[Cancer Research 66, 4512-4515, April 15, 2006]
© 2006 American Association for Cancer Research
Epidemiology and Prevention |
Duration but not Intensity of Alcohol and Tobacco Exposure Predicts p16INK4A Homozygous Deletion in Head and Neck Squamous Cell Carcinoma
Kim S. Kraunz1,
Michael D. McClean3,
Heather H. Nelson2,
Edward Peters4,
Henry Calderon1 and
Karl T. Kelsey1
Departments of 1 Genetics and Complex Diseases and 2 Environmental Health, Harvard School of Public Health; 3 Department of Environmental Health, Boston University School of Public Health, Boston, Massachusetts; and 4 Louisiana State University Health Sciences Center, School of Public Health, New Orleans, Louisiana
Requests for reprints: Karl T. Kelsey, Department of Genetics and Complex Diseases, Harvard School of Public Health, Room 607, Building I, 665 Huntington Avenue, Boston, MA 02115. Phone: 617-432-3313; Fax: 617-432-0107; E-mail: kelsey{at}hsph.harvard.edu.
 |
Abstract
|
|---|
In tobacco-associated solid tumors, evidence suggests that the pattern of carcinogen exposure is related to the nature of somatic gene inactivation within crucial pathways, including the retinoblastoma (Rb) pathway. One somatic event in this pathway, homozygous deletion of the p16INK4A gene, is commonly observed in head and neck squamous cell carcinoma (HNSCC). Alcohol and tobacco are both well-established risk factors for HNSCC but there has been little characterization of the relationship of exposure to these carcinogens and inactivation of the p16INK4A gene. Hypothesizing that p16INK4A homozygous deletion is associated with tobacco and alcohol exposure, we investigated 330 consecutive HNSCC tumors. The odds ratio (OR) for p16INK4A homozygous deletion among alcohol consumers in the upper tertile (>43 years used) was 5.2 [95% confidence interval (95% CI), 2.1-12.8] as compared with those with 
43 years of alcohol consumption. Intensity of alcohol exposure, measured as average alcoholic drinks per week, was not associated with gene deletion. When we examined the distribution of duration of tobacco use, the OR for p16INK4A homozygous deletion was 1.3 (95% CI, 0.5-3.0) and 1.9 (95% CI, 0.9-4.0) for 29 to 39 years and >39 years of tobacco smoking, respectively, as compared with those that smoked 28 years. As in the case of alcohol use, intensity of tobacco exposure (measured as packs per day) was not associated with gene deletion. Hence, the duration of alcohol use and duration of smoking, but not intensity of either, significantly predicted p16INK4A homozygous deletion in HNSCC. (Cancer Res 2006; 66(8): 4512-5)
|
Introduction
|
|---|
Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide and the tenth most common cancer in the United States (1, 2). It consists of cancers of epithelial origin of the upper aerodigestive tract, including the oral cavity. Despite the diverse anatomy and mixture of tissues represented by the head and neck region, all HNSCCs share common risk factors and are similar in their epidemiology, treatment, and prognosis and, hence, are often studied together (3). Tobacco usage and alcohol consumption are the main risk factors for developing HNSCC in the United States. Furthermore, the combination of alcohol and tobacco exposure synergistically enhances the risk of HNSCC (4).
There is mounting evidence that the pattern of somatic alterations that result in clonal selection in carcinogenesis reflects the nature of carcinogen exposure. In non–small-cell lung cancer, another tobacco-related disease, the occurrence of p16INK4A methylation is associated with increased tobacco exposure in a dose response–related fashion (5, 6). In HNSCC, loss of heterozygosity (LOH) of the whole arm of chromosome 3p is associated with increased tobacco and alcohol exposure (7) and LOH at chromosome 22q13 has been associated with increased alcohol exposure (8). However, the relationship between p16INK4A homozygous deletion and carcinogen exposure in HNSCC remains largely unexplored. Inactivation of the gatekeeper Rb pathway by homozygous deletion of the p16INK4A gene occurs frequently in HNSCC (9–12). Because the Rb pathway is inactivated in the majority of HNSCCs (11, 13), it provides a model amenable to examining the role that tobacco and alcohol exposure play in selecting for deletion of p16INK4A.
