Towards a tumor-specific de-intensification strategy in oropharyngeal squamous cell carcinomas: Using tumor immunophenotypes to predict patient outcomes

Presentation: AHNS026
Topic: Pharynx / Larynx Cancer
Type: Oral
Date: Wednesday, April 18, 2018
Session: 2:20 PM - 3:10 PM Larynx/Pharynx
Authors: Wesley H Stepp, PhD1, Douglas Farquhar, MD1, Angela Mazul, PhD1, Jose Zevallos, MD2
Institution(s): 1University of North Carolina at Chapel Hill, 2Washington University in St. Louis

Purpose/Objectives: Many of the clinical trials that de-intensify treatment for patients with suspected HPV-positive oropharyngeal squamous cell carcinoma (OPSCC) use p16 expression alone to positively identify HPV-mediated tumors. While p16 immunohistochemistry is widely available and has a strong correlation with HPV-positive OPSCC, approximately 12% of p16-positive cases have non-HPV16-positive and HPV-negative tumors. In these patients, treatment de-intensification based on p16 immunohistochemistry alone could adversely affect this patient population who have more frequent recurrent disease and may not receive treatment consistent with the standard of care if included in trials with de-intensified radiotherapy. In this study, we show that OPSCC that are HPV-positive have a unique genetic signature with respect to gene expression and tumor-specific mutations, even in the context of p16-positive immunohistochemistry.

Materials/Methods: Formalin-fixed, paraffin-embedded p16-positive or p16-negative OPSCC samples were obtained from the university pathology core facility. Samples were sectioned onto slides and examined by a university pathologist who marked tumor boundaries. Tumor tissue was microdissected and RNA from tumor tissue was harvested. RNA was either hybridized directly to a Nanostring PanCancer Immune molecular RNA array or reverse-transcribed into cDNA and assayed for HPV16 mRNA targets (E1, E2, E5, E6/E7 and E1^E4) using quantitative, real-time PCR.

Results: Using NanoString molecular array technology, we have identified a pattern of immunoregulatory and cancer-associated gene expression in HPV-positive OPSCCs that clusters patients into HPV-positive, low-risk and HPV-positive, high-risk patients. Further, when stratified by clinical characteristics such as smoking status, we find that HPV-positive, never smokers have a distinct gene expression profile from those patients who are HPV-positive and have ever smoked.  Additionally, we find that HPV-mediated OPSCC tumors susceptible to recurrent disease may be identified with increased accuracy by combining viral mRNA qPCR and DNA-seq reads in combination with a tumor-derived genetic signature.

Conclusions: The use of p16-IHC alone to identify OPSCC candidates eligible for radiotherapy de-intensification may be insufficient. Using a combination of host, tumor and viral genetics, we may be able to limit the number of cases where radiation therapy is de-intensified in the setting of p16-positive immunohistochemistry. This could prevent recurrence of more aggressive subtypes of HPV-positive, p16-positive OPSCC. Further, the use of molecular profiling of tumors from multiple angles (e.g. host and oncovirus transcription) may aid in the clinical typing of OPSCCs, promote the discovery of drug targets, and improve our ability to safely de-intensify radiotherapy of HPV-positive OPSCCs.