Interventional radiology in the management of complications after pancreatic surgery: a single-center experience.

PURPOSE: To assess the impact of interventional radiology (IR) as first-line rescue therapy for major complications after pancreaticoduodenectomy and to quantify its effect on surgical re-exploration and 90-day mortality.

MATERIALS AND

METHODS

In this retrospective single-center cohort, 200 consecutive patients undergoing pancreaticoduodenectomy (2014-2025) were reviewed. Complications were recorded using ISGPS/ISGLS definitions; postoperative pancreatic fistula (POPF) included biochemical leak. Patients were grouped by initial management strategy (IR-first vs. primary surgery); crossover to the alternative treatment was recorded and interpreted within a step-up management framework rather than uniformly as treatment failure. Outcomes included need for re-intervention, technical success, length of stay, and 90-day mortality. Fisher's exact test compared mortality between strategies within clinically relevant POPF (CR-POPF) and postpancreatectomy hemorrhage (PPH) subgroups.

RESULTS

Overall, 126/200 patients (63%) developed postoperative complications and 75/126 (59.5%) required early re-intervention. IR accounted for 63/75 reinterventions (84.0%). POPF occurred in 79/200 (39.5%); CR-POPF in 49/200 (24.5%). Among CR-POPF, 33/49 (67.3%) underwent image-guided drainage and 15/49 (30.6%) required re-exploration; 90-day mortality was lower with IR-first than surgery-first management in unadjusted analysis (5/33, 15.2% vs. 8/15, 53.3%; p = 0.012). PPH occurred in 26/200 (13.0%); embolization was performed in 16/26 (61.5%) with low step-up to surgery (1/16, 6.3%). Mortality did not differ significantly between embolization-only and re-exploration-only pathways (3/16, 18.8% vs. 4/8, 50.0%; p = 0.182). Biliary complications were managed primarily with IR in 32/33 patients (97.0%).

CONCLUSION

A substantial proportion of major post-pancreaticoduodenectomy complications can be controlled with minimally invasive IR techniques, limiting surgical re-exploration. IR-first strategy was associated with lower 90-day mortality in unadjusted analysis; however, this association was attenuated after adjustment for clinical severity. Prospective multicenter studies are needed to further refine selection criteria and step-up treatment algorithms.

LEVEL OF EVIDENCE: Level 3, therapeutic study.