内田 雄一郎

Objective: We evaluated the prognostic significance and implications of a major pathologic response (MPR) after preoperative treatment for pancreatic ductal adenocarcinoma (PDAC). Summary Background Data: Preoperative treatment is increasingly used for PDAC to improve oncologic outcomes. The pathologic response represents a potential indicator of treatment efficacy; however, its prognostic value in PDAC remains unclear. Methods: We retrospectively analyzed 739 patients who underwent pancreatectomy for PDAC after preoperative treatment at 21 institutions in Japan. The pathologic response was graded using Evans’ classification, with MPR defined as Evans grade III/IV. Survival outcomes and prognostic factors were evaluated, and factors associated with achieving MPR were analyzed to develop a predictive model. Results: MPR was achieved in 11.5% of patients. The MPR group had a significantly longer median overall survival (71.5 vs. 40.9 mo) and recurrence-free survival (55.5 vs. 15.2 mo) than the non-MPR group. Multivariate analysis identified MPR as an independent prognostic factor for overall survival. In the MPR subgroup, neither overall nor recurrence-free survival differed according to adjuvant chemotherapy administration; multivariate analysis did not identify adjuvant therapy as an independent prognostic factor. Predictive factors for achieving MPR included chemoradiotherapy, preoperative duration ≥6 months, normal carbohydrate antigen 19-9 after preoperative treatment, and complete or partial radiologic response. Finally, we developed a simplified predictive model for achieving MPR. Conclusions: MPR was independently associated with favorable survival in PDAC. The prognostic impact of adjuvant chemotherapy was not observed among patients who achieved MPR, suggesting that MPR may inform individualized postoperative management and warrants prospective validation.

BACKGROUND: Stapler transection during robotic distal pancreatectomy (RDP) has limitations, including device cost, difficulty in thick pancreas, and challenges in pathological assessment of the margin. We developed the crush and clip (CC) technique, in which the parenchyma is crushed using Maryland forceps and the main pancreatic duct is clipped without stump reinforcement. This study aimed to describe the CC technique and assess its non-inferiority to stapler transection regarding postoperative pancreatic fistula (POPF). METHODS: We retrospectively analyzed 127 RDP cases at a high-volume center (CC: 23; stapler: 104) between 2010 and 2025. Procedures used included da Vinci Xi, da Vinci SP, or hinotori™ systems. The primary endpoint includes clinically relevant POPF (ISGPS grade B/C). Body mass index, stump thickness, and robotic platform were used to estimate the propensity score, and overlap weighting was applied. Non-inferiority was prespecified as a risk difference (CC - stapler) of less than + 5% with a 90% bootstrap confidence interval (CI). RESULTS: POPF occurred in 13% of CC and 26% of stapler cases (p = 0.280). The weighted analysis revealed incidences of 16.0% and 32.0%, respectively. The weighted risk difference was - 16.0% (90% CI, - 34.0% to + 4.3%), thereby meeting the non-inferiority margin. POPF was significantly lower with CC (6% vs. 47%, p = 0.013) when the pancreatic stump thickness was ≥ 14 mm. Major complications (Clavien-Dindo ≥ III) occurred in 14% of stapler cases but in none of the CC cases (p = 0.071). CONCLUSIONS: The CC technique was feasible and safe and statistically non-inferior to stapler transection for POPF, while providing technical advantages.