Clinical landscape of cell free DNA alterations in patients with advanced biliary tract cancer treated with targeted therapies.

Abstract

626 Background: Although 40% of patients (pts) with advanced biliary tract cancer (BTC) harbor actionable alterations amenable to targeted therapy (TT), tumor genotyping can be challenging due to limited tissue and other factors. Next generation sequencing (NGS) of cell free DNA (cfDNA) may aid in prognostication, identification of TT opportunities, and characterization of acquired resistance mechanisms to TT. Methods: cfDNA was collected prospectively in pts with BTC treated at MSK from 2016-2023. Samples were analyzed using a CLIA-approved, cfDNA targeted NGS assay (MSK-ACCESS). When available, matched tumor samples were analyzed using an FDA-authorized targeted NGS assay (MSK-IMPACT). Objectives included evaluating association of cfDNA genomic alteration variant allele frequency (VAF) with outcomes and description of genomic features identified on cfDNA at resistance to TT. Results: N=170 BTC pts (intrahepatic cholangiocarcinoma [CCA], n=121; extrahepatic CCA, n=29, gallbladder cancer, n=20) underwent cfDNA genotyping, comprising 270 samples overall. A majority of pts (146/170, 86%) had locally advanced or metastatic disease. OncoKB level 1/2 alterations were identified in 19% of pts overall. Pts with treatment-naïve locally advanced or metastatic BTC whose baseline detected cfDNA genomic alterations were VAFhigh (defined as > median VAFmax across all samples, n=60) had significantly worse median PFS (4.5 (95% CI 2.2-6.5) vs. 11.8 (95% CI 5.4-18.4) months, (HR 2.7, p=0.002) and OS (14.1 (95% CI 5.4-19.9) vs. 32.0 (95% CI 13.8-39.9) months, HR 3.1, p=0.002)) to first line treatment. Serial cfDNA was available from 28 pts who received TT. Emergent RAS isoform alterations were identified at resistance in 7/25 (28%) pts who received BRAF-, FGFR-, or HER2-directed TT. These included 3/5 pts who received BRAF/MEK TT for BRAF V600E+ BTC (emergent KRAS G12D in 1, NRAS Q61K in 1, polyclonal KRAS G12D, G12V, G13D in 1), 2/13 pts who received FGFR TT for FGFR2 fusion+ BTC (KRAS Q61H in 1, NRAS Q61R in 1), and 2/7 pts who received HER2 TT for ERBB2-amplified BTC (KRAS G12C in 1, KRAS amp in 1). Serial cfDNA also revealed canonical FGFR2 resistance mutations in 3 pts who received FGFR TT, as well as loss of index ERBB2 amp in 2, acquired ERBB2 resistance mutation in 1, acquired MYC amp in 2, and MET amp in 1 pt who received HER2 TT. Conclusions: cfDNA NGS in pts with BTC may assist with prognostication in advanced disease and can identify novel patterns of resistance, including frequent emergence of RAS alterations at resistance across multiple TTs. These may identify therapeutic opportunities for KRAS- and NRAS-directed therapeutics in development, and underscore the importance of serial cfDNA profiling.