We have tested the hypothesis that p16INK4A homozygous deletion in HNSCC is associated with carcinogen exposure, determining the deletion frequency in a series of 330 cases.
|
Materials and Methods
|
|---|
HNSCC case series study. We examined 330 consecutive primary HNSCC from subjects who were enrolled in a population-based case-control study. Patients were identified in concert with the Rapid Case Identification Core of the Harvard Comprehensive Cancer Center. The overall response rate for the study was 83%. Paraffin specimens were collected from patients who underwent surgical resection at the Dana-Farber Cancer Institute, Brigham & Women's Hospital, Beth Israel Deaconess Hospital, Mass General Hospital, Mass Eye and Ear Infirmary, Boston Medical Center, Boston Veterans Administration Medical Center, and New England Medical Center from December 1999 through December 2004. Specimens were pathologically reviewed for selection of paraffin samples containing primarily tumor. Information about smoking status, years smoked, packs consumed per day, alcoholic drinks consumed per week, and other sociodemographic characteristics such as age and gender were obtained by completion of a detailed interviewer-administered questionnaire. Alcohol and tobacco usage patterns were assessed (separately), such that we addressed usage over each individual's lifetime. For those that were not able to complete the long version of the questionnaire (n = 258), a shorter version of the questionnaire was administered that captured alcohol and tobacco usage history without the lifetime specific exposure data (n = 25). A sensitivity analysis showed response to these questions not to differ based on data garnered from the long or short version of the questionnaire. For those for whom neither the detailed nor shorter questionnaire was available (n = 47), smoking and drinking data were collected from medical charts.
Extraction of DNA from paraffin-embedded samples. Three 20-µm sections were cut from each fixed, paraffin-embedded tumor sample and transferred into microcentrifuge tubes. The paraffin was dissolved using Histochoice Clearing Agent (Sigma-Aldrich, St. Louis, MO) followed by two washes with 100% ethanol and one wash with PBS. The samples were then incubated in SDS-lysis solution [50 mmol/L Tris-HCl (pH 8.1), 10 mmol/L EDTA, 1% SDS] with proteinase K (Qiagen, Valencia, CA) overnight at 55°C. De-crosslinking was done by adding NaCl (final concentration, 0.7 mol/L) and incubating at 65°C for 4 hours. DNA was recovered using the Wizard DNA clean-up kit (Promega, Madison, WI) according to the protocols of the manufacturer.
p16INK4A homozygous deletion. Homozygous gene deletion of p16INK4A was determined using a TaqMan-based real-time PCR method as previously described (14). Briefly, exon 2 of p16INK4A and the ß-actin gene were amplified in a multiplex assay. The primer sequences for detecting p16INK4A exon 2 were 5'-GGCTCTACACAAGCTTCCTTTCC-3' (sense) and 5'-TCATGACCTGCCAGAGAGAACA-3' (antisense). The primer sequences for detecting ß-actin were 5'-AGCGCGGCTACAGCTTCA-3' (sense) and 5'-CGTAGCACAGCTTCTCCTTAATGTC-3' (antisense). The probe sequence for p16INK4A was 5'-CCCCCACCCTGGCTCTGACCA-3' and was labeled with FAM, whereas the probe sequence for ß-actin was 5'-ATTTCCCGCTCGGCCGTGGT-3' and was labeled with VIC (both probes manufactured by Applied Biosystems, Foster City, CA). The reaction was done in a final volume of 50 µL. The final concentrations of primers and probes were as follows: p16INK4A primers, 50 ng/µL; p16INK4A probe, 200 nmol/L; ß-actin primers F, 200 ng/µL; and ß-actin probe, 200 nmol/L. Each reaction contained 50 ng of DNA as template and the TaqMan universal master mix (Applied Biosystems) was used. The thermal cycling conditions of the ABI PRISM 7000 instrument were set to 2 minutes at 50°C, 10 minutes at 95°C followed by 40 cycles of 15 seconds at 95°C alternating with 1 minute at 60°C. The efficiency of the multiplex reaction was verified, testing amplification of different concentrations of DNA (20-250 ng). DNA from H460 cells that have homozygous deletion at p16INK4A was used as a negative control and DNA from lymphocytes isolated from blood was used as a positive control. All reactions were done in duplicate.
Statistical analysis. Data were analyzed using SAS statistical software. The 
2 test was used for the unadjusted univariate analysis. The categories for the exposure variables were based on the distribution among cases, with the exception of packs per day, which was categorized by less than one pack per day, one pack per day, or more than one pack per day. To consider the effect of multiple variables on p16INK4A homozygous deletion status, logistic regression was used to estimate odds ratios (OR). Model selection was dependent on overall significance of the model (
= 0.05) using the log likelihood test and inclusion of known risk factors and potential confounders. Cigarette use was modeled using duration of use (years) and intensity (packs per day). Alcohol use was modeled using duration of use (years) and intensity (average drinks per week). Potential confounders included age, gender, and tumor location.
