The Prognostic and Predictive Value of Soluble Type IV Collagen in Colorectal Cancer: A Retrospective Multicenter Study.

PURPOSE To investigate the prognostic and predictive biomarker value of type IV collagen in colorectal cancer. EXPERIMENTAL DESIGN Retrospective evaluation of two independent cohorts of patients with colorectal cancer included prospectively in 2004-2005 (training set) and 2006-2008 (validation set). Plasma samples were available from 297 (training set) and 482 (validation set) patients. Type IV collagen determinations were performed using an ELISA. From the training set, 222 tumors were available for IHC. Clinical and follow-up data were retrieved from patient files and national registries. RESULTS High levels of type IV collagen showed independent prognostic significance in both cohorts with hazard ratios (HRs; for a one-unit change on the log base 2 scale) of 2.25 [95% confidence intervals (CIs), 1.78-2.84; P < 0.0001] and 2.24 (95% CI, 1.75-2.86; P < 0.0001) for the training and validation set, respectively. The prognostic impact was present both in patients with metastatic and nonmetastatic disease. The predictive value of the marker was investigated in stage II and III patients. In the training set, type IV collagen was prognostic both in the subsets of patients receiving and not receiving adjuvant antineoplastic therapy. However, in the validation set, the prognostic effect of the marker vanished when looking at patients who received adjuvant antineoplatic therapy (HR 0.90; 95% CI, 0.42-1.93) but was still present in the group not receiving adjuvant chemotherapy (HR 2.88; 95% CI, 1.98-4.21). CONCLUSIONS The results indicate clinical validity of type IV collagen as a prognostic biomarker in colorectal cancer, although the suggested predictive role of the marker should be validated. Clin Cancer Res; 22(10); 2427-34. ©2015 AACR.