We thank Ms. Margaret Lung, Ms. Angeli Fairley, Ms. Liliana Mugartegui, and Ms. Kathryn Tipton for assistance with patient recruitment, and Dr. Chong Zhao and Ms. Yingdong Li for laboratory assistance.
Mark E. Zafereo MD,
Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas
The first and third authors contributed equally to this article.
Because of the structural and biochemical similarities between the antitumor p53 and p73 proteins, the authors hypothesized that individuals who carry high-risk genotypes of p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms have a higher risk of developing second primary malignancy (SPM) after index squamous cell carcinoma of the head and neck (SCCHN).
A cohort of 1269 patients with index cases of SCCHN was recruited between May 1995 and January 2007 at The University of Texas MD Anderson Cancer Center and followed for SPM development. Patients were genotyped for p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms. A log-rank test and Cox proportional hazard models were used to compare SPM-free survival and SPM risk among different risk groups with the combined risk genotypes of the 2 polymorphisms.
The data demonstrated that patients with p53 WP + PP and p73 GC/GC genotypes had a worse SPM-free survival and an increased SPM risk compared with the corresponding p53 WW and p73 GC/AT + AT/AT genotypes. After combining the 2 polymorphisms, a borderline significantly or significantly reduced SPM-free survival and increased SPM risk were observed in the medium-risk group (p53 WW and p73 GC/GC or p53 P carriers and p73 AT carriers) and high-risk group (p53 P carriers and p73 GC/GC) compared with low-risk group (p53 WW and p73 AT carriers), respectively.
Squamous cell carcinoma of the head and neck (SCCHN) is among the most common cancers worldwide.1 SCCHN is characterized by highly aggressive tumor growth and results in significant morbidity, commonly in the form of disfigurement and loss or impairment in the ability to speak and swallow.2 These medical and psychosocial consequences are exacerbated by relatively stagnant survival rates over the past 30 years despite advances in treatment.3 The survival advantages afforded by new treatment modalities are undermined by poor prognosis of SCCHN, which is to some extent because of the increased likelihood of developing second primary malignancy (SPM).4
SPMs are estimated to occur in about 15% of SCCHN patients and are a significant cause of post-treatment morbidity and mortality.4 Although continued use of alcohol and tobacco,5, 6 and well as some cancer treatments,7, 8 have been determined to play a role in the development of SPM, these factors alone do not explain the risk of SPM. Many patients, including smokers and drinkers, never develop SPM, suggesting that genetic susceptibility may also contribute to SPM etiology.9 Determining a genetically susceptible risk group would allow better identification of high-risk SPM subgroups from cancer survivors. Identifying markers of risk for SPM would make possible improved initial treatment management and increased secondary prevention, currently limited to basic clinical post-treatment screenings.
Cell cycle control is paramount in maintaining normal growth and differentiation of cells. Both p53 and p73 are important tumor suppressor genes that regulate the cell cycle via apoptosis and cell cycle arrest. The p53 protein plays an important role in the prevention of carcinogenesis in that upon DNA damage from various agents it mediates pathways leading to DNA repair, cell cycle arrest, and apoptosis.10 Down-regulation of p53 leads to diminished DNA repair and poor cell cycle control, ultimately resulting in cellular malignancy.11 Furthermore, p53 has been shown to be mutated in most cancers, and approximately half of all SCCHNs exhibit such mutations.12, 13 Although the p73 protein does not function as a traditional tumor suppressor gene, its high level of sequence homology with the DNA-binding domains of p53 enables p73 to transactivate p53-response genes, resulting in cell cycle arrest, DNA repair, and apoptosis. Thus, the 2 proteins, p53 and p73, are interrelated and are considered members of the same family.14-16 In human malignancies involving p53 mutations, p73 expression has been found to be increased, suggesting an additional role for p73 as a compensator for p53 in the event of dysfunctional p53 mutations.17-20
A polymorphism of the p53 consisting of either proline or arginine at amino-acid position 72 has been found in a proline-rich domain necessary for full induction of apoptosis.21 Of the 2 amino acids, the Arg72 type has been shown to induce apoptosis with faster kinetics and suppresses transformation more efficiently than the p53 Pro72 variant.22 It has been proposed that this increased apoptotic ability is because of an increased ability of Arg72 to localize to the mitochondria, resulting in cytochrome c release into the cytosol and subsequent apoptosis.21 Research has suggested an association with the p53 codon 72 polymorphism with risk of several cancers and survival outcomes.23-26 Although findings suggest that p73 mutations are rare,17, 27 it is possible that genetic variation of p73 may lead to differences in susceptibility to cancer. Specifically, it is believed that the 2 linked, noncoding polymorphisms at exon 2 of p73 at positions 4 (G→A) and 14 (C→T) (the p73 G4C14-to-A4T14 polymorphism) affect p73 function by altering gene expression.28 Previous studies have documented the role of this polymorphism on risk of several cancers including SCCHN and survival outcomes.29-33
More recently, we have reported that each of p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms alone was associated with risk of SPM in patients after index SCCHN.34, 35 However, because these proteins do not function in isolation from each other, a combined analysis of both p53 and p73 polymorphisms has not been performed to determine the joint effects on risk of SPM in patients with index SCCHN. To test whether individuals who carry a higher number of risk genotypes of both p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms have a higher risk of SPM after index SCCHN, we analyzed the combined effect of these 2 polymorphisms in a cohort of 1269 index cases of SCCHN to compare SPM-free survival and SPM risk between different risk groups with the combined risk genotypes.
MATERIALS AND METHODS
This research was approved by the institutional review board of The University of Texas MD Anderson Cancer Center. Details and response rate for this study have been previously published.34, 35 For this combined analysis, the cases with index SCCHN were recruited through the Head and Neck Clinic at The University of Texas MD Anderson Cancer Center between May 1995 and January 2007 as part of an ongoing molecular epidemiological study.
At our institution, SCCHN patients are typically followed and monitored through their treatment and post-treatment courses with regularly scheduled clinical and radiographic examinations. On the basis of modified criteria of Warren and Gates,36 second lesions were considered SPMs if they were of different histopathologic type, occurred >5 years after treatment for the index tumor, and/or were clearly separated by normal epithelium based on clinical and radiographic assessment. Pulmonary lesions were considered to be SPM if they had a nonsquamous histology, or if they were isolated squamous lesions appearing >5 years after the initial SCCHN and felt to be SPM by the thoracic oncologist and thoracic surgeon. If there was a discrepancy or difference of opinions regarding the origin of the tumor (ie, recurrence vs SPM), the second lesion was classified as a local recurrence rather than an SPM.
Genomic DNA was isolated from patients' peripheral leukocyte pellets according to the manufacturer's instructions (Qiagen, Valencia, Calif). Genotyping of p53 and p73 polymorphisms was performed as previously described.30, 34 More than 10% of the samples were randomly selected and retested, and the results were 100% concordant.
Software used for analysis was Statistical Analysis System software (SAS version 9.1.3; SAS Institute, Cary, NC). Statistical significance was set at P < .05, and all tests were 2-sided. Chi-square tests were used to assess differences in demographic and clinical variables, as well as genotype distributions between the groups of patients who developed SPM and those who remained SPM free.
Kaplan-Meier methods were used to determine whether there were significant differences (P < .05) in SPM-free survival between different risk groups with the combined genotypes. Both univariate and multivariate Cox proportional hazards regression models were used for assessment as previously published.34, 35 Details of building the multivariate proportional hazards model were described previously.34, 35 After a stepwise search strategy was used in building the multivariate proportional hazards model, the final, fully adjusted Cox regression models included age, sex, ethnicity, and smoking and alcohol status.
