Acute pancreatitis linked to poorer survival in checkpoint inhibitor cohort

In a retrospective single-center cohort study of patients with malignant tumors who developed pancreatic enzyme elevation while receiving immune checkpoint inhibitors, acute pancreatitis was associated with substantially poorer overall survival and more frequent interruption or permanent discontinuation of immunotherapy. Among 104 patients with immune checkpoint inhibitor–related pancreatic injury, 32 (30.8%) fulfilled the criteria for acute pancreatitis. Patients with pancreatitis had a higher mortality rate than those with isolated enzyme elevation, and the association remained significant in multivariable and sensitivity analyses.

The study included patients treated with immune checkpoint inhibitors between January 2020 and October 2025 who developed new-onset pancreatic enzyme elevation without an alternative cause. The investigators classified patients as having immune checkpoint inhibitor–related pancreatic injury without pancreatitis or immune checkpoint inhibitor–related acute pancreatitis according to whether they met the Revised Atlanta criteria. The cohort consisted of people with malignant tumors, and the study compared clinical features, treatment patterns and survival outcomes between the two groups.

Patients who developed acute pancreatitis were older, and a larger proportion had non-hepatocellular carcinoma tumors or stage IV disease than patients with isolated pancreatic enzyme elevation. The pancreatic event was also followed by more changes in cancer treatment. Interruption of immune checkpoint inhibitor therapy occurred in 75.0% of patients with pancreatitis, compared with 18.1% of those without pancreatitis. Permanent discontinuation occurred in 68.8% and 12.5% of the respective groups. Both comparisons had P values below 0.001, showing that treatment disruption was more frequent among patients with the pancreatitis phenotype.

Overall survival was analyzed with pancreatitis treated as a time-varying exposure, using Simon–Makuch curves and time-dependent Cox models. Pancreatitis was associated with worse overall survival in the unadjusted analysis, with a hazard ratio of 3.17 (95% confidence interval, 1.86–5.40; P < 0.001). The association was also significant after multivariable adjustment, with a hazard ratio of 3.49 (95% confidence interval, 2.05–5.94; P < 0.001). These estimates describe an association within the cohort and do not establish that pancreatitis caused the deaths.

The association remained significant in a restricted cohort of patients with hepatocellular carcinoma and in a Cox model stratified by tumor type. Landmark analyses at 3 and 6 months also showed significantly poorer subsequent survival among patients who had developed pancreatitis before the landmark time. Across these analyses, the results consistently linked the pancreatitis phenotype with worse observed survival, while the retrospective design and single-center scope limit the conclusions that can be drawn from the findings.

Immune checkpoint inhibitor–related pancreatic injury can range from isolated elevation of pancreatic enzymes to overt acute pancreatitis, and the study indicates that these phenotypes differ in their clinical course and observed outcomes. Acute pancreatitis was associated with more frequent interruption or discontinuation of immunotherapy and with markedly lower subsequent survival than isolated enzyme elevation. The findings do not establish that every patient with enzyme elevation will develop pancreatitis, nor do they provide an individual estimate of absolute risk. The investigators concluded that acute pancreatitis may represent a clinically important phenotype within pancreatic immune-related adverse events.

Subject of Research: Survival and treatment disruption in immune checkpoint inhibitor–associated acute pancreatitis

Article Title: Pancreatitis and survival outcomes in immune checkpoint inhibitor–treated patients with pancreatic enzyme elevation: a single-center cohort study

Article References: Liu, J., Lu, X., Jiang, W., Chen, G., Zhang, K., Zhang, Z., Deng, Y., & Weng, C. (2026). Pancreatitis and survival outcomes in immune checkpoint inhibitor–treated patients with pancreatic enzyme elevation: a single-center cohort study. BMC Cancer. https://doi.org/10.1186/s12885-026-16901-7

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16901-7

Keywords: immune checkpoint inhibitors, acute pancreatitis, pancreatic enzyme elevation, immune-related adverse events, overall survival, cancer immunotherapy, time-dependent Cox analysis, retrospective cohort study, Pancreatitis, survival, outcomes, immune