Mitral Annulus Disjunction: A Comprehensive Cardiovascular Magnetic Resonance Phenotype and Clinical Outcomes Study.

Background: Clinical importance of mitral annulus disjunction (MAD) is not well established.

Purpose: Characterize a population of MAD all-comers diagnosed by cardiac magnetic resonance imaging (MRI).

Study type: Retrospective.

Population: MAD confirmed in 222 patients, age of 49.2 ± 19.3 years, 126 (56.8%) males.

Field strength/sequence: 1.5 T and 3 T/steady-state free precession and inversion recovery.

Assessment: Clinical history, outcomes, imaging, and arrhythmia data. MAD defined as a separation ≥2 mm between left ventricular myocardium and mitral annulus. Presence and pattern of late gadolinium enhancement (LGE) were analyzed. LGE in the papillary muscles and adjacent to MAD were identified as MAD related. Ventricular arrhythmias (VA) were grouped into non-sustained ventricular arrhythmias (NSVA) or sustained. Cardiovascular death assessed.

Statistical tests: Differences between baseline characteristics were compared. Univariate regression was used to investigate possible associations between ventricular arrhythmia and cardiovascular death with characteristics associated with the severity of MAD. A multivariable logistic regression included significant variables from the univariate analysis and was performed for MAD-related and global LGE.

Results: MAD extent 5.0 ± 2.6 mm. MV annulus expanded during systole for MAD ≥6 mm. Systolic expansion associated with prolapse, billowing, and curling. LGE present in 82 patients (36.9%). Twenty-three patients (10.4%) showed MAD-related LGE by three different observers. No association of LGE with MAD extent (P = 0.545) noted. Follow-up 4.1 ± 2.4 years. No sustained VA observed. In univariable analysis, NSVA was more prevalent in patients with MAD ≥6 mm (33.3% vs. 9.9%), but this was attenuated on multivariate analysis (P = 0.054). The presence of NSVA was associated with global LGE but not MAD-related LGE in isolation (P = 0.750). Three patients died of cardiovascular causes (1.4%) and none had MAD-related LGE. None died of sudden cardiac arrest.

Conclusion: In patients referred for cardiac MRI, mitral valve dysfunction was associated with MAD severity. Scar was not related to the extent of MAD, but associated with NSVA. The risk of sustained arrhythmias and cardiovascular death was low in this population.

Evidence level: 4 TECHNICAL EFFICACY: Stage 2.