The demographics and clinical variables for the study patients are shown in Table 1. Overall, a total of 1269 SCCHN patients were included in the study. Of those recruited, 1160 patients remained SPM free, and 109 developed SPMs. SPM-free patients and those who developed SPM appeared to have no significant differences in sex, ethnicity, and alcohol drinking status (P = .704, P = .100, and P = .124, respectively); however, patients with SPM were more likely to be older (P < .001) and smokers (P = .021). Compared with the SPM-free group, patients who developed SPM had similar characteristics with respect to index cancer site (P = .220), index cancer stage (P = .866), and treatment (P = .910).
Table 1. Distribution of Selected Characteristics of the Patient Cohort (N = 1269)
Combined Effects of the p53 and p73 Polymorphisms on Risk of SPM
In this study, we examined the distribution of the combined p53 and p73 genotypes among the patients who developed SPM and those who remained SPM free, and the associations with risk of SPM (Table 2). Because both p53 and p73 variant homozygous genotypes were relatively uncommon, we combined the variant homozygous with the heterozygous genotypes for the final analyses. As previously reported,34, 35 the p53 WP + PP variant genotypes were more common among patients with SPM than among patients who remained SPM free (P = .008) and was associated with approximately 60% increased risk for SPM compared with the WW genotype (adjusted hazard ratio [HR], 1.58; 95% confidence interval [CI], 1.07-2.34); the p73 GC/GC genotype was more common in patients who developed SPM (P = .019) and was associated with approximately 70% increased risk for SPM compared with the GC/AT + AT/AT variant genotypes (HR, 1.68; 95% CI, 1.12-2.52). Because p53 and p73 share a common pathway, we used a meaningful combination of the 2 polymorphisms to determine whether the combined risk genotypes modified the risk of SPM. The patients carrying p53 WW and p73 GC/AT + AT/AT genotypes were placed in the low-risk group; the patients with p53 WW and p73 GC/GC or p53 WP + PP and p73 GC/AT + AT/AT were placed in the medium-risk group; and the patients with p53 WP + PP and p73 GC/GC were placed in the high-risk group. Our data demonstrated that the patients had significant differences in SPM-free survival among the 3 different risk groups (overall log-rank, P = .0008; specifically, P = .0004 for high-risk vs low-risk; P = .0860 for medium-risk vs low-risk; and P = .0096 for high-risk vs medium-risk, respectively) (Fig. 1). After adjusting for age, sex, ethnicity, tobacco smoking, and alcohol drinking, the patients in the medium-risk and high-risk groups had an approximately 1.7-fold and 2.7-fold elevated risk for developing an SPM compared with those in the low-risk group, respectively (adjusted HR, 1.66; 95% CI, 1.00-3.06 for the medium-risk group and adjusted HR, 2.69; 95% CI, 1.44-5.00 for the high-risk group). Furthermore, a dose-response relationship was observed among the 3 risk groups with different numbers of risk genotypes of the 2 polymorphisms (Ptrend = .0007).
Table 2. SPM Risk Associated With p53 and p73 Polymorphisms After Index Squamous Cell Carcinoma of the Head and Neck
Chi-square test for differences in the distribution of p73 genotypes between the patients who developed SPM and the patients who did not.
Adjusted for age, sex, ethnicity, tobacco smoking, and alcohol drinking in a Cox model.
WP + PP
GA + AA
Combined risk genotypes
Stratification Analysis of the Combined p53 and p73 Genotypes With Risk of SPM
To further evaluate risk of SPM for specific subgroups, the data were further stratified by age, sex, ethnicity, smoking status, drinking status, treatment, index tumor stage, and tumor site (Table 3). In each subgroup except females, the patients in the medium-risk group had an increased risk for SPM compared with those in the low-risk group, although the increased risk was only statistically significant for males (adjusted HR, 2.50; 95% CI, 1.12-5.56). There was an increased risk of SPM for all subgroups in the high-risk group, and the increased risk was statistically significant for patients older than 57 years (adjusted HR, 3.46; 95% CI, 1.52-7.86), males (adjusted HR, 3.32; 95% CI, 1.47-7.52), non-Hispanic whites (adjusted HR, 2.56; 95% CI, 1.28-5.13), smokers (adjusted HR, 3.0; 95% CI, 1.49-6.06), and drinkers (adjusted HR, 2.53; 95% CI, 1.28-4.99). Furthermore, the patients with late stage index SCCHN (3 or 4) at time of diagnosis, those with DNA-damaging treatments (radiotherapy or chemotherapy), and those with index nonoropharyngeal cancer had a significantly pronounced SPM risk (adjusted HR, 2.79; 95% CI, 1.37-5.70, adjusted HR, 3.17; 95% CI, 1.57-6.39, and adjusted HR, 2.92; 95% CI, 1.33-6.41, respectively).