|
Results
|
|---|
Demographic and exposure characteristics and p16INK4A homozygous deletion. We determined the prevalence of p16INK4A homozygous deletion in 330 HNSCC cases. Among the 330 tumors, 40% (132 of 330) were deleted at this locus. Table 1
summarizes the demographic and exposure characteristics of the population by p16INK4A deletion status. There was no statistically significant difference in p16INK4A homozygous deletion by age, gender, or tumor site. When alcohol consumption was evaluated by comparing ever drinkers to never drinkers, there was no significant difference in the occurrence of deletion. However, when alcohol consumption was evaluated by duration (years) of drinking, the duration of alcohol consumption was found to be significantly higher among patients with homozygous p16INK4A deletion (Table 1; P = 0.003). Interestingly, there was no association of intensity of alcohol use (measured as the average alcoholic drinks per week) with p16INK4A loss. In our initial unadjusted analysis, there was no statistically significant difference in p16INK4A homozygous deletion by any measure of tobacco usage (i.e., intensity measured in packs per day or duration measured in years of use). Models that did not include packs per day of smoking or drinks per week did not differ from that shown in Table 2
.
p16INK4A homozygous deletion is associated with increased years of drinking and smoking. We further examined the relationship between tobacco and alcohol exposure and p16INK4A homozygous deletion using a logistic regression analysis. This was done to control for potential confounding variables including age, gender, and tumor site while also evaluating the duration and intensity of alcohol and tobacco exposure as individual predictors of p16INK4A homozygous deletion. The analysis was limited to cases with a history of alcohol and tobacco consumption in an effort to understand the potential differential action of these carcinogens on deletion at the p16INK4A locus. Comparison of the ever drinkers to never drinkers yielded a nonsignificant OR of 1.6 for homozygous p16INK4A deletion. Because the association of years of drinking with p16INK4A deletion was most prominent in the upper third of drinkers in the univariate analysis (Table 1), we evaluated alcohol duration by comparing the highest tertile of drinking duration versus the lower two thirds. This resulted in a significantly elevated (but imprecise) OR for predicting p16INK4A homozygous deletion of 4.9 [95% confidence interval (95% CI), 2.0-12.1] for >43 years of alcohol drinking as compared with 43 years of alcohol drinking (Table 2). While controlling for both duration and intensity of drinking and intensity of smoking, age, gender, and tumor site, there was an apparent dose-response relationship between years of smoking and p16INK4A homozygous deletion (Table 2). Years of smoking, assessed in tertiles, yielded ORs of 1 (1-28 years of smoking), 1.3 (29-39 years), and 1.9 (>39 years). In this model, neither intensity of smoking (packs per day) nor intensity of alcohol use (average alcohol drinks consumed per week) was significantly associated with gene deletion. There was an imprecise decreased risk for p16INK4A homozygous deletion with age, indicating that the deletion was more likely to occur in younger patients.
Next, we tested whether there was a synergistic effect between alcohol and tobacco duration using an interaction term in the logistic regression model. Although alcohol remained a significant predictor of p16INK4A deletion, the interaction between alcohol and tobacco was not a significant predictor of deletion (P = 0.69).
|
Discussion
|
|---|
We detected homozygous deletion of the p16INK4A gene in 40% (132 of 330) of HNSCC tumors examined, a frequency similar to that previously reported (9–12). Although alternative mechanisms of inactivation of p16INK4A (methylation silencing and mutation) can occur, homozygous gene deletion is the most common type of genetic aberration at this locus. Alcohol exposure, one of the main risk factors for HNSCC, was a significant predictor of p16INK4A homozygous deletion, but only when measured as duration and not intensity. The majority of studies of the potentially genotoxic action of alcohol have focused on the intensity of alcohol exposure using average alcohol drinks per week as a means to measure alcohol as a risk factor for HNSCC. When intensity and duration of alcohol exposure have been examined separately in large epidemiologic studies, both have been found to contribute to cancer risk (15). Because these measures are colinear, it has been difficult to distinguish the independent action of the differing exposure metrics. Castellsague et al. (15) observed the cancer risk associated with more intense alcohol use (drinks per day) to be greater than that associated with duration (years of drinking), but both were significant HNSCC risk factors. The data that we collected enabled us to analyze the independent effects of both duration and intensity in the same model. Here we report the first evidence, to our knowledge, that duration of alcohol exposure, independent of intensity, is the primary predictor of a somatic event in HNSCC.
