|Home | About | Journals | Submit | Contact Us | Français|
Both p53 tumor suppressor and murine double minute 2 (MDM2) oncoprotein are crucial in carcinogenesis. We hypothesized that MDM2 promoter single nucleotide polymorphism (SNP)309, A2164G, and p53 codon 72 SNP are associated with risk and age at onset of squamous cell carcinoma of head and neck (SCCHN). We genotyped these SNPs in a study of 1,083 Caucasian SCCHN cases and 1,090 cancer-free controls. Although none of these SNPs individually had a significant effect on risk of SCCHN, nor did their combined putative risk genotypes (i.e. MDM2 SNP309 GT + GG, 2164 AA, and p53 codon 72 CC), we found that individuals with 2–3 risk genotypes had significantly increased risk of non-oropharyngeal cancer (OR = 1.42; 95% CI=1.07–1.88). This increased risk was more pronounced among young subjects, men, smokers, and drinkers. In addition, female patients carrying the MDM2 SNP309 GT and GG genotypes showed a 3-year (56.7 years) and 9-year (51.2 years) earlier age at onset of non-oropharyngeal cancer (Ptrend = 0.007), respectively, compared with those carrying the TT genotype (60.1 years). The youngest age (42.5 years) at onset of non-oropharyngeal cancer was observed in female patients with the combined MDM2 SNP309 GG and p53 codon 72 CC genotypes. The findings suggest that MDM2 SNP309, A2164G, and p53 codon 72 SNPs may collectively contribute to non-oropharyngeal cancer risk and that MDM2 SNP309 individually or in combination with p53 codon 72 may accelerate the development of non-oropharyngeal cancer in women. Further studies with large sample sizes are warranted to validate these results.
Squamous cell carcinomas of the head and neck (SCCHN) including cancers of the oral cavity, pharynx, and larynx are the most common malignancies in the world, representing more than 90% of all head and neck cancers. Although tobacco use and alcohol consumption have been well recognized as the major risk factors for SCCHN, only a fraction of individuals exposed to tobacco or alcohol develop SCCHN, suggesting that genetic factors may also play an important role in the etiology of SCCHN.
The p53 tumor suppressor plays a key role in organizing cellular responses to various types of stresses, including DNA damage and oncogene activation with apoptosis, cell cycle arrest, senescence, DNA repair, cell metabolism or autophagy [1–3]. Malfunction and mutations of p53 have been found in most of human cancers, leading to deregulated p53 activity that allows to proliferate and survive unchecked . The activity of p53 is regulated by many important proteins, and one of the most extensively studied regulators of p53 is the murine double minute 2 (MDM2) onconprotein. MDM2 regulates the activity of p53 in at least three ways [5–8]. Firstly, MDM2 directly binds to the p53 transactivation domain, thus inhibiting its transcriptional activity. Secondly, MDM2 promotes ubiquitin-dependent proteasomal degradation of p53 by functioning as an E3 ubiquitin ligase. Finally, MDM2 shuttles p53 out of the nucleus to the cytoplasm of the cell, promoting the degradation of p53. Importantly, MDM2 forms a negative-feedback loop in regulating p53 activity, in which p53 induces transcription of MDM2, and, in turn, the MDM2 protein inhibits p53 activity .
Because MDM2 is a key negative regulator of p53 activity, over-expression of MDM2 results in the inhibition of p53-mediated-transcriptional activation, thereby contributing to human carcinogenesis[9,10]. In fact, over-expression and amplification MDM2 commonly occur in a wide variety of human cancers including SCCHN [11–16]. It has been reported that MDM2 transgenetic mice expressing approximately four times higher levels of MDM2 in various tissues compared to non-transgentic mice all develop spontaneous tumors, suggesting that over-expression of MDM2 affect tumor susceptibility in mice [9,10]. Taken together, MDM2 may play a critical role in carcinogenesis, and therefore it is biologically plausible that functional genetic variants in the p53 pathway may have an effect on cancer development in the general population.
