Stroke in mitral valve prolapse: risk factors and left atrial function in cryptogenic versus non-cryptogenic ischemic subtypes

BACKGROUND AND PURPOSE: Mitral valve prolapse (MVP) has been associated with an increased risk of ischemic stroke. Older age, thicker mitral leaflets, and significant mitral regurgitation (MR) leading to atrial fibrillation have been traditionally considered risk factors for ischemic stroke in MVP. However, specific risk factors for MVP-stroke subtypes are not well defined. The aim of this study is to evaluate clinical and echocardiographic parameters, including left atrial (LA) function, in MVP with cryptogenic (C) vs. non-cryptogenic (NC) stroke. METHODS: In this case-control matched study, MVPs were identified in consecutive echocardiograms obtained after a stroke from January 2013 to December2016 at the University of California, San Francisco. MVP was defined as leaflet displacement ≥2 mm in the parasternal long-axis view at end-systole. Age/gender matched MVPs without stroke and healthy controls without MVP were also identified. We analyzed LA end-systolic/diastolic volume index, emptying fraction (LAEF), function index (LAFI), and global longitudinal strain in all MVPs and controls. We also measured left ventricular (LV) volume indexes, mass index, ejection fraction (EF), degree of MR and leaflet thickness. RESULTS: We identified a total of 30 MVPs (age 70 ± 12, 50% females) with stroke (11 with C- and 19 with NC-stroke), 20 age/gender matched MVPs without a stroke and 16 controls. MVPs without stroke had lower BMI, less hypertension but more MR (≥moderate in 45% vs. 17%), more abnormal LA function (lower LAEF, LAFI) and larger LV volumes/mass (all p < 0.05) when compared to MVPs with stroke. Leaflet thickness was overall mild (<3 mm) and similar in the 2 groups. Within the MVP stroke group, NC-stroke had higher BMI, more hypertension and more atrial fibrillation compared to C-stroke. In the variables tested, patients with C-stroke did not differ from controls. CONCLUSIONS: MVP-related MR may be protective against stroke despite abnormal LA function. Risk of NC-stroke in MVP is related to common stroke risk factors rather than mitral valve leaflet thickness. The etiology of C-stroke in MVP warrants further studies.