94MO Advanced extrahepatic cholangiocarcinoma: Post-hoc analysis of the ABC-01, -02 and -03 clinical trials

Prior data suggest that patients with iCCA have a better prognosis compared to other biliary tract tumours (BTC) (Lamarca et al, JNCI 2019). However, the outcome for eCCA is not well described and some of the current studies focused on this patient population may be challenging to contextualise. The aim of this post-hoc analysis is to provide reference survival data for patients with advanced eCCA treated with first-line cisplatin-gemcitabine (CisGem) chemotherapy within the prospective, randomised Advanced Biliary tract Cancer (ABC)-01, -02 and -03 studies. Individual level data from patients with eCCA recruited to these studies were retrieved.Survival analysis was performed using univariate and multivariable Cox Regression. All statistical tests were two-sided. Of 534 patients recruited into the ABC-01, -02 and -03 studies,179 (33.5%) had eCCA (107 distal (dCCA), 72 hilar (hCCA)). Of these, 117 were treated with CisGem and eligible for analysis. Most patients (n=82; 70.1%) had metastatic disease at the time of recruitment. Objective radiological response was achieved in 28 (23.9%) of eCCA: 19 dCCA (27.9%) and 9 (18.4%) hCCA. The median progression-free survival (PFS) for eCCA, dCCA and hCCA were 8.4 (95% CI 7.36-9.33), 8.3 (95% CI 6.3-9.4) and 8.4 months (6.5-10.6), respectively; median overall survival (OS): 12.8 months (95%CI 10.4-16.1), 14.3 (95% CI 8.8-17.4) and 12.2 (95% CI 8.3-16.1), respectively. eCCA with locally advanced disease had longer median PFS (13.1 (95% CI 8.8-16.6) and OS (17.4 (95% CI 14.3-21.1) . Multivariable survival analysis for OS confirmed worse OS for eCCA with performance status of 2 (vs 0; HR 45.877 (95% CI 9.55-219.45)) and high baseline CA125 (HR 1.0007 (95% CI 1.0001-1.001; p-value 0.017) when adjusted for other prognostic factors identified in the univariate analysis (stage, Ca 19.9). These remained significant when adjusted for site of eCCA (data not shown). Patient outcomes were similar regardless of location of eCCA but worse than those reported historically for iCCA (specifically OS). Performance status and baseline tumour markers (CA125) were strongly associated with survival in eCCA. Outcomes with first-line CisGem + immunotherapy should also be explored.