Fax: (713) 794-4662
p73 G4C14-to-A4T14 polymorphism and risk of human papillomavirus-associated squamous cell carcinoma of the oropharynx in never smokers and never drinkers
Article first published online: 5 NOV 2008
Copyright © 2008 American Cancer Society
Volume 113, Issue 12, pages 3307–3314, 15 December 2008
How to Cite
Chen, X., Sturgis, E. M., Etzel, C. J., Wei, Q. and Li, G. (2008), p73 G4C14-to-A4T14 polymorphism and risk of human papillomavirus-associated squamous cell carcinoma of the oropharynx in never smokers and never drinkers. Cancer, 113: 3307–3314. doi: 10.1002/cncr.23976
- Issue published online: 4 DEC 2008
- Article first published online: 5 NOV 2008
- Manuscript Accepted: 28 JUL 2008
- Manuscript Revised: 18 JUL 2008
- Manuscript Received: 10 APR 2008
- University of Texas M. D. Anderson Cancer Center
- National Institutes of Health (NIH) Head and Neck Specialized Program of Research Excellence. Grant Number: P50CA097007
- Career Development Award
- The University of Texas M. D. Anderson Cancer Center
- NIH. Grant Number: ES 11740
- Clinician Investigator Award. Grant Number: K-12 CA88084
- NIH Cancer Center Support. Grant Number: CA 16672
- M. D. Anderson Cancer Center. Grant Numbers: CA135679-01, CA133099-01A1
- p73 polymorphism;
- cancer risk;
- genetic susceptibility;
- human papillomavirus;
- molecular epidemiology;
- squamous cell carcinoma
The p53 tumor suppressor protein homolog p73 can be inactivated by oncoprotein E6 of human papillomavirus (HPV). Variation in p73 may alter the interaction between the E6 protein and p73 and, thus, alter the risk for HPV-associated carcinogenesis. It is believed that the p73 G4C14-to-A4T14 polymorphism affects p73 function by altering gene expression; however, whether that polymorphism also alters the risk of HPV type 16 (HPV-16)-associated squamous cell carcinoma of the oropharynx (SCCOP) is unknown.
The current case-control study included a case group of 188 non-Hispanic white patients with newly diagnosed SCCOP and a control group of 349 healthy individuals. Logistic regression analyses were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for cases and controls stratified by p73 genotype, age, sex, smoking status, alcohol use, and HPV-16 status. The effects of p73 genotypes on the risk of HPV-16-associated SCCOP were explored with further stratification by smoking and drinking status.
HPV-16 seropositivity was associated with an increased risk of SCCOP (adjusted OR, 5.98; 95% CI, 3.89-9.20), especially among never smokers (adjusted OR, 13.8; 95% CI, 5.91-32.1), never drinkers (adjusted OR, 14.9; 95% CI, 5.24-42.4), and individuals with p73 variant genotypes (GC/AT and AT/AT; adjusted OR, 7.96; 95% CI, 3.83-16.5). Moreover, the risk of HPV-16-associated SCCOP for individuals who had p73 variant genotypes was particularly high in never smokers and never drinkers.
The p73 G4C14-to-A4T14 polymorphism may modulate the risk of HPV-16-associated SCCOP, and the p73 variant genotypes may be markers of genetic susceptibility to HPV-16-associated SCCOP, particularly in never smokers and never drinkers. Cancer 2008. © 2008 American Cancer Society.
Despite declining smoking prevalence in the United States, the incidence of squamous cell carcinoma of the oropharynx (SCCOP) has increased in recent decades, particularly in young adults.1 It is estimated that, in the United States, approximately 12,000 new cases of SCCOP will be diagnosed and that 2200 deaths will result from those cancers in 2008.2 Epidemiologic and experimental evidence clearly indicates that high-risk human papillomavirus (HPV) plays a causative role in the etiology of SCCOP, especially in the population that lacks the known risk factors of tobacco and alcohol consumption, accounting for the continually increasing incidence rate in the general population.3–7
Of the known HPV types, oncogenic HPV type 16 (HPV-16) is the most frequent type, accounting for approximately 90% of HPV-associated SCCOP.7–9 HPV-16 antibodies, as markers of past exposure to HPV-16, have been associated positively with an increased risk of cervical cancer10–13 and SCCOP.9, 14, 15 Although HPV infection may be a major risk factor for SCCOP,8, 16 only a small fraction of the population with a long period of high-risk HPV exposure develops SCCOP, suggesting that host genetic variants in genes, which are involved in cell cycle control and apoptosis and interact with HPV oncoprotein E6 or E7, may contribute to interindividual variation in the susceptibility to HPV-associated SCCOP.
