Relative utility of magnetic resonance imaging and right ventricular angiography to diagnose arrhythmogenic right ventricular cardiomyopathy

Abstract

Introduction: Arrhythmogenic right ventricular cardiomyopathy (ARVC) is characterized by fibrofatty replacement of the RV myocardium. Two imaging techniques used to assess patients suspected of having ARVC are magnetic resonance imaging (MRI) and right ventricular angiography (RVA). Traditionally, RVA has played a central role in the diagnosis of ARVC, but the non-invasive nature of MRI and its unique ability to detect fatty tissue infiltration has increased its popularity as a diagnostic tool. The objective of this study was to assess the relative diagnostic accuracy of MRI and RVA for ARVC.

Methods and results: Seventeen patients (9 men, 8 women; ages 42 +/- 17 [range 16-78] years) with documented ventricular arrhythmias were investigated for ARVC. A positive diagnosis of ARVC was based on criteria set forth by the ISFC Working Group on Cardiomyopathies and Dysplasia. ECG-gated spin-echo and gradient-echo MR images in multiple planes and RAO/LAO RV angiograms were compared for diagnostic concordance. Based on working group criteria, 7 patients were diagnosed with ARVC. In ten patients, MRI suggested ARVC. The remaining 7 patients had no MRI findings suggestive of the disease. Four patients with MRI findings of ARVC were incorrectly diagnosed based on Task Force criteria. Conversely, 1 patient with a normal MRI met Task Force criteria for the diagnosis of ARVC. Based on RV angiograms, 7 patients had findings suggestive of ARVC. The 10 patients without AVRD (based on RVA) also did not meet the necessary criteria for diagnosis of ARVC using Task Force standards. RVA was 100% specific and 100% sensitive compared to MRI that was only 86% sensitive and 60% specific. MRI proved to be most reliable when the images demonstrated gross, lipomatous infiltration, evidenced by a large area of hyperintensity. When the results of MRI and RVA were congruent, the diagnosis was always accurate.

Conclusion: RVA is more sensitive and specific to diagnose ARVC diagnosis than MRI, at least until MRI protocols are better developed. MRI results are most robust when indicators of ARVC are grossly apparent. False-positive diagnosis by MRI was primarily related to perceived motion abnormalities that were not seen by RVA. One of its greatest potential assets (fat detection) did not enhance diagnostic specificity.