Identification of antigens for a multivalent vaccine for head and neck squamous cell carcinoma

Presentation: AHNS-QS-097
Topic: Systemic Therapy / Immunotherapy
Type: Quickshot
Date: Thursday, May 2, 2019
Session: 5:30 PM - 6:00 PM
Authors: Harrison A Cash, MD1, Jeffrey Houlton, MD1, Laura Riolobos, PhD2, Mary L Dises, MD2
Institution(s): 1University of Washington, Department of Otolaryngology, 2University of Washington, Cancer Vaccine Institute

IMPORTANCE: Head and neck squamous cell carcinoma (HNSCC) has poor outcomes secondary to late presentation and poor therapies available for advanced-stage disease. Treatment paradigms have remained largely unchanged in this regard over the last several decades with very few improvements in survival for patients with this form of cancer. More recently, immunotherapies have become of increasing interest for treatment of many malignancies, including HNSCC. It is known that an effective immune response against malignancy requires induction of a predominantly Th1 effector-type response, while Th2 type responses are associated with tumor tolerance. It has been traditionally challenging to develop vaccines which were able to induce a Th1 specific immune response, however our lab has pioneered development of Th1 selective epitope based vaccines. Such vaccines have already proven to be effective at decreasing tumor development and reducing tumor burden in mouse models of lung, breast and colon cancer. Bringing such immunotherapies into the realm HNSCC would broaden the arsenal of available therapies against this disease and potentially improve stagnant trends in overall survival.

OBJECTIVES: 1. Identify antigens that are overexpressed in HNSCC and determine their functional relevance to cancer progression by siRNA screening in HNSCC cell lines. 2. In silico prediction to identify epitopes from the antigens able to elicit Th1 selective CD4+ T cell response.

METHODS AND RESULTS: Target genes were identified from curated open access data sets of human array data as overexpressed in human HNSCC but not in normal mucosa. Initial database query resulted in 3 data sets (GSE31056, GSE3524, and GSE6631) which contained a total of 72 normal mucosa samples and 84 tumor samples. The top 300 over-expressed genes for each dataset were selected and then cross-compared between datasets for those with >3 log-fold increased expression with p<0.005 between all data sets analyzed. From within these data sets, 9 over-expressed genes were identified which met our criteria: AGRN, MFAP2, COL4A1, AURK, FN1, SOX4, BUB1B, GTF2E1, and COL1A2. Systematic review of the literature was performed to investigate a role for these genes and their proteins in HNSCC and a pathogenic or prognostic role was revealed for each gene. We have used an in silico prediction algorithm to identify 15-mer peptides sequences (epitopes) which display high affinity binding across multiple HLA-II subtypes, which we now correlate with those able to induce Th CD4+ T cell responses. These predicted Th epitopes will be our first candidates to evaluate Th1 and Th2 responses. Finally, to narrow our list of potential targets for the vaccine, we will evaluate these 9 antigens in siRNA screening to identify those required for survival of HNSCC cell lines in comparison with a control normal epithelial cell line.

CONCLUSION: We have identified 9 genes which are implicated in HNSCC and whose peptide products may serve as ideal anti-cancer vaccine candidates after further immunologic testing. A vaccine able to elicit a strong Type I immune response could prevent disease recurrence or synergize with current anti-PD1/PD-L1 or other therapies to improve responses.