Head And Neck Squamous Cell Carcinoma Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Head and neck squamous cell carcinoma is a heterogeneous group of tumors with different histopathological parameters and it is likely that each subtype also has its distinct molecular profile. Most tumors share to some extent the same multistep carcinogenic pathways, which include a wide variety of genetic and epigenetic changes. Epigenetic alteration represent all inherited changes in gene expression patterns that do not alter the actual DNA sequence. Now, it has become clear that silencing of cancer related genes is not exclusively a result of genetic changes such as mutations or deletions, but is also regulated on epigenetic level, mostly by means of gene promoter hypermethylation. Results from recent studies have demonstrated that DNA methylation patterns contain tumor type specific signatures, which could serve as biomarkers for clinical outcome in the near future. The main focus of this review is oral and oropharyngeal squamous cell carcinoma (OSCC), which we considered as one entity. The main objective is to analyse the available data on promoter hypermethylation of the cell cycle regulatory proteins p16INK4a and p14ARF, and to investigate their clinical significance as novel biomarkers in OSCC. The hypermethylation of both genes seems to possess predictive properties for several clinicopathological outcomes, and we conclude that the methylation status of p16 might serve as a candidate biomarker for predicting the clinical course of OSCC, especially recurrence-free survival.


Head and neck cancer is one of the most prevalent malignancies and causes each year a significant burden of morbidity and mortality, accounting for over half a million new cases worldwide, mostly men1. Traditionally this cancer encloses a wide variety of malignant tumors with squamous cell carcinoma (SCC) being the most common type. Each subtype is distinctive in its anatomical location in the head and neck area (e.g. oral cavity, nasopharynx, pharynx, larynx) and different pathobiological features2,3.

Tobacco and alcohol consumption are major aetiological risk factors in head and neck squamous cell carcinoma (HNSCC)4,5. These two factors are responsible for the development of HNSCC, especially in elderly men and in women due to increasing smoking rates. Furthermore, additional risk factors are studied in a new group of non-smoking, non-drinking young adults who have shown increased incidence rates of oral and oropharyngeal cancer6. In recent years, human papillomavirus (HPV) infection is recognized as a main determinant for oropharyngeal cancer6,7. Other predisposing factors for HNSCC are radiation exposure, betel nut chewing and environmental toxins8.

HNSCCs have in common that they are preceded by pre-cancerous lesions9, and share to some extent the same multistep carcinogenic pathways, subsequently leading to invasive squamous cell carcinoma3. This multistep process includes a wide variety of genetic changes. In the past decade, it has become clear that silencing of tumor suppressor genes, (one of the central principles of carcinogenesis), is not exclusively a result of genetic changes such as mutations or deletions, but can also be regulated on epigenetic level10-12.

In contrast to genetic DNA alterations, epigenetics encompass all inherited changes in gene expression patterns that do not alter the actual DNA sequence. One of the most extensively studied epigenetic modification is gene promoter hypermethylation, which usually results in transcriptional inactivation of the gene. More interesting, gene silencing by means of promoter hypermethylation might occur more frequently than by DNA mutations in cancer13. Unlike genetic perturbations (such as gene mutations or deletion), DNA hypermethylation is much more dynamic and often reversible in nature11,14 and are, therefore, an attractive target for new therapeutic agents.

Considering HNSCC as a highly inhomogeneous collection of tumors regarding their aetiology, histology, clinical course and prognosis it is quite appealing to investigate whether this heterogeneity can be attributed to differences in molecular basis, especially variations in DNA hypermethylation3,15. Thus, it is of major importance to select tumor subsites that are considered as one entity based on their biological commonalities. A distinct epigenetic profile, or an 'epigenome', for DNA hypermethylation, is interesting as it offers new opportunities to develop new specific biomarkers for each disease entity improving screening, early diagnosis and therapeutic decision making16.

We have chosen for oral and oropharyngeal squamous cell carcinoma (OSCC) as the main focus of this review, because of the alarming increase in the incidence rates of this specific subtype6. Furthermore, the need for novel, more specific biomarkers in HNSCC is endorsed by the fact that despite the rapid advances in the sphere of diagnosis and therapy, the prognosis, in particular for oral and oropharyngeal squamous cell carcinoma (OSCC), has not been improved in the past decades, remaining unchanged at a five-year survival of 50%17.

