Evaluation of the inflammation-based modified Glasgow Prognostic Score (mGPS) as a prognostic and predictive biomarker in patients with metastatic colorectal cancer receiving first-line chemotherapy: a post hoc analysis of the randomized phase III XELAVIRI trial (AIO KRK0110)

Abstract

The inflammation-based modified Glasgow Prognostic Score (mGPS) combines serum levels of C-reactive protein and albumin and was shown to predict survival in advanced cancer. We aimed to elucidate the prognostic impact of mGPS on survival as well as its predictive value when combined with gender in unselected metastatic colorectal cancer (mCRC) patients receiving first-line chemotherapy in the randomized phase III XELAVIRI trial. In XELAVIRI, mCRC patients were treated with either fluoropyrimidine/bevacizumab followed by additional irinotecan at first progression (sequential treatment arm; Arm A) or upfront combination of fluoropyrimidine/bevacizumab/irinotecan (intensive treatment arm; Arm B). In the present post hoc analysis, survival was evaluated with respect to the assorted mGPS categories 0, 1 or 2. Interaction between mGPS and gender was analyzed. Out of 421 mCRC patients treated in XELAVIRI, 362 [119 women (32.9%) and 243 men (67.1%)] were assessable. For the entire study population a significant association between mGPS and overall survival (OS) was observed [mGPS= 0: median 28.9 months, 95% confidence interval (CI) 25.9-33.6 months; mGPS= 1: median 21.4 months, 95% CI 17.6-26.1 months; mGPS= 2: median 16.8 months, 95% CI 14.3-21.2 months; P < 0.00001]. Similar results were found when comparing progression-free survival between groups. The effect of mGPS on survival did not depend on the applied treatment regimen (P= 0.21). In female patients, a trend towards longer OS was observed in Arm A versus Arm B, with this effect being clearly more pronounced in the mGPS cohort 0 (41.6 versus 25.5 months; P= 0.056). By contrast, median OS was longer in male patients with an mGPS of 1-2 treated in Arm B versus Arm A (20.8 versus 17.4 months; P= 0.022). We demonstrate the role of mGPS as an independent predictor of OS regardless of the treatment regimen in mCRC patients receiving first-line treatment. mGPS may help identify gender-specific subgroups that benefit more or less from upfront intensive therapy.