Development of a prognostic risk model for colorectal cancer based on microsatellite stability-associated genes

Abstract Background Colorectal cancer (CRC) patients have a high recurrence rate, impacting survival. Microsatellite instability (MSI) is strongly linked to CRC development, making the MSI-related prognostic genes crucial for diagnosis and treatment. Methods This study used CRC datasets, including TCGA-CRC, GSE17537, GSE39582, and GSE18088. We analyzed differential expression between CRC and control samples, and between MSS and MSI-H samples. Key genes were identified through a co-expression network and used to develop a prognostic risk model. The model's performance was validated in GSE17537, and independent prognostic factors were identified to construct a survival nomogram. We also explored pathways linked to the risk groups and their association with the tumor immune microenvironment, and predicted potential therapeutic agents for CRC. Results We identified 11 prognostic genes ( CHGB , FABP4 , PLIN4 , PLIN1 , RPRM , C7 , AQP8 , C2CD4A , APLP1 , ADH1B , and CD36 ) and developed a CRC risk model that showed significant survival differences in the TCGA-CRC cohort and GSE17537, with AUCs over 0.6 at 3, 5, and 7 years. Independent prognostic factors included risk score, age, tumor stage, and pathological N, and a nomogram was created for survival prediction. The identified genes may influence CRC through various pathways and are linked to immune responses. Bleomycin emerged as a potential treatment, with CHGB and RPRM regulated by non-coding RNAs and transcription factors, possibly affecting CRC development. Conclusions Our analysis of microsatellite stability-associated genes in CRC highlights their impact on TIME, clinicopathological features, and prognosis, providing new insights into predicting prognosis and developing personalized treatments.