Abstract
Breast cancer is one of the most prevalent malignancies among women. Although the impact of age on breast cancer progression is well documented, the role of aberrant alternative splicing events in older adults breast cancer patients remains poorly understood. Here, we identified that older adults breast cancer patients exhibit a higher frequency of aberrant alternative splicing events compared to younger patients using public database, a finding that was further validated by data from FUSCC cohorts. A total of 390 high-variability-specific splicing events were observed exclusively in older adults patients. The unsupervised clustering analysis revealed the existence of three distinct subtypes of older adults patients, each displaying significantly different immune cell infiltration profiles and prognostic outcomes. To identify the key regulatory factors of these splicing subtypes, we conducted AS activity score analysis and identified 68 RNA-binding proteins as potential modulators. Subsequently, a machine learning approach using SelectKBest-SVM was employed to construct a predictive model, which demonstrated optimal performance in predicting the prognosis of older adults breast cancer patients, with a high AUC and validation on an independent test set. The developed predictive model offers a promising tool for personalized treatment strategies and accurate prognostication, advancing precision medicine for older adults breast cancer patients.