ctDNA-based clinicogenomic analysis of advanced head and neck cancer patients treated with immune checkpoint inhibitors.

Abstract

6036 Background: Head and Neck cancer (HNC) is among the diverse group of malignancies affecting the head and neck region including the oral cavity. Being the most prevalent malignancy in Southeast Asia, it has a high mortality rate. Despite the advancement in treatment, 5 years survival rate for HNC remains below 50%, and the majority of Pts receiving frontline therapy experience locoregional or incurable metastatic relapse. Immune checkpoint inhibitors (ICI) are recommended for relapsed patients, but only 20% of patients show measurable response. Currently, no predictive biomarkers are available to predict ICI response and there is an urgent need for genomic markers to predict ICI outcomes. Here we report comprehensive genomic profiling (CGP) of advanced HNC patients receiving ICI. Methods: ctDNA from 69 advanced HNC patients receiving combinational immune-chemotherapy were serially profiled at the baseline (BL) and post-treatment (Tx) by targeted, hybridization-based CGP using OncoIndx comprehensive gene panel (CGP) comprising 1080 genes. The ctDNA differential features at BL and post-Tx as well as among responders (R) and non-responders (NR) were correlated with Progression-free survival (PFS) and Overall survival (OS) using Kaplan-Meier statistics and multivariate analysis. Results: Among total patients, 58% (40/69) were responders (R) while the remaining were non-responders (NR). At the population level, HRR pathway tumor suppressors and epigenetic modifiers were the most frequent pathogenic variants. At BL, the NR population was enriched with oncogenic gene mutations compared to the R population. TP53 and BRCA pathway mutations (mTP53 + BRCA) showed a strong association with progression-free survival (PFS) and overall survival (OS). Pts with cooccurring mTP53 + BRCA had significantly lower PFS (median PFS: 2.77 months for mTP53 + BRCA pathway vs 9.1 months for wt TP53 + BRCA pathway. P=<0.0001, HR=3.2-11.6) and OS (median OS: 4.67 months for mTP53 + BRCA pathway vs 12.63 months for wtTP53 + BRCA pathway. P=<0.0001, HR = 11.18-55.27). NOTCH 1 or 2 variants were enriched in R population, with a beneficial effect on survival outcomes. Elevated ctDNA alterations and Tumor fraction (TF) concentrated in the NR population disproportionately contained subclonal potential drivers of immunotherapy resistance including NF1, STAT5 B, and STK11 mutations, and were associated with short survival. Univariate and multivariate analysis suggested that ctDNA mutations, TF, and high mutational heterogeneity emerged as risk factors for shorter PFS and OS. In contrast, total Indel burden and NOTCH mutations had beneficial effects on PFS and OS. Conclusions: Minimally invasive plasma ctDNA CGP showed heterogenous actionable mutations at BL and post Tx and identified immunotherapy resistance conferring genomic markers for stratifying potential responders for immunotherapy guidance.