Here, we briefly review some promising epigenetic factors in OSCC. The main emphasis will be promoter hypermethylation of the cell cycle regulatory proteins p16INK4a and p14ARF, both encoded by the CDKN2A gene, one of the most widely investigated genes in HNSCC. The aim of this review is to address the clinical significance of p16 and p14 hypermethylation in OSCC and whether they can serve as prognostic biomarkers. Finally, we will discuss the interaction between p16 promoter hypermethylation and HPV infection.

Brief introduction into gene promoter hypermethylation

Epigentics mostly refers to promoter hypermethylation, although other less abundant alterations such as histone deacethylation, global genomic hypomethylation and histone remodelling are present. DNA hypermethylation represents the covalent addition of a methyl group to a cytosine nucleotide resulting in 5-methylcytosine. This modification is catalysed by the DNA methyltransferase enzyme family (DNMTs) with S-adenosyl-methionine acting as a methyl donor11. Cytosine methylation occurs in CpG dinucleotides which have an asymmetrical distribution throughout the genome. However, a small proportion of the CpG dinucleotides is clustered together in a 500 base pair long regions, called 'CpG-islands', where they take up more than half of the nucleotides. These CpG-islands are known to be located in promoter regions of approximately 50% of mammalian genes18.

The promoter sequence is a gene control region where general transcription factors and RNA polymerases bind, before the DNA transcription is initiated. Deletions, loss-of-function mutations and promoter hypermethylation can all lead to loss or decreased gene expression. According to Knudson's two-hit-hypothesis, a tumor suppressor gene is silenced when both alleles are knocked out. Promoter hypermethylation can cause loss of heterozygosity by inducing the second hit, after a mutational first hit. In sporadic cancer, two point mutations are rarely responsible for bi-allelic inactivation of tumor suppressor genes, while promoter hypermethylation of both alleles is more common11,19. This indicates that aberrant hypermethylation is an early event in the development of tumors16.

Under physiological conditions there is a basic methylation pattern, sometimes referred to as 'methylotype'20. In the basic pattern the majority of genes have promoter regions with unmethylated CpG-islands, whereas methylation of CpG dinucleotides outside the CpG-islands is abundantly present. This is thought to be part of a natural defence mechanism which safeguards the integrity of the genome during replication, on one hand by imposing transcriptional repression on large parts of mainly noncoding DNA, which may contain harmful sequences, and on the other hand supporting transcription of coding DNA through gene promoter hypomethylation3.

Compared to normal cells, there is a significant shift in the basic pattern of DNA methylation in neoplastic cells. Promoter regions are no longer hypomethylated, but increased CpG methylation does occur in promoters of specific genes, mainly involved in DNA repair, apoptosis, cell cycle regulation and tumor suppression. Simultaneously, loss of methylation in otherwise silenced regions takes place, a process named 'global hypomethylation', which is characterised by an increase in overall gene expression level due to weakened transcriptional repression. These changes indeed affect genetic stability and contribute to cancerization of the cell. Aberrant promoter hypermethylation is present is almost all types of cancer, including HNSCC, and the pattern of methylation appears to be tumor type specific16,21. In addition to that, the possible reversibility of abnormal methylation patterns makes DNA hypermethylation an appealing target for new cancer specific therapy. In fact, several chemotherapeutic agents have been found to possess demethylating properties which can reverse the transcriptional silencing of oncogenes22.

Promoter hypermethylation in oral and oropharyngeal squamous cell carcinoma

Epigenetic DNA alterations play a prominent role in cancer, including OSCC, the main focus of this review. Numerous studies have investigated epigentic alterations in OSCC and have found that promoter hypermethylation of multiple genes is highly prevalent, although various methylation rates for each gene have been reported. The silenced genes are typically tumor suppressor genes. Table 1 is a summary of candidate genes with substantial evidence for being hypermethylated in OSCC, and their reported clinicopathological associations. The inclusion criterion was as follows; hypermethylation, proven in more than one study was considered as substantial evidence. Furthermore, we made a selection of candidate genes and classified them according to their potential biomarker application as indicated by reported associations (table 2). The main emphasis of this review is the p16INK4a oncogene, because its role has been well-studied in oral and oropharyngeal squamous cell carcinoma (table 1). We also discuss the hypermethylation of the p14ARF oncogene.