Cell cycle-related genes play a role in modulating cellular DNA repair, cell cycle control, cell growth, and apoptosis to maintain genomic stability by monitoring the order and integrity of cell division events, such as p53, which is 1 of the critical cell cycle-regulatory tumor suppressor genes.17–20 The loss of p53 function may result in loss of cell cycle control and checkpoint integrity. The chiefly oncogenic E6 protein of HPV-16 binds to the p53 protein of the host cell, ending in p53 degradation through the ubiquitination pathway.21–23 It was demonstrated that the polymorphism of p53 at codon 72 altered the susceptibility of p53 to oncogenic HPV E6-mediated degradation24 and was associated significantly with oncogenic HPV infection in cervical cancer.24, 25 Furthermore, in case-control analyses, the Pro allele was associated with an increased risk of HPV-associated SCCOP,26 and the risk appeared to be particularly high in never smokers.27
A member of the p53 family,p73 activates the promoters of several p53-responsive genes that participate in DNA repair, cell cycle control, and apoptosis, and it inhibits cell growth in a p53-like manner by inducing apoptosis or G1 cell cycle arrest.28–31 Inactivation of p73 by oncogenic HPV E6 appears to be analogous to that of p53 without modulating DNA binding activity.32 Therefore, p73 may act as a tumor suppressor with some of the same functions as p53 and may compensate for the loss of p53 function. However, p73, unlike p53, is resistant to degradation by HPV-16 E6 and can suppress cell growth and induce apoptosis in HPV-16 E6-expressing cells.33 It is possible that p73 variation may alter the interaction between E6 protein and p73 and, thus, alter the risk of HPV-16-associated carcinomas.
It is believed that the 2 linked, noncoding exon 2 polymorphisms of p73 at positions 4 (G;→A) and 14 (C→T;) (the p73 G4C14-to-A4T14 polymorphism) affect p73 function by altering gene expression, perhaps by altering the efficiency of translational initiation.30 It is plausible that genetic variation of p73 may lead to interindividual variation in susceptibility to HPV-associated SCCOP. To the best of our knowledge, the association of this polymorphism with the risk of HPV-16-associated SCCOP has not been investigated previously, particularly among never smokers and never drinkers.
To test the hypothesis, we evaluated the relation between the p73 G4C14-to-A4T14 polymorphism and HPV-16 serologic status for the risk of SCCOP and explored the joint effects of p73 variant genotypes and HPV-16 serologic status in subgroups of individuals stratified by smoking and drinking status in a case-control study of a case group of 188 patients with newly diagnosed SCCOP and a control group of 349 cancer-free individuals.
MATERIALS AND METHODS
Consecutive patients with newly diagnosed, histopathologically confirmed, and untreated SCCOP were recruited between May 1996 and January 2001 through the Head and Neck Center at the University of Texas M. D. Anderson Cancer Center in Houston, Texas, as part of a molecular epidemiologic study of squamous cell carcinoma of the head and neck. The accrual rate was 81% for the cases. The controls included 2 groups of cancer-free individuals. One group was comprised of 160 healthy individuals (45.8%) who were selected from a control pool of enrollees at the Kelsey-Seybold Clinic, a multispecialty physician practice with multiple clinics throughout the Houston metropolitan area. The overall response rate was approximately 75%. The other controls were 189 healthy visitors (54.2%) who were accompanying cancer patients to the outpatient clinics at M. D. Anderson Cancer Center but who were genetically unrelated to the cases. The response rate for this M. D. Anderson Cancer Center control group was approximately 80%. Both control groups had no previous histories of any cancers, were not on therapies or treatment for any diseases, and were frequency-matched to the cases on age (±5 years), sex, smoking status, and drinking status. To avoid confounding because of ethnic characteristics, we included only non-Hispanic whites in the case group and the control groups.
