Cardiac T2* Magnetic Resonance Imaging (MRI) in Iron Overload and Ventricular Diastolic Function.

Accumulation of iron in the heart leads to myocardial dysfunction and in severe cases heart failure. Myocardial hemosiderosis occurs in many hematological diseases treated with blood transfusions, including sickle cell anemia (SCA) and beta thalassemia (β-thal). MRI T2* can quantify myocardial iron non-invasively and can be compared to functional studies (e.g., echocardiography). Multiple MRI readers can introduce variation in the measurement. We investigated the relationship of myocardial MRI T2* with echo measurements of both systolic and diastolic function, and the variability introduced by different MRI reviewers. Patients were selected if they received ≥ 18 transfusions, or had a serum ferritin ≥ 1000 ng/mL. Study participants underwent 1.5 Tesla MRI T2* and echocardiography testing within 30 days. Echo measurements included left ventricular ejection fraction (LVEF), mitral annular tissue Doppler (TDV) e’ velocity (a measure of LV diastolic function), left ventricular myocardial performance index (a measure of both systolic and diastolic function - LVMPI), and left ventricular mass index (LVMI). Regions of interest (ROI) were drawn in a homogeneous area of the intraventricular septum. Three independent reviewers, blinded to the patients’ clinical status and the other 2 reviewers’ results, performed the ROI analysis. Agreement among the 3 raters was assessed using interclass correlation coefficient (ICC). Echo findings were associated with heart T2* both as continuous and categorical (normal vs abnormal) variables. Forty-seven patients, median age 14 years (range 7 – 37), participated; 24 (51.1%) were female. Thirty-five (74.5%) had SCA, 8 (17%) had β-thal (major or intermedia), and 4 (8.5%) had bone marrow failure syndromes. Mean (±1SD) echo results were LVEF (±5.8) %, LVMI 86 (±20.4) g/m2, LVMPI 0.3 (±0.1), and TDV e’ velocity 16.2 (±3.4) cm/sec. Only one patient had low LVEF, but 8 (17%) had high LVMPI and 20 (42%) had low TDV e’ velocity. All patients had normal LVMI. The mean myocardium T2* ranged among the 3 reviewers from 27 to 33 (±6.9 to 8.7) msec. Agreement among the 3 raters had an ICC = 0.64. Median T2* values were significantly lower among patients with lower LVMI (32 versus 42 msec, p= 0.019). No other comparisons among echo and T2* results were significant. Median echo findings were compared with T2* values above and below 20 msec: lower LVMI (71 vs 89 g/m2, p=0.04), and higher LVMPI (0.48 versus 0.28, p=0.015) were found among patients with T2* < 20 msec. Mean serum ferritin was 2917 (±2239) ng/mL. Serum ferritin was significantly negatively associated with T2* values (ρ=−0.33, p=0.031). We conclude:

1. Most patients had normal LV systolic function, but 42% had signs of diastolic dysfunction,

2. LVMPI was associated with myocardial iron loading (T2*>20msec), suggesting diastolic dysfunction to be an early sign of myocardial dysfunction in cardiac hemosiderosis,

3. Heart T2* was weakly associated with serum ferritin,

4. The agreement among reviewers of the heart T2* technique was acceptable, but not high, likely reflecting technical difficulties of accurate myocardial ROI determination on “bright blood” T2* maps drawn on thin intraventricular septa.

Dark blood techniques should therefore be investigated to reduce blood artifacts, potentially improving accuracy by eliminating artifacts from bright blood signals and thus reducing inter-rater variability in the ROI analysis of myocardial iron.