The evolving Use of prognostic factors after resection of colorectal liver metastases

Georgios Karagkounis, Michael A. Choti

Research output: Contribution to journalArticlepeer-review


The expanding use of surgical therapy for patients with colorectal liver metastases (CRLM) and developments in chemotherapeutic regimens have led to a significant improvement in survival. However, outcomes can vary substantially and criteria for determining the prognosis of individual patients are lacking. Traditionally, clinicopathologic factors, for example primary tumor stage, number and size of liver metastases, preoperative carcinoembryonic antigen levels, presence of extrahepatic disease, and others, have been used to determine which patients are more likely to experience recurrence and poor survival. However, these factors, both separately and as part of scoring systems, have been inconsistent and conflicting in determining prognosis. Recently, molecular and biological indicators have emerged as potential prognosticators for CRLM patients undergoing hepatic resection. Tumor response to chemotherapy, both on imaging and on pathology, mutation status of such oncogenes as KRAS and BRAF, and expression patterns of proliferative markers including MACC1 and Ki-67, have all furnished promising results in prediction of patient outcomes. Moreover, circulating tumor cells and tumor DNA may not only be a useful prognostic instrument but also an excellent means of screening for early detection of recurrence.

Original languageEnglish (US)
Pages (from-to)218-226
Number of pages9
JournalCurrent Colorectal Cancer Reports
Issue number2
StatePublished - Jun 2014


  • BRAF
  • Colorectal liver metastases
  • Hepatectomy
  • KRAS
  • Molecular
  • Prognostic factors
  • Resection
  • Scoring systems
  • Survival

ASJC Scopus subject areas

  • Hepatology
  • Oncology
  • Gastroenterology


Dive into the research topics of 'The evolving Use of prognostic factors after resection of colorectal liver metastases'. Together they form a unique fingerprint.

Cite this