Premium
Liver Iron Content Determination Using a Volumetric Breath‐Hold Gradient‐Echo Sequence With In‐Line R 2 * Calculation
Author(s) -
Wunderlich Arthur P.,
Schmidt Stefan A.,
Mauro Valeria,
Kneller Lena,
Kannengießer Stephan,
Beer Meinrad,
Cario Holger
Publication year - 2020
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.27185
Subject(s) - medicine , intraclass correlation , nuclear medicine , correlation , mathematics , geometry , clinical psychology , psychometrics
Background Liver iron overload is a serious condition occurring in patients requiring blood transfusions (eg, in thalassemia and different forms of anemia) or with dysfunctional iron resorption, since there is no physiological mechanism to excrete iron. Above a certain level of iron concentration, chelation therapy is indicated. To monitor therapy success, liver iron content should be assessed regularly. A noninvasive method is important for patient management. Existing MRI methods suffer from long acquisition times and cost. Purpose To study the correlation of liver iron content (LIC) reference values to liver R 2 * determined using a 3D breath‐hold multigradient echo (GRE) MRI sequence, employing accelerated acquisition by parallel imaging and in‐line R 2 * calculation. Study type Prospective. Population In all, 117 patients (22.1 ± 14.1 years, 66 men) suspected of iron overload. Sequence GRE. Field Strength 1.5T. Assessment For comparison, a regulatory‐approved method with a considerably longer scan time was used, providing LIC reference values. Participants were divided into a calibration group (65 participants), analyzed independently by two observers, and a validation group (52 participants). Statistical Tests Linear correlation parameters were evaluated for R 2 * values with LIC reference values, and for LIC determined from R 2 * for validation group participants with LIC reference values. Sensitivity/specificity for clinical relevant LIC thresholds were analyzed. Interobserver variability was determined by intraclass correlation coefficient (ICC). Results Interobserver agreement was excellent, with an ICC of 0.99, P < 0.001. Good correlation (R 2 = 0.89) and congruence of LIC values obtained with our method to LIC reference values was found, and almost identical diagnostic accuracy. Sensitivity/specificity were 0.98/0.67 for the diagnostic relevant LIC threshold of 4.5 mg/g and 1.0/0.95 for the threshold of 7 mg/g. Data Conclusion MRI acquisition times for determination of LIC can be significantly reduced by the use of comprehensive in‐line R 2 * map generation without compromising diagnostic accuracy. Level of Evidence 1 Technical Efficacy Stage 2