Journal: Scientific reports
This study developed and validated a serum tumor marker-based predictive model for early breast cancer diagnosis using data from 1,366 patients and 1,186 healthy controls.
Eight markers were assessed, with AFP, CA242, CA15-3, and NSE identified as independent diagnostic predictors through multivariate logistic regression.
The four-marker panel demonstrated high diagnostic accuracy, with an AUC of 0.90 in the development cohort and 0.82 in validation, alongside strong sensitivity and specificity.
The model effectively differentiated:
- tumor stages
- lymph node involvement
- metastatic status
- molecular subtypes
Additionally, longitudinal changes in risk scores reflected treatment response, decreasing during remission and increasing with disease progression.
This model offers a promising adjunct tool for early detection and dynamic monitoring of breast cancer.