It has been reported that a T to G change of a single-nucleotide polymorphism (SNP) in the Sp1-binding site within the intronic promoter region of MDM2 (SNP309, rs2279744) may increase the affinity of the Sp1 transcriptional factor, which results in higher levels of MDM2 mRNA and protein and attenuates the p53 pathway, both in vitro and in vivo, thereby increasing tumor susceptibility [10,17]. Consistently, the MDM2 SNP309 GG genotype was found to be associated with at least an average nine-year earlier onset of disease in both hereditary and sporadic cancers in humans . Likewise, a common SNP at codon 72 in exon 4 of p53 (codon 72; rs1042522; G > C) has also been found to be of functional significance, with the C-allele having reduced apoptotic potential compared to the G-allele, which may modulate cancer risk, progression and/or response to treatment [11,12]. However, in a preliminary study with 304 cases and 333 cancer-free controls and a following-up study with 1,111 cases and 1,130 cases, we did not find an association between p53 codon 72 SNP and SCCHN risk [13,14], although the p53 codon 72 C-allele was initially found to be associated with an earlier age of onset of SCCHN . These early studies inspired us to further investigate a possible synergistic role of SNPs in MDM2 and p53 in modulating risk and age of SCCHN onset. It is known that gene regulatory regions, such as the promoter regions, are important to gene function through binding to specific transcription factors and regulating the initiation of gene transcription . Since p53 and MDM2 physically and functionally interact in the p53 pathway, we selected the MDM2 promoter SNPs (SNP309 and A2164G) together with the well-studied p53 codon 72 (rs1042522) SNP to test our hypothesis that these SNPs, through synergistic effects, are associated with risk and age at onset of SCCHN.
The study population included 1,083 patients with newly diagnosed SCCHN between October 1999 and October 2007, including cancers of the oral cavity, oropharynx, hypopharynx and larynx, identified at The University of Texas M. D. Anderson Cancer Center. Patients with second primary tumors, primary tumors of the skin, nasopharynx, sinonasal tract, and/or any histopathologic diagnoses other than squamous cell carcinoma were excluded. All patients were Caucasians and had not received any treatment at the time of recruitment. Approximately 93% of the eligible patients contacted chose to participate in this study. The 1,090 cancer-free controls we recruited in the same period were genetically unrelated visitors or companions of patients seen at The M. D. Anderson Cancer Center clinics and were frequency matched to the cases by age (± 5 years), sex and ethnicity. The response rate of the eligible controls was approximately 85%. Having signed a written informed consent during the interview, all subjects enrolled in the study were interviewed to gather demographic data and history of tobacco and alcohol use. Each subject donated one-time 30-mL of blood, of which 1 ml was used for genomic DNA extraction with a DNA blood Mini Kit (Qiagen, Valencia, CA) according to the manufacture’s instructions. The research protocol was approved by the M.D. Anderson Institutional Review Board.
Genotyping data of p53 codon 72 polymorphism were available from our previously published study . The polymerase chain reaction (PCR)-restriction fragment length polymorphism (RFLP) assay for the p53 codon 72 polymorphism has been described elsewhere .
Except for the MDM2 SNP309, we also identified another three SNPs, i.e., G1797C (rs937282), A2164G (rs937283), and C2326 T (rs2870820) that have a minor allele frequency (MAF) > 10% within the promoter region of MDM2 in a published SNP database (http://egp.gs.washington.edu/data/mdm2/mdm2x.csnps.txt). The linkage disequilibrium (LD) analysis by the computational tool of LD Tag SNP Selection (Tag SNP) (http://manticore.niehs.nih.gov/snptag.htm) revealed that rs937283 (A2164G) tags both rs937282 (G1797C) and rs2870820 (C2326T) with LD coefficient r2 = 0.90 in the European population of the NCBI/dbSNP database. Thus, two MDM2 promoter SNPs, i.e., MDM2 SNP309 and A2164, were selected to genotype in the present study.
