Baseline left atrial reservoir strain predicts subclinical left ventricular dysfunction in breast cancer patients undergoing chemotherapy

Background Cancer therapy–related cardiac dysfunction (CTRCD) is traditionally defined by reductions in left ventricular ejection fraction (LVEF) or global longitudinal strain (LV-GLS). However, ventricular measures alone such as baseline LVEF may inadequately capture early myocardial vulnerability. Left atrial reservoir strain (LASr), reflecting atrial–ventricular mechanical coupling, may provide incremental prognostic information beyond LVEF. Purpose To evaluate whether baseline LASr independently predicts 12-month LVEF deterioration and improves risk stratification beyond baseline LVEF and LV-GLS in breast cancer patients receiving chemotherapy. Methods In this prospective cohort study, breast cancer patients underwent serial echocardiography at baseline and 12 months post chemotherapy. Patients without complete imaging data or receiving early cardioprotective therapy were excluded. The primary endpoint was 12-month change in LVEF (ΔLVEF) calculated as LVEF at 12 months – LVEF at baseline. Sequential multivariable linear regression models were constructed: Model 1 included baseline LVEF; Model 2 added LV-GLS; Model 3 further incorporated LASr. Incremental value was assessed by change in R² and nested model comparison. Multivariable logistic regression examined predictors of CTRCD at 12 months. Results A total of 132 patients were included (mean age 50.6±10.6 years; 23% hypertension; 11% diabetes), with preserved baseline systolic function (LVEF 63.1±5.8%; LV-GLS −23.4±2.8%; LASr 55.1±18.9%). Baseline LVEF alone explained 26.4% of the variability in 12-month LVEF change. Addition of LV-GLS significantly improved model performance (ΔR²=0.075; p<0.001), and further incorporation of LASr provided additional incremental value (ΔR²=0.020; p=0.048), yielding a final R² of 0.359. In the fully adjusted model, baseline LVEF, LV-GLS and LASr were independently associated with ΔLVEF, with higher baseline LASr associated with less LVEF decline. At 12 months, 14 patients (10.6%) developed CTRCD. In multivariable logistic regression, baseline LV-GLS independently predicted CTRCD (OR 0.79 per 1% increase; 95% CI 0.62–0.99; p=0.041), whereas LASr was not independently associated with overt cardiotoxicity. Conclusion Sequential incorporation of strain parameters significantly enhances prediction of left ventricular systolic deterioration beyond baseline LVEF alone. LASr provides incremental information beyond LV-GLS and appears to capture early myocardial susceptibility preceding clinically defined CTRCD. These findings support integration of atrial strain into multiparametric echocardiographic risk assessment in cardio-oncology.