Early versus late recurrence of intrahepatic cholangiocarcinoma after resection with curative intent.

Abstract

The objective of this study was to investigate the characteristics, treatment and prognosis of early versus late recurrence of intrahepatic cholangiocarcinoma (ICC) after hepatic resection. Patients who underwent resection with curative intent for ICC were identified from a multi-institutional database. Data on clinicopathological characteristics, initial operative details, timing and sites of recurrence, recurrence management and long-term outcomes were analysed. A total of 933 patients were included. With a median follow-up of 22 months, 685 patients (73·4 per cent) experienced recurrence of ICC; 406 of these (59·3 per cent) developed only intrahepatic disease recurrence. The optimal cutoff value to differentiate early (540 patients, 78·8 per cent) versus late (145, 21·2 per cent) recurrence was defined as 24 months. Patients with early recurrence had extrahepatic disease more often (44·1 per cent versus 28·3 per cent in those with late recurrence; P < 0·001), whereas late recurrence was more often only intrahepatic (71·7 per cent versus 55·9 per cent for early recurrence; P < 0·001). From time of recurrence, overall survival was worse among patients who had early versus late recurrence (median 10 versus 18 months respectively; P = 0·029). In multivariable analysis, tumour characteristics including tumour size, number of lesions and satellite lesions were associated with an increased risk of early intrahepatic recurrence. In contrast, only the presence of liver cirrhosis was independently associated with an increased likelihood of late intrahepatic recurrence (hazard ratio 1·99, 95 per cent c.i. 1·11 to 3·56; P = 0·019). Early and late recurrence after curative resection for ICC are associated with different risk factors and prognosis. Data on the timing of recurrence may inform decisions about the degree of postoperative surveillance, as well as help counsel patients with regard to their risk of recurrence.