Background: Traditionally, survival estimates have been reported solely as survival from the time of surgery, but future survival probability likely changes based on the survival time already accumulated after therapy-otherwise known as conditional survival (CS). We sought to assess the comparative performance of various colorectal liver metastasis prognostic scoring systems, as well as to investigate the CS of patients who underwent resection of colorectal liver metastasis.
Study design: Between 1982 and 2008, 949 patients who underwent colorectal liver metastasis resection were identified from an international multi-institutional database. Various prognostic scoring systems were evaluated using Cox proportional hazards models and calculated concordance index (c). CS estimates were calculated as CS = S((x+5))/S((x)).
Results: Overall survival after liver resection was 65% at 3 years and 45% at 5 years, with a median survival of 52 months. All of the prognostic scoring systems had poor-to-moderate prognostic discriminatory ability (Fong c = 0.57, Nordlinger c = 0.56, Memorial Sloan-Kettering Cancer Center nomogram c = 0.58). Using CS, the probability of surviving an additional 5 years, given that the patient had already survived 1, 3, or 5 years, was 41%, 40%, or 50%, respectively. The inadequate performance of the prognostic scoring systems was explained by the fact that as survival from liver resection increased from 0 to 5 years, the 5-year observed CS improved substantially for patients who were initially predicted to have poor survival at the time of surgery.
Conclusions: Colorectal liver metastasis prognostic scoring systems have fair-to-moderate performance. CS can provide more accurate prognostic information for patients and physicians after colorectal liver metastasis resection and should be incorporated into the quantification of survival.
Copyright 2010 American College of Surgeons. Published by Elsevier Inc. All rights reserved.