MDM2 SNP309 was genotyped by the PCR- RFLP assay. The primers used in the PCR-RFLP for MDM2 SNP309 was: (forward) 5′-CGGGAGTTCAGGGTAAAGGT -3′ and (reverse) 5′-AGCAAGTCGGTGCTT ACCTG -3′. The PCR reactions profile consisted of an initial melting step of 95°C of 5 min, 35 cycles of 95°C for 30s, 55°C for 45s and 72°C for 1 min and a final extension step of 72°C for 10 min. The 352-bp PCR product was digested with MspA1 I (New England Biolabs, Beverly, MA) at 37°C overnight, The wild-type allele (TT) produced three bands of 233, 88, and 31bp; wild-type/variant allele (TG) produced 233, 187, 88, 46 and 31bp and the variant allele (GG) produced 187, 88, 46 and 31bp. PCR-RFLP was conducted and the results were evaluated without knowledge of the subjects’s case-control status. Approximately 10% of the samples were randomly selected and repeated with PCR-RFLP, and the results were 100% concordant.
MDM2 A2164G (rs937283) was genotyped using the Applied Biosystems TaqMan genotyping platform according to the manufacture’s recommendations. Briefly, the reactions were prepared by using TaqMan Universal Master Mix, 40×SNP Genotyping Assay Mix, DNase-free water, and 5-ng genomic DNA in a final volume of 5 μL per reaction. Both negative and positive controls and three samples with known genotypes were included in each plate to ensure the accuracy of the genotyping. The PCR amplification was run and the plate was read using a TaqMan 7900 HT sequence detection system (Applied Biosystems, Foster City, CA). The analyzed fluorescence results were then auto-called into the genotypes using the built-in SDS2.3 software of the system.
The differences in the distributions of categorical variables, including demographic characteristics, tobacco smoking and alcohol use, and genotype/allele frequencies of selected SNPs between the patients and controls were evaluated by Chi-square tests. The deviation from Hardy-Weinberg equilibrium was tested by chi-square goodness-of-fit test. Crude or adjusted (for age, sex, smoking and drinking status) odds ratios (ORs) and 95% confidence intervals (CIs) were obtained from unconditional univariate and multivariable logistic regression analyses to evaluate associations between SNPs and SCCHN risk in the case-control analysis, which was further stratified by age, sex, smoking and drinking status. Student t-test and general linear models were used to evaluate the differences in age at onset between different genotype groups in patients and the trend in association with the age at onset, respectively. Finally, we calculated the false positive report probability (FPRP) for observed statistically significant associations in a range of prior probabilities, and used a threshold of noteworthiness of FPRP ≤0.2 in the present study. All statistics tests were two-sided with a 0.05 significance level and all data were analyzed with SAS statistical software (SAS version 9.2; SAS Institute Inc., Cary, NC).
In this study, 1,083 SCCHN patients and 1,090 controls of Caucasians were included from our ongoing studies whose DNA samples were available. The frequency distribution of selected characteristics of the patients and controls is presented in Table 1. Because of frequency matching by design, there was no statistical difference in the distributions of age and sex between cases and controls: The mean age was 57.1 years for the cases (± 11.2 years; range, 18–90 years) and 56.7 years for the controls (± 11.0 years; range, 20–87 years) (P = 0.615), and 75.2% and 24.8% of the cases and 76.3% and 23.7% of the controls were men and women, respectively (P = 0.525). However, the cases were more likely to be smokers and drinkers than the controls (P < 0.001 for both). Therefore, smoking and drinking were adjusted in the subsequent multivariate logistic regression analyses. Of the 1,083 cases, 318 (29.3%) had cancer of the oral cavity, 550 (50.8%) of oropharynx, 43 (4.0%) of hypopharynx and 172 (15.9%) of larynx.
