Associations among smoking, MGMT hypermethylation, TP53-mutations, and relapse in head and neck squamous cell carcinoma

Abstract

Background

Epigenetic silencing of the O6-methylguanine-DNA methyltransferase (MGMT) DNA repair enzyme via promoter hypermethylation (hmMGMT) may increase mutations in the TP53 oncosuppressor gene and contribute to carcinogenesis. The effects of smoking, which is a risk factor for head and neck squamous cell carcinoma (HNSCC), were investigated to determine whether they up- or down-regulate hmMGMT. Additionally, the impact of hmMGMT and disruptive TP53-mutations on relapse was investigated in patients with HNSCC.

Methods

This study included 164 patients with HNSCC who were negative for both p16 protein expression and human papilloma virus infection. The association of smoking and hmMGMT was investigated using multiple logistic regression analysis. Competing risk regression was used to evaluate the effects of hmMGMT and TP53-mutations in exon 2 to 11 on relapse of HNSCC.

Results

hmMGMT was observed in 84% of the 164 patients. TP53-mutations, specifically, G:C>A:T transition, were more frequent in patients with hmMGMT (32%) than in those without hmMGMT (8%). The frequency of disruptive TP53-mutations was not significantly different between groups. Compared with nonsmoking, heavy smoking of 20 pack-years or more was significantly associated with decreased hmMGMT (adjusted odds ratio, 0.08; 95% CI, 0.01 to 0.56; P = 0.01). Patients who had both hmMGMT and disruptive TP53-mutations showed a significantly higher relapse rate than all other patients (subdistribution hazard ratio, 1.77; 95% CI, 1.07 to 2.92; P = 0.026).

Conclusions

It was found that hmMGMT was suppressed by heavy smoking, and hmMGMT combined with disruptive TP53-mutations may indicate a poor prognosis in patients with HNSCC.

About the Authors:

Shinichi Matsuda

Roles Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing

Affiliations Division of Molecular Epidemiology, The Jikei University School of Medicine, Tokyo, Japan, Real World Data Science Department, Chugai Pharmaceutical Co. Ltd., Tokyo, Japan

ORCID logo http://orcid.org/0000-0003-1822-1090

Aki Mafune

Roles Investigation, Writing – original draft

Affiliations Division of Molecular Epidemiology, The Jikei University School of Medicine, Tokyo, Japan, Division of Kidney and Hypertension, Department of Internal Medicine, The Jikei University School of Medicine, Tokyo, Japan

ORCID logo http://orcid.org/0000-0002-6160-2879

Nagisa Kohda

Roles Investigation

Affiliation: Division of Molecular Epidemiology, The Jikei University School of Medicine, Tokyo, Japan

Takanori Hama

Roles Conceptualization, Data curation, Investigation

Affiliations Division of Molecular Epidemiology, The Jikei University School of Medicine, Tokyo, Japan, Department of Oto-Rhino-laryngology, The Jikei University School of Medicine, Tokyo, Japan

Mitsuyoshi Urashima

Roles Conceptualization, Formal analysis, Methodology, Supervision, Writing – review & editing

* E-mail: [email protected]

Affiliation: Division of Molecular Epidemiology, The Jikei University School of Medicine, Tokyo, Japan

Introduction

Head and neck squamous cell carcinoma (HNSCC) is one of the most common cancers worldwide. Approximately 550,000 cases of HNSCC are newly diagnosed each year in the world, and only 40–50% of patients with HNSCC survive for 5 years [1]. It is well known that HNSCC is a multifactorial disease with contributing etiologies including tobacco smoking, alcohol consumption, and infection with the human papillomavirus (HPV) [2, 3]. In addition, some reports have shown the importance of epigenetic mechanisms in the development and progression of HNSCC and other cancers [4, 5].

O6-methylguanine DNA methyltransferase (MGMT) is one of the DNA repair enzymes that protects genes from mutations by directly removing cytotoxic alkyl adducts from the O6 position of guanine [6]. The expressions of MGMT RNA and protein are decreased by methylation of a CpG island in its promoter region [7, 8]. Thus, aberrant hypermethylation of the MGMT promoter region (hmMGMT) may hamper its DNA repair function, allowing mutations of G:C>A:T transition in TP53, as well as other carcinogenic genes, in various cancers [9–12].