Table 3. Stratification Analysis of Combined Association of p53 and p73 Polymorphisms With SPM Risk
Abbreviations: CI, confidence interval; HR, hazard ratio; Non-HW, non-Hispanic white; Ref., reference; SPM, second primary malignancy; TA, tobacco associated.
Adjusted for age, sex, ethnicity, tobacco smoking, and alcohol drinking in a Cox model.
Early (1 or 2)
Late (3 or 4)
The role of both p53 and p73 proteins in modulating carcinogenesis has been well established, and there is an apparent difference between the different polymorphic forms. In our study of 1269 SCCHN patients, we analyzed the 2 well-known polymorphisms, p53 codon 72 and p73 G4C14-to-A4T14, and their associations with the risk of SPM. Our previous studies have shown that the p53 and p73 polymorphisms individually modify the risk of SPM34, 35; however, no study has been done to assess the joint effect of the 2 polymorphisms on risk of SPM. In this study, we found that the p53 and p73 polymorphisms jointly borderline significantly (medium-risk group) or significantly (high-risk group) increased the risk of SPM, and such joint effect on risk of SPM was more pronounced in certain subgroups, suggesting that the p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms may jointly modify the risk of SPM after index SCCHN.
It is biologically plausible that p53 and p73 play a role in the development of SPM because these 2 proteins have similar biological properties, and each may play a similar role in the regulation of cell cycle control, DNA repair, and apoptosis. Furthermore, members of the p53 family, including p53 and p73, have been shown to interact in development of human cancers. In malignancies associated with loss of p53 expression, an increased expression of p73 has been observed in malignant tissues compared with adjacent normal tissues, providing evidence that p73 may compensate for the loss of p53 function.17-20 Thus, these 2 proteins may also have a combined effect on SPM risk.
The 2 polymorphic forms of p53 may result in a marked alteration of the primary structure of the protein, thereby modifying its biochemical properties and effects.36 The Pro72 variant interacts more effectively with elements of the transcriptional machinery and is capable of inducing higher levels of transcriptional activity than the Arg72 form. It also induces G1 arrest and more effectively activates DNA repair system.21, 22, 37, 38 However, Arg72 has demonstrated apoptotic induction with faster kinetics and suppresses transformation more efficiently than the Pro72 variant.22, 37 Thus, the differences in these biological activities caused by each of 2 polymorphic variants may result in a different effect modification of SPM risk. Conversely, the p73 protein, similar to p53, also plays a role in DNA repair, cell cycle regulation, and apoptosis, and therefore influences tumor development and progression. The location of the p73 G4C14-to-A4T14 polymorphism is upstream of the initiating AUG of exon 2 and may have a role in the formation of a stem-loop structure, which may result in an alteration of gene expression by altering the initiation of translation,39 thereby modifying the risk of human cancer, including SPM. However, further studies are needed to confirm these biological functions of the 2 polymorphisms.
Either p53 or p73 polymorphism has been reported to be associated with risk for several human malignancies, including SCCHN.30, 32, 40 Recently, we also found that each of the p53 codon 72 and the p73 G4C14-to-A4T14 polymorphisms moderately modified the risk of SPM after an index SCCHN.34, 35 Because there is considerably biological interaction between p53 and p73 proteins, a recent case-control study found a combined effect of p53 and p73 polymorphisms on risk of head and neck cancer in an Italian population.41 Similarly, we therefore undertook the current study with a combined analysis for the p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms and their association with risk for SPM. We found that the results from the current study were consistent with the notion that these 2 polymorphisms may jointly increase risk of SPM. Moreover, the risk of the combined risk genotypes of the 2 polymorphisms was more pronounced in several subgroups, including older patients, males, non-Hispanic whites, smokers, drinkers, patients with late stage index SCCHN, those with DNA damaging treatments, and patients with nonoropharyngeal cancer, in each of which we found similar results for each of the 2 polymorphisms.34, 35 It is possible that these polymorphisms may affect the DNA repair capacity of damage induced by tobacco and alcohol carcinogens, DNA-damaging therapy, or reduced DNA repair capacity by aging. Although how ethnicity affects the SPM risk is not clear, it is possible that certain behaviors and other genetic factors may play a role in development of SPM. In addition, the late stage patients with more extensive treatment modalities including chemotherapy and/or radiotherapy may have more extensive DNA damage.
Our study has some limitations. First, although the study was performed in a large cohort of SCCHN patients, approximately 85% of the patients were non-Hispanic white. Nevertheless, ethnicity was adjusted for in the multivariate analyses. Second, although demographics, exposure, and clinical data for the cohort were collected prospectively, the clinical outcomes such as SPM were collected retrospectively. Therefore, follow-up time was limited, and patients may not have had enough time to develop SPM or could have been lost to follow-up. Also, the prevalence of never-smokers and late stage index cancer patients, and our strict criteria for determining SPM, resulted in an SPM rate that was lower than expected. Therefore, the low rate of SPM limited statistical power for the analysis, particularly for the stratified analysis. Finally, data on human papillomavirus (HPV) status, 1 of the major risk factors for SCCHN, was not taken into account. Although the major risk factor for SCCHN is exposure to tobacco and/or alcohol, currently sufficient evidence concludes that there is strong and consistent association between oncogenic HPV (principally type 16 and occasionally type 18) and a distinct subset of head and neck cancers (ie, soft palate, palatine tonsil, and base of tongue/lingual tonsil).42-44 Despite declining smoking rates in the United States, the rising incidence of oral cavity and pharyngeal cancer within certain sites, particularly the base of tongue, tonsil, and oropharynx, among white men born since the mid-1940s appears to be caused by the increasingly prevalent infection of oncogenic subtypes of HPV and may reflect changes in sexual practices since the mid-1960s.45 Molecular studies have shown that oncogenic E6 and E7 proteins of HPV have a high binding affinity for p53 and RB, promoting the ubiquitination and complete degradation of these tumor suppressor genes, leading to the deregulation of cell cycle control and subsequent tumor development.42-44 Studies also have shown that HPV-positive patients appear to be a distinct epidemiologic, clinical, and molecular subgroup that exhibits unique clinical behaviors and treatment responses compared with HPV-negative patients.43, 44 In the current study, the absence of HPV status did not allow us to evaluate its potential influence on the development of SPMs in patients with index SCCHN. Thus, we will closely monitor the role of HPV in the outcomes of SCCHN patients in our future studies when a much larger patient cohort with HPV-associated tumors becomes available.
In conclusion, our results show that p73 and p53 polymorphisms jointly significantly increase the risk of SPM development after an index SCCHN; such risk was more pronounced in several subgroups. This study provides evidence that simultaneous presence of the p53 codon 72 and p73 G4C14-to-A4T14 polymorphisms may have joint effects on increased risk of SPM, and the combination of the 2 polymorphisms may provide more comprehensive and accurate estimates of the risk of SPM than the single polymorphism alone.
Funded by a Research Training Award, the American Laryngological, Rhinological, and Otological Society (E.M.S.); The University of Texas MD Anderson Cancer Center Start-up Funds (E.M.S.); National Institutes of Health grant K-12 88084 (E.M.S., faculty trainee; R. C. Bast, principal investigator); National Institute of Environmental Health Sciences grant R01 ES-11740 (Q.W.); and NIH grants CA135679-01 (G.L.) and CA133099-01A1 (G.L.).