Abstract
Background: The correlation of type 2 diabetes mellitus (T2DM) with colorectal cancer (CRC) has garnered considerable attention in the scientific community. Despite this, the molecular mechanisms underlying the interaction between these two diseases are yet to be elucidated. Hence, the present investigation aims to explore the shared gene signatures, immune profiles, and drug sensitivity patterns that exist between CRC and T2DM. Methods: RNA sequences and characteristics of patients with CRC and T2DM were retrieved from The Cancer Genome Atlas and Gene Expression Omnibus databases. These were investigated using weighted gene co-expression network analysis (WGCNA) to determine the co-expression networks linked to the conditions. Genes shared between CRC and T2DM were analyzed by univariate regression, followed by risk prognosis assessment using the LASSO regression model. Various parameters were assessed through different software such as the ESTIMATE, CIBERSORT, AND SSGSEA utilized for tumor immune infiltration assessment in the high- and low-risk groups. Additionally, pRRophetic was utilized to assess the sensitivity to chemotherapeutic agents in both groups. This was followed by diagnostic modeling using logistic modeling and clinical prediction modeling using the nomogram. Results: WGCNA recognized four and five modules that displayed a high correlation with T2DM and CRC, respectively. In total, 868 genes were shared between CRC and T2DM, with 14 key shared genes being identified in the follow-up analysis. The overall survival (OS) of patients in the low-risk group was better than that of patients in the high-risk group. In contrast, the high-risk group exhibited higher expression levels of immune checkpoints The Cox regression analyses established that the risk-score model possessed independent prognostic value in predicting OS. To facilitate the prediction of OS and cause-specific survival, the nomogram was established utilizing the Cox regression model. Conclusion: The T2DM + CRC risk-score model enabled independent prediction of OS in individuals with CRC. Moreover, these findings revealed novel genes that hold promise as therapeutic targets or biomarkers in clinical settings.