Recently, genotoxic stressors such as tobacco smoking have been investigated with regards to their possible involvement in the regulation of MGMT [13]. Several studies have demonstrated an increase in MGMT activity/expression in the normal/tumor tissue of smokers compared to non-smokers, suggesting the possible role of tobacco smoking in regulating MGMT protein expression in the tissue [14–16]. However, studies of the effect of smoking on hypermethylation of the MGMT promoter (hmMGMT) reported conflicting results. In HNSCC, there has been only one study that showed that hmMGMT was unchanged by smoking [17]. Regarding other types of cancers, hmMGMT was reported to be upregulated in lung adenocarcinoma [18], downregulated in non-small cell lung cancer [19], or unchanged in non-small cell lung cancer [20] by smoking. One possible explanation for this discrepancy may be the differences in the analysis methods used. For example, one study defined smoking status as a binary characteristic (i.e. nonsmoker and smoker), whereas the other study defined smoking status based on the degree of smoking (i.e. pack-years). In addition, because the numbers of patients in the previous studies were relatively small, the association between smoking and hmMGMT was often evaluated by means of a simple chi-squared test, or evaluated by adjustment for only limited confounders. Thus, there were no studies that considered enough confounders, such as cancer stage, primary site of cancer, differentiation, and degree of alcohol consumption. In the present study, the aim was to clarify whether smoking enhances or suppresses hmMGMT in HNSCC by performing multivariate adjustment for potential confounders. Additionally, the effects of hmMGMT and TP53 mutations on relapse in patients with HNSCC were analyzed.

Materials and methods

Ethics statement

The study protocol was reviewed and approved by the Ethics Committee for Biomedical Research of the Jikei Institutional Review Board. Written, informed consent was obtained from all patients enrolled in the study.

Study design

This study was a post hoc analysis of our prospective cohort study [21, 22], which was conducted at Jikei University Hospital from March 2006 to November 2012. The entire process of study design, data monitoring, and analyses was performed at the Division of Molecular Epidemiology. Eligible participants were Japanese patients with HNSCC (oropharyngeal, hypopharyngeal, laryngeal, oral cavity, and sinonasal cancer) aged 20 years and over, who had newly diagnosed or recurrent disease, and who had surgical resection with curative intent before chemoradiotherapy. Clinical information was obtained from clinical and surgical charts. The tumor node metastasis (TNM) classification and cancer stages were determined according to the 6th Union for International Cancer Control TNM classification and stage groupings.

Based on the above cohort, this study excluded patients with high-risk HPV infections (16/18/31/33/35/52b/58) and patients who tested positive for p16, because this subpopulation is known to have a different etiology and pathogenesis from smoking/alcohol-induced HNSCC [1]. HPV infection was detected using multiplex polymerase chain reaction (PCR) with the TaKaRa Human Papillomavirus Typing Set #6603 following the manufacturer’s protocol (Takara Bio Inc., Shiga, Japan). Positive p16 expression, which was defined as strong and diffuse nuclear and cytoplasmic staining in at least 70% of tumor cells was detected by immunohistochemistry using a rabbit monoclonal antibody to p16 (Anti-CDKN2A/p16INK4a antibody [EPR1473]): Abcam plc, Science Park, Cambridge, England).

Smoking and alcohol consumption

Patients were divided into the following three groups based on smoking status prior to diagnosis of HNSCC: (1) nonsmokers, defined as patients who had never used tobacco or had stopped using tobacco for more than 20 years; (2) moderate smokers, defined as current or past smokers who smoked less than 20 pack-years within the last 20 years; and (3) heavy smokers, defined as current or past smokers who had smoked 20 pack-years or more within the last 20 years. This definition of heavy smokers is consistent with the study that reported that a cumulative dose corresponding to 20 cigarettes per day over 10–20 years or 10–20 pack-years is associated with a clinically relevant increase in morbidity [23, 24].

