Abstract
The extracellular matrix (ECM) plays a central role in the formation of the tumor microenvironment. The deposition of the ECM is associated with poor prognosis in a variety of tumors. Aberrant ECM deposition could undermine the effect of chemotherapy and immunotherapy. However, there is no systematic analysis on the relationship between the ECM and prognosis or chemotherapy effect. In the present study, we applied the gene set variation analysis (GSVA) algorithm to score 2199 canonical pathways in 2125 cases of probe or sequencing data and identified the core matrisome as the driving factor in gastric cancer progression. We classified gastric cancer samples into three clusters according to the composition of the ECM and evaluated clinical and multi-omics characterization of ECM phenotypes. The ECM score was evaluated by GSVA score of core matrisome and a higher ECM score predicted poor prognosis of gastric cancer [Hazard Ratio (HR), 2.084; p-value < 2 × 10-16]. In The Cancer Genome Atlas (TCGA) cohort and KUGH, YUSH, and KUCM cohorts, we verified that patients with a low ECM score could benefit from chemotherapy. By contrast, patients with a high ECM score did not achieve satisfactory response from chemotherapy. Determining the characteristics of the ECM microenvironment might help to predict the prognosis and chemotherapy response of patients with gastric cancer, and help to resolve the enigma of chemoresistance acquisition, as well as providing inspiration to develop combination therapy.