Surgery for hilar cholangiocarcinoma: a multi-institutional update on practice and outcome by the AFC-HC study group

Abstract

Introduction: Surgical resection is the only option for long-term survival in patients with hilar cholangiocarcinoma (HC), but it is associated with high morbidity and mortality. The aim of the present study was to prospectively assess the perioperative management and short-term outcomes of surgical treatment of HC in a recent, multi-institutional study with a short inclusion period.

Methods: Between January and December 2008, a register prospectively collected data on patients operated on for HC (exploratory or curative surgery) in eight tertiary centers. The register focused on perioperative management, resectability, surgical procedures employed, morbidity, and mortality. The study cohort consisted of 56 patients (40 men and 16 women) with a median age of 63 years (range, 33-83 years).

Results: Among the 56 patients, 47 (84%) were jaundiced and 42 (75%) tumors were classified as Bismuth-Corlette type III-IV. Nine patients (16%) underwent staging laparoscopy and four (7%) received neoadjuvant chemotherapy. Preoperative biliary drainage (endoscopy, 42%) was performed in 38 (81%) jaundiced patients and portal vein embolization (right side, 83%) was performed prior to surgery in 18 patients (32%). Among these 56 patients, curative resection was achieved in 39 (70%). All underwent major liver resection (>3 segments), bile duct resection, and lymphadenectomy. Thirteen patients (36%) underwent portal vein resection, one of whom also required pancreaticoduodenectomy. Eighty-two percent of resected patients (n = 32) had no proof of malignancy prior to hepatectomy. Clear surgical margins were obtained in 77% (n = 30). The postoperative mortality was 8% and complications occurred in 72% of the resected patients. Seven (25%) patients required reoperation, and 15 (54%) patients required percutaneous drainage. In a univariate analysis, the risk factors for morbidity were intraoperative blood transfusion (p = 0.009) and vascular clamping (p = 0.006). The median length of hospitalization was 20 ± 13 days.

Conclusion: Curative resection for HC is associated with a high rate of R0 resection. However, surgery is associated with high levels of morbidity and mortality, despite intensive perioperative management.