Extra-hepatic (ehCCA) and intra-hepatic cholangiocarcinomas (iCCA): A genomic landscape comparison of peri-hilar (phCCA), common bile duct (cbdCCA) and gallbladder (gbCCA) with iCCA.

Abstract

630 Background: Metastatic CCAs have poor prognosis. Comprehensive genomic profiling (CGP) was used to compare genomic alterations (GA) in phCCA, with cbdCCA, gbCCA and iCCA. Frequencies of GA in therapy targets such as FGFR2, IDH, ERBB2 & BRAF, & Immuno-oncology (IO) drug response predictors (MSI status, TMB & PDL1) were examined. Methods: DNA extracted from 110 cases of phCCA, 743 cases of cbdCCA, 2983 cases of gbCCA and 9178 cases of ihCCA underwent hybrid capture based CGP to evaluate classes of genomic alterations (GA), MSI status, TMB levels, genomic ancestry, genomic signature, HRD score and germline status. phCCA cases were confirmed at surgery and pathology as originating from either left, right or common hepatic ducts. cbdCCA cases were confirmed by cbd biopsies or Whipple procedures. PD-L1 expression was determined by IHC using the Dako 22C3 assay using tumor proportional score (TPS) system. Results: The pt age distributions were similar; female gender was more frequent in gbCCA. gbCCA featured higher GA/tumor (P=.0001). The frequencies of FGFR2 & IDH1 targetable GA in phCCA, cbdCCA and gbCCA were uncommon especially when compared iCCA (P<.0001). GA in BRAF, CDKN2A, MTAP and PIK3CA in gbCCA were like phCCA and cbdCCA. At 16.5%, the frequency of GA in ERBB2 were highest in gbCCA and in all ehCCA when compared with iCCA (P<.0001). The frequencies of KRAS GA including KRAS G12C were greater in both phCCA and cbdCCA when compared with gbCCA (P=.0015) and iCCA (P.0001). TP53 GA were highest in gbCCA (P=.0007). IO biomarker frequencies were similar in all. The frequencies of HRD positive (gLOH) status and MMR genomic signature were similar in all. Conclusions: ehCCA subtypes feature characteristic GA profiles that have potential to influence therapy selection with lower frequencies of FGFR2 & IDH1 when compared with iCCA. In contrast, targetable ERBB2 & KRAS G12C are more frequent in ehCCA than iCCA, and highest in gbCCA & cbdCCA. peri-hilarCCA cbdCCA P-value (phCCA vs cbdCCA) gbCCA P-value (phCCA vs gbCCA) ihCCA P-value (phCCA vs iCCA) BRAF 3.6% 4.7% NS 2.4% NS 5.1% NS ERBB2 14.5% 10.6% NS 16.5% NS 5.1% 0.002 FGFR2 0.9% 2.0% NS 1.5% NS 11.7% 0.001 IDH1 3.6% 0.8% NS 0.6% NS 13.9% 0.006 KRAS 29.1% 46.2% 0.015 14.2% 0.0015 19.8% 0.071 TP53 45.5% 63.8% 0.011 65.7% 0.0007 33.9% 0.061 MSI-high 1.8% 1.3% NS 1.2% NS 1.8% NS TMB≥10 mut/mb 3.6% 4.2% NS 5.2% NS 3.6% NS PD-L1 low positive 21.1% 31.3% NS 23.5% NS 23.7% NS