient in marijuana, in male rats can act centrally to affect circulating levels of testosterone, follicle-stimulating hormone, and luteinizing hormone.13, 18-20 Two cannabinoid receptor subtypes exist in humans: the brain-type receptors (CB1) and the spleen-type receptors (CB2).21-23 These two subtypes are part of the G protein-coupled receptor family and influence a variety of biological responses. CB1 and CB2 are expressed in the testes and sperm as well as in the brain, heart, uterus, embryo, spleen, and immune cells.19 Laboratory evidence demonstrates that cannabis and cannabis-like compounds target cannabinoid receptors in Leydig and Sertoli cells and influence testosterone and pituitary gonadotropin release as well as Sertoli cell survival. Studies have shown that endogenous cannabinoid-like (endocannabinoid) lipid mediators (anandamide, specifically) suppressed luteinizing hormone and testosterone levels in wild-type but not CB1 knockout mice, providing evidence that the endocannabinoid system acts to alter testosterone and pituitary gonadotropin concentrations.24 Furthermore, there is evidence that cannabinoids can inhibit testosterone activity by impairing androgen binding to receptors.25
Although our results support the findings by Daling et al,10 there are several limitations to our study. The study relied on the self-reported use of marijuana, which is an illicit drug. Individuals with a serious disease, such as cancer, may more accurately report the use of an illegal substance than individuals without a serious medical condition. To address this concern, we compared marijuana use in controls with publicly available national data and found no significant difference. The specificity of our finding—that the association between marijuana use and TGCTs was primarily limited to nonseminoma—may be due to the limited numbers of seminoma and mixed germ cell tumors.
Use of friend controls as the referent group in any study raises the concern that controls are too similar to cases. If the controls, in fact, were too similar to cases in terms of their marijuana use patterns, then it is likely that our estimates of risk would have underestimated the true relation between marijuana use and TGCT. In our study, however, the controls were older and reported higher incomes than the patients, suggesting that overmatching was not present, at least for these factors. While we attempted to match on age, the cases tended to nominate friend controls who were generally older than they were, so we adjusted for age in all of our analyses. We also evaluated income as a potential confounding factor and found that it had no effect in a model adjusting for age, race, history of cryptorchidism, cigarette smoking and alcohol consumption. Thus, income was not included in the final models.
Because the use of friend controls can bias the results of retrospective studies, we evaluated the extent to which reporting of marijuana use among controls was consistent with data from the 1996 population-based National Survey on Drug Use and Health (NSDUH; known as the National Household Survey on Drug Abuse prior to 1999). We compared the observed number of controls who reported ever using marijuana with the expected number based on age- and race-specific proportions in the survey data26; we did not find a significant difference, suggesting that selection or reporting bias was not an explanation for the observed associations (among 128 white men of all ages in the control group, 68 reported any marijuana use compared with 70.9 expected based on NSDUH data [observed-to-expected ratio, 0.96; 95% CI, 0.74-1.22]). The NSDUH did not capture information on lifetime frequency of use; however, it did collect information on current frequency of use. Using the NSDUH variable on current frequency of use, we did not find a significant observed-to-expected difference when we compared the observed number of controls who reported daily marijuana use with the expected number based on age- and race- specific proportions of current daily use (among 123 white men of all ages in the control group, 32 reported any marijuana use compared with 32.5 expected based on NSDUH data [observed-to-expected ratio, 0.98; 95% CI, 0.67-1.3]).
Our finding of an association between frequent marijuana use and TGCT, particularly among men with nonseminoma, is consistent with the findings of a previous report.10 The biologically active components of marijuana may directly affect TGCT risk by altering gonadotropin and hormone levels during puberty; however, these components may function through pathways other than the endocannabinoid system. Additional studies of marijuana use and TGCTs are warranted, especially studies evaluating the role of endocannabinoid signaling and cannabinoid receptors in TGCTs.
CONFLICT OF INTEREST DISCLOSURES
Supported by The University of Texas M. D. Anderson Cancer Center Education Program in Cancer Prevention (National Cancer Institute grant R25-CA-57730) and the Intramural Research Program of the National Cancer Institute.