Preoperative imaging of hilar cholangiocarcinoma: surgical evaluation of standard practises

It was the goal of this study to compare the results of the preoperative diagnostic workup (ERC, MRC, and PTC) with the tumor extent of the surgical specimen in patients with hilar cholangiocarcinoma (hilCC).

Between 9/97 and 12/2002 82 patients with hilCC were treated at our institution. In 59 patients tumor resection was feasible. Preoperative ERC, MRC and PTC - blinded for the idendity of the patients - were analysed retrospectively and compared with the surgical specimen.

PTC resulted in significantly superior visualization of the bile ducts including the hilar lesion compared to ERC and MRC (p < 0.01). ERC, MRC and PTC were correct in predicting tumor extent in 29, 36 and 53 % of cases, respectively. The extent of the tumor was overestimated in 42 % (ERC), 41 % (MRC) and 31 % (PTC). Underestimation or wrong assessment or no decision at all occurred in 31, 23 and 16 %, respectively. In 20 patients results of all three diagnostic methods were available allowing a statistical comparison regarding the resection to be performed: PTC was superior to ERC (McNemar test:p < 0.01), but not to MRC. In the patients with overestimated tumor extent both the rate of curative resections and survival were similar to the other resected patients.

In contrast to most reports in the literature, ERC and MRC were found to be of limited reliability regarding the assessment of tumor extent. PTC proved to be the most reliable approach. Overestimation of the tumor extent, which may lead to exclude the patient from potentially curative surgery, was the most common mistake in each diagnostic modality.