Impact of immune checkpoint inhibitors (ICI) in biliary tract cancer (BTC): Real-world outcomes and predictive factors.

Abstract

e14653 Background: ICI with gemcitabine and cisplatin (GC-ICI) has become a first-line (FL) standard of care in BTC following the success of the TOPAZ and KEYNOTE-966 trials. A clear understanding of the factors impacting ICI efficacy in BTC will help selecting optimal treatment strategies. This single institute retrospective study is aimed at achieving this. Methods: Patients with BTC receiving ICI from 1/2017 and 6/2023 at the Ohio State University were evaluated. The patient's baseline characteristics (BL, at the first dose of ICI) were extracted with details of immune-related adverse events (irAE) from chart review. Descriptive statistics, Cox proportional hazard (for univariate (UVA) and multivariate (MVA) analysis), and Wilcoxon test (compare survival) were used for survival outcomes that assess time to progression (TTP, first ICI dose to progression (PD) or death) and overall survival (OS). Results: The study cohort comprised 41 patients, with a median age of 64 (range: 39–88), 51% male, and predominantly Caucasian (88%). Nineteen patients received FL ICI alone or ICI-based therapy, while the remaining received second-line (SL) or later therapy. Treatment regimens (TR) included single-agent ICI (Sg, n = 25), GC-ICI (n = 13), and ICI with other drug combinations (Ot, n = 3). 9 MSI-H, 4 FGFR2, and 1 IDH1-positive patients were present in the cohort. Durvalumab was the ICI agent for GC-ICI in all patients. Other key BLs are the median neutrophil-to-lymphocyte ratio (NLR) of 4.77 and albumin (Alb) level of 3.6 mg/dL. IrAE was observed in 13 patients (5 ≥ grade 3). The group's median TTP and OS were 3 (range: 2-5) and 18 (15-26) months, respectively. Factors significantly associated with TTP in UVA included baseline neutrophil count (NC), white cell count, Alb, and hemoglobin, while trends were noted for NLR and irAE incidence (p = 0.06–0.7). For OS, significant factors in UVA included the line of ICI use, stage at diagnosis, NLR, and TR, with NLR remaining significant in MVA. Survival comparisons for select factors are described in the table below. At PD, notable changes included a rise in NC by 2.09 (p = 0.06) and NLR by 8.9 (p = 0.01). In the irAE group, a significant rise in NC by 4.2 (p = 0.04) was observed at irAE onset. Conclusions: This study highlights the role of NLR, Alb, and irAEs as potential prognostic biomarkers in BTC patients treated with ICIs. Patients receiving GC-ICI in FL settings demonstrated modest PFS and OS compared to other regimens. Notable survival disparities based on TR and BL factors underscore the need for personalized treatment strategies to optimize outcomes in BTC. Further research is warranted to validate these findings. TTP* OS* GC-ICI vs. Ot vs. Sg 3 vs. 4 vs. 3 (0.9) 9 vs. 40 vs. 25 (0.0002) SL vs. FL 3 vs. 5 (0.4) 34 vs. 11 (0.0006) irAE, yes vs. no 3 vs. 6 (0.01) 15.5 vs. 25 (0.1) NLR (≤4.77 vs. > 4.77) 4 vs. 2.5 (0.04) 22 vs. 15.5 (0.1) Alb (<3.6 vs. ≥3.6) 2 vs. 4 (0.03) 15 vs. 26 (0.3) *in months (p-value).