Article Text
PDF
Abstract
Introduction We sought to identify echo parameters that could predict outcomes in patients with severe aortic stenosis (AS) and normal left ventricular (LV) systolic function.
Methods We identified patients with an echo result consistent with severe AS (aortic valve area (AVA) <1 cm2 or aortic Vmax >4 m/s) and normal LV systolic function (LVEF >55% or visually estimated as normal). The outcome was a composite of any valve intervention and death. Clinical records were reviewed to record demographics and co-morbidities. Principal component analysis (PCA) was used to identify echo parameters that could potentially identify subpopulations in the cohort. Multivariate Cox regression analysis was used to determine if these parameters independently predicted the outcome in a model that included traditional echo predictors (AVA, AV Vmax, markers of left ventricular hypertrophy) and a second model that also included co-morbidities.
Results We identified 191 patients with the characteristics displayed in table 1.
Abstract 72 Table 1 Continuous variables displayed as mean (standard deviation) and categorical variables represented as percentages. COPD- chronic obstructive pulmonary disease; IVSd- interventricular septum thickness in diastole; LVIDd- left ventricular internal dimension in diastole; TAPSE- tricuspid annular peak systolic excursion; GFR- glomerular filtration rate.
During follow up 3 patients underwent balloon valvuloplasty, 30 patients underwent aortic valve replacement, 84 underwent transcatheter aortic valve implantation and 15 died. The key echo variables derived from PCA were AV mean gradient and left atrial area (LAA). In a multivariate Cox regression analysis of traditional echo parameters (AVA, AV Vmax, and IVSd), along with the two selected by PCA, two parameters were associated with the outcome: AVA (HR 0.084 [0.014–0.511], p=0.007) and LAA (HR 1.070 [1.027–1.115], p=0.001). When these two echo parameters were combined in another multivariate Cox regression with the co-morbidities listed in Table 1, they both remained significant predictors of outcome; AVA (HR 0.088 [95% CI 0.013–0.594], p=0.013) and LAA (HR 1.058, [95% CI 1.011–1.107] p=0.016).
Conclusions Enlarged LAA is an important predictor of outcome in patients with aortic stenosis but one that does not receive a lot of attention. Physiologically it is likely to reflect the impact of aortic stenosis on left ventricular filling pressures and may possibly occur prior to a noticeable reduction in LV ejection fraction.