Tissue Characterization
Shahrad Shadman, MD
Clinical Fellow/ Cardiologist
National Heart, Lung, and Blood Institute, National Institutes of Health, United States
Shahrad Shadman, MD
Clinical Fellow/ Cardiologist
National Heart, Lung, and Blood Institute, National Institutes of Health, United States
Joao G. Ramos, MD, PhD
Research Fellow
National Heart, Lung, and Blood Institute, National Institutes of Health, United States
Nathan M. Kattapuram
Special Volunteer
National Heart, Lung, and Blood Institute, National Institutes of Health
Bethesda, Maryland, United States
Dong-Yun Kim, PhD
Statistician
National Heart, Lung, and Blood Institute, National Institutes of Health, United States
Eric E. Morgan, MD, PhD
Radiologist/Clinical fellow
National Heart, Lung, and Blood Institute, National Institutes of Health
Bethesda, Maryland, United States
Charles W. Benton, RT
MRI Technologist
National Heart, Lung, and Blood Institute, National Institutes of Health
North Bethesda, Maryland, United States
Gaby Weissman, MD
Section Chief, Clinical Cardiology
Medstar Heart and Vascular Institute
Washington, District of Columbia, United States
Ana Barac, MD, PhD
Staff Clinician/Cardiologist
National Heart, Lung, and Blood Institute, National Institutes of Health, District of Columbia, United States
Marcus Carlsson, MD, PhD
Professor, Head of Department
Karolinska Institute, Clinical Physiology, United States
Bright-blood late gadolinium enhancement (LGE) is considered the reference standard for the non-invasive detection of myocardial scar. However, detection of subendocardial and papillary muscle scar is improved by using dark blood LGE.1 This study aims to compare scar and fibrosis detectability using dark-blood technique compared to bright-blood in a heterogenous cohort of patients with ischemic and nonischemic cardiomyopathy.
Methods:
This prospective study included 95 subjects who underwent cardiac MRI at 1.5T (MAGNETOM Sola, Siemens Healthcare, Germany) and all had bright-blood and dark-blood LGE acquired in random order to avoid systematic bias. Dark-blood was acquired by nulling blood to a shorter inversion time(TI). Left ventricle (LV) short axis stacks were acquired using both sequences. Apical segments were excluded from the analysis. LGE image stacks were reviewed by an experienced reader ( >20 years of experience) in a blinded fashion, and scored per segment for presence and characteristics of the LGE findings (ischemic vs. nonischemic). Furthermore, we used the previously validated Expectation Maximization with Weighting (EWA)2 semiautomated algorithm in Segment (Medviso AB) to quantify the amount of LGE detected by each sequence.
Results:
We analyzed 3008 segments from 94 patients, one patient was excluded due to poor image quality. Dark blood technique detected 223/1504 (15%) LGE-positive segments, compared to 162/1504 (10%) segments detected by bright-blood technique (p=.034). There was moderate agreement between bright-blood and dark-blood for scar tissue characterization (Cohen’s kappa 0.62), with enhanced detectability of ischemic and nonischemic segments using the dark-blood technique. Dark blood identified 29 (31%) patients with identifiable LGE in papillary muscles, compared to 6 (6%) patients using the bright-blood technique, which corresponded to a low agreement between the two sequences (Cohen’s kappa 0.26). Furthermore, Bland-Altman analysis showed somewhat larger extent of scar/fibrosis using the dark-blood technique compared to BB (1.2 + 2.9 ml).
Conclusion:
We showed that dark-blood LGE increased detection of scar/fibrosis not only in ischemic but also nonischemic patients, as shown by visual scoring and semiautomated quantification. Moreover, dark blood LGE notably improved detection of papillary muscle fibrosis. These findings support further use of dark-blood LGE for detection of scar burden in patients with cardiomyopathies and thus may be of potential value in patient management.