Validation Of Selected Head And Neck Cancer Prognostic Markers From The Pathology Atlas In An Oral Tongue Cancer Cohort

Simple SummaryPrognostic markers are used to predict the aggressiveness of a cancer and to help decide the best treatment for individual patients. Despite intense research, reliable prognostic markers for oral cancer are still few. The aim of the present study was to validate selected prognostic markers for head and neck cancer identified by unbiased approaches in oral tongue cancer, a specific subsite of head and neck cancer. From a list of 790 markers, we selected three based on reported prognostic value as well as expression pattern and availability of validated antibodies. These were analyzed on transcriptional and protein level in a cohort of 121 oral tongue cancers. Only one of the markers showed significant prognostic value when controlling for established prognostic parameters. Our study highlights the need to evaluate prognostic markers in homogeneous groups of cancers and to control for established prognostic parameters.The Pathology Atlas is an open-access database that reports the prognostic value of protein-coding transcripts in 17 cancers, including head and neck cancer. However, cancers of the various head and neck anatomical sites are specific biological entities. Thus, the aim of the present study was to validate promising prognostic markers for head and neck cancer reported in the Pathology Atlas in oral tongue squamous cell carcinoma (OTSCC). We selected three promising markers from the Pathology Atlas (CALML5, CD59, LIMA1), and analyzed their prognostic value in a Norwegian OTSCC cohort comprising 121 patients. We correlated target protein and mRNA expression in formalin-fixed, paraffin-embedded cancer tissue to five-year disease-specific survival (DSS) in univariate and multivariate analyses. Protein expression of CALML5 and LIMA1 were significantly associated with five-year DSS in the OTSCC cohort in univariate analyses (p = 0.016 and p = 0.043, respectively). In multivariate analyses, lymph node metastases, tumor differentiation, and CALML5 were independent prognosticators. The prognostic role of the other selected markers for head and neck cancer patients identified through unbiased approaches could not be validated in our OTSCC cohort. This underlines the need for subsite-specific analyses for head and neck cancer.