Evidence regarding the efficacy and tolerability of chemotherapy in older patients with pancreatic cancer remains insufficient. This study aimed to determine the optimal chemotherapy regimen and relative dose intensity (RDI) in older patients with pancreatic cancer.

This retrospective analysis included patients aged ≥ 75 years with metastatic pancreatic cancer treated with gemcitabine plus nab-paclitaxel (GnP) or gemcitabine (Gem) monotherapy, using data from the Tokushukai Real-World Data project. Patients were categorized by RDI into high (≥ 85%), middle (> 50, < 85%), and low (≤ 50%) groups. Hazard ratios with 95% confidence intervals for overall survival were estimated using multivariate Cox proportional hazards regression models to evaluate the impact of each covariate on overall survival.

A total of 153 patients were identified. More patients in the Gem group had a performance status of ≥ 2 (13.4% vs. 5.4%). There were slightly more patients aged ≥ 80 years in the Gem group, but no significant differences in sex, nutritional indices, body mass index, smoking history, or initial chemotherapy dosage. The median overall survival was longer in the GnP group (9.1 months vs. 6.1 months, p = 0.02). Patients in the GnP-middle RDI group had the best overall survival (median, 11.0 months, p = 0.01). Cox regression identified GnP-middle RDI, performance status, modified Glasgow Prognostic Score, and high body mass index as significant factors for prolonged overall survival.

Maintaining a medium RDI of GnP resulted in better overall survival than maintaining a high RDI GnP or Gem monotherapy in older patients with metastatic pancreatic cancer. These findings emphasize balancing treatment intensity with tolerability.