Preoperative portal vein embolization is not associated with increased morbidity after major hepatectomy: A study of the national surgical quality improvement database

Abstract

Background: Portal vein embolization (PVE) is selectively performed in the preoperative setting to induce hypertrophy of the future liver remnant prior to major liver resection and avoid post-hepatectomy liver insufficiency. The primary aim of this study was to evaluate the safety of PVE among patients undergoing major hepatectomy by measuring the association of PVE with liver-specific and overall postoperative complication rates. Methods: A retrospective cohort study of patients who underwent major hepatectomy, including right or left hepatectomy and trisectionectomy, from 2014-2016 within the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) hepatectomy-specific module was performed. Patient demographic and clinical data was collected, and postoperative complication rates were compared. Multivariable (MV) regression models were constructed to investigate the association of PVE and liver-specific and overall morbidity. Results: A total of 3912 patients were identified, of which 9.9% (N = 388) underwent PVE. The majority of resections were performed for either primary hepatobiliary tumors (34.0%, N = 1332) or metastatic disease (45.9%, N = 1795). Patients who underwent PVE were slightly older (59.1 vs. 57.7 years, p = 0.03) and more often male (61.3% (N = 238) vs. 50.4% (N = 1776), p < 0.001). Most patients in the PVE cohort underwent right hepatectomy (51.8%, N = 201) or trisectionectomy (46.1%, N = 179), compared to right (49.3%, N = 1738) and left hepatectomy (29.6%, N=1042) in the non-PVE cohort (p<0.001). Median operative time was longer in the PVE group (310 vs. 276 minutes, p<0.001). Post-hepatectomy liver insufficiency was more common among patients undergoing PVE (18.6%, N = 72), as was bile leak (17.3%, N = 67) when compared to patients who did not undergo PVE (9.9%, N = 350 and 12.2%, N = 428, respectively). Overall complication rates were also higher among patients who underwent PVE (45.9% (N = 178) vs. 34.0% (N = 1199), p < 0.001). However, on MV analysis controlling for patient and technical factors, PVE remained associated with an increased risk of liver-specific complications [OR 1.33 (95% CI 1.01-1.74)] but was not associated with an increase in overall morbidity [OR 1.17, 95% CI 0.92-1.50), Table]. There was no difference in 30-day mortality (PVE 2.8% (N = 11) vs. non-PVE 2.7% (N = 95), p = 1.0). Conclusion: Within this large, national cohort, patients treated with PVE are older and undergo a more extensive liver resection, with a longer operative time. When controlling for patient and technical factors, PVE is associated with an increase in liver-specific complication rates. However, PVE is not associated with an increase in overall morbidity, suggesting that PVE is a safe procedure that can be used selectively in patients undergoing major hepatectomy.