Pharmacogenomic prediction of immune-related adverse events from immune checkpoint inhibitors among Asian patients.

Abstract

2547 Background: Immune checkpoint inhibitors (ICIs) have ushered in a unique entity known as immune-related adverse events (IrAEs) that can be debilitating and challenging for physicians. Given that genomic variation underlies both disease susceptibility and drug response, there is reasonable cause to believe that genomic markers are predictive of IrAEs. We perform a pioneering pharmacogenomic study to uncover genomic biomarkers associated with IrAEs from ICIs. Methods: Since March 2018, we recruited cancer patients treated with ICIs from the National University Cancer Institute Singapore and Tan Tock Seng Hospital. IrAEs were clinical characterized and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 5. DNA was extracted and genotyped by Infinium Global Screening Array (700K markers). Statistical analyses were performed using SVS/HelixTree. Genetic association was performed by logistic regression. Bonferroni corrected P < 7.1E-08 was considered statistically significant. Results: We conducted a pilot pharmacogenomic genome-wide association study (GWAS) on 307 patients of Asian Ancestry. Median age was 62. Majority were male (68.1%), Chinese (75.9%), ECOG PS 0-1 (91.6%), stage IV cancer at diagnosis (64.6%). Non-small cell lung cancer (36.2%), renal cell carcinoma (12.4%) and hepatocellular carcinoma (7.2%) were the three most common cancers. Top four ICIs used were pembrolizumab, nivolumab, atezolizumab and durvalumab, respectively (44.6%, 26.1%, 10.1% and 10.1% respectively). Nine percent of patients received dual ICIs concurrently. Median duration of treatment was 152 days and median follow up was 212 days. IrAEs were seen in 50.5% of patients. Skin (21.8%), endocrine (6.8%) and hepatotoxicity (5.9%) were the most common IrAEs. Eight percent of patients had CTCAEv5 grade ≥3 toxicity, of which hepatotoxicity (2.6%), skin (1.6%) and pulmonary toxicities (1.3%) were the most common. A preliminary pharmacogenomic investigation revealed one potential novel genetic locus associated with IrAEs: LOC105373202 rs5915369; Unadjusted P= 6.6E-08, OR (95%CI) = 27.8 (3.7-206.5), minor allele = A, 10.7% in cases vs 0.4% in controls. We also identified an additional 3 independent SNPs (rs167609, rs2341687, DMDrs5928214) with nominal significance (7.1E-0-8≤ P< 5.0E-7). Conclusions: This pilot pharmacogenomics GWAS uncovered 4 potential novel genetic loci predictive of IrAEs from ICIs amongst Asian patients. Further pharmacogenomic discovery/replication and functional validation studies are currently on-ongoing to identify specific genomic biomarkers that predispose individual patients to IrAEs.