HEARTOX-P: a novel multiparametric score for predicting cardiotoxicity in HER2-positive breast cancer patients undergoing chemotherapy

Introduction Cardiotoxicity is a complication of chemotherapy in breast cancer patients, especially associated with anthracyclines and trastuzumab. Consensus definitions encompass symptomatic and laboratory criteria, and imaging-based criteria, including a left ventricular ejection fraction (LVEF) reduction of >10 percentage points to a value <50%, or a >15% decline in global longitudinal strain (GLS) from baseline. Current guidelines recommend cardiac evaluation prior to initiating treatment. However, existing tools for predicting cardiotoxicity remain limited. Improved risk stratification could facilitate the early implementation of cardioprotective measures. Purpose This study aimed to identify variables with predictive capacity for cardiotoxicity in breast cancer pts and to develop a risk stratification score to aid in clinical decision-making. Methods A retrospective analysis was conducted on HER-2 positive breast cancer pts who underwent echocardiographic assessment before chemotherapy initiation between 2020 and 2024. Clinical, electrocardiographic, and echocardiographic parameters were evaluated. Group comparisons were performed using the independent t-test and Chi-square test, with statistical significance set at p<0.05. Logistic regression was used to assess associations between variables and cardiotoxicity. The predictive performance of the developed score was evaluated using ROC curve analysis and the corresponding AUC. Results A cohort of 60 females were analyzed. The mean age was 59.0 ± 13.0 years. Of these, 11 (18.3%) developed cardiotoxicity, according to ESC criteria. Categorical variables with statistically significant differences included the presence of diabetes (p=0.01), overweight (p=0.034), significant valvular disease (p=0.026), arrhythmia (p=0.026), prior trastuzumab-induced toxicity (p=0.033), and previous anthracycline exposure (p=0.028). Continuous variables with significant differences included pre-chemotherapy LVEF (<0.001), indexed left ventricular end-diastolic volume (p=0.043), GLS (p=0.004), and QTc interval (p=0.045). Continuous variables were dichotomized to maintain statistical significance, and a predictive score ranging from 0 to 10 was constructed. Patients who developed cardiotoxicity had a higher HEARTOX-P score (2.5 ± 1.9) compared to those who did not (0.9 ± 1.0) (p<0.001). The HEARTOX-P score demonstrated predictive capability, with an odds ratio of 2.2 (p=0.042; 95% CI: 1.331–3.651). ROC curve analysis yielded an AUC of 0.782 (p=0.004; 95% CI: 0.644–0.920), indicating a moderate discriminatory capacity. Conclusions The HEARTOX-P score is a simple, multiparametric tool incorporating ten variables collected through established surveillance protocols. It identifies patients at higher risk of developing cardiotoxicity during treatment and may guide early cardioprotective interventions. However, further studies are required to validate the score and determine its clinical applicability.