Participants who had smoked >100 cigarettes in their lifetimes were categorized as ‘ever smokers,’ and the remaining participants were categorized as ‘never smokers.’ Participants who had drunk alcoholic beverages at least once a week for more than 1 year were categorized as ‘ever drinkers,’ and the remaining participants were categorized as ‘never drinkers.’ After signing informed consent forms, which had been approved by the institutional review boards of both the M. D. Anderson Cancer Center and the Kelsey-Seybold Clinic, study participants completed a questionnaire regarding demographic and relevant risk factors and donated 30 mL of blood.
Human Papillomavirus Type 16 Serologic Testing
HPV-16 L1 virus-like particles generated from recombinant baculovirus-infected insect cells were used to test for antibodies against HPV-16 in the plasma of study participants by using a standard enzyme-linked immunosorbent assay, as described previously.10, 34 Control sera known to be positive and negative also were tested in parallel with the study samples in duplicate on each plate. The cutoff level, above which optical density (OD) values were considered positive and below which OD values were considered negative for HPV-16, was based on the absorbance value of a standard pooled serum known to be at the threshold of detection. Samples that were within 15% of the cutoff level were tested twice more, and samples that were positive in all 3 runs were considered positive. We also randomly selected 10% of the samples to retest for confirmation of the original findings. To eliminate potential binding interference by heparin, we treated the plasma samples with 43 U/mL heparinase I (Sigma, St. Louis, Mo) before testing.35 We tested heparinized plasma, as well as serum, obtained from 3 individuals and did not detect discernible differences between the reactions of the serum samples and the heparinized plasma samples treated with heparinase.
We extracted genomic DNA from a leukocyte cell pellet, which was obtained from the buffy coat by centrifugation of 1 mL of whole blood, by using the QIAGEN DNA Blood Mini Kit (QIAGEN Inc., Valencia, Calif) according to the manufacturer's instructions. We typed for the p73 G4C14-to-A4T14 genotypes by polymerase chain reaction (PCR) with confronting 2-pair primers, which makes genotyping possible by electrophoresis without restriction digestion.36 The A4T14 allele was amplified with primers forward 1 (F1: 5′-CCACGGATGGGTCTGATCC-3′) and reverse 1 (R1: 5′-GGCCTCCAAGGGCAGCTT-3′), which produced a 270-base pair (bp) fragment, and the G4C14 allele was amplified with primers F2 (5′-CCTTCCTTCCTGCAGAGCG-3′) and R2 (5′-TTAGCCCAGCGAAGGTGG-3′), which amplified a 193-bp fragment. F1 and R2 also produced a common 428-bp fragment in each PCR. The PCR reaction was performed in a 10-μL volume that contained approximately 20 ng of genomic DNA, 0.1 mM each deoxyribonucleotide triphosphate, 1 × PCR buffer (50 mM KCl, 10 mM Tris HCl and 0.1% Triton X-100), 1.5 mM MgCl2, 0.5 U Taq polymerase (Sigma-Aldrich Biotechnology, St. Louis, Mo), and 2 pmol each of the 4 primers. The amplification conditions included 10 minutes of initial denaturation at 95°C, 35 cycles of 1 minute at 95°C, 45 seconds at 62°C, 1 minute at 72°C, and a final 5-minute extension at 72°C. All PCR products were observed on a 2% agarose gel that contained 0.25 mg/mL of ethidium bromide. More than 10% of samples were selected randomly to perform the repeated assays, and the results were 100% concordant.
