BACKGROUND:
Elevated aortic valve gradients are common after transcatheter aortic valve implantation for degenerated surgical aortic valve replacement bioprostheses, but their clinical impact is uncertain.
METHODS:
A total of 12 122 patients who underwent transcatheter aortic valve implantation-in-surgical aortic valve replacement from November 2011 to December 2019 in the Society of Thoracic Surgery/American College of Cardiology Transvalvular Therapeutics Registry were included. The primary outcome was a composite of 1-year all-cause mortality, stroke, myocardial infarction, or valve reintervention. Secondary outcomes included 1-year all-cause mortality, readmission, and change from baseline 12-question self-administered Kansas City Cardiomyopathy Questionnaire-Overall Summary Score. Due to nonlinearity observed with restricted cubic splines analysis, a Cox regression analysis with aortic valve mean gradient modeled as a spline-continuous variable (with 20 mm Hg as a cutoff) was used to study the 1-year composite outcome and mortality.
RESULTS:
The composite outcome occurred most frequently in patients with aortic valve mean gradient ≥30 and <10 mm Hg, as compared with those with 10 to 20 and 20 to 30 mm Hg ranges (unadjusted rates, 13.9%, 12.1%, 7.5%, and 6.5%, respectively;
P
=0.002). When the mean aortic valve gradient was ≥20 mm Hg, higher gradients were associated with greater risk of the 1-year composite outcome (adjusted hazard ratio, 1.02 [1.02–1.03] per mm Hg;
P
<0.001) and 1-year mortality (adjusted hazard ratio, 1.02 [1.00–1.03] per mm Hg;
P
=0.007). Whereas when the mean aortic valve gradient was <20 mm Hg, higher gradients were not significantly associated with the composite outcome (adjusted hazard ratio, 0.99 [0.98–1.003] per mm Hg;
P
=0.12) but were associated with lower 1-year mortality (adjusted hazard ratio, 0.98 [0.97–0.99] per mm Hg;
P
=0.007).
CONCLUSIONS:
The relationship between postprocedural aortic valve mean gradient after transcatheter aortic valve implantation-in-surgical aortic valve replacement and clinical outcomes is complex and nonlinear, with relatively greater adverse events occurring at low and high gradient extremes. Further study of factors mediating the relationship between postprocedural gradients and clinical outcomes, including low-flow states, is