Hilar Cholangiocarcinoma: patterns of spread, the importance of hepatic resection for curative operation, and a presurgical clinical staging system.

To determine the resectability rate for hilar cholangiocarcinoma, to analyze reasons for unresectability, and to devise a presurgical clinical T-staging system. Ninety patients with hilar cholangiocarcinomas seen between March 1, 1991, and April 1, 1997, were evaluated. Accurate patterns of disease progression and therapy were evaluable. Disease was staged in 87 patients using extent of ductal tumor involvement, portal vein compromise, and liver atrophy. In 21 patients, disease was deemed unresectable for cure at presentation. In 39 patients, disease was found to be unresectable at laparotomy, 23 secondary to nodal (N2) or distant metastases. Unresectability was the result of metastases in 52% and of locally advanced disease in 28%. Thirty patients (33%) had resection of all gross disease, and 25 of these (83%) had negative histologic margins. Twenty-two patients underwent partial hepatectomy. The 30-day mortality rate was 7%. Projected survival is greater than 60 months in those with a negative histologic margin, with a median follow-up of 26 months. A presurgical T-staging system allows presurgical selection for therapy, predicts partial hepatectomy, and offers an index of prognosis. In half the patients, unresectability is mainly the result of intraabdominal metastases. Presurgical imaging predicts unresectability based on local extension but is poor for assessing nodal metastases. In one third of patients, disease can be resected for cure with a long median survival. Curative resection depends on negative margins, and hepatic resection is necessary to achieve this. The T-staging system correlates with resectability, the need for hepatectomy, and overall survival.