The proteins p16INK4a and p14ARF are two alternative splice variants of the CDKN2A gene, located on chromosome 9p21. Both proteins function as inhibitors of cell cycle progression (figure 1). The

p16INK4a protein promotes senescence and differentiation by interfering in the retinoblastoma (Rb) pathway. It prevents entry into S phase by inhibiting the CDK4/6-cyclin D1 complexes, thereby preventing the phosphorylation of Rb proteins. As a result, E2F transcription factors are inactivated, as the Rb-E2F complex remains intact. Accordingly, dysfunctional p16INK4a results in uncontrolled cell cycle progression. The p14ARF protein activates the tumor suppressor gene p53 by inhibiting MDM2, an ubiquitin ligase that marks p53 for degradation. Overexpression of p14ARF enhances the p53 function which in turn leads to cell cycle arrest or apoptosis in cells3,52.

In the last decade, aberrant promoter hypermethylation of p16 and p14 has been observed in oral and oropharyngeal cancer tissue (table 1) as well as premalignant oral lesions53-57 and histologically healthy mucosa surrounding the tumor38,39,58. More important, p16 and p14 gene inactivation is primarily due to promoter methylation30. In one of the earliest studies focussing on OSCC as a distinct entity, Wu et al (1999), demonstrated that the vast majority of the tumors (>80%) had loss of p16 expression34. Interestingly, p16 promoter hypermethylation appeared to be more common than point mutation (23% and 7%, respectively). In a Brazilian cohort of 45 patients with resected primary OSCC tumors, Crodeiro-Silva at al. (2012) investigated the methylation status of four genes and found high rates of hypermethylation for CDKN2A (p16 and p14), EDNRB, RUNX3 and SFN. They also reported more CDKN2A and EDNRB promoter hypermethylation in subjects with lymph node metastases40. In an Indian cohort, Kaur et al. (2010) selected four genes and evaluated their methylation status in a sample of 92 OSCC patients39. EDNRB, KIF1A, p16 and DCC were found to be highly methylated in tumor tissue, and p16INK4a methylation was associated with nodal involvement. In another study a semi-quantitative approach (pyrosequencing) was adopted in order to quantify the promoter hypermethylation of five genes in OSCC samples and analysed whether there is a correlation between the quantitative methylation index and clinicopathological variables from the patients33. No such correlation was observed, methylation of the genes p16INK4a, CYGB and CYCA1 however, was highly tumor specific, because healthy resections margins contained significantly less abnormal methylation for these genes. Also, hypermethylation of ECAD and RARβ was observed in tumor tissue and adjacent healthy mucosa. No significant hypermethylation is observed in healthy control tissue33,39.

Several studies have evaluated the presence of promoter hypermethylation in oral premalignant lesions. In an early study, loss of p16 function was reported in a small number of patients (17/37) with leukoplakia, accounting for 5 out of 8 patients who developed malignant transformation53. Also, increasing p16 methylation rates were reported for mild to severe dysplastic lesions, 30% and 82% respectively54. In patients with severe dysplastic epithelial lesions, Kresty et al (2002) detected a p16 methylation rate of 57%, while p14 was methylated in 3.8% of the samples55. An association was found for p16 hypermethylation with loss of heterozygosity and lesions of the tongue and floor of the mouth. Two studies investigated the prognostic significance of p16 hypermethylation in oral epithelial dysplasia56,57. In the first study, a significant proportion of patients with malignant transformation of epithelial dysplasia had p16 hypermethylation compared to patients with no malignant transformation (57% vs 8%, p=0.002). However, p16 did not correlate with time of onset of transformation56. Cao et al (2009), supported these findings when they described a significantly higher progression rate for oral dysplasia to OSCC in p16 hypermethylated cases (43.8% vs 17.4%; OR=3.7)57. This effect was more evident in patients aged above 60 years (OR=12.0, p=0.013) and subjects with moderate epithelial dysplasia (OR=15.6, p=0.022). These findings suggest that p16 hypermethylation is a powerful marker for selecting patients with precancerous lesions who are at risk for progression to malignant disease.

