Fatty acid metabolism-related molecular subtypes and a novel model for predicting prognosis in bladder cancer patients

Abstract

This study aims to develop fatty acid metabolism-related molecular subtypes and construct a fatty acid metabolism-related novel model for bladder cancer (BCa) by bioinformatic profiling. Genome RNA-seq expression data of BCa samples from the TCGA database and GEO database were downloaded. We then conducted consensus clustering analysis to identify fatty acid metabolism-related molecular subtypes for BCa. Univariate and multivariate Cox regression analysis were performed to identify a novel prognostic fatty acid metabolism-related prognostic model for BCa. Finally, we identified a total of three fatty acid metabolism-related molecular subtypes for BCa. These three molecular subtypes have significantly different clinical characteristics, PD-L1 expression levels, and tumor microenvironments. Also, we developed a novel fatty acid metabolism-related prognostic model. Patients with low-risk score have significantly preferable overall survival compared with those with high-risk score in the training, testing, and validating cohorts. The area under the ROC curve (AUC) for overall survival prediction was 0.746, 0.681, and 0.680 in the training, testing and validating cohorts, respectively. This model was mainly suitable for male, older, high-grade, cluster 2–3, any TCGA stage, any N-stage, and any T-stage patients. Besides, we selected FASN as a hub gene for BCa and further qRT-PCR validation was successfully conducted. In conclusion, we developed and successfully validated a novel fatty acid metabolism-related prognostic model for predicting outcome for BCa patients.