Molecular subtypes of head and neck cancer – Prognostic impact and correlations with morphological characteristics

Aim Treatment options for patients with recurrent or metastatic head and neck squamous cell carcinoma (R/M-HNSCC) are limited. Therefore, prognostic and predictive biomarkers for the application of personalized therapies are urgently needed. To this end, we designed, applied and evaluated a NanoString gene panel which revealed the subtype in R/M-HNSCC in order to identify potential treatment-relevant conditions and correlations with histomorphological characteristics. Material and Methods 130 formalin-fixed, paraffin-embedded samples from 95 treated CeFCiD trial patients with R/M-HNSCC were sequenced and analyzed. A customized NanoString panel composed of 231 genes was utilized to determine molecular subtypes and gene expression levels of potential predictive genes. For 67 patients, clinical data and H&E stained slides were available for additional comprehensive histomorphological analyses (e.g., histotype, grading, tumor cell budding (TCB), nuclear size, stroma content). Results 114 out of 130 (88%) samples passed quality control and we could successfully determine their molecular subtype in 103 of 114 (90%) cases. Gene expression levels of AREG were significantly higher for the basal and classical subtype compared to the mesenchymal subtype. Cases assigned to the classical subtype showed superior progression-free and overall survival rates compared to non-classical subtypes. Additionally, among histomorphological parameters high TCB was associated with poor outcomes. Conclusion The highly compact NanoString panel was able to determine the molecular subtype of R/M-HNSCC. The implications of molecular subtypes on prognosis and therapy prediction might aid in treatment planning of R/M-HNSCC patients. The prognostic significance of the histomorphological parameters TCB could be confirmed.