A real-world study on the prognosis of patients with pancreatic cancer: A prospective stationary survey from 2007-2020.

Objective Pancreatic cancer (PC) is one of the most refractory malignancies with a poor prognosis. However, improvements in the PC prognosis over time have not yet been thoroughly evaluated.We herein present a retrospective analysis of a prospectively maintained hospital-based cancer registry over 14 years. Methods The patients were prospectively enrolled in a Japanese government study. The period from 2007 to 2020 was divided into two seven-year periods to assess a prognostic improvement. The most recent censorship date is December 2024. Prognostic trends were compared across three treatment modalities: surgery, chemotherapy, and best supportive care (BSC). Results A total of 26,379 patients with all cancer types were prospectively enrolled over 14 years. Among these, 711 (2.7%) were enrolled for PC. Dichotomizing the period into the first and second halves, both the surgery group (18.7 months vs. 48.5 months, HR 0.45, P <0.001) and the chemotherapy group (9.1 months vs. 12.0 months, HR 0.58, P <0.001) had significantly better outcomes in the second half than in the first half. In the surgery group, the tumors were significantly smaller (P = 0.001) and detected earlier in the second half than in the first half (P = 0.007). In the chemotherapy group, the first-line treatment shifted from gemcitabine to gemcitabine plus nab-paclitaxel (GnP) in the second half (P <0.001). No significant improvement in the prognosis was observed in the BSC group (P = 0.597). Conclusions Prospective real-world data from 2007 to 2020 revealed that improvements in the early detection and adoption of GnP were temporally associated with an improved survival in patients with PC.