Although not fully convincing due to small sample size, the feasibility of p16 hypermethylation in surgical resection margins was confirmed by Goldberg and colleagues demonstrating hypermethylation in margins of three patients with SCC of the tongue59, and in a later study, positive margins were reported in 4 OSCC patients of which two developed a local recurrence23. Recently, a prospective study including a larger number of Indian patients with SCC of the tongue, showed that 43% (13/30) of histologically tumor free margins contained p16 hypermethylation and further analysis showed a 6.3 fold increased risk for local recurrence for this margins58. Still, the p16 methylation status did not affect the overall survival rate. A more recent study in OSCC resection margins also could not establish a correlation between p16 hypermethylation and overall survival42. To our knowledge these two studies are the first to evaluate the significance of p16 hypermethylation as a predictive and prognostic marker in surgical resection margins. Further research is needed to investigate whether intraoperative p16 hypermethylation analysis in surgical margins results in more accurate resection with less recurrence compared to the conventional histopathological assessment.

The fact that aberrant promoter hypermethylation of p16 and p14 is detected in both peritumoral tissue and premalignant lesions, suggests that these epigenetic alterations are an early event, making tissue more prone to cancerization. These findings are in concordance with the concept of 'field cancerization', originally proposed by Slaughter et al. in 1953 to explain the high recurrence rate of head and neck cancers60. They hypothesised that multiple acquired genetic defects in large patches of mucosa in the upper aerodigestive tract, make morphologically normal epithelium prone to dysplastic or malignant transformation3,60. This so called 'fields' are not limited to the boundaries of the malignancy, but extend into surgical resection margins and increase the risk of local relapse or a second primary tumor.

Promoter hypermethylation and clinicopathological associations

Many studies have described correlations between p16 and p14 promoter hypermethylation and different clinical outcomes in OSCC. The available data on survival rates reveal a trend towards favourable survival in OSCC with p14 hypermethylation, whereas p16 hypermethylation is associated with poor survival (figure 2). However, reports on other clinicopathological associations are more inconsistent. In a cohort of 96 OSCCs aberrant methylation of p16 and p14 was observed in 29% and 14% of the tumors respectively28. Younger age and tumor size (lower T-stage) were significantly associated with p16 methylation, remarkably p14 methylation was significantly associated with a longer overall survival time. Similar results by Ishida et al (2005) related hypermethylation of p14 with tobacco and alcohol consumption, increased lymph node invasion and higher clinical stage29. p16 hypermethylation correlated with increased tumor size and higher clinical stage, although association did not reach significance. In another study, Sailasree et al (2008), verified these findings in an Indian cohort of 116 OSCC patients31. They found that cases with hypermethylation of p16 were three times more at risk for disease recurrence (RR=3.3), whereas p14 methylation was significantly associated with reduced recurrence rate (RR=0.109). Hypermethylation of both markers did not show any correlation with overall survival rates, yet, high p16 protein expression was associated with reduced residual disease after treatment (RR=0.351) and increased overall survival during follow-up (RR=0.318). More recently, a study in buccal SCC, one of the most frequent types of OSCC, indicated a relation between p16 hypermethylation and lymph node metastasis and poor overall survival61. However, in concordance with previous studies such relation did not appear in the multivariate analysis.

In conclusion, the available data from recent studies in OSCC carefully suggest that promoter hypermethylation of p16 and p14 is tumor specific, since transcriptional silencing of both genes by hypermethylation is highly prevalent in tumor tissue and rather lacking in healthy controls. Based on reported clinical associations, we conclude that there is sufficient evidence in OSCC for p16 hypermethylation as a predictive marker for a less favourable clinical course. Surprisingly, high p16 expression levels, proven by immunochemistry, are associated with improved prognosis62, which supports our previous conclusion. However, clinical impact of p16 hypermethylation on overall survival remains inconclusive. Therefore, we acknowledge the potential application of p16 hypermethylation as a biomarker in OSCC, at least as a marker for prognosis. There is a need for larger prospective studies in order to overcome the current inconsistency in the literature and to investigate whether p16 hypermethylation is applicable as a fully independent prognostic marker and whether patients will clinically benefit from these developments. First results regarding p14 hypermethylation are promising, still, the available evidence in OSCC is too limited to draw conclusions. Aberrant p14 hypermethylation certainly deserves more attention and hopefully more research will be dedicated to the clinical significance of this oncogene.