The differences in the distributions of selected demographic variables, tobacco smoking, alcohol use, and p73 allele and genotype frequencies between cases and controls were evaluated using the chi-square test of association. Both univariate and multivariate logistic regression analyses were used to calculate odds ratios (ORs) and 95% confidence interval (CIs) for cases and controls stratified by age, sex, p73 genotype, smoking, drinking, and HPV-16 status. In the multivariate logistic regression model, the ORs and 95% CIs were adjusted by age, sex, smoking, and alcohol use. Because only a small number of individuals (14 cases and 20 controls) were homozygous for the AT allele, which may have precluded meaningful subgroup analyses, p73 genotype data were dichotomized according to a dominant model in which homozygosity for GC/GC was coded as 0, and both the heterozygosity for GC/AT and the homozygosity for AT/AT were coded as 1. All tests were 2-sided, and a P value of .05 was considered the cutoff for statistical significance. All statistical analyses were performed with Statistical Analysis System software (version 9.1; SAS Institute, Cary, NC).
Table 1 lists the distributions of demographic variables and risk factors for the study population of 188 cases and 349 cancer-free controls. All study participants were non-Hispanic whites, and the cases and controls appeared to be adequately frequency-matched for age, sex, smoking, and alcohol consumption; however, these factors also were adjusted further in later analyses to control for any residual effects. The distributions of p73 genotypes among the controls were in agreement with Hardy-Weinberg equilibrium (P = .349). When comparing p73 genotype distributions between cases and controls, no significant differences were observed (P = .738). The p73 AT allele was present in 39.4% of cases and in 38.4% of cancer-free controls (P = .893), and the p73 AT allele frequency was 23.4% in cases and 22.1% in controls (P = .682). We also observed that the distribution of p73 AT carriers and noncarriers between cases and controls was not statistically significant (P = .826), but HPV-16 seropositivity was significantly more common in cases than in controls (P < .001).
|Characteristic||Cases n=188||Controls n=349||P*|
|GC/AT and AT/AT||74||39.4||134||38.4||.826|
Table 2 summarizes the distributions of p73 genotypes, age, sex, smoking, and alcohol use stratified by HPV-16 status and their association with SCCOP risk. Overall, HPV-16 seropositivity was associated with an approximately 6-fold risk of SCCOP (adjusted OR, 5.98; 95% CI, 3.89-9.20) after adjusting for age, sex, smoking status, and alcohol use. HPV-16 seropositivity among those with the p73 homozygous wild-type genotype (GC/GC) was associated with an elevated risk of SCCOP (adjusted OR, 5.47; 95% CI, 3.14-9.54); whereas, among individuals with the p73 variant genotypes (GC/AT and AT/AT), HPV-16 seropositivity was associated with an even higher risk of SCCOP (adjusted OR, 7.96; 95% CI, 3.83-16.5).
|Characteristic||HPV-16 Status||Cases, n=188||Controls, n=349||Adjusted OR (95% CI)*|
|GC/AT and AT/AT||−||37||50||119||88.8||1.00|
To investigate the effects of other factors on the risk of HPV-16-associated SCCOP, we further stratified the associations between the HPV-16 status and the cancer risk by age, sex, smoking status, and alcohol use. We observed that the risk of SCCOP associated with HPV-16 seropositivity was evident for all subgroups, particularly among individuals aged <56 years (adjusted OR, 7.94; 95% CI, 4.26-14.8), men (adjusted OR, 6.65; 95% CI, 4.09-10.8), never smokers (adjusted OR, 13.8; 95% CI, 5.91-32.1), and never drinkers (adjusted OR, 14.9; 95% CI, 5.24-42.4). However, the interaction between HPV-16 status (seropositive vs seronegative) and p73 polymorphism (AT carriers vs AT noncarriers), age, and sex was not statistically significant (P = .408 for p73 polymorphism; P = .169 for age; and P = .386 for sex).