Patients were divided into the following three categories based on average daily alcohol consumption during the 20 years preceding diagnosis of HNSCC: (1) non-drinkers, defined as non-drinkers or light drinkers who consumed less than one drink per day; (2) moderate drinkers, defined as drinkers who consumed at least one but less than two drinks per day; and (3) heavy drinkers, defined as drinkers who consumed two or more drinks per day. One drink was defined as containing approximately 10 g of alcohol, which is equal to 30 mL of hard liquor, 100 mL of wine containing 12% alcohol, or 360 mL of beer.

Samples

With each patient’s consent, tumor and margin samples from the primary site, but not metastatic sites, were collected. These samples were rapidly frozen and stored at -80 °C after excision. The cancer tissue was divided into two specimens: one for pathological confirmation in which the sample was composed of >70% cancer cells and the other for DNA extraction. DNA was extracted and purified using the QIAamp DNA Micro Kit 50 (QIAGEN, Tokyo, Japan), and the DNA concentration of the samples was measured using NanoVue plus (General Electric Healthcare Japan, Tokyo, Japan). Samples were then frozen at -80 °C until use.

Methylation-specific PCR for detection of hypermethylation of the MGMT promoter

The methylation-specific polymerase chain reaction (MSP) was used to distinguish between hmMGMT and non-hmMGMT. Briefly, DNA samples extracted from tumor tissues were treated with bisulfite using MethylEasy Xceed (Takara Bio Inc.) according to the manufacturer’s protocol. Subsequently, MSP was carried out using the EpiScope MSP kit (Takara Bio Inc.). The primers used for PCR were described previously [25]. PCR was carried out in a 50-μL volume containing 4 μg of bisulfite-treated DNA, 25 μL of 2xMSP buffer, 1.2 μL of MSP enzyme, 0.5 μL of 100xSYBR Green I, 16.3 μL of nuclease-free water, and 1.5 μL of each of the two primers. The reaction was incubated at 95 °C for 30 sec, followed by 35 cycles at 98 °C for 5 sec, 55 °C for 30 sec, and 72 °C for 1 min, with a final incubation step at 16 °C.

Analysis of TP53-mutation

Exons 2 to 11 of the TP53-gene were amplified using PCR with purchased primers following the manufacturer’s protocol (Nippon Gene Co. Ltd., Tokyo, Japan), cloned, and then sequenced using the ABI PRISM 3700 Genetic Analyzer (Applied Biosystems, Foster City, CA). Disruptive TP53-mutations are defined as non-conservative mutations located inside the key DNA-binding domain (L2-L3 region) or stop codons in any region [26]. The missense changes (V31I, P36P, P47S, P72R, R72R, R158R, R213R, V217M, P222P, T312S, and G360A) reported as single nucleotide polymorphisms [27] were not included in the total TP53 mutations.

Statistical analysis

Patients’ characteristics and TP53-mutation status were compared between the hmMGMT and non-hmMGMT groups using Pearson’s chi-squared test, Student’s t-test, and the Mann-Whitney test, as appropriate. To clarify whether smoking enhances or suppresses hmMGMT, odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using the following multivariate logistic regression models: Model I was adjusted only by smoking status (nonsmoker, moderate smoker, heavy smoker); Model II was adjusted by age, sex, alcohol consumption (none, moderate, heavy), and primary site of tumor, in addition to the variables in model I; Model III was adjusted by tumor stage (stage I to IV), in addition to the variables in model II; Model IV was adjusted by tumor cell differentiation status (well, moderately, poorly differentiated), in addition to the variables in model III.

In survival analyses, the time from surgery to relapse was used to calculate the relapse-free ratio. To evaluate the effects of hmMGMT and TP53-mutations on the relapse of HNSCC, cumulative incidence functions (CIFs) were applied by considering patients’ death by causes other than the relapse as the competing risk. Competing risk regression was performed by the Fine and Gray subdistribution hazard model [28]. A value of P < 0.05 was considered significant. All statistical analyses were performed using STATA 14.2 (STATA Corp., College Station, TX).