Human papillomavirus infection and p16: clinical and molecular implications on OSCC

There is a significant increase in the incidence of oropharyngeal squamous cell carcinoma since the late 1980s, with base of the tongue and tonsils as predilection sites7,63. HPV infection has been repeatedly identified as a risk factor for this type of cancer, especially HPV type 16 and to a lesser extent type 1812. Both types are widely accepted as causative agents in the pathogenesis of cervical cancer64, and there is now evidence that HPV plays a similar role in oropharyngeal cancers65.

The viral DNA genome of human papillomaviruses encodes two essential oncogenic proteins, E6 en E7, both responsible for immortalization of infected cells66. In presence of viral E6 and E7 two key cell cycle control mechanisms are disrupted (figure 1). The protein E6 interferes in the proapoptotic p53 pathway through an interaction with p53 and inhibits its function by degradation. In contrast, E7 inhibits the retinoblastoma (Rb) pathway; it binds to the Rb protein and releases the E2F transcription factors. As both pathways are inhibited, cell cycle regulation is lost and infected epithelial cells gain unlimited replicative potential66. At the molecular level, HPV related HNSCCs are, in general, characterised by wild-type p53, decreased protein levels of Rb and p16 overexpression, whereas in HPV-negative HNSCCs p53 is typically mutated, Rb proteins are up regulated and p16 levels are low3. This finding indicates that both cancers have different molecular fingerprints and, therefore, should be considered as distinct entities.

Apparent differences in clinicopathological features are also noticed. Patients with HPV-positive oropharyngeal carcinoma are predominantly young adults with no history of smoking or alcohol consumption, the traditional risk factors of HNSCC67,68. Moreover, patients with HPV-positive tumors have a significantly better prognosis in terms of therapy response and survival69-71. In a recent meta-analysis, pooled results from 42 studies showed that HPV-positive HNSCCs were associated with 54% better overall survival (HR=0.46) and 72% decrease in disease specific mortality (HR=0.28), compared to HPV-negative tumors72. Also, reduced progression and recurrence rates were observed (>50%)72. HPV infected cancers of the oral cavity (only two studies pooled), showed an improvement of 68% in overall mortality (HR=0.32). It is uncertain whether these differences can fully be attributed to HPV status, considering that better prognosis might be partially explained by confounding associations, such as age and therapy, for which several studies did not adjust for. In their analysis of 489 tonsillar cancers, Hong et al (2012) found a modifying interaction between HPV-positivity and N-stage and T-stage, locoregional recurrence and survival73.

In contrast to oropharyngeal cancers, a causative role of HPV infection in SCC of the oral cavity is debatable. A large number of studies have detected HPV infection in pre-cancerous and malignant lesions of the oral cavity, with prevalence rates ranging from 0% to 95%74. This wide range is mainly due the relative small sample size of the studies, and differences in quality of analysed tumor tissue as well as differences in sensitivity of employed detection assays. In addition, some variation is attributed to differences in genetic predisposition or risk factor exposure according to geographic location74-77. For example, betel quid chewing in combination with smoking is a major risk factor for oral carcinoma in South Asia78. Several reports have addressed the aetiological role of HPV infection in oral carcinoma. Pooled data from 94 studies including 4680 samples showed that HPV is present in approximately half of the carcinoma samples and that it is 2 to 3 times more likely to detect HPV in pre-cancerous lesions, compared to normal mucosa (10% HPV infected)75. Kreimer et al (2005) found overall HPV prevalence in oral carcinoma to be 23.5%, with a rate of 16% in Europe and North-America, increasing to 33% in Asia76. HPV-16 and HPV-18 were the most frequent detected viruses (68.2% and 34% respectively). In a meta-analysis of case-control studies, the pooled odds ratio for HPV prevalence was estimated at 3.98 for oral SCC and 3.87 for precancerous lesions, suggesting a near quadruple increase of HPV detection77. Albeit, the authors did not rule out confounding factors, such as age, smoking, alcohol consumption and sexual behaviour, they did not adjust for due to incomplete data extraction from inadequate reports.