Precisely how alcohol acts in producing large-scale genetic deletion events is unclear. Alcohol (ethanol) itself is not a potent carcinogen (16); similarly, its metabolite, acetaldehyde, is mutagenic but its potency is also regarded as relatively mild (17). When alcohol is applied locally to the oral mucosa in model systems, it increases the occurrence of tumors (18). Consistent with our finding of deletion associated with drinking, Reis et al. (8) reported alcohol increases deletion events on chromosome 22q13 in HNSCC. Thus, the combination of these data suggests that alcohol exerts its carcinogenic effects in generating large-scale DNA breakage events. The precise mechanism behind this is not clear. It might be attributable to a potentially genotoxic metabolite of alcohol (such as acetaldehyde), to the inflammatory effects of irritation, or to breakdown of normal membranous barriers that act to protect cells from the advent of genotoxic events.
We also found evidence for the presence of a dose response between years of smoking and p16INK4A homozygous deletion. The magnitude of this association was small as compared with that of alcohol, but it is consistent with previous reports in tobacco-related disease (5, 19–21). Interestingly, similar to alcohol, duration of tobacco exposure, and not intensity (packs per day), was the better predictor of deletion. Thus, duration of exposure to either of the major carcinogens in HNSCC was the primary determinant of p16INK4A homozygous deletion. Although it is not clear why duration is such a powerful predictor of homozygous deletion of p16INK4A, it might be that continued, chronic exposure is most effective in promoting clonal selection and expansion in this disease. Finally, although smoking and drinking interact to cause HNSCC, we did not observe the exposures to interact in inducing p16INK4A loss. This suggests that the interaction in causing the disease may be the result of the concomitant effects of alteration of multiple loci, rather than interaction of the exposures in induction of loss of a single locus.
Interestingly, our finding that homozygous p16INK4A deletion is associated with longer duration of carcinogen exposure in HNSCC is strikingly different from our prior similar investigation in non–small-cell lung cancer (22). In surgically excised non–small-cell lung cancer, homozygous p16INK4A deletion occurred in patients who reported a less intense history of cigarette smoking. This difference in the association of carcinogen exposure and gene deletion in these anatomically adjacent solid tumors is dramatic and puzzling. It is of interest, at the same time, to note that our data in HNSCC suggest that, controlled for smoking, homozygous p16INK4A deletion is associated with a younger age. Hence, both data sets suggest that deletion events at this locus may have some association with measures of susceptibility to these cancers. This also is consistent with the observations that cytogenetic sensitivity (measured with chromosomal breakage) of cells from HNSCC patients to the in vitro action of carcinogens is significantly greater than control, cancer-free individuals (23). Despite this, in the upper aerodigestive tract, smoking promotes clonal p16INK4A alteration by deletion, whereas in the lung, increased smoking is associated with clonal selection via inactivation of p16INK4A by promoter methylation and epigenetic silencing (5, 6). This difference may be attributable to intrinsic differences in the target tissues, to the components of cigarette smoke that reach the different anatomic locations, or perhaps to the nature of the type of changes needed to immortalize the cells in these tissues. Whatever the mechanism, these observations illustrate that there are profound exposure-related differences in the nature of clonal selection in solid tumors from adjacent but distinct sites.
It should be noted that our sample size, although moderately large, may limit a precise estimation of the point estimates of associations and also limit our ability to detect interaction of smoking and drinking. Furthermore, the detailed questionnaire that collected specific alcohol and tobacco data was not completed by 22% of the cases, further limiting our sample. In sum, our data suggest that the chronic temporal pattern of duration of alcohol and tobacco exposure, independent of intensity, is remarkably important in determining the character of the somatic alteration in the Rb pathway at the p16INK4A locus in HNSCC.
|
Acknowledgments
|
|---|
Grant support: NIH grants R01 CA100679, R01 CA078609, T32ES07155, and P30 ES00002.
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.
We thank the following surgeons, physicians, and professors for their advice and contributions: Dr. John Clark, Dr. Greg Grillone, Dr. Karl Munger, Dr. Marshall Posner, Dr. James Rocco, Dr. Sook Bin Woo, and Dr. Zhi-Min Yuan.
Received 10/17/05.
Revised 2/ 3/06.
