Contrast-free detection of myocardial fibrosis in hypertrophic cardiomyopathy patients with diffusion-weighted cardiovascular magnetic resonance

Christopher Nguyen, Minjie Lu, Zhaoyang Fan, Xiaoming Bi, Peter Kellman, Shihua Zhao, Debiao Li

Previous studies have shown that diffusion-weighted cardiovascular magnetic resonance (DW-CMR) is highly sensitive to replacement fibrosis of chronic myocardial infarction. Despite this sensitivity to myocardial infarction, DW-CMR has not been established as a method to detect diffuse myocardial fibrosis. We propose the application of a recently developed DW-CMR technique to detect diffuse myocardial fibrosis in hypertrophic cardiomyopathy (HCM) patients and compare its performance with established CMR techniques. HCM patients (N = 23) were recruited and scanned with the following protocol: standard morphological localizers, DW-CMR, extracellular volume (ECV) CMR, and late gadolinium enhanced (LGE) imaging for reference. Apparent diffusion coefficient (ADC) and ECV maps were segmented into 6 American Heart Association (AHA) segments. Positive regions for myocardial fibrosis were defined as: ADC > 2.0 μm(2)/ms and ECV > 30 %. Fibrotic and non-fibrotic mean ADC and ECV values were compared as well as ADC-derived and ECV-derived fibrosis burden. In addition, fibrosis regional detection was compared between ADC and ECV calculating sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) using ECV as the gold-standard reference. ADC (2.4 ± 0.2 μm(2)/ms) of fibrotic regions (ADC > 2.0 μm(2)/ms) was significantly (p < 0.01) higher than ADC (1.5 ± 0.2 μm(2)/ms) of non-fibrotic regions. Similarly, ECV (35 ± 4 %) of fibrotic regions (ECV > 30 %) was significantly (p < 0.01) higher than ECV (26 ± 2 %) of non-fibrotic regions. In fibrotic regions defined by ECV, ADC (2.2 ± 0.3 μm(2)/ms) was again significantly (p < 0.05) higher than ADC (1.6 ± 0.3 μm(2)/ms) of non-fibrotic regions. In fibrotic regions defined by ADC criterion, ECV (34 ± 5 %) was significantly (p < 0.01) higher than ECV (28 ± 3 %) in non-fibrotic regions. ADC-derived and ECV-derived fibrosis burdens were in substantial agreement (intra-class correlation = 0.83). Regional detection between ADC and ECV of diffuse fibrosis yielded substantial agreement (κ = 0.66) with high sensitivity, specificity, PPV, NPV, and accuracy (0.80, 0.85, 0.81, 0.85, and 0.83, respectively). DW-CMR is sensitive to diffuse myocardial fibrosis and is capable of characterizing the extent of fibrosis in HCM patients.