Unlike oropharyngeal SCC, literature on the clinical impact of HPV infection in SCC of the oral cavity is rather ambiguous. Several studies have reported favourable survival outcomes for HPV-positive oral SCC79-81. For example, Schwartz et al. (2001), assessed the HPV-16 status of 254 oral SCC samples and found improved overall survival rate for patients with HPV infection79. Other studies have associated HPV-positive oral cancer with more advanced disease and poor prognosis, in terms of higher recurrence rate, increased incidence of distant metastases and decreased disease-free and overall survival82-84. Regarding therapy response, HPV-16 infected patients with advanced oral cancer were three times more at risk for distant metastases and had a two- to threefold increase in mortality, despite radical surgery and adjuvant chemoradiation83. Therefore, the authors hypothesized that there is a subgroup of HPV related oral SCCs, particularly cases with extracapsular spread, who might need a more intensive therapeutic approach. Nonetheless, we recommend future studies, in de first place, to throw more light on the clinical relevance of HPV status in oral SCC and then address this issue.

As mentioned earlier, HPV positive tumors are likely to overexpress p1685, which is a well-described finding in cervical cancer86. The E7 oncoprotein disrupts the Rb-E2F complex, leading to increased levels of free E2F, which in turn upregulates the transcription of p1687 (fig. 1). Expression of p16, proven by immunohistochemistry, has been suggested as a surrogate marker for HPV infection and is in combination with viral DNA PCR assay a powerful tool to detect oncogenic HPV85. In addition, different studies have correlated p16 expression with the clinical course of HPV related HNSCCs62,88-90. Smith et al (2008), found significant associations between p16 expression and oropharyngeal cancer, higher disease stage, improved disease-specific and overall survival, and longer recurrence-free time88. Another study described similar results90; they analysed pre-treatment samples of 156 patients, predominantly oropharyngeal cancers (47%), who received radiotherapy as only treatment. Tumors with p16 expression were correlated to improved locoregional control (OR=0.26, p=0.001), decreased disease-specific mortality (OR=0.30) and better overall survival (OR=0.22). Notably, patients with more advanced disease were more likely to benefit from p16 expression90. These results indicate that analysis of p16 expression might be a potent predictor for prognosis in oropharyngeal SCC.

There is some evidence that HPV infection could be involved in DNA hypermethylation. The HPV oncoprotein E7 has been linked to increased activity of DNA methyltransferases, which are responsible for CpG methylation. Abnormal methylation patterns probably driven by overexpression of DNMTs have been associated with carcinogenesis10. HPV E7 oncoprotein has been shown to bind and upregulate DNMT1, which is crucial for post-replicative maintenance of methylation patterns, leading to increased enzymatic activity91. In other reports, p16 hypermethylation is significantly associated with HPV infection and increased expression levels of DNMT3b, a methyltransferase important for de novo methylation34,92,93.

To our knowledge, there are few studies that address a possible link between p16/14 promoter hypermethylation and HPV infection in OSCC. In a small cohort of 24 oral and oropharyngeal SCC samples, HPV 16 positive tumors seemed to correlate with p16 overexpression, but not with p16 hypermethylation94. Such association was also absent in a Brazilian cohort predominated by SCC of the oral cavity (90%)40. However, one recent case-control study found a significantly higher prevalence rate (69.2%) for p16 hypermethylation in HPV 16 infected oral epithelial dysplasia samples, compared to non-infected samples (20.8%)95. DNMT1 and DNMT3b levels did not correlate to HPV status or p16 hypermethylation. Since HPV positive tumors are more likely to overexpress p16, the authors hypothesize that the observed association between HPV 16 and p16 hypermethylation could explain the unusual low p16 protein expression in a subset of HPV positive tumors with less favourable prognosis. More clinical evidence is needed to back up this hypothesis. For future research, it would be appealing to investigate the molecular impact of HPV infection on epigenetic regulation, with particular attention for p16 promoter hypermethylation, and if there is a modifying relation to elucidate the molecular advantages for HPV in adopting a mechanism that downregulates p16.


