• Cardiac CT Angiography for the Diagnosis of Mitral Valve Prolapse: Comparison with Echocardiography

    Radiology: Volume 254: Number 2—February 2010

    Gudrun M. Feuchtner, MD Hatem Alkadhi, MD Christoph Karlo, MD Ammar Sarwar, MD Andreas Meier, MD Wolfgang Dichtl, MD, PhD Sebastian Leschka, MD Ron Blankstein, MD Juerg Gruenenfelder, MD Paul Stolzmann, MD Ricardo C. Cury, MD

    Purpose: To evaluate the diagnostic performance of coronary com¬puted tomographic (CT) angiography for the diagnosis of mitral valve prolapse (MVP).

    Materials and Methods: The retrospective case-controlled multicenter study proto­col was approved by the institutional review boards. The U.S. part of the study was HIPAA compliant. One hun­dred twelve patients who underwent electrocardiographi-cally gated 64-section coronary CT angiography (n = 60) or dual-source coronary CT angiography (n = 52) and trans­thoracic echocardiography (TTE) were included. Fifty-three patients with MVP were matched for age and sex with 59 patients without MVP. CT images were analyzed on three-, two-, and four-chamber (CH) views by two independent observers. MVP was defined as a greater than 2-mm dis­placement of leaflets below the annulus plane and was sub-classified as "billowing" (bowing) or "flail leaflet" (free leaflet margin displacement). Leaflet thickness was measured and defined as thickened if it was greater than 2 mm.

    Results: The diagnostic performance of CT when three- and two-CH views were combined for the diagnosis of MVP was as follows: sensitivity, 96%; specificity, 93%; posi­tive predictive value (PPV), 93%; and negative predictive value, 96%. On four-CH views, the excursion of billow­ing was higher than it was on three-CH views (P <.001), and the PPV of the four-CH view for diagnosis of MVP was 89%. The correlation between CT and TTE for ex­cursion of billowing was high (r = 0.80-0.91). In a sub­set of 32 patients, the agreement between CT and TTE for differentiation of billowing (n = 13) and flail leaflet (n = 2) was 100%. Leaflet thickening was more prevalent in patients with MVP than it was in those without (71% vs 20%, P <.001), and correlation with TTE was good (r = 0.81 [anterior leaflet] and 0.77 [posterior leaflet]).

    Conclusion: The combined use of three- and two-CH views allows an accurate diagnosis of MVP at coronary CT angiography.