PURPOSE

The clinical risk scores (CRSs) of Fong and Nordlinger are used to predict the outcome of patients with colorectal liver metastases (CRLMs). This study investigated whether CRSs could predict the overall survival of patients with CRLM treated with or without neoadjuvant chemotherapy prior to resection.

METHODS

Patients with CRLM undergoing liver resection were analyzed retrospectively. The primary outcome measure was overall survival with vs. without neoadjuvant chemotherapy.

RESULTS

Between August 2002 and October 2011, 300 patients underwent liver resection for CRLMs at a large university hospital in Germany. Group A comprised 117 patients who received neoadjuvant chemotherapy and group B comprised 71 patients who did not. The Fong score predicted overall survival for patients who did not receive chemotherapy (p = 0.02), but not for those treated with chemotherapy (p = 0.69). The Nordlinger score was not predictive for either of the groups (p = 0.71 vs. p = 0.08 for groups A and B, respectively). Subgroup analysis of the Nordlinger score identified better overall survival in the high-risk group treated with chemotherapy (p = 0.05). Multivariate analysis identified a resection margin of <1 cm [OR 0.622 (95% CI: 0.17-2.31); p = 0.044], age >60 years [OR 0.535 (95% CI: 0.16-1.77); p = 0.022] and number of metastases >4 [OR 0.189 (95% CI: 0.06-0.61); p = 0.018] as independent prognostic factors for overall survival.

CONCLUSION

Thus, CRSs were not reliable prognostic tools for patients treated with neoadjuvant chemotherapy before liver resection in this analysis.