The allele and genotype distributions of the studied SNPs in cases and controls and their associations with SCCHN risk are summarized in Table 2. All observed genotype distributions among controls were in agreement with the Hardy-Weinberg equilibrium (all P > 0.05). Linkage disequilibrium (LD) analysis revealed that the minor variant alleles of MDM2 SNP309 and MDM2 A2164G were not in LD (r2 = 0.29 and D′ = 0.88) among the controls. No significant difference in the genotype frequencies of the studied SNPs was found between the cases and controls (P = 0.638 for MDM2 SNP309, P = 0.580 for MDM2 A2164G and P = 0.193 for p53 codon 72). Although none of the variant genotypes alone was associated with significantly altered risk, the MDM2 SNP309 GT and GG genotypes tended to be associated with non-significantly increased SCCHN risk (adjusted OR = 1.11; 95% CI = 0.92–1.34 for GT, and adjusted OR = 1.08; 95% CI = 0.81–1.43 GG), the AG and GG genotypes of MDM2 A2164G tended to be associated with non-significantly reduced SCCHN risk (OR = 0.98; 95% CI = 0.81–1.20 for AG, and OR = 0.83; 95% CI = 0.64–1.07 for GG), and the p53 codon 72 CC genotype tended to be associated with non-significantly increased SCCHN risk (adjusted OR = 1.30; 95% CI = 0.91–1.86). In the haplotype analysis of MDM2 SNP309 and MDM2 A2164G, we did not observe any risk of SCCHN associated with specific common haplotypes (frequencies > 5%) of these two SNPs (data not shown). In addition, we performed a meta-analysis of MDM2 SNP309 with our and other six published studies of 1595 cases and 2254 controls (Figure 1). Consistently, we found that the MDM2 SNP309 was not significantly associated with risk of SCCHN (the pooled OR = 1.05; 95% CI = 0.93–1.18 for GT/GG genotype; P heterogeneity = 0.69).
Considering a possible combined effect from the two SNPs in MDM2 and p53 codon 72 SNP on risk of SCCHN, we combined them by the number of the putative risk genotypes (i.e., MDM2 SNP309 GT/TT, MDM2 2164 AA, and p53 codon72 CC) to assess their possible combined effect on risk of SCCHN, and we did not found the combined risk genotypes were statistically associated with increased overall risk of SCCHN (Table 2).
Because SCCHN is a heterogeneous group of tumors with different primary sites, we further evaluated the combined effect of polymorphisms of MDM2 SNP309, A2164G and p53 codon 72 on the risk of SCCHN stratified by tumor site. As shown in Table 2, the combined risk genotypes were found to be associated with increased risk of SCCHN at non-oropharyngeal sites but not at oropharyngeal site. There were more individuals with two or three risk genotypes and fewer individuals with one risk genotype among the cases of non-oropharyngeal cancer (35.1% and 32.0%, respectively) than among the controls (27.5% and 37.6%, respectively), and these differences were statistically significant (P = 0.005). The combined genotype with 2–3 risk genotypes was associated a statistically significantly increased risk of non-oropharyngeal cancer compared with those with 0 risk genotypes (OR = 1.42; 95% CI=1.07–1.88) after adjustment for age, sex, smoking, and drinking status. In the further stratified analysis by age, sex, smoking and drinking status, an increased risk associated with the combined genotype with 2–3 risk genotypes was more pronounced in younger subjects(≤57 years old) (OR = 1.61; 95% CI = 1.05–2.47), men (OR = 1.41; 95% CI = 1.01–1.98), ever smokers (OR = 1.43; 95% CI = 1.02–2.02) and ever drinkers (OR = 1.47; 95% CI = 1.03–2.11). However, there was no statistical evidence for interactions among these variables in further multivariate logistic regression models (data not shown).
Finally, because multiple tests had been performed, we estimated the false positive reporting probability (FPRP) for all the statistically significant results of non-oropharyngeal cancer. As shown in Table 4, the association of the combined genotype with 2–3 risk genotypes with the risk of non-oropharyngeal cancer yielded an FPRP value of 0.188 for a prior probability of 0.1, assuming that the OR for non-oropharyngeal cancer risk was 1.50 and had a statistical power of 0.78, which was a noteworthy finding. The FPRP values for a prior probability of 0.1 for the subgroup analyses by age, sex, smoking status and drinking status was 0.444, 0.311, 0.353 and 0.334, respectively, assuming that the OR for the combined genotype was 1.50 with a post-hoc statistical power of 0.54, 0.73, 0.67, and 0.68, respectively, suggesting some possible biases in these findings.
We further evaluated age at onset of SCCHN by MDM2 SNPs and p53 codo72 SNP in the case-only analysis with 1083 patients. The mean ages of SCCHN onset for the TT, TG, and GG genotypes of MDM2 SNP309 were 57.7 (±11.0), 56.8 (±11.0), and 56.3 (±12.2) years old, respectively. These differences in age of disease onset were not significantly different. Similar results were also shown for MDM2 A2164G and p53 codon 72 (data not shown). Then we further evaluated age at onset of oropahryngeal cancer (550 paptients) and non-oropharyngeal cancer (533 patients) for these three SNPs, respectively, and no statistically significant association with age at onset were observed (data not shown). However, there were distinct differences in age at onset of non-oropharyngeal cancer between sexes.
We found that MDM2 SNP309 genotype was associated with an earlier age at onset of non-oropharyngeal cancer in an allele-dependent manner (i.e., a younger age at onset with increasing number of G alleles, Ptrend = 0.007, Figure 2A). Specifically, female patients with TG and GG genotypes showed on average a 3-year (56.7 years) and 9-year (51.2 years) earlier age at onset of non-oropharyngeal cancer, respectively, compared to those with the wild-type TT (60.1 years). However, such findings were not observed in male patients (data not shown). When the genotypes of MDM2 SNP 309 and p53 codon 72 SNP were combined, there was also a trend of age at onset in a genotype-does dependent manner (Ptrend = 0.004, Figure 2B), which was not observed in male patients (data not shown). Notably, female patients with the combined MDM2 SNP309 GG and p53 codon 72 CC genotypes had the youngest age at onset (42.5 years), compared with those with the combined MDM2 TT and p53 codon 72 GG/CG genotypes (60.4 years). However, these differences in female patients were not observed for MDM2 A2164G SNP alone or in combination with p53 codon 72 SNP (data not shown).
In a study population of 1,083 SCCHN patients and 1,090 cancer-free controls of Caucasians, although none of the three SNPs in MDM2 and p53 (MDM2 SNP309, A2164G, and p53 codon 72) was alone associated with risk of overall SCCHN, a joint effect of these three SNPs may contribute to risk of SCCHN at non-oropharygenal sites. Furthermore, it appeared that the MDM2 SNP309 G-allele was associated with an early age at onset of non-oropharyngeal cancer in an allele-dose response manner in female patients, and this age effect could be further amplified by the p53 codon 72 C-allele. In contrast, we did not find similar effects for MDM2 A2164G and p53 codon 72, either individually or in combination. Given the role of MDM2 as a key inhibitor of p53, it is biologically plausible that these two SNPs in the MDM2 promoter region and p53 codon 72 SNP may collectively modulate risk of non-oropharyngeal cancer.
It had been demonstrated that the G-allele of SNP309 located in the promoter region of MDM2 increases the affinity of the transcriptional activator Sp1, resulting in high levels of MDM2 mRNA and MDM2 protein, thereby affecting p53 tumor suppressor activity and potentially cancer development in humans. Recently, in mouse models carrying either the polymorphic MDM2SNP309G or MDM2SNP309T allele , Post et al. found that MDM2SNP309G/G cells exhibit elevated MDM2 levels, reduced p53 levels, and decreased p53-dependent apoptosis in response to DNA damage; importantly, those mice with two MDM2SNP309G alleles have an attenuation of the p53 pathway resulting in a decreased latency to tumor formation and decreased survival. In addition, over-expression of MDM2 could be involved in downregulation of other important cellular proteins such as pRB, E2F1, and p19ARF, further promoting carcinogenesis [17,20]. Therefore, it is likely that MDM2 SNP309 may be a rate-limiting event in carcinogenesis .
A number of studies have investigated associations between SNPs of MDM2 SNP309 and p53 codon 72 and SCCHN risk, but the results are controversial [12,16,21–27]. A recent meta-analysis reported no association between the p53 codon 72 polymorphism and risk of oral carcinoma . So far, six published studies with relatively small sample sizes have investigated the association between MDM2 SNP309 and the risk or age at onset of SCCHN [28–33]. Our mini meta-analysis using these published data and ours in the present study showed that MDM2 SNP309 was not significantly associated with risk of SCCHN. It is possible that the effect of the MDM2 SNP309 on SCCHN risk may be modest and could not be detected in the present study, or the effect can be modified by other SNPs of other genes. Indeed, we did find that those who carried the two to three risk genotypes (i.e., MDM2 SNP309 GT/TT, 2164 AA, and p53 codon72 CC) appeared to have an increased risk of non-oropharyngeal cancer, and this risk was more pronounced among ever smokers and ever drinkers. Human papillomavirus (HPV)-related SCCHN occurs predominantly in the oropharynx, and HPV is an established etiologic agent of oropharyngeal cancer. It has been reported that HPV-related oropharyngeal cancer is most likely among those who are less smoking and drinking. In contrast, non-HPV related SCCHN (oral cavity, hypopharyngeal, and larynx) is more commonly linked with heavy tobacco and alcohol use [34–36]. Hence, our findings in non-oropharyngeal cancer suggest that genetic variants in MDM2 and P53 could modulate non-HPV-induced carcinogenesis through gene-environment interaction. However, we did not find statistical evidence for interactions among these variables in the current study, which could be possible that the limited sample size in the group of non-oroparyngeal cancer (530 cases) may not have enough power to detect such interactions. Further studies with large sample size are warranted to confirm our findings in non-oropharygeal cancer.
Although MDM2 SNP309 and p53 codon 72 SNPs have been demonstrated to be functional, the function of MDM2 promoter A2164G SNP has not yet been reported. MDM2 A2164G has been predicted to be functional by the computation tool of SNP Function Prediction (FuncPred, http://manticore.niehs.nih.gov/snpfunc.htm). Hence, based on its putative function and our data reported herein, further functional studies need to be performed on MDM2 A2164G and other untyped putative functional polymorphisms that are in LD with this SNP.
With respect to the association between MDM2 SNP309 and the age at onset of SCCHN in female patients, two published studies found that the GG genotype was associated with an early age at onset of SCCHN in an Asian population. A Japanese case-only study with 119 SCCHN patients found that the SCCHN patients with the GG genotype had a nine-year earlier tumor onset in comparison with those with the TT genotype . Huang et al. found that the MDM2 SNP309 G-allele was associated with at least a three-year earlier age of tumor onset in a Taiwanese population . In contrast, another study showed that patients with the GG genotype showed a five-year later age of tumor onset than those with the TT genotype in women in a Malaysian population . In the current case-only analysis with 1083 SCCHN patients, we found that the MDM2 SNP309 G-allele associated with an early age at onset of SCCHN at non-oropharygeal sites in female patients, which could be as early as 42.5 years old for women who also carried the p53 codon 72 C-allele.
It is likely that primarily female-specific hormones, such as estrogen, could allow for the SNP309 G-allele to accelerate tumor formation in women [37,38]. More recent studies reported that estrogen receptors were expressed in the majority of SCCHN tumors and cell lines [39–41]. It had been shown that estrogen signaling could regulate MDM2 expression levels [42,43]. SNP309 is located only 10 bp away from the potential estrogen receptor responsive region of the MDM2 promoter. Since SNP309 enhances the binding affinity of Sp1, a co-transcriptional activator of hormone receptor, it may affect the transcriptional regulation of MDM2 by estrogen signaling. Indeed, Hu et al. demonstrated that estrogen preferentially stimulated transcription of the MDM2 gene from the SNP309 G-allele and increased the levels of MDM2 protein in estrogen-responsive cells homozygous for SNP 309 (G/G), leading to further attenuation of the p53 pathway .