Results

Associations among hmMGMT, TP53-mutations, and patient characteristics

A total of 164 patients who were negative for both HPV gene and p16 were analyzed. The patients’ characteristics stratified by the methylation status of MGMT (hmMGMT vs. non-hmMGMT) are summarized in Table 1. Eighty-four percent of the study population was classified as hmMGMT. There were no significant differences in age, sex, primary site of tumor, differentiation, smoking, or drinking status between hmMGMT and non-hmMGMT patients. On the other hand, cancer stages were more advanced in hmMGMT patients than in non-hmMGMT patients (P < 0.001).

[Figure omitted. See PDF.]

Table 1. Patients’ characteristics according to the methylation status of MGMT.

https://doi.org/10.1371/journal.pone.0231932.t001

Frequencies of TP53-mutation spectra were compared between the hmMGMT and non-hmMGMT groups. The proportion of patients who had at least one TP53-mutation was significantly higher in the hmMGMT (76%) than in the non-hmMGMT (50%) group (P = 0.007). Patients with hmMGMT tended to have a greater number of TP53-mutations per patient than patients with non-hmMGMT (P = 0.003). Regarding the type of TP53 mutation, G:C>A:T transition was significantly more common in hmMGMT (32%) than in non-hmMGMT (8%) (P = 0.012). Frequencies of disruptive TP53-mutations were 26% and 15%, respectively, and they were not significantly different (P = 0.244). No other type of mutation was significantly different between hmMGMT and non-hmMGMT.

Association between smoking and hmMGMT

To determine whether smoking is associated with hmMGMT, multivariate logistic regression analyses were performed (Table 2). In Model I, smoking status did not show significant associations with hmMGMT. In contrast, in models II, III, and IV, heavy smoking was significantly associated with a reduced frequency of hmMGMT (in model IV, adjusted OR, 0.08; 95% CI, 0.01 to 0.82; P = 0.03).

[Figure omitted. See PDF.]

Table 2. Effects of smoking on hmMGMT using logistic regression models.

https://doi.org/10.1371/journal.pone.0231932.t002

Association between hmMGMT and relapse

The impact of hmMGMT and disruptive TP53-mutations on relapse of HNSCC was investigated by competing risk regression. Patients with hmMGMT did not show a significantly higher risk of relapse compared with those without hmMGMT (subdistribution hazard ratio [SHR], 1.73; 95% CI, 0.83 to 3.59; P = 0.141). In addition, patients with disruptive TP53-mutations did not show a significantly higher risk of relapse compared with those without the mutations (SHR, 1.48; 95% CI, 0.89 to 2.45; P = 0.129). In contrast, the subgroup of patients who were positive for both hmMGMT and disruptive TP53-mutations showed a significantly higher risk of relapse than all of the other patients (SHR, 1.77; 95% CI, 1.07 to 2.92; P = 0.026) (Fig 1).

[Figure omitted. See PDF.]

Fig 1. Competing risk regression for relapse of HNSCC.

https://doi.org/10.1371/journal.pone.0231932.g001

Discussion

This study demonstrated that heavy smoking (20 pack-years and more) was inversely associated with hmMGMT after adjusting for possible confounders. The previous studies [17–20] did not adjust for several confounders such as tumor stage and alcohol consumption. In the present study, the degree of smoking (non-, moderate-, heavy smokers) was considered, which has not been previously distinguished but was considered a binary characteristic (never smokers or smokers) in other studies.

Previously, smoking was reported to upregulate MGMT protein expression and activity [13], which is consistent with the present result that heavy smoking downregulates the methylation of MGMT because this is expected to result in upregulation of MGMT RNA and protein expressions. A similar association between lower levels of hmMGMT and smoking was reported in colorectal adenoma [29]. In the present study, as a kind of biological defense mechanism, it was speculated that smoking might trigger a variety of gene mutations and simultaneously upregulate MGMT expression through demethylation of the MGMT promoter region in order to repair G:C>A:T transition. Additional studies are necessary to investigate the reasons for the conflicting results obtained from different types of cancers previously. As of now, there would be possibilities of effects from ethnic diversity or different cancer pathogeneses.

