Background: It has been reported that the combination of inflammation parameters, such as
albumin and C-reactive protein, in the modified Glasgow prognostic score (m-GPS) is a
poor prognostic indicator in several malignancies. Here, we quantify the prognostic impact
of this score and assess its value in colorectal cancer.
Methods: A systematic review of electronic databases was conducted to identify publications exploring the
association of m-GPS with outcome in colorectal cancer. Overall survival (OS) was the primary outcome, and
cancer-specific survival (CSS), progression-free survival, and disease-free survival were secondary outcomes.
Data from studies reporting a hazard ratio (HR) and 95% confidence interval (CI) were included in a metaanalysis.
Pooled HRs were computed and weighted using generic inverse-variance and random effects modeling.
All statistical tests were two-sided.
Results: Nine studies, which included a total of 2,227 patients, were included in the analysis. Overall, according
to multivariate analysis, m-GPS ≥1 was independently associated with an HR for OS of 1.69 (95%
CI=1.4–2.04; P<0.00001), an effect observed in all stages of disease. Six studies including a total of 1,751 patients
reported HR for CSS. Overall, a high m-GPS was associated with an HR for CSS of 1.84 (95%
Conclusions: A high m-GPS is associated with poor OS in colorectal cancer. The m-GPS is a cheap and easily
evaluable biomarker, and its incorporation into known prognostic scores for clinical decision making warrants
further investigation in this setting.