Early Detection of Myocardial Involvement in Thalassemia Intermedia Patients: Multiparametric Mapping by Magnetic Resonance Imaging.

Abstract

Background: No study has assessed myocardial T1 and T2 values in patients with beta-thalassemia intermedia (β-TI).

Purpose: To assess the prevalence of myocardial involvement in β-TI patients by T2* relaxometry and native T1 and T2 mapping and to determine the correlation of myocardial relaxation times with demographic and clinical parameters.

Study type: Prospective matched-cohort study.

Subjects: 42 β-TI patients (27 females, 39.65 ± 12.32 years), enrolled in the Extension-Myocardial Iron Overload in Thalassaemia Network, and 42 age- and sex-matched healthy volunteers (27 females, 40.01 ± 11.36 years) and thalassemia major (TM) patients (27 females, 39.27 ± 11.57 years).

Field strength/sequence: 1.5 T/multi-echo gradient echo, modified Look-Locker inversion recovery, multi-echo fast-spin-echo, cine balanced steady-state-free precession, and late gadolinium enhancement (LGE) sequences.

Assessment: Hepatic, pancreatic, and left ventricular (LV) T2* values, LV native T1 and T2 values, biventricular ejection fractions and volumes, and presence and extent of replacement myocardial fibrosis.

Statistical tests: Comparisons between two groups were performed with two-sample t tests, Wilcoxon's signed rank tests, or χ² testing. Correlation analysis was performed using Pearson's or Spearman's test. P < 0.05 was considered statistically significant.

Results: β-TI patients had significantly higher LV T2 values than healthy subjects (56.84 ± 4.03 vs. 52.46 ± 2.50 msec, P < 0.0001) and significantly higher LV T1 values than TM patients (1018.32 ± 48.94 vs. 966.66 ± 66.47 msec, P < 0.0001). In β-TI, female gender was associated with significantly increased LV T1 (P = 0.041) and T2 values (P < 0.0001), while splenectomy and presence of regular transfusions were associated with significantly lower LV T1 values (P = 0.014 and P = 0.001, respectively). In β-TI patients, all LV relaxation times were significantly correlated with each other (T2*-T1: P = 0.003; T2*-T2: P = 0.003; T1-T2: P < 0.0001). Two patients with a reduced LV T2* also had a reduced LV T1, while only one had a reduced LV T2. Three patients had a reduced LV T1 but a normal LV T2*; 66.7% of the patients had an increased LV T2. All LV relaxation times were significantly correlated with pancreas T2* values (T2*: P = 0.033; T1: P < 0.0001; T2: P = 0.014). No LV relaxation time was associated (P > 0.05) with hepatic iron concentration, biventricular function parameters, or LGE presence.

Conclusion: The combined use of all three myocardial relaxation times has potential to improve sensitivity in the detection of early/subclinical myocardial involvement in β-Tl patients.

Level of evidence: 2 TECHNICAL EFFICACY: Stage 2.