LAUSR

Introduction : Immune check point inhibitors(ICIs) are widely used to treat various solid tumors, but there is a paucity of biomarkers that reliably predict response to such therapy. We sought to identify somatic alterations (SAs) that can predict response to ICI. Further, we investigated the effect of these SAs on clinical outcome. Methods: We retrospectively reviewed medical records of all patients (pts) that received ICI at our institution (2012-2017) and had CGP performed on pre-treatment biopsies. DNA was extracted from formalin fixed paraffin embedded clinical specimens and CGP was performed on hybrid-capture, adaptor ligation based libraries to a mean coverage depth of >600 unique reads utilizing the Foundation Medicine platform (315 gene panel). Twenty-four SAs, occurring in at least 5% of the pts were correlated with response using the Wilcoxon-Mann-Whitney test and p-values were adjusted for multiple testing using Benjamini-Hochberg9s false-discovery rate (FDR) method. Cox proportional hazards regression was used to investigate the effect of baseline covariates on progression-free survival (PFS); The effect of SAs was analyzed using the exact log rank test. The effect of response on progression and death was investigated using a 4-state model with “ICI therapy”, “response”, “progression”, and “death” as possible states. Results: Among the 76 pts that met criteria, 71 (lung-25, urothelial-6, esophageal-6, gynecologic-6, renal-5, sarcoma-5, melanoma-4, colorectal-4, prostate-3, head and neck-3, other-4) had an evaluable response and were included in the analyses. Median age was 62 years, 6 (8.5%) pts received ICI as first line therapy, while 65 (91.5%) pts received a median of 1 line of therapy prior to ICI; median of 5 doses of ICI therapy were administered. Complete Response (CR), Partial Response (PR), Stable Disease (SD), and Progressive Disease (PD) were noted in 3 (4.2%), 11 (15.5%), 10 (14.1%) and 47 (66.2%) pts respectively. SAs in RBM10 (p = 0.0024), PIK3CA (p = 0.0027), ARID1A (p = 0.0039), and SMARCA4 (p = 0.0047) correlated with response in a statistically significant fashion (adjusted p value = 0.028). PFS and overall survival (OS) of the entire cohort was 3.7 and 11.4 months respectively. Pts with SAs in RBM10 (p = 0.0011), ARID1A (p = 0.0102), and PIK3CA (p = 0.0013) demonstrated a statistically significant improvement in PFS (median not met, not met, and 9.3 months, respectively). Response to ICI significantly reduced the hazard of progression (HR=0.11, p=0.0006), but not the hazard of non-relapse mortality (NRM) (HR=0.41, p=0.53). Conclusion: CGP was able to identify SAs predictive of response to ICI and improved PFS. The significant reduction in the hazard of progression with response to ICI emphasizes the predictive value of the identified biomarkers. The validity and putative mechanistic relevance of these predictive SAs need further elucidation. Citation Format: Ravi K Narra, Arun K Singavi, Jonathan Thompson, Smitha Menon, James Thomas Thomas, Carolyn Oxencis, Mathew Riese, Paul Ritch, Deepak Kilari, Aniko Szabo, Ben George. Application of comprehensive genomic profiling (CGP) to predict therapeutic response to immune checkpoint inhibitors (ICI) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 654.