51. Role of biomarkers in prognosis and treatment of head and neck squamous cell carcinomas

Abstract

Background Head and neck squamous cell cancers (HNSCCs) account for 5% of all newly diagnosed cancer cases and are the fifth most common cancer worldwide. HPV infection, tobacco smoking and alcohol consumption are established risk factors for HNSCCs. Recently an increased incidence have been documented in younger men who are non-smokers, and non-alcoholics. Several studies have revealed HPV infection, particularly HPV16, as an aetiological factor for HNSCCs. Data also shows that HPV related HNSCCs have a better prognosis and clinical outcome than HPV negative disease. The primary objective of this study was to determine the frequency of HPV infection, p16 and p53 tumour profile and mutations in EGFR, K-RAS and B-RAF genes and correlate with clinical outcome and histomorphological parameters. Materials and methods 60 of 726 randomly selected patients with squamous cell carcinomas (SCCs) of head and neck diagnosed at SCGH over the period from 1996 to 2008 were selected. Data on demographics, smoking, alcohol, staging, treatment, recurrences, outcomes, survival, cause of death and histopathology were recorded. The tumours were further classified as tonsillar and non-tonsillar, and based on histomorphology separated as keratinising and non-keratinising. Other parameters assessed include type of biopsy, in situ or invasive tumour, margins, perineural invasion (PNI), lymphovascular space invasion (LVSI) and node status. High risk (HR) HPV PCR typing including HPV16, was performed by PCR based method and correlated with immunohistochemistry for p16 and p53. The frequency of mutations of EGFR (exons 18, 19, 20, 21), B-RAF (V600E and V600K) and K-RAS (exons 2, codons 12 and 13) were detected by Flourescence Single Strand Polymorphism (F-SSCP) and direct DNA sequencing analysis. Results Ofthe 60 cases studied, 20 (33%) were tonsillar squamous cell carcinoma and 40 (67%) were non-tonsillar. 45 (75%) were keratinising SCC and 15 (25%) were non-keratinising. 11 (18%) showed LVSI, 4 had detectable PNI (6.6%), 13 (21.6%) had positive margins and 13 (21.6%) of the tonsillar cases were smokers. 22 of the 60 (37%) cases studied were HR-HPV DNA positive. 19 (88%) of these were tonsillar and only 3 (12%) were non-tonsillar. Of the 20 tonsillar SCCs, 19 were HR HPV DNA positive (95%). Only 3 (7.5%) cases of the 40 non-tonsillar SCCs were HR HPV DNA positive. 9 of the 19 (47%) HPV DNA+ tonsillar cases were p16 positive and none (0%) of the 3 HR HPV DNA+ non-tonsillar SCCs were p16 positive. 11 of 20 (55%) tonsillar cases were p53 positive and 24 of the 40 (60%) non-tonsillar cases were p53 positive. BRAF (1.6%), KRAS (0%) and EGFR (1.6%) gene mutations were relatively rare in HNSCCs; however, 73% of cases had silent EGFR exon 20 mutation or single nucleotide polymorphism at codon 787. Clinical outcome analysis revealed statistically significant correlation between HPV+ tumours, tonsillar location of tumour, younger patient age, no history of smoking and better survival ( p = p = Conclusion According to our study, HR HPV DNA and p16 expression were significantly higher in tonsillar than non-tonsillar HNSCCs. p16+ tonsillar tumours were statistically significantly associated with HR HPV-DNA, poorly differentiated histology, frequent LVI, higher stage and grade, location of the tumour and lymph node involvement. 9 (47%) of tonsillar HPV-DNA+ cases had positive p16 immunoprofile whilst only two cases of non-tonsillar tumours were positive for p16 but were negative for HPVDNA. 10 (52%) of HR HPV DNA positive tonsillar cases were negative for p16 by immunohistochemistry, therefore although p16 is a good surrogate marker for HPV status, due to its lower sensitivity should be used in combination with HR HPV PCR studies. The overall frequency of mutations in BRAF, KRAS and EGFR genes is low, therefore it seems that currently there is no role for routine testing for mutational status of the above genes for HNSCCs. Coincidentally, we have detected a high frequency of heterozygous and homozygous single nucleotide polymorphism of EGFR-TK both in tonsillar and non-tonsillar HNSCCs. The significance of this finding is not clear and further larger studies, including matching with control non-smoker or smoker cases would be required to further elucidate the significance of these results. In summary, our findings reinforce the role of HPV infection in tonsillar SCCs and association of positive p16/HPV DNA tumour status with a more favourable clinical course. The study provides a rationale for routine p16 immunohistochemistry and/or HR-HPV DNA testing of tonsillar SCCs.