Evaluation of clinical benefit and progression-free survival ratio of targeted treatments in a prospective cohort study of patients with biliary tract cancers.

590 Background: Biliary Tract Cancers (BTC) have poor prognosis and limited therapeutic options. There is mounting evidence for biomarker-directed therapy for BTC, with several potentially actionable molecular targets reported in up to ~50% of patients and significant biological differences between intrahepatic (ICC) and extrahepatic (ECC) cholangiocarcinoma and Gallbladder cancer (GBC). ESMO and ASCO recommends the use of tumour multigene next generation sequencing (NGS) for CCA. Methods: This was a monocentric study assessing real-life clinical actionability of molecular alterations identified with expanded NGS for patients (pts) affected by advanced BTCs treated at Fondazione IRCCS Istituto Nazionale dei Tumori of Milan from January 2016 to July 2022. NGS was performed either via the “Hotspot Cancer Panel, Ion Torrent”, the “Oncomine Comprehensive Assay Plus” or the FoundationOne CDx panel; FGFR2 testing was also performed via fluorescence in situ hybridisation. Molecular alterations were classified according to ESMO Scale for Clinical Actionability of Molecular Targets (ESCAT) and correlated with targeted treatments administered and efficacy endpoints. Benefit from targeted treatment was evaluated by means of progression-free survival (PFS) ratio, as defined as the ratio of each patient’s PFS2 (i.e. PFS on molecularly informed therapy) to the PFS1 (PFS on his/her most recent previous treatment), with d for efficacy equal to 1. Results: Out of 340 pts with adequate tissue for NGS, 231 (68%) were affected by ICC, 61 (32%) by ECC and 48 (14%) by GBC. Actionable alterations as per ESCAT I-III were found in 33 % of pts (N= 113, including IDH1 mutation, FGFR2 fusions/rearrangement/mutations, MSI-H, BRAF V600E mutations, ERBB2 mutations/amplifications, BRCA1/2 mutations); 88% of these alterations were found in pts with ICC. Targeted therapy was actually started for 48 pts (14 %), with 9 pts treated with ivosidenib, 32 with FGFR2 inhibitors, 2 with PD-L1 inhibitor, 6 with BRAF+MEK inhibitor and 1 with an HER2 inhibitor). Overall response rate and disease control rate were 18% and 62%, respectively.Median PFS was 5.4 months(95% CI 3.0 - 6.4) overall and 10.12 months on BRAF+MEK inhibitors, 5.55 months on FGFR2 inhibitors, 2.92 months on ivosidenib and 6.23 months on PD-L1 inhibitors. Median PFS ratio was 1.1 (95%CI 0.7 - 1.4), with benefit rate of 51%. Of note, 5 (10%) received targeted therapy as > III line. Median overall survival of targeted therapy was 10 months. Conclusions: Expanded sequencing for BTCs is feasible in real-life and can improve treatment strategy. Analysis of PFS ratio and of treatment outcomes shows clinically meaningful benefit of targeted treatment in pretreated patients.