In agreement with previous reports [30, 31], there was a higher frequency of TP53 mutation in patients with hmMGMT than in those with non-hmMGMT. Regarding the type of TP53 mutation, G:C>A:T transition was significantly more common in patients with hmMGMT than in those with non-hmMGMT. The obtained result is theoretically plausible, since decreased expression of MGMT protein through hmMGMT allows O6-alkylguanine adducts to pair with thymine during DNA replication, resulting in a G:C>A:T transition mutation [9]. The present findings showed that hmMGMT is associated with a high frequency of TP53 mutations, particularly with G:C>A:T transitions in HNSCC.

A previous study showed that the prognostic value of TP53-mutation varied by the prediction method used, and Poeta rules, which is a prediction algorithm based on whether the mutation is disruptive or non-disruptive, did not significantly predict the prognosis of HNSCC [32]. In the present study, patients who were positive for hmMGMT and disruptive TP53-mutations showed a higher relapse rate. hmMGMT may allow mutations not only in the TP53 gene, but also in the genes of other oncosuppressors and oncogenes [33]. Thus, these patients that had both hmMGMT and disruptive TP53-mutations might have shown a poor prognosis.

This study has several limitations. First, the study included not only patients with newly diagnosed, but also those with recurrent HNSCC. Thus, the previous therapies might have affected the methylation status of MGMT in patients with recurrent HNSCC. In addition, because all patients in this study underwent surgery, the proportion of hmMGMT obtained from this study may not indicate the proportion in the whole HNSCC population. Second, the effects of primary sites of HNSCC could not be adequately explored due to the limited sample size. Third, most patients in the present study had advanced stage III to IV disease (74%). Therefore, generalization of the present findings to patients in earlier stages would be limited. Last, MGMT RNA and protein expressions were not measured. However, several reports have shown a significant correlation between MGMT methylation status and its protein expression in patients with HNSCC [31, 34, 35].

In conclusion, hmMGMT was suppressed by heavy smoking, and hmMGMT, combined with disruptive TP53-mutations, may be associated with a poor prognosis in patients with HNSCC.

Supporting information

[Figure omitted. See PDF.]

S1 Data. Data set of the present study.

https://doi.org/10.1371/journal.pone.0231932.s001

(XLSX)

Acknowledgments

The authors would like to thank Hiroaki Suga and Takeshi Mimura for processing the samples and Chikako Sakanashi for sequencing.

Citation: Matsuda S, Mafune A, Kohda N, Hama T, Urashima M (2020) Associations among smoking, MGMT hypermethylation, TP53-mutations, and relapse in head and neck squamous cell carcinoma. PLoS ONE 15(4): e0231932. https://doi.org/10.1371/journal.pone.0231932

References

1. Leemans CR, Braakhuis BJ, Brakenhoff RH. The molecular biology of head and neck cancer. Nature reviews Cancer. 2011;11(1):9–22. Epub 2010/12/17. pmid:21160525.

2. Choudhury JH, Ghosh SK. Gene–environment interaction and susceptibility in head and neck cancer patients and in their first-degree relatives: a study of Northeast Indian population. Journal of Oral Pathology & Medicine. 2015;44(7):495–501.

3. Fotopoulos G, Pavlidis N. The role of human papilloma virus and p16 in occult primary of the head and neck: A comprehensive review of the literature. Oral oncology. 2015;51(2):119–23. pmid:25467774

4. Hansen KD, Timp W, Bravo HC, Sabunciyan S, Langmead B, McDonald OG, et al. Increased methylation variation in epigenetic domains across cancer types. Nature genetics. 2011;43(8):768–75. pmid:21706001

5. Rodriguez-Paredes M, Esteller M. Cancer epigenetics reaches mainstream oncology. Nat Med. 2011;17(3):330–9. Epub 2011/03/10. pmid:21386836.