Thanks to advances in the field of epigentics, our understanding of the molecular origin of cancer has changed rapidly. There is now sufficient and well established evidence that epigenetic DNA alteration play a decisive role in the development of cancer by regulating the transcription of many (tumor suppressor) genes. Furthermore, it has become clear that in many tumor types specific epigenetic features can be distinguished and that DNA hypermethylation is a major determinant of the 'epigenome'. The key question remains whether extended knowledge of cancer epigentics will result in a new molecular classification of this disease in well-defined and more uniform subcategories.

In this review we have focussed on oral and oropharyngeal squamous cell carcinoma as a subcategory of HNSCC, likely to have its own distinct epigenetic profile. How these differences do occur is not clear, but they might be explained by the alternative aetiologies of head and neck tumors since risk factor exposure is different for age, ethnicity and geographic location.

Concerning the objective of this review to address the feasibility of aberrant promoter hypermethylation of p16INK4a and p14ARF as biomarkers in OSCC, one can draw several conclusions based on the available reports. Aberrant hypermethylation of p16INK4a and p14ARF is a cancer-specific finding, since it is a significantly observed in OSCC, and is not likely to be found in healthy control tissue. Promoter hypermethylation of p16INK4a and p14ARF occurs early in the process of cancerization, as both are detected in oral precancerous lesions as well as peritumoral tissue. Last but not least, a growing body of evidence has confirmed the predictive value of p16 promoter hypermethylation for several clinicopathological parameters, including progression of premalignant lesions to OSCC, advanced disease, local recurrence and disease-specific survival. The methylation status of p16 is definitely interesting as a candidate biomarker for predicting the clinical course of OSCC. However, more prospective studies are needed to affirm the prognostic power of p16 hypermethylation in larger groups of patients. Early assessment of p16 hypermethylation might enable the identification of subgroups of patients with poor prognosis, who might require a different therapeutic approach. Therefore, we also recommend future research to explore the position of this biomarker in the clinical management of OSCC and to evaluate whether it can contribute to individualised treatment strategies. Moreover, the molecular relationship between HPV infection and promoter hypermethylation is poorly understood. Addressing this issue would certainly contribute to a deeper understanding of HPV positive oral and oropharyngeal cancer.

Figure 1. Cell cycle arrest by CDKN2A. The CDKN2A gene encodes two alternatively spliced transcripts, p16INK4A and p14ARF, which differ in their first exon. The p16INK4A protein inhibits the CDK4/6-cyclin D1 complexes, keeping the retinoblastoma (Rb) proteins in a dephosphorylated state, enable to bind and inactivate the E2F transcription factors. Free E2F ensures the transcription of various proteins, most of them necessary for progression to S phase. P16INK4A is also upregulated by E2F. In contrast, p14ARF stabilizes and thus activates the tumor suppressor gene p53 by inhibiting MDM2, which inactivates p53 by ubiquitin-mediated degradation. Active p53 induces the expression of p21, a negative cell cycle regulator which is an inhibitor of the CDK1-Cyclin A/B complexes, thereby preventing the progression from G2-phase to metaphase. The human papillomavirus oncoproteins E6 and E7 interfere in the Rb pathway and in the p53 pathway, in order to bypass the cell cycle checkpoints. The E7 oncoprotein promotes the progression to S phase. It binds the Rb proteins and thereby releases the E2F transcription factors. The E6 protein targets p53 and induces loss of function by degradation. (Green arrow = activating interaction; red arrow: inhibiting interaction)

Figure 2. Cumulative survival rates. A few number of studies have evaluated the relation between p16INK4A/p14ARF promoter hypermethylation and survival outcome. Despite differences in reported survival variable, there is a tendency towards favourable survival in p14ARF hypermethylated OSCC, and poor survival in p16INK4A hypermethylated OSCC.