Nomogram For Predicting Survival in Breast Cancer Patients Based on Lncrna Expression and Clinical Characteristics

Abstract

Objective: Current prognostic models for breast cancer (BC) are largely dependent on clinical factors and immunohistochemical markers, or in some cases, a limited number of gene signatures. These approaches have certain limitations in accuracy and clinical applicability. This study aimed to construct a predictive nomogram that integrates molecular signatures of long non-coding RNAs (lncRNAs) with conventional clinical factors, thereby providing a more comprehensive and individualized tool for survival prediction in BC patients.

Methods: Using The Cancer Genome Atlas (TCGA) database, RNA-sequencing data and clinical information of breast cancer patients were retrieved. Differentially expressed genes were identified with the DESeq2 R package, followed by univariate and multivariate Cox regression analyses to identify prognostic lncRNA biomarkers. A 9-lncRNA risk score model was then established and validated. Independent prognostic factors were further integrated with clinical variables, and a predictive nomogram was constructed. Model performance was evaluated using the concordance index (C-index), Kaplan–Meier survival analysis, ROC curves, and calibration plots.

Results: A total of 1208 transcriptome profiles were analysed, including 1096 breast cancer and 112 normal tissue samples. From these, 2100 differentially expressed genes were identified. Nine lncRNAs (AC068858.1, AC000067.1, LINC00460, LINC02408, AC136475.5, AC023043.4, AC073359.1, AC244502.1, and COL4A2-AS1) were significantly associated with overall survival (OS). Four acted as risk factors (HR > 1), whereas five served as protective factors (HR < 1). The 9-lncRNA signature stratified patients into high- and low-risk groups with significant prognostic differences (p < 0.001). Time-dependent ROC curves demonstrated strong predictive accuracy, with AUC values ranging from 0.72–0.92 across different datasets and follow-up periods. Multivariate Cox analysis confirmed that age and the lncRNA model were independent prognostic predictors. A nomogram combining these two factors was constructed, achieving a C-index of 0.81 and demonstrating excellent calibration for 1-, 3-, and 5-year OS predictions.

Conclusion: The 9-lncRNA-based prognostic model, integrated with clinical risk factors such as age, provides a robust and individualized tool for predicting breast cancer survival. This nomogram may serve as a valuable reference for clinical decision-making and personalized management strategies in breast cancer patients.