Hunting the end of the rainbow : prognostic biomarkers and human papillomavirus in tonsillar and base of tongue cancer

Abstract

Aim: The aim of this thesis was to hunt for additional prognostic biomarkers in HPV positive TSCC and BOTSCC to identify patients that could if they wished, be enrolled in clinical trials for de-escalation of treatment.

Background. Tonsillar squamous cell carcinoma (TSCC) and base of tongue squamous cell carcinoma (BOTSCC), which account for most oropharyngeal squamous cell carcinomas (OSCC) have been increasing in the last decades and now comprise almost half of all head and neck cancers (HNSCC) in Sweden. The main risk factors for HNSCC were originally considered to be smoking and alcohol however in 2007 the International Agency for Research on Cancer (IARC) acknowledged HPV infection as a risk factor for OSCC. It has been shown that patients with HPV positive OSCC and especially TSCC and BOTSCC have a much better clinical outcome (80% 3 year disease free survival, DFS) as compared to for patients with the corresponding HPV negative tumors (40%). Due the generally poor prognosis for HNSCC, treatment for HNSCC including TSCC and BOTSCC has been intensified, leading to more serious side effects for the patients. It has been proposed that due to the good prognosis of patients with HPV positive TSCC and BOTSCC such an aggressive treatment might be unnecessary and de-escalation of treatment would be of benefit. However, since 20% of patients with HPV positive tumors do not do well, additional prognostic biomarkers are needed for selecting patients that would be the best responders to less intensive treatment.

Results: In paper I, high numbers of CD8+ tumor infiltrating lymphocytes (TILs) were shown to correlate with a favorable clinical outcome in both HPV DNA+ and HPV DNA- TSCC and BOTSCC. In addition, HPV DNA+ tumors were significantly more infiltrated than the HPV DNA- ones. In paper II, absence of HLA class I in HPV DNA+ OSCC was shown to be associated with increased survival of the patients while the opposite was true for normal expression. In paper III, the HLA-A*02 allele, common in the Scandinavian population, was demonstrated to be a negative prognostic factor for patients with HPV DNA+ TSCC and BOTSCC. Following that study, in papers IV and V, the expression of components of the antigen processing machinery (APM) were examined for nuclear and cytoplasmic staining and the absence of LMP10 nuclear staining as well as low LMP7 expression were disclosed to be correlated with increased survival of patients with HPV DNA+ TSCC and BOTSCC. Finally, in paper VI, biomarkers CD8+ TILs, HLA class I, HLA-A*02 and LMP10 were combined with clinical characteristics of age, stage of the patients with HPV DNA+ TSCC and BOTSCC in order to develop an algorithm allowing for prediction of the outcome of the patients and CD8+ TILs, age and stage of the patients were found to be most significant prognostic contributors.

Conclusion: Investigating and combining our best prognostic biomarkers, we have developed an algorithm which makes use of different biomarker and clinical characteristics allowing for prediction of patients with good clinical outcome that will enable de-escalation of treatment