Because we observed a significant interaction between HPV-16 status (seropositive vs seronegative) and smoking status (ever smoking vs never smoking) on the risk of SCCOP (P = .029), we also investigated the effects of p73 genotypes on the risk of SCCOP stratified by HPV-16 serologic status and smoking status (with adjustment for age, sex, and alcohol use) (Table 3). Those never smokers who were HPV-16 seropositive and had p73 variant genotypes exhibited an approximately 29-fold greater risk of SCCOP (adjusted OR, 28.6; 95% CI, 5.53-148.2) than those never smokers who were HPV-16 seronegative and had the wild-type genotype. However, HPV-16 seropositivity in never smokers with the wild-type genotype conferred an approximately 15-fold increased risk of SCCOP (adjusted OR, 14.6; 95% CI, 5.24-40.5). Ever smokers who had p73 variant genotypes and were HPV-16 seropositive exhibited only a 4-fold higher risk of SCCOP (adjusted OR, 4.43; 95% CI, 2.09-9.42) compared with ever smokers who had the wild-type genotype and were HPV-16 seronegative. Ever smokers with the wild-type genotype who were HPV-16 seropositive exhibited a less increased risk of SCCOP (adjusted OR, 3.35; 95% CI, 1.71-6.55) compared with ever smokers with the wild-type genotype who were HPV-16 seronegative.
|Smoking and Drinking Status: p73 Genotype||HPV-16 Status||No. (%)||OR [95%CI]|
|GC/GC||−||12 (19.1)||57 (53.3)||1.00||1.00|
|+||24 (38.0)||8 (7.4)||14.3 [5.17–39.3]||14.6 [5.24–40.5]|
|GC/AT and AT/AT||−||15 (23.8)||40 (37.4)||1.78 [0.75–4.21]||1.66 [0.68–4.08]|
|+||12 (19.1)||2 (1.9)||28.5 [5.63–144.2]||28.6 [5.53–148.2]|
|GC/GC||−||52 (41.6)||130 (53.7)||1.00||1.00|
|+||26 (20.8)||20 (8.3)||3.25 [1.67–6.32]||3.35 [1.71–6.55]|
|GC/AT and AT/AT||−||22 (17.6)||79 (32.6)||0.70 [0.39–1.23]||0.67 [0.38–1.19]|
|+||25 (20)||13 (5.4)||4.81 [2.29–10.1]||4.43 [2.09–9.42]|
|GC/GC||−||12 (31.6)||55 (60.4)||1.00||1.00|
|+||16 (42.1)||6 (6.6)||12.2 [3.96–37.7]||13.7 [4.27–43.8]|
|GC/AT and AT/AT||−||6 (15.8)||29 (31.9)||0.95 [0.32–2.79]||0.98 [0.32–3]|
|+||4 (10.5)||1 (1.1)||18.3 [1.88–179]||23.1 [2.01–265.4]|
|GC/GC||−||52 (34.7)||132 (51.2)||1.00||1.00|
|+||34 (22.7)||22 (8.5)||3.92 [2.10–7.33]||3.83 [2.04–7.21]|
|GC/AT and AT/AT||−||31 (20.6)||90 (34.9)||0.87 [0.52–1.47]||0.86 [0.50–1.45]|
|+||33 (22)||14 (5.4)||5.98 [2.96–12.1]||5.55 [2.73–11.3]|
Similarly, because a borderline significant interaction between HPV-16 status (seropositive vs seronegative) and drinking status (ever drinking vs never drinking) was observed on the risk of SCCOP (P = .068), we also investigated the effects of p73 genotypes on risk of SCCOP stratified by HPV-16 serologic status and drinking status (with adjustment for age, sex, and smoking status) (Table 3). Those never drinkers who were HPV-16 seropositive and had p73 variant genotypes exhibited an approximately 23-fold greater risk of SCCOP (adjusted OR, 23.1; 95% CI, 2.01-265.4) than those never drinkers who were HPV-16 negative and had the wild-type genotype. However, HPV-16 seropositivity in never drinkers with the wild-type genotype conferred an approximately14-fold increased risk of SCCOP (adjusted OR, 13.7; 95% CI, 4.27-43.8). Ever drinkers who had p73 variant genotypes and were HPV-16 seropositive exhibited only a 5.5-fold higher risk of SCCOP (adjusted OR, 5.55; 95% CI, 2.73-11.3) compared with ever drinkers who had the wild-type genotype and were HPV-16 seronegative. Ever drinkers with the wild-type genotype who were HPV-16 seropositive exhibited a less increased risk of SCCOP (adjusted OR, 3.83; 95% CI, 2.04-7.21) compared with ever drinkers with the wild-type genotype who were HPV-16 seronegative. Although these ORs differed between groups, the 95% CIs overlapped widely, and the differences were not statistically different (P > .05). The estimates of association may have biased in part because of small sample sizes within the strata of each stratification variables, misclassification of HPV serologic status, or other confounding factors. The findings from this study need to be confirmed in future, larger, and well designed studies.
