Abstract
ABSTRACT
Background
Anthracycline cardiotoxicity is often silent at the subclinical stage. Left ventricular ejection fraction (LVEF) detects injury late, and global longitudinal strain (GLS) is sensitive but usually used alone. This study aimed to develop and internally validate a baseline model combining speckle‐tracking GLS, conventional echocardiography, and clinical factors to identify women at high risk of subclinical myocardial dysfunction before doxorubicin (DOX) chemotherapy.
Methods
In a single‐center prospective study, 200 breast cancer patients scheduled for DOX‐based chemotherapy underwent baseline and 1‐month assessment. The primary outcome was subclinical myocardial dysfunction, defined as > 15% relative GLS reduction with preserved LVEF.
P
in univariable analysis entered multivariable logistic regression for nomogram construction. Discrimination, calibration, and decision curve analysis used 1000‐bootstrap validation.
Results
Subclinical myocardial dysfunction occurred in 60 of 200 patients (30.0%). Independent predictors were baseline GLS, LDL‐C, trastuzumab use, planned cumulative dose, LAVI, and hypertension. The model reached an AUC of 0.803, higher than baseline GLS, a clinical‐only model, and baseline LVEF. The bootstrap‐corrected C‐index was 0.777.
Conclusion
A baseline nomogram combining GLS, conventional echocardiography, and clinical factors may help identify patients at high risk of early subclinical myocardial dysfunction before chemotherapy, outperforming single parameters and supporting risk‐stratified monitoring.