Multigene profiling to identify clinically relevant actionable mutations in head and neck cancers: An Indian study

Background: Head and neck squamous cell carcinoma (HNSCC) represents approximately 5%–10% of malignancies worldwide. The most appropriate treatment approach for HNSCC varies with the disease stage and disease site in the head and neck (H&N). Radiotherapy (RT) combined with chemotherapy has become the standard of care for patients having locally advanced tumors. However, there is significant morbidity associated with these treatments, and recurrent or metastatic diseases will occur in 50%–60% of patients. Moreover, the detection of residual viable tumor at the end of therapy remains an important issue. It is therefore an unmet need to improve the outcome of therapy by identifying predictive (prognostic) indicators at the molecular level and radioresistance that will enable the clinicians to select the logical treatment modality. Materials and Methods: Fifty H&N cancer patients aged 27 to 85 years diagnosed at HCG between April 2015 and 2017 were screened using Illumina's TSCAP panel and MiSeq technology for hotspot mutations in 48 cancer-related genes. All the cases had histopathological reviews and comprised tumors from the following sites – oral, nasopharynx, throat, hypopharynx, larynx, thyroid, or nasal cavity and paranasal sinuses. The average coverage across 220 hotspots was >1000X. Data were processed using Strand Avadis NGS™. Mutations identified in the tumor were assessed for “actionability,” i.e., response to therapy and impact on prognosis. Results: Somatic variants were detected in 65% of cases with direct impact on therapy and/or prognosis. Genetic aberrations were identified in major RAS/RAF signaling pathways in nearly 15% of H&N cancers, out of which HRAS activating mutations were the most common (n = 5). HRAS was also found to be co-mutated with phosphatidylinositol 3-kinase (n = 3) and PTEN deletions (n = 3). In contrast to the MAPK signaling pathways, mutant HRAS is able to signal exclusively through PI3K-AKT, reducing the response to cetuximab and increasing the response to MEK inhibitors including selutinib and tramatinib. Based on the results, cetuximab was discontinued in two patients who had presented with metastatic HNSCC. Other targetable mutations included PIK3CA (n = 3), EGFR (n = 1), cKIT (n = 1), RB1 (n = 1), and PTEN (n = 3) were reported. Further, disruptive and nondisruptive mutations in TP53 alone were found in 45% of H&N cancers, varying widely among different histologies, indicating a poor response to cisplatin- and 5FU-based chemotherapy. Interestingly, all metastatic/recurrent patients treated with cisplatin presented with very short progression-free survival of 9–12 months were found to have TP53. TP53 was also found to be co-mutated with ATM gene (n = 1), an important prognostic marker indicating poor response to chemotherapy and RT. Conclusion: This study validates the utility of multigene profiling in H&N cancer patients, both early diagnosed and advanced cases, to stratify based on their molecular profile that could potentially benefit/not benefit from targeted therapy and chemoradiation. Few ongoing prospective studies and randomized clinical trials may help us confirm the independent prognostic and therapeutic value of the mutations in a larger cohort of Indian population.