In this hospital-based case-control study of a non-Hispanic white population, we observed that HPV-16 seropositivity was associated with an increased risk of SCCOP. HPV-16 seropositivity among patients with p73 variant genotypes (GC/AT and AT/AT) was associated with a slightly greater risk for developing SCCOP compared with the risk among individuals with the p73 wild-type genotype (GC/GC), but the risk was relatively greater in never smokers and never drinkers. This finding suggests that the p73 G4C14-to-A4T14 polymorphism may modulate the genetic susceptibility to HPV-16-associated SCCOP.
Although, to our knowledge, the precise mechanism by which the p73 G4C14-to-A4T14 polymorphism plays a role in the development of HPV-16-associated SCCOP has not been clarified to date, these findings are biologically plausible. p73 Shares structural and functional similarities to p53 and can be inactivated functionally through physical interaction with E6 oncoprotein, possibly analogous to that of p53.28–32 Currently, no reported studies have investigated the association between this p73 polymorphism and the risk of HPV-16-associated SCCOP; however, lit has been demonstrated that the p53 codon 72 polymorphism alters the susceptibility of p53 to oncogenic HPV E6-mediated degradation24 and is associated with an increased risk of HPV-associated cancers.24–27 These findings suggest that p73 may have an effect similar to that of p53 on the development of HPV-16-associated SCCOP.
Genetic variation may lead to different splice transcripts and may have functional consequences, which may contribute to interindividual difference and disease susceptibility.37, 38 The p73 G4C14-to-A4T14 polymorphism lies upstream of the initiating AUG of exon 2 and may affect the differential use of p73 promoters, leading to different functional transcripts with variable 5′-untranslated regions. This p73 polymorphism also may form a stem-loop structure, which may result in alteration of gene expression, possibly by altering the efficiency of translational initiation.30 In addition, this p73 polymorphism also may be in linkage disequilibrium with other functional polymorphisms or adjacent susceptibility loci of the gene, thereby affecting p73 gene expression and activity, leading to altered interaction between E6 protein and p73, and, thus, may modulate the risk of HPV-associated carcinogenesis.37 However, more studies will be are needed to confirm these hypotheses.
In the current study, a higher risk of HPV-16-associated SCCOP was observed among never smokers than among ever smokers, and a higher risk was observed among never drinkers than among ever drinkers. Moreover, a significant interaction was identified between HPV-16 status and smoking status, and a near significant interaction was identified between HPV-16 status and drinking status. For instance, in individuals with the variant p73 genotypes (GC/AT and AT/AT) who were seropositive for HPV, there was an approximately 7-fold increased risk of SCCOP among never smokers; whereas, in individuals with the wild-type p73 genotype (GC/GC) who were seropositive for HPV, there was an approximately 4-fold increased risk of SCCOP compared with the individuals with HPV-16 seronegativity and the wild-type p73 genotype (GC/GC). These findings suggest that this p73 polymorphism may have a stronger interaction with HPV-16 among never smokers and never drinkers than among ever smokers and ever drinkers or that this p73 polymorphism may be less important in smokers and alcohol drinkers, because non-HPV-associated SCCOP already has a heavy carcinogen/mutation burden driving the development of SCCOP. Therefore, the p73 G4C14-to-A4T14 polymorphism may play a role in the development of HPV-16-associated SCCOP among never smokers and never drinkers in the general population. However, smoking or drinking and p73 variants may not be cofactors in the etiology of HPV-16-associated SCCOP, suggesting that it is imperative to control for the confounding effects of smoking and alcohol use while assessing the role of this p73 polymorphism as a risk factor for HPV-16-associated SCCOP. However, these findings need to be tested further in studies with larger sample sizes.