6. Pegg AE. Repair of O 6-alkylguanine by alkyltransferases. Mutation Research/Reviews in Mutation Research. 2000;462(2):83–100.

7. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer research. 1999;59(4):793–7. Epub 1999/02/24. pmid:10029064.

8. Esteller M. Relevance of DNA methylation in the management of cancer. The lancet oncology. 2003;4(6):351–8. pmid:12788407

9. Esteller M, Risques RA, Toyota M, Capella G, Moreno V, Peinado MA, et al. Promoter hypermethylation of the DNA repair gene O(6)-methylguanine-DNA methyltransferase is associated with the presence of G:C to A:T transition mutations in p53 in human colorectal tumorigenesis. Cancer research. 2001;61(12):4689–92. Epub 2001/06/19. pmid:11406538.

10. Wu JY, Wang J, Lai JC, Cheng YW, Yeh KT, Wu TC, et al. Association of O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation with p53 mutation occurrence in non-small cell lung cancer with different histology, gender, and smoking status. Annals of surgical oncology. 2008;15(11):3272–7. Epub 2008/08/21. pmid:18712569.

11. Watanabe T, Katayama Y, Komine C, Yoshino A, Ogino A, Ohta T, et al. O6-methylguanine-DNA methyltransferase methylation and TP53 mutation in malignant astrocytomas and their relationships with clinical course. International journal of cancer Journal international du cancer. 2005;113(4):581–7. Epub 2004/09/30. pmid:15455376.

12. Wang K, Wang YY, Ma J, Wang JF, Li SW, Jiang T, et al. Prognostic value of MGMT promoter methylation and TP53 mutation in glioblastomas depends on IDH1 mutation. Asian Pacific journal of cancer prevention: APJCP. 2014;15(24):10893–8. Epub 2015/01/22. pmid:25605197.

13. Christmann M, Kaina B. O(6)-methylguanine-DNA methyltransferase (MGMT): impact on cancer risk in response to tobacco smoke. Mutation research. 2012;736(1–2):64–74. Epub 2011/06/29. pmid:21708177.

14. Mattern J, Koomägi R, Volm M. Smoking-related increase of O6-methylguanine-DNA methyltransferase expression in human lung carcinomas. Carcinogenesis. 1998;19(7):1247–50. pmid:9683184

15. Drin I, Schoket B, Kostic S, Vincze I. Smoking-related increase in O6-alkylguanine-DNA alkyltransferase activity in human lung tissue. Carcinogenesis. 1994;15(8):1535–9. pmid:8055630

16. Nozoe T, Korenaga D, Kabashima A, Sugimachi K. Smoking-related increase of O 6-methylguanine-DNA methyltransferase expression in squamous cell carcinoma of the esophagus. Cancer letters. 2002;184(1):49–55. pmid:12104047

17. Puri SK, Si L, Fan CY, Hanna E. Aberrant promoter hypermethylation of multiple genes in head and neck squamous cell carcinoma. American journal of otolaryngology. 2005;26(1):12–7. Epub 2005/01/07. pmid:15635575.

18. Liu Y, Lan Q, Siegfried JM, Luketich JD, Keohavong P. Aberrant promoter methylation of p16 and MGMT genes in lung tumors from smoking and never-smoking lung cancer patients. Neoplasia. 2006;8(1):46–51. pmid:16533425

19. Pulling LC, Divine KK, Klinge DM, Gilliland FD, Kang T, Schwartz AG, et al. Promoter hypermethylation of the O6-methylguanine-DNA methyltransferase gene more common in lung adenocarcinomas from never-smokers than smokers and associated with tumor progression. Cancer research. 2003;63(16):4842–8. pmid:12941804

20. Toyooka S, Maruyama R, Toyooka KO, McLerran D, Feng Z, Fukuyama Y, et al. Smoke exposure, histologic type and geography-related differences in the methylation profiles of non-small cell lung cancer. International journal of cancer. 2003;103(2):153–60. pmid:12455028

21. Suda T, Hama T, Kondo S, Yuza Y, Yoshikawa M, Urashima M, et al. Copy number amplification of the PIK3CA gene is associated with poor prognosis in non-lymph node metastatic head and neck squamous cell carcinoma. BMC cancer. 2012;12(1):1.

