Steroid hormone biosynthesis pathway indicates platinum chemoresistance: Insights From Genomics and Metabolomics

Background Platinum-based chemotherapy is commonly used in first-line cancer treatments. However, because of drug resistance, some patients cannot benefit from this treatment. This study aimed to identify platinum-resistance-related biological pathways and validate them in metabolomics data. Methods Using the Transcriptome data of pan-cancer cell lines, random forest and Diffrank algorithms were utilized to identify differential pathways between cisplatin-resistant and -sensitive groups. Untargeted metabolomics measured from colorectal cancer plasma was used to examine the selected pathway. Results Steroid hormone biosynthesis pathway was selected as the most differential pathway between platinum-sensitive and resistant groups. In the colorectal cancer metabolic dataset, eight metabolites in this pathway were enriched, among which the three highest predictive metabolites were used for calculating prognosis index. Prognosis indexs for predicting progression free survival and overall survival achieved area under receiver operating characteristics curve values of 0.6509 and 0.8088 in preoperative plasma and 0.9698 and 0.9709 in postoperative plasma. Kaplan-Meier curves showed progression free survival was significantly different between high- and low-level prognosis index groups in both preoperative ( P  = 0.0405) and postoperative plasma ( P  = 0.0079). Conclusions Using biological pathways as a bridge, drug-sensitivity-related pathways and metabolic biomarkers can be identified. Thus, this scheme is promising for personalized treatment.