Multi-disease transcriptomic analysis of sex hormone genes reveals a novel prognostic model for thyroid cancer with breast cancer correlations

Background There is a potential bidirectional pathogenicity between thyroid and breast cancers. The association between sex hormones and two types of malignant tumors has emerged as a topic of intense academic debate in recent years. However, the role of sex hormone metabolism-related genes in thyroid cancer still needs to be further explored. Methods We obtained thyroid and breast cancer transcriptome data from the TCGA database and sex hormone metabolism-related gene sets from the MSigDB database, thus screening for sex hormone metabolism-related genes linked to the two malignant tumors. Univariate cox regression analysis was used for the screening of disease-free survival (DFS)-associated genes. The TCGA-THCA patients were classified as two categories via a consistent clustering algorithm, and the differential genes between the two categories were subsequently screened. A sex hormone metabolism-related prognostic model (TBSMRPM) of thyroid cancer versus breast cancer consisting of 10 genes was developed by Cox regression analyses and least absolute shrinkage with selection operator (LASSO) cox regression analysis. Finally, we performed clinicopathological subgroup analyses to analyze the correlation between TBSMRPM and clinical characteristics, immune infiltration, tumor mutation burden (TMB), and chemosensitivity, and verified the expression of TBSMRPM signature genes by qRT-PCR. Results We identified 2 clusters correlated with sex hormone metabolism, and screened 10 prognostic differential genes related to thyroid cancer, breast cancer and sex hormone metabolism. After establishing the two risk groups for thyroid cancer originated from TBSMRPM, the results showed that the high-risk group exhibited the shorter DFS ( P <0.05). In further clinical stratification analysis, immune infiltration analysis, TMB and drug sensitivity analysis, the two TBSMRPM groups showed significant differences. The qRT-PCR results showed that C2CD4A , CERS1 , MMP9 , SLC5A1 , HORMAD2 were highly expressed in the IHH4, KTC-1, and TPC-1 cell lines, while SLITRK2 , ARHGEF37 , PLP1 , RNF223 , and F3 were lowly expressed. Conclusion The TBSMRPM established in this study has a certain value for the prognosis of thyroid cancer and contributes to refine clinicians’ treatment protocols.