Performance of a Simplified Dichotomous Phenotypic Classification of Bicuspid Aortic Valve to Predict Type of Valvulopathy and Combined Aortopathy.

2017 
Background A simplified classification of bicuspid aortic valve (BAV) morphology using only the orientation of fused cusps was recently proposed. The aim of this study was to test whether it is useful for showing an association with the type of valvulopathy or aortopathy. Methods BAV phenotype was retrospectively classified in 681 patients (mean age, 59 ± 12 years; 424 men) who underwent aortic valve surgery. Each BAV was classified using both dichotomous (right and left coronary cusp fusion [CCF] vs mixed cusp fusion [MCF]) and conventional methods, and its association with the dominant valvulopathy (aortic stenosis [AS] vs regurgitation) and concomitant aortic surgery was analyzed. Four cardiologists individually reviewed transthoracic echocardiographic images of 100 randomly selected patients to compare the feasibility and accuracy of the two classification methods. Results The frequencies of BAV CCF and MCF were 53% ( n  = 361) and 47% ( n  = 320), respectively. AS was the predominant cause of surgery ( n  = 546 [80%]), and concomitant aortic surgery was done in 31% ( n  = 214). Patients with BAV MCF showed a higher frequency of AS (89% vs 73%, P P P P  = .001). The feasibility of the classification methods did not differ, but dichotomous classification revealed higher accuracy than conventional (87% [95% CI, 84.1%–90.7%] vs 70% [95% CI, 65.0%–74.3%]) for all four examiners, with higher κ coefficients representing interrater agreement (κ = 0.73 ± 0.06 to 0.83 ± 0.06 [dichotomous method] vs 0.51 ± 0.06 to 0.73 ± 0.06 [conventional method]). Conclusions The dichotomous classification method is useful for showing the association with the type of valvulopathy or aortopathy, with better diagnostic performance than the conventional method.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    30
    References
    12
    Citations
    NaN
    KQI
    []