We also observed that the risk of HPV-16-associated SCCOP was more evident among younger individuals (aged <56 years) and men, but there was no evidence of a significant interaction between HPV-16 status and age or sex. These findings are consistent with previous studies indicating that HPV-associated carcinoma had a trend toward younger age at onset that may be related to sexual practices (ie, orogenital sexual contact).6, 39, 40 Overall, squamous cell carcinomas of the head and neck occur more frequently in men (approximately 70%) than in women.41
The current study is the first to our knowledge to investigate the relation between the p73 G4C14-to-A4T14 polymorphism and the risk of HPV-16-associated SCCOP. A Japanese study42 that included 112 patients, 320 healthy women, and 122 women outpatients without cancer investigated the role of this p73 polymorphism in the risk of cervical cancer, which is caused chiefly by HPV-16 infection. In that study, the p73 variant genotypes were associated with a borderline increased risk of cervical cancer (adjusted OR, 1.51; 95% CI, 0.98-2.35). In our previous study,43 we explored the association of this p73 polymorphism with the risk of squamous cell carcinoma of the head and neck and observed no significant association with the risk of SCCOP after we stratified the analysis by cancer sites. We expect that the estimated risk in these 2 studies may have been confounded by HPV-16 status, smoking, and drinking status. Therefore, it will be necessary to stratify the data further by HPV-16 status, smoking status, and drinking status in future studies that assess the effect of this p73 polymorphism on the risk of HPV-associated cancers.
In the current hospital-based case-control study, there may have been some selection bias, because our cases and controls may have had different population bases. SCCOP cases were enrolled from among M. D. Anderson Cancer Center outpatients, and the controls were recruited from both outpatient clinic visitors at M. D. Anderson Cancer Center and a control pool of enrollees at the Kelsey-Seybold Clinic throughout the Houston metropolitan area. However, there was no statistically significant difference in the frequency of p73 genotypes between cases and controls, as we reported previously for the same control populations.43 In addition, our stratified analyses included a limited number of individuals in some subgroups; therefore, our results may be chance findings and should be confirmed in larger studies. Moreover, because our study included only non-Hispanic white individuals, it is uncertain whether these results are generalizable to other ethnic populations. However, the cases and controls were frequency-matched for age, sex, smoking status, and drinking status, and the effects of any confounding demographic factors may have been minimized. Finally, HPV-16 seropositivity may not reflect actual tumor HPV-16 status, leading to some misclassification; ie, some patients may have been classified as seronegative, although their tumors actually were HPV-16 DNA-positive. This is because individual patients may have had differences in their immune response to HPV-16 infection or antibody instability. This misclassification could result in a major selection bias for the estimates of the association. We may assess the HPV-16 specificity and sensitivity for SCCOP patients to explore the discordance between HPV-16 serologic and tumor status in our future study with larger sample sizes when the tumor tissues become available. However, an early study confirmed a reasonable concordance between HPV-16 seropositivity and HPV-16 DNA positivity of tumor tissues9; and, more important, the use of serologic status allows for the inclusion of a cancer-free control group.
In conclusion, our findings suggest that the p73 G4C14-to-A4T14 polymorphism may modulate the risk of HPV-16-associated SCCOP and that the p73 variant genotypes may be markers of genetic susceptibility to HPV-16-associated SCCOP, particularly in never smokers and never drinkers. Advanced biologic studies will be needed to validate these results. Further investigations with larger sample sizes of different populations also are warranted. To advance these findings, currently, we are testing other functional polymorphisms of genes involved in cell cycle and apoptosis pathways to elucidate the roles of these genetic variants in the development of HPV-16-associated SCCOP.
The authors thank Angelique Siy for article editing; Margaret Lung, Kathryn Patterson, Liliana Mugartegui, and Angeli Fairly for their help with recruiting participants; and Li-E Wang for laboratory management
- 37The promise of cancer genetics. Lancet. 1998; 351( suppl 2): 1–8., .