Myocardial and hepatic iron overload assessment by region-based and pixel-wise T2* mapping analysis: technical pitfalls and clinical warnings.

Objective: The aim of this study was to compare myocardial T2* assessment with region-based (RB) T2* multiecho technique (CMRtools) with the pixel-wise (PW) inline myocardial T2* mapping (Siemens) in patients with thalassemia major for myocardial iron characterization.

Materials And Methods: Forty-three thalassemia major patients were examined on a 1.5-T scanner using conventional gradient multiecho sequence. All the images were analyzed using both RB and PW T2* mapping. Coefficients of reproducibility (CRs) were used to assess the interoperator and intraobserver variability of each software.

Results: The mean (SD) myocardial T2* values using RB and PW software resulted significantly different (30.7 [15] milliseconds [range, 4.8-52.6 milliseconds] vs 24.3 [10.5] milliseconds [range 4.6-38.2 milliseconds]; P < 0.0001). Interestingly, we found that SD had exponential relationship with T2* with evidence of increase in SD for T2* values greater than 20 milliseconds. For myocardial T2* values less than 20 milliseconds, intraobserver CR was 1.2 milliseconds for RB and 1.8 milliseconds for PW T2* mapping, and the interoperator CR was 3.4 and 1.6 milliseconds for RB and PW T2* mapping, respectively. Comparing iron overload classification by both software, we found that 7 patients (16%) were differently categorized using the standard T2* thresholds.

Conclusions: Our data show that RB and PW T2* mapping can be used interchangeably to measure severe myocardial and hepatic iron overload, whereas for borderline T2* values, we observed differences among the 2 methods causing different categorization.