Accepted 2/17/06.
|
References
|
|---|
- Parkin DM, Laara E, Muir CS. Estimates of the worldwide frequency of sixteen major cancers in 1980. Int J Cancer 1988;41:184–97.[Medline]
- Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA Cancer J Clin 2004;54:8–29.[Abstract/Free Full Text]
- Dobrossy L. Epidemiology of head and neck cancer: magnitude of the problem. Cancer Metastasis Rev 2005;24:9–17.[Medline]
- Tuyns AJ, Esteve J, Raymond L, et al. Cancer of the larynx/hypopharynx, tobacco and alcohol: IARC international case-control study in Turin and Varese (Italy), Zaragoza and Navarra (Spain), Geneva (Switzerland) and Calvados (France). Int J Cancer 1988;41:483–91.[Medline]
- Kim DH, Nelson HH, Wiencke JK, et al. p16(INK4a) and histology-specific methylation of CpG islands by exposure to tobacco smoke in non-small cell lung cancer. Cancer Res 2001;61:3419–24.[Abstract/Free Full Text]
- Toyooka S, Suzuki M, Tsuda T, et al. Dose effect of smoking on aberrant methylation in non-small cell lung cancers. Int J Cancer 2004;110:462–4.[CrossRef][Medline]
- Schantz SP, Huang Q, Shah K, et al. Mutagen sensitivity and environmental exposures as contributing causes of chromosome 3p losses in head and neck cancers. Carcinogenesis 2000;21:1239–46.[Abstract/Free Full Text]
- Reis PP, Rogatto SR, Kowalski LP, et al. Quantitative real-time PCR identifies a critical region of deletion on 22q13 related to prognosis in oral cancer. Oncogene 2002;21:6480–7.[CrossRef][Medline]
- Reed AL, Califano J, Cairns P, et al. High frequency of p16 (CDKN2/MTS-1/INK4A) inactivation in head and neck squamous cell carcinoma. Cancer Res 1996;56:3630–3.[Abstract/Free Full Text]
- Namazie A, Alavi S, Olopade OI, et al. Cyclin D1 amplification and p16(MTS1/CDK4I) deletion correlate with poor prognosis in head and neck tumors. Laryngoscope 2002;112:472–81.[CrossRef][Medline]
- Nakahara Y, Shintani S, Mihara M, et al. High frequency of homozygous deletion and methylation of p16(INK4A) gene in oral squamous cell carcinomas. Cancer Lett 2001;163:221–8.[CrossRef][Medline]
- Shintani S, Mihara M, Nakahara Y, et al. Expression of cell cycle control proteins in normal epithelium, premalignant and malignant lesions of oral cavity. Oral Oncol 2002;38:235–43.[CrossRef][Medline]
- Wu CL, Roz L, McKown S, et al. DNA studies underestimate the major role of CDKN2A inactivation in oral and oropharyngeal squamous cell carcinomas. Genes Chromosomes Cancer 1999;25:16–25.[CrossRef][Medline]
- Carter TL, Reaman GH, Kees UR. INK4A/ARF deletions are acquired at relapse in childhood acute lymphoblastic leukaemia: a paired study on 25 patients using real-time polymerase chain reaction. Br J Haematol 2001;113:323–8.[CrossRef][Medline]
- Castellsague X, Quintana MJ, Martinez MC, et al. The role of type of tobacco and type of alcoholic beverage in oral carcinogenesis. Int J Cancer 2004;108:741–9.[CrossRef][Medline]
- Ketcham AS, Wexler H, Mantel N. Effects of alcohol in mouse neoplasia. Cancer Res 1963;23:667–70.[Abstract/Free Full Text]
- Poschl G, Stickel F, Wang XD, et al. Alcohol and cancer: genetic and nutritional aspects. Proc Nutr Soc 2004;63:65–71.[CrossRef][Medline]
- Seitz HK, Poschl G, Simanowski UA. Alcohol and cancer. Recent Dev Alcohol 1998;14:67–95.[CrossRef][Medline]
- Pfeifer GP, Denissenko MF, Olivier M, et al. Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking-associated cancers. Oncogene 2002;21:7435–51.[CrossRef][Medline]
- Hollstein MC, Peri L, Mandard AM, et al. Genetic analysis of human esophageal tumors from two high incidence geographic areas: frequent p53 base substitutions and absence of ras mutations. Cancer Res 1991;51:4102–6.[Abstract/Free Full Text]
- Wang X, Christiani DC, Wiencke JK, et al. Mutations in the p53 gene in lung cancer are associated with cigarette smoking and asbestos exposure. Cancer Epidemiol Biomarkers Prev 1995;4:543–8.[Abstract]
- Kraunz KS, Nelson HH, Lemos M, et al. Homozygous deletion of p16INK4a and tobacco carcinogen exposure in non small cell lung cancer. Int J Cancer 2006;118:1369–9.
- Zhu Y, Spitz MR, Zheng YL, Hong WK, Wu X. BPDE-induced lymphocytic 3p21.3 aberrations may predict head and neck cancer risk. Cancer 2002;95:56343–8.
Top
Abstract
Introduction
