Perioperative risk factors and postoperative outcomes following pelvic exenterations for gynecologic versus non-gynecologic malignancies

Objectives: Pelvic exenteration (PE) is an aggressive, radical procedure associated with high rates of postoperative complications (POC). Although PE was first described for the palliation of gynecologic malignancies, this has been adopted for the management of colorectal and genitourinary cancers. The objective of this study is to evaluate the perioperative risk factors and 30-day surgical outcomes following PE for gynecologic and non-gynecologic cancers. Methods: Data was obtained from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database. All women who underwent PE from 2012-2018 were identified using appropriate CPT codes and gender. Comparative analyses were performed and stratified by admission status to evaluate demographics, preoperative and intraoperative variables, and surgical outcomes. Statistical tests were performed with R Studio v.1.1.456. Results: A total of 2,069 women who had undergone a PE were identified from 2012-2018. Of these, 73 were performed for an unknown indication and 109 were for a nonmalignant indication, and these were excluded. The remaining 1,887 were used for our analysis: 473 (25.1%) had a PE for a gynecologic cancer, 973 (51.6%) for a genitourinary cancer, 427 (22.6%) for a colorectal cancer, and 14 (0.74%) for an unspecified cancer. PE performed for gynecologic malignancy were associated with longer operative times (median 385 vs 350 minutes, p<0.001) and longer hospital stays (median 10 vs 8 days, p<0.001). There was a higher POC rate in gynecologic cancer patients (75.8% vs 66.8%, p<0.001). There were significantly higher rates of blood transfusions, surgical site infections, wound disruptions, venous thromboemboli, and reoperations in the gynecologic malignancy group (Table 1). There were no significant differences in urinary tract infections and 30-day readmission rates. Among the entire cohort, 277 (13.4%) patients had a concurrent myocutaneous flap procedure. Higher POC rates were seen in patients who also had a flap performed (82.7% vs 65.8%, p<0.001), and this included more surgical site infections in flap patients (29.6% vs 18.0%, p<0.001). More patients undergoing PE for a gynecologic cancer had a flap performed (21.1% vs 10.3%, p<0.001) compared to patients following PE for non-gynecologic cancer. When adjusting for type of cancer and blood transfusions on multivariate analysis, concurrent flap procedures remained the only independent risk factor for higher POC rates (OR = 1.73, 95% CI 1.04-2.85, p=0.03). Conclusions: PE performed for gynecologic malignancies are associated with higher morbidity compared to PE done for non-gynecologic cancers. Flaps are performed more often during PE for gynecologic cancers, which may contribute to the increased morbidity seen in these patients following PE. Data on prior radiation therapy which is not available through the NSQIP database, could be evaluated in future studies. Pelvic exenteration (PE) is an aggressive, radical procedure associated with high rates of postoperative complications (POC). Although PE was first described for the palliation of gynecologic malignancies, this has been adopted for the management of colorectal and genitourinary cancers. The objective of this study is to evaluate the perioperative risk factors and 30-day surgical outcomes following PE for gynecologic and non-gynecologic cancers. Data was obtained from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database. All women who underwent PE from 2012-2018 were identified using appropriate CPT codes and gender. Comparative analyses were performed and stratified by admission status to evaluate demographics, preoperative and intraoperative variables, and surgical outcomes. Statistical tests were performed with R Studio v.1.1.456. A total of 2,069 women who had undergone a PE were identified from 2012-2018. Of these, 73 were performed for an unknown indication and 109 were for a nonmalignant indication, and these were excluded. The remaining 1,887 were used for our analysis: 473 (25.1%) had a PE for a gynecologic cancer, 973 (51.6%) for a genitourinary cancer, 427 (22.6%) for a colorectal cancer, and 14 (0.74%) for an unspecified cancer. PE performed for gynecologic malignancy were associated with longer operative times (median 385 vs 350 minutes, p<0.001) and longer hospital stays (median 10 vs 8 days, p<0.001). There was a higher POC rate in gynecologic cancer patients (75.8% vs 66.8%, p<0.001). There were significantly higher rates of blood transfusions, surgical site infections, wound disruptions, venous thromboemboli, and reoperations in the gynecologic malignancy group (Table 1). There were no significant differences in urinary tract infections and 30-day readmission rates. Among the entire cohort, 277 (13.4%) patients had a concurrent myocutaneous flap procedure. Higher POC rates were seen in patients who also had a flap performed (82.7% vs 65.8%, p<0.001), and this included more surgical site infections in flap patients (29.6% vs 18.0%, p<0.001). More patients undergoing PE for a gynecologic cancer had a flap performed (21.1% vs 10.3%, p<0.001) compared to patients following PE for non-gynecologic cancer. When adjusting for type of cancer and blood transfusions on multivariate analysis, concurrent flap procedures remained the only independent risk factor for higher POC rates (OR = 1.73, 95% CI 1.04-2.85, p=0.03). PE performed for gynecologic malignancies are associated with higher morbidity compared to PE done for non-gynecologic cancers. Flaps are performed more often during PE for gynecologic cancers, which may contribute to the increased morbidity seen in these patients following PE. Data on prior radiation therapy which is not available through the NSQIP database, could be evaluated in future studies.