Implications of clinical risk score to predict outcomes of liver-confined metastasis of colorectal cancer

Sang Joon Shin, Joong Bae Ahn, Jin Sub Choi, Gi Hong Choi, Kang Young Lee, Seung Hyuk Baik, Byung Soh Min, Hyuk Hur, Jae Kyung Roh, Nam Kyu Kim

Research output: Contribution to journalReview articlepeer-review

6 Citations (Scopus)

Abstract

Objective/background: We investigated the usefulness of a clinical risk scoring system (CRS) for guiding management and defining prognosis for patients with colorectal liver met"astases (CLM). Method: We retrospectively analyzed data about the correlation between outcomes and Fong's CRS from 1989 to 2010 for patients treated for CLM at the Severance Hospital. Results: Of 566 patients, 232 received adjuvant treatment after liver resection. Of these patients, 185 (81%) had a low CRS (0-2) and 47 (19%) had a high CRS (3-5). Stratification into high and low CRS allowed significant distinction between Kaplan-Meier curves for outcome. The 5-year survival rate was 88.5% and 11.5% among patients with a low and high CRS, respectively (P < 0.001). Seventy patients with initially unresectable CLM underwent liver resection after tumor downsizing by induction chemotherapy. Shifting of the CRS from high to low (8 patients; 11.4%) improved disease-free survival and overall survival. Conclusion: High CRS is associated with worse survival after resection in resectable and unresectable disease. The CRS may be used for risk assessment when recommending oncological surgical timing in initially unresectable disease and treatment options for perioperative or adjuvant treatment in resectable disease.

Original languageEnglish
Pages (from-to)e125-e130
JournalSurgical Oncology
Volume21
Issue number3
DOIs
Publication statusPublished - 2012 Sep

All Science Journal Classification (ASJC) codes

  • Surgery
  • Oncology

Fingerprint Dive into the research topics of 'Implications of clinical risk score to predict outcomes of liver-confined metastasis of colorectal cancer'. Together they form a unique fingerprint.

Cite this