|| Checking for direct PDF access through Ovid
Although recent findings of cancer biology research indicate that prognostic power arises from genes expressed by stromal cells rather than epithelial cells, desmoplastic reaction (DR) has not been completely examined as a prognostic marker for colorectal cancer. A pathologic review of 821 stage II and III patients who underwent R0 resection for colorectal cancer at 4 independent institutions was conducted. DR was classified as mature, intermediate, or immature based on the existence of hyalinized keloid-like collagen and myxoid stroma at the extramural desmoplastic front. Totally, 325, 282, and 214 patients were classified as having mature, intermediate, and immature DR, respectively. DR significantly influenced the recurrence rate in the liver, lung, and peritoneum (P≤0.0001 to 0.01). Five-year relapse-free survival (RFS) rate was the highest in the mature group (85.7%), followed by the intermediate (77.3%) and immature (50.4%) groups. A significant adverse impact of immature stroma on RFS was observed in subset analyses of the 4 institutions. Multivariate analysis revealed that DR, along with T and N stages, is an independent prognostic factor. On the basis of Harrell’s concordance index, the prognostic power of DR categorization (0.67) in stratifying RFS was greater than any other conventional prognostic factors, including TNM (0.64), N (0.62) and T stages (0.59), venous invasion (0.59), and tumor grade (0.54). Characterizing DR based on the histologic products of activated fibroblasts is valuable for evaluating prognostic outcomes. To our knowledge, this is the first study reporting a greater prognostic power of histology of the fibrotic stroma than that of tumor factors.