Human papillomavirus (HPV) induced malignancies and markers of immunogenicity.

Abstract

e23083 Background: Approximately 5% of all cancers result from HPV, commonly leading to head and neck (HN) and anogenital (AG) cancers. Immune therapy is used in many cancers including HN. Positive PDL1 and high tumor mutation load (TML) have been proposed as potential markers of immunogenicity among many tumors, but little is known specifically among HPV driven cancers. Methods: A retrospective analysis of patients with HPV related cancers was performed based on tissue availability. TP53 wild type status was used as a surrogate for HPV positivity. Among oropharynx cases with TP53 wt, p16 testing was performed for confirmation. Next generation Sequencing (NGS Sanger) was used for tumor mutation load (TML) calculation. High TML cut off was defined as ≥17 mutations/megabase based on prior publications. PDL1 status was reported based on IHC ≥5%. A 2-tail Fisher’s exact test was utilized for statistical analysis. Results: A total of 737 patients with HN and AG (cervical, anal, vulvar, penile) squamous cell carcinoma were identified. Of these, 251 had complete 592 gene panel NGS data. HPV + TP53 Wt disease was detected in oropharynx (70%), cervical (88%), and anal (82%) cancer. HPV+ HNSCC had a trend to a higher PDL1 compared to HPV neg (50% vs 31%, p = 0.16) but there were no significant differences among PDL1 between HPV+ and neg patients across the entire combined disease cohorts (41% vs 36%, p = 0.56). TML-high patients showed a trend to increased PDL1 positivity (60% vs 38%, p = 0.09). However, high TML was less common in HPV+ HNSCC vs HPV neg (2.2 vs. % 9.7% p = 0.19). Unlike some other malignancies, high TML was uncommon across HPV+ tumors in general (8%). Conclusions: Despite fewer TML high cases among HPV + patients, there was no difference among PDL1 expression. It is possible that a lower threshold may need to be used when determining TML high vs. low in viral-driven cancers. It is also possible that in view of viral immunogenicity, certain genomic alterations contribute more to higher PDL1 expression rather than the sum of total genomic alterations. We are currently pursuing both lines of investigation. These findings are hypothesis generating for future trial design rather than treatment decision-making, and need to be confirmed in large prospective cohorts.