To systematically describe the clinical implementation of four-stage progressive pelvic floor reconstruction following extended pelvic exenteration (EPE), including baseline characteristics, intraoperative parameters, postoperative recovery, and the distribution of complications related to empty pelvis syndrome (EPS). A total of 389 cases from a single center were included, so as to provide comprehensive overview and evidence for standardized pelvic floor reconstruction.

This retrospective descriptive study extracted data from a prospective Chinese pelvic exenteration database. A total of 389 consecutive pelvic tumor patients receiving EPE with en bloc pelvic floor resection at the Second Affiliated Hospital of Naval Medical University from November 2021 to January 2026 were included. Reconstruction was categorized into four stages according to pelvic floor repair, pelvic cavity obliteration, wound management and gastrointestinal reconstruction for radiated patients. Demographics, preoperative baseline, intraoperative indicators, postoperative recovery and complication data were collected. Given its descriptive nature, no intergroup hypothesis tests between four stages were performed.

The four-stage technique was implemented in a stepwise manner: biological mesh reconstruction→greater omental packing→staged negative-pressure wound therapy→prophylactic ileocecal resection for high-risk patients receiving radiotherapy. The median operative duration across the four stages ranged from 520 to 560 min, median blood loss ranged from 740 to 880 ml, and R₀ resection rate ranged from 80.0% to 86.2%. Perineal wound management evolved from primary closure to open staged negative-pressure therapy. The median time to first postoperative flatus across the four stages was 60 to 72 hours, drainage tube indwelling duration was 50 to 80 days, hospital stay was 38 to 75 days, and time to complete healing of the empty pelvis was 62 to 82 days. The proportion of complications with Clavien-Dindo grade<Ⅲ ranged from 52.8% to 68.0%. Perioperative mortality was only observed in the first two stages. In Stage 4 (n=36), no enterocutaneous fistula, perineoenteric fistula, pelvic hemorrhage or flap-related complications were observed (0/36). This finding remains to be validated due to limited sample size; deep pelvic collection, bowel obstruction and severe urological complications still existed.

The four-stage progressive strategy achieves layered optimization of pelvic floor repair, wound management and individualized intervention for radiotherapy-high-risk patients. Reduced severe EPS-associated complications were noted, which is conducive to improving perioperative safety and wound healing quality, and offering clinical reference for precise EPS prevention and enhancedrecovery.