Development and validation of a prognostic prediction model for cervical cancer patients treated with radical radiotherapy: a study based on TCGA database.

Background: Radiotherapy or concurrent chemoradiotherapy is the standard treatment for patients with locally advanced or inoperable cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC). However, treatment failure for CESC patients treated with radical radiotherapy still occurs due to local recurrence and distant metastasis. The previous prediction models were focused on all CESC patients, neglecting the prognostic differences under different treatment modalities. Therefore, there is a pressing demand to explore novel biomarkers for the prognosis and sensitivity of radiotherapy in CESC patients treated with radical radiotherapy. As a single biomarker has limited effect in stratifying these patients, our objective was to identify radioresponse-related mRNAs to ameliorate forecast of the prognosis for CESC patients treated with radical radiotherapy. Methods: Sample data on CESC patients treated with radical radiotherapy were obtained from The Cancer Genome Atlas (TCGA) database. We randomly separated these patients into a training and test cohorts using a 1:1 ratio. Differential expression analysis was carried out to identify radioresponse-related mRNA sets that were significantly dysregulated between complete response (CR) and radiographic progressive disease (RPD) groups, and univariate Cox regression analyses, least absolute shrinkage and selection operator (LASSO) method and multivariate Cox regression were performed to identify the radioresponse-related signature in the training cohort. we adopted survival analysis to measure the predictive value of the radioresponse-related signature both in the test and entire cohorts. Moreover, we developed a novel nomogram to predict the overall survival (OS) of CESC patients treated with radical radiotherapy. In addition, immune infiltration analysis and Gene Set Enrichment Analysis (GSEA) were conducted to preliminarily explore possible mechanisms. Results: This study included a total of 92 CESC patients subjected to radical radiotherapy. We developed and verified a risk score model based on radioresponse-related mRNA. The radioresponse-related mRNA signature and International Federation of Gynecology and Obstetrics (FIGO) stage were served as independent prognostic factors for CESC patients treated with radical radiotherapy. Moreover, a nomogram integrating radioresponse-related mRNA signature with FIGO stage was established to perform better for predicting 1-, 3-, and 5-year survival rates. Mechanically, the low-risk group under the risk score of this model had a better survival status, and the distribution of CD4 T cells was potentially involved in the regulation of radiotherapy response in CESC, leading to a better survival outcome in the low-risk group. Conclusions: This study presents a new radioresponse-related mRNA signature that shows promising clinical efficacy in predicting the prognosis of CESC patients treated with radical radiotherapy.