Carcinoembryonic antigen as an independent prognostic factor of recurrence and survival in patients resected for colorectal liver metastases: a systematic review.

PURPOSE: We attempted to answer the question of whether serum levels of carcinoembryonic antigen provide prognostic information, in terms of survival, in patients resected for colorectal liver metastases, independently of that provided by other commonly used radioclinical and pathologic factors.
METHOD: We performed a systematic review, without meta-analysis, of the biomedical literature using the methodology recommended by the Committee on Evidence-Based Laboratory Medicine of the International Federation of Clinical Chemistry and Laboratory Medicine.
RESULTS: Despite the absence of sufficient details about the methods used to measure serum carcinoembryonic antigen in the 14 studies reviewed, strong arguments exist to include preoperative carcinoembryonic antigen measurements in future trials on the subject. In particular, preoperative carcinoembryonic antigen was found to be significant in the two studies with the greatest number of patients having a preoperative carcinoembryonic antigen assay, in the four studies with the most recent series of patients, in the study in which preoperative carcinoembryonic antigen was used as a continuous variable, and in the study in which preoperative carcinoembryonic antigen was used in terms of doubling time. Postoperative carcinoembryonic antigen was found to have a prognostic significance in the only two studies that evaluated this variable.
CONCLUSION: Taking into account the possible reasons for disagreements regarding carcinoembryonic antigen prognostic value between the 14 studies reviewed, we propose some recommendations to improve the reproducibility and the quality of future studies in this field. In particular, we stress the need for a higher degree of multidisciplinary collaboration in clinical trials.