Abstract 2415: Resistance to immune checkpoint inhibitor (ICI) therapy in metastatic melanoma is revealed by deep circulating tumor DNA tracking in concert with tumor immune profiling

Charles W. Abbott,Laura Keller,Isabel Heidrich,Julian Kött,Glenn Geidel,Daniel J. Smit,Ronald Simon,Stefan Schneider,Jason Pugh,Sean M. Boyle,Richard O. Chen,Klaus Pantel,Christoffer Gebhardt

DOI: https://doi.org/10.1158/1538-7445.am2024-2415

IF: 11.2

2024-03-28

Cancer Research

Abstract:Monitoring clinically actionable and emerging mutations in circulating tumor DNA (ctDNA) may be crucial for guiding treatment, but adoption has been slow due to the narrow scope and reproducibility of current approaches. Here, we investigate melanoma patients receiving ICI over several years, beginning with extensive tumor mutational and immune profiling followed by deep, whole-exome ctDNA monitoring.Here, 66 plasma samples from 23 advanced stage melanoma patients that were collected throughout ICI treatment (up to 40 months) were profiled using the NeXT Liquid Biopsy® platform (NeXT LB). NeXT LB monitors exome-wide ctDNA by leveraging a patient's tumor, PBMCs and plasma samples, detecting both tumor-derived and emerging mutations. In addition to ctDNA tracking, we also used the ImmunoID NeXT Platform® to immunologically profile tumor tissue; including gene expression quantification, HLA profiling (typing, mutation, and loss of heterozygosity), T-cell receptor and tumor microenvironment profiling, and neoantigen prediction. ctDNA dynamics and tumor immunity were correlated with clinical outcome. Patients who progressed had lower tumor mutational burden (TMB, Mann-Whitney (MW); p=0.03), and lower neoantigen burden (MW; p=0.04). Responding tumors were enriched for interferon gamma and inflammatory signaling (Wald; adjusted p<0.01) and myeloid dendritic cells (MW; p=0.01), as measured by baseline RNA expression profiling. With an average coverage of 2260X SD +/-157, plasma samples yielded from 4 to 4081 variants. The number of on-treatment tumor variants detected in plasma predicted increased OS hazard (HR=2.95, Wald p=0.03; Kaplan-Meier (KM) p=0.0003) and progression hazard (HR=1.74, Wald p=0.04; KM p=0.0003). Overall, 10% (4277) of all variants were novel mutations observed only in the plasma. Novel variants were less likely to emerge in genes conferring sensitivity to ICI therapy (mixed-model; p<0.001). Dynamic shifts in allelic fraction between baseline and the first available on-treatment timepoint predicted increased OS hazard (HR=3.02, Wald p=0.05) and progression hazard (HR=2.90, Wald p=0.02). In ctDNA clonal phylogenies, 100% of BRAF and NRAS mutations belonged to the dominant clone. By combining immune and ctDNA features, a multivariable model predicted resistance to ICI treatment (log-likelihood p=0.002; c-index=0.89). Using a single liquid biopsy platform, we profiled tumor immunity and tracked existing and emerging ctDNA variants over the course of ICI treatment. We demonstrate that, even though counts of on-treatment ctDNA variants were sufficient for significant survival models, deep profiling of both tumor immunity and ctDNA dynamics provided deeper understanding of ICI resistance. Subsequent studies will determine if ctDNA-guided intervention can improve clinical outcomes. Citation Format: Charles W. Abbott, Laura Keller, Isabel Heidrich, Julian Kött, Glenn Geidel, Daniel J. Smit, Ronald Simon, Stefan Schneider, Jason Pugh, Sean M. Boyle, Richard O. Chen, Klaus Pantel, Christoffer Gebhardt. Resistance to immune checkpoint inhibitor (ICI) therapy in metastatic melanoma is revealed by deep circulating tumor DNA tracking in concert with tumor immune profiling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular s); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl) nr 2415.

oncology