| Purpose:Sarcomatoid renal cell carcinoma is a specific dedifferentiated form of kidney cancer with an extremely poor prognosis,and the molecular mechanisms involved are not well studied.The prognosis of renal cell carcinoma itself is extremely variable,and prognostic models are needed to stratify patients and thus guide precise treatment.Methods:Data on clear cell renal cell carcinoma with sarcomatoid differentiation(sRCC)and clear cell renal cell carcinoma without sarcomatoid differentiation(ccRCC)as a control were downloaded from the TCGA database,and the two groups of cases were matched using propensity scores,followed by transcriptomics and other comparisons of differences.Prognostic stratification models were constructed and validated by Lasso-COX regression using the genes associated with sarcomatoid differentiation obtained from the analysis.Results:A total of 409 clear cell carcinoma cases and 45 sarcomatoid renal cell carcinoma cases were collated,and propensity score matching resulted in 21 pairs of case data.Differential analysis of up-and down-regulated transcriptomics in sRCC was performed and functional enrichment was performed.18 immune checkpoint genes were found to be highly expressed in sRCC.Non-negative matrix decomposition unsupervised clustering was performed using sarcoma-like differentiation-associated genes,and 454 cases were better classified into 4 categories.A prognostic stratification model and a nomogram were constructed by downscaling sarcoma-like differentiation-related genes,and the area under the working curve of the nomogram was up to 0.875 for one-year subjects.Conclusion:A detailed comparison of multi-omic differences between sRCC and ccRCC was performed,and a prediction model developed using sRCC-related genes could well predict the prognosis of patients with clear cell renal cell carcinoma with/without sarcomatoid differentiation. |
PDF Full Text Request