Although our results are somewhat inconsistent with those previously published findings on SCCHN [31–33], our findings are consistent with the biological function of the MDM2 SNP309 SNP and similar to previous reports on other cancer types including colorectal cancer, diffuse large B-cell lymphoma, soft tissue sarcoma, breast cancer and ovarian cancer, in which acceleration of tumor formation was shown to be associated with the G-allele, particularly in female patients [26,37,38,44]. For example, Bond G. et al. also demonstrated that the GG genotype was associated with a 13-year, 14-year, and a 10-year earlier age at onset of diffuse large B-cell lymphoma, soft tissue sarcomas and colorectal cancer, respectively, only in women [37,38]. A Finland study found that GG carriers showed a 2.7-year earlier age at onset of colorectal cancer than TT carries only in women but not in men ; however, that study found that MDM2 SNP309 was neither associated with risk nor with the age at onset of SCCHN, and further stratified analysis by sex was not reported in that study because of it small sample size (157 SCCHN patients and 185 controls). The discrepancies between our present study and those reported studies could be due to differences in ethnicity and study designs. Indeed, genotype distributions of MDM2 SNP309 may vary by ethnicity. For example, the frequencies of the GG, TG, and TT genotypes of MDM2 SNP309 among our Caucasians population were 11.9%, 43.3%, and 44.8%, respectively, compared with 27.5%, 41.7%, and 30.8% of 120 Japanese subjects in the study by Nakashima et al., and 27.1%, 51.3%, and 21.5% of 1,272 Taiwanese subjects in the study by Huang et al.. Therefore, these ethnic variations in genotype distributions and their influence on the risk of cancer need to be further investigated.
To the best of our knowledge, the present study is the first one using a large patient cohort to investigate the association between the risk and age at onset of SCCHN and MDM2 promoter SNPs, individually and in combination with p53 codon 72 SNP. Furthermore, our study population is ethnically homogeneous for Caucasians, decreasing possible biases from population stratification. Nevertheless, some potential limitations in the current study should be considered. Firstly, three SNPs were included in the analyses and the possibility of false-positive associations could not be ruled out because of multiple tests. The FPRP analysis showed that some possible biases may exist in our findings. Secondly, although our sample size was relatively large, the small sample size in subgroup analyses may have limited statistical power. Hence, our findings need to be confirmed by studies with larger sample sizes. Finally, HPV is another etiologic factor, in addition to tobacco and alcohol, for SCCHN, particularly for oropharyngeal cancer. However, when we evaluated the modification effects of polynorphisms in MDM2 and p53 on the risk of SCCHN, HPV status was not taken into account because only a small subset of SCCHN patients included in this study had available data on HPV infection. This limitation should be overcome in our future studies. Despite these limitations, our findings are biologically plausible. Our results provide some novel clues for the role of MDM2 and p53 SNPs in the development of non-orophryngeal cancer and age at onset of the disease.
In summary, we found that the combined risk genotypes of three SNPs in MDM2 and p53 (i.e., MDM2 SNP309 GT+GG, 2164AA, and p53 codon 72 CC) were significantly associated with risk of non-oropharyngeal cancer. We also found that the MDM2 SNP309 G-allele was associated with an earlier age at onset of non-oropharyngeal cancer in women but not in men, and this age effect could be amplified by the p53 codon 72 C-allele. Our findings indicate that MDM2 promoter SNPs, i.e., MDM2 SNP309 and MDM2 A2164G, as well as p53 codon 72 SNP may collectively contribute to the development of non-oropharyngeal cancer and that MDM2 SNP309 individually or in combination with p53 codon 72 may accelerate non-oropharyngeal cancer development in women. Further independent studies with large sample sizes and relative functional studies are needed to validate these findings.
We thank Ana Neumann, Margaret Lung, Jessica Fiske and Shara Challa for their assistance in recruiting the subjects and gathering the questionnaire information, Yawei Qiao, Jianzhong He, Kejing Xu and Min Zhao for laboratory assistance, and Dakai Zhu for his technical support. This work was partly supported by the National Institute of Health grants R01 CA131274 and R01 ES011740 (Q. Wei), ES015587 (D.G. Johnson), P50 CA097007 (Scott Lippman), and P30 CA016672 (The University of Texas M. D. Anderson Cancer Center). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.