State of the Art of Immune Checkpoint Inhibitors in Unresectable Pancreatic Cancer: A Comprehensive Systematic Review

Abstract

Immune checkpoint inhibitors (ICIs) have transformed the therapeutic landscape for several malignancies, but their efficacy in unresectable pancreatic adenocarcinoma remains uncertain. This systematic review aimed to evaluate the effectiveness and safety of ICIs in this context, focusing on overall survival (OS), progression-free survival (PFS), objective response rate (ORR), disease control rate (DCR), and toxicity. A comprehensive search of MEDLINE, EMBASE, CENTRAL, and Scopus identified 34 eligible studies, including randomized controlled trials and observational cohorts. Quantitative synthesis involved 21 studies comprising 937 patients, with additional qualitative analyses on biomarker-driven subgroups and early-phase trials. The median OS across studies was 8.65 months, while the median PFS was 2.55 months. The ORR and DCR were 16.2% and 50.3%, respectively, with grade ≥3 treatment-related adverse events occurring in 22% of patients. Promising outcomes were observed in MSI-H/dMMR populations, although these represented only 1-2% of cases. Combination strategies with chemotherapy demonstrated synergistic potential but lacked definitive evidence due to heterogeneity and the absence of phase III trials. ICIs showed a manageable toxicity profile, highlighting their feasibility in selected patients. Future research should focus on overcoming tumor microenvironment barriers and identifying biomarkers to optimize responsiveness and expand the applicability of ICIs in pancreatic cancer.

Keywords: immune checkpoint inhibitors; immunotherapy; microsatellite instability; pancreatic adenocarcinoma.

Figure 1

Prisma 2020 flow diagram [14].

Figure 2

Forest plot displaying the meta-analysis results for OS. Callahan, 2023 (A), (B) [42]; Parikh, 2022 [40]; Xie, 2020 (A), (B), (C), [34]; Yang, 2024 (A) [35]; Chen, 2023 [23]; Gong, 2022 [24]; Liu, 2024 [39]; Ma, 2019 [26]; Sun, 2018 [31]; Wang, 2024 [33]; Yang, 2024 (B), (C) [35]; Zhang, 2022 [36].

Figure 3

Forest plot displaying the meta-analysis results for PFS. Callahan, 2023 (A), (B) [42]; Parikh, 2022 [40]; Xie, 2020 (A), (B), (C), [34]; Yang, 2024 (A) [35]; Chen, 2023 [23]; Gong, 2022 [24]; Liu, 2024 [39]; Ma, 2019 [26]; Sun, 2018 [31]; Yang, 2024 (A), (B), (C) [35]; Zhang, 2022 [36].

Figure 4

(a) Funnel plot illustrating the trim-and-fill method for the OS analysis. (b) Funnel plot illustrating the trim-and-fill method for the PFS analysis. The gray dots represent potential missing studies identified by the method, while the black dots correspond to the included studies.

Figure 5

(a) Plots of leave-one-out analyses for the OS; Callahan, 2023 (A), (B) [42]; Parikh, 2022 [40]; Xie, 2020 (A), (B), (C), [34]; Yang, 2024 (A) [35]; Chen, 2023 [23]; Gong, 2022 [24]; Liu, 2024 [39]; Ma, 2019 [26]; Sun, 2018 [31]; Yang, 2024 (B), (C) [35]; Zhang, 2022 [36]; Wang, 2024 [33]. (b) plots of leave-one-out analyses for the PFS; Callahan, 2023 (A), (B) [42]; Parikh, 2022 [40]; Xie, 2020 (A), (B), (C), [34]; Yang, 2024 (A) [35]; Chen, 2023 [23]; Gong, 2022 [24]; Liu, 2024 [39]; Ma, 2019 [26]; Sun, 2018 [31]; Yang, 2024 (A), (B), (C) [35]; Zhang, 2022 [36].