The Petersen prognostic index revisited in Dukes B colon cancer--Inter-institutional differences

Abstract

A prognostic index (Petersen index, PI) was created for patients with pT3-4 pN0 M0 (Stage II, Dukes' B) colon cancers to distinguish between patients with better and worse outcome, and to help in recommending adjuvant chemotherapy for high risk patients in this stage. The prognostic value of the PI was evaluated in two independent retrospective series of stage II (Dukes' B) colon cancer patients. The parameters defining the PI (venous invasion, peritoneal involvement, circumferential margin involvement, perforation through the tumour) and performance of the PI were compared in two institutions. The two series of patients consisted of 127 and 87 patients. Venous invasion was more frequently detected at one of the centres (p<0.01) and tumour perforation was more frequent at the other (p<0.01). There were no significant differences in the 5-year survival estimates of all patients (p=0.19), and of either the low PI value groups (p=0.52) or that of the high PI value groups (p=0.99) between the two sites. In contrast, there were significant differences in the survival estimates between patients of the low PI category and those of the high PI category altogether (p<0.01) and in either centre. Although, it was expected that differences in the frequency of the parameters involved in the PI would influence its performance, this was not confirmed by the data. Our results suggest that using the PI may be of value in prognostic factor based therapy selection of colon carcinoma patients.

Keywords: Colon cancer; Dukes’ B; Prognostic index; Stage II.