22. Urashima M, Hama T, Suda T, Suzuki Y, Ikegami M, Sakanashi C, et al. Distinct effects of alcohol consumption and smoking on genetic alterations in head and neck carcinoma. PloS one. 2013;8(11):e80828. Epub 2013/11/28. pmid:24278325.

23. Kamholz SL. Pulmonary and cardiovascular consequences of smoking. The Medical clinics of North America. 2004;88(6):1415–30, ix–x. Epub 2004/10/07. pmid:15464105.

24. Neumann T, Rasmussen M, Heitmann BL, Tonnesen H. Gold standard program for heavy smokers in a real-life setting. International journal of environmental research and public health. 2013;10(9):4186–99. Epub 2013/09/12. pmid:24022655.

25. Russo AL, Thiagalingam A, Pan H, Califano J, Cheng KH, Ponte JF, et al. Differential DNA hypermethylation of critical genes mediates the stage-specific tobacco smoke-induced neoplastic progression of lung cancer. Clinical cancer research: an official journal of the American Association for Cancer Research. 2005;11(7):2466–70. Epub 2005/04/09. pmid:15814621.

26. Poeta ML, Manola J, Goldwasser MA, Forastiere A, Benoit N, Califano JA, et al. TP53 mutations and survival in squamous-cell carcinoma of the head and neck. The New England journal of medicine. 2007;357(25):2552–61. Epub 2007/12/21. pmid:18094376.

27. International Agency of Research on Cancer Website. IARC TP53 Database. http://p53.iarc.fr/TP53GeneVariations.aspx. Accessed 30 November 2018.

28. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. Journal of the American statistical association. 1999;94(446):496–509.

29. Paun BC, Kukuruga D, Jin Z, Mori Y, Cheng Y, Duncan M, et al. Relation between normal rectal methylation, smoking status, and the presence or absence of colorectal adenomas. Cancer. 2010;116(19):4495–501. pmid:20572039

30. Nakamura M, Watanabe T, Yonekawa Y, Kleihues P, Ohgaki H. Promoter methylation of the DNA repair gene MGMT in astrocytomas is frequently associated with G:C—> A:T mutations of the TP53 tumor suppressor gene. Carcinogenesis. 2001;22(10):1715–9. Epub 2001/09/29. pmid:11577014.

31. Scesnaite A, Jarmalaite S, Mutanen P, Anttila S, Nyberg F, Benhamou S, et al. Similar DNA methylation pattern in lung tumours from smokers and never-smokers with second-hand tobacco smoke exposure. Mutagenesis. 2012;27(4):423–9. Epub 2012/01/06. pmid:22217548.

32. Masica DL, Li S, Douville C, Manola J, Ferris RL, Burtness B, et al. Predicting survival in head and neck squamous cell carcinoma from TP53 mutation. Human genetics. 2015;134(5):497–507. Epub 2014/08/12. pmid:25108461.

33. Nagy E, Gajjar KB, Patel II, Taylor S, Martin-Hirsch PL, Stringfellow HF, et al. MGMT promoter hypermethylation and K-RAS, PTEN and TP53 mutations in tamoxifen-exposed and non-exposed endometrial cancer cases. British journal of cancer. 2014;110(12):2874–80. Epub 2014/05/24. pmid:24853176.

34. Scesnaite A, Jarmalaite S, Mueller M, Agaimy A, Zenk J, Hartmann A, et al. Prognostic value of O-6-methylguanine-DNA methyltransferase loss in salivary gland carcinomas. Head & neck. 2014;36(9):1258–67. Epub 2014/09/10. pmid:25201059.

35. Zuo C, Ai L, Ratliff P, Suen JY, Hanna E, Brent TP, et al. O6-methylguanine-DNA methyltransferase gene: epigenetic silencing and prognostic value in head and neck squamous cell carcinoma. Cancer epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2004;13(6):967–75. Epub 2004/06/09. pmid:15184253.