To Each Imaging Modality, Their Own MAD

Abstract

Purpose

The clinical significance of mitral annular disjunction (MAD) is uncertain. Imaging modality might impact the prevalence of MAD. We aimed to assess MAD prevalence at transthoracic echocardiography (TTE) and cardiac magnetic resonance (CMR) as well as their inter-modality agreement.

Methods

This observational retrospective study included patients undergoing TTE and CMR within 6 months. MAD was defined as ≥1 mm systolic separation between the left atrial wall-mitral leaflet and the left ventricular (LV) wall. The maximum MAD longitudinal extent was measured. The inter-modality agreement for MAD diagnosis was evaluated.

Results

One hundred twenty four patients (59 ± 17 years; 62% male) were included. MAD was detected in 60 (48%) using CMR and in 10 (8%) using TTE. All patients with MAD on TTE had MAD on CMR. The inter-modality agreement was low (Cohen's kappa = 0.17) but improved when the diagnostic cut-off was increased from 1 to 5 mm (Cohen's kappa = 0.66). The median longitudinal length of MAD was 2.0 mm (25th–75th percentiles: 1.5–3.0) by CMR and 4.0 mm (25th–75th percentiles: 2.7–4.5) by TTE with moderate agreement (intraclass correlation coefficient = 0.66).

Conclusion

MAD of limited extent is common on CMR and more than two thirds of patients showing MAD on CMR did not have MAD on TTE. The inter-modality agreement between TTE and CMR increased when the diagnostic threshold for MAD was increased from 1 to 5 mm. Methodological discrepancies impact MAD assessment and contribute to the discordant prevalence and clinical significance reported in the literature.