Key result
A 6-variable clinical prediction score identified low-risk CTRCD patients with high negative predictive value (98.2% derivation, 92.9% validation) and good internal discrimination (AUC 0.822).
Why the study?
Cancer therapy-related cardiac dysfunction (CTRCD) is a major concern in patients undergoing anticancer therapy, and a validated clinical prediction tool was needed for early identification and basal risk stratification to guide follow-up and management.
Does a clinical prediction score based on baseline clinical, biomarker, and imaging variables accurately predict and stratify the risk of symptomatic or moderate/severe asymptomatic cancer therapy-related cardiac dysfunction in patients undergoing anticancer therapy?
Cohort (n=1,068)
Does a clinical prediction score based on baseline clinical, biomarker, and imaging variables accurately predict and stratify the risk of symptomatic or moderate/severe asymptomatic cancer therapy-related cardiac dysfunction in patients undergoing anticancer therapy?
A six-variable clinical risk score demonstrates high specificity and negative predictive value for ruling out moderate-to-severe or symptomatic cancer therapy-related cardiac dysfunction, potentially allowing less intensive cardiological follow-up for low-risk patients.
May support biomarker-based risk stratification in anthracycline recipients; leaves open whether monitoring alters CTRCD outcomes.
Background Cancer therapy-related cardiac dysfunction (CTRCD) is a major concern in patients undergoing anticancer therapy. Risk assessment tools aid clinical practice in patient-risk stratification, guiding follow-up and management. Purpose The aim of this study is to develop and externally validate a clinical prediction tool for the early identification of CTRCD, allowing stratification of patients according to their basal risk. Methods Model development was derived on 921 patients from the CARDIOTOX registry, largely exposed to anthracyclines, and externally validated in 147 patients with HER2-positive breast cancer from the CAMAHER cohort. Patients were stratified by degree of CTRCD into two categories: no or mild asymptomatic cardiotoxicity and symptomatic or moderate/severe asymptomatic cardiotoxicity. Independent predictors were identified through Fine-Gray competing risk regression, and a clinical score was performed. Internal and external validation was assessed by calibration and discrimination power. Results During follow-up, 50 patients (5.4%) from the CARDIOTOX registry and 9 patients (6.1%) in the CAMAHER cohort developed symptomatic or moderate/severe CTRCD. Independent predictors most strongly associated with the outcome included NT-proBNP (sHR: 3.46; 95% CI 1.55-7.71; 2 points), creatinine (sHR: 4.27; 95% CI 1.57-11.61; 2 points), left ventricular ejection fraction (sHR: 7.00; 95% CI 2.58-18.97; 3 points), previous cancer (sHR 3.21; 95% CI 1.21-8.51; 2 points), smoking (sHR 2.15; 95% CI 0.97-4.77; 1 point) and abnormal electrocardiogram (sHR 2.37; 95% CI 1.03-5.45; 1 point). Four risk categories were established based on these variables: Low (0-1 point), Moderate (2 points), High (3-4 points) and Very High (≥ 5 points). Seventy percent of patients without CTRCD or with mild asymptomatic were correctly classified into the Low-Risk category (Figure 1). Internal validation with 1000 bootstrap samples showed good discrimination (AUC: 0.822; 95% CI: 0.717-0.925), while external validation showed an AUC of 0.630 (95% CI: 0.395-0.862). Goodness of fit was confirmed by the Hosmer-Lemeshow test in both the derivation (χ2= 1.33, p= 0.515) and validation cohorts ((χ2= 1.23, p= 0.528) (Figure 2). High specificity (85.52% and 85.87%) and negative predictive values (98.22% and 92.94%) in both cohorts highlight the model´s utility in ruling out patients who do not require cardiological monitoring. Conclusions The clinical score reliably identifies patients at low risk of CTRCD, allowing a less intensive follow-up and optimizing cardio-oncology care.
No takes yet. Share an insight, caveat, or question.
Gonzalez-Juanatey et al. (2026) conducted a cohort in Cancer therapy-related cardiac dysfunction (CTRCD) in patients undergoing anticancer therapy (n=1,068). Clinical prediction score for CTRCD (based on NT-proBNP, creatinine, LVEF, previous cancer, smoking, and abnormal ECG) vs. Derivation vs. external validation cohort risk stratification was evaluated on Symptomatic or moderate/severe asymptomatic cancer therapy-related cardiac dysfunction (CTRCD) (AUC 0.822, 95% CI 0.717-0.925). A 6-variable clinical prediction score identified low-risk CTRCD patients with high negative predictive value (98.2% derivation, 92.9% validation) and good internal discrimination (AUC 0.822).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: