Assessment of Myocardial Iron Overload and Strain Abnormalities in Pediatric β-Thalassemia Using Multiparametric CMR.

Background/

Objectives

Myocardial iron overload is a major contributor to adverse cardiac outcomes in pediatric patients with transfusion-dependent β-thalassemia (TDT). Cardiovascular magnetic resonance (CMR), including T2* and T1 mapping, allows quantification of myocardial iron and early detection of cardiac dysfunction. Artificial intelligence (AI)-assisted CMR feature tracking (CMR-FT) provides a sensitive and reproducible approach for assessing myocardial deformation, even in patients with preserved left ventricular ejection fraction (LVEF). This study aimed to evaluate the utility of AI-based CMR-FT and its relationship with multiparametric CMR biomarkers, including myocardial strain (GCS, GLS, GRS), tissue characteristics (T2*, T1), and left ventricular (LV) geometry in pediatric TDT patients.

Methods

In this retrospective study, 68 pediatric patients with β-thalassemia major and 20 age-matched healthy controls underwent CMR with T2*, T1 mapping, and FT-based strain analysis. Myocardial iron overload was defined as T2* < 20 ms. Strain parameters were compared between groups, and correlations with tissue characteristics and LV geometry were assessed using Pearson's correlation.

Results

Patients with myocardial iron overload have significantly reduced GCS compared to controls (-17.4 ± 1.6% vs. -19.3 ± 4.5%, p < 0.01). GCS correlated with T2* (r = -0.33, p = 0.007) and T1 (r = -0.45, p < 0.001). GLS was sensitive to LV geometric changes, particularly concentric remodeling, correlating with LV mass/EDV ratio (r = -0.449, p < 0.001).

Conclusions

AI-based CMR-FT combined with multiparametric tissue imaging enhances early detection of subclinical myocardial dysfunction in pediatric TDT, offering diagnostic insights beyond conventional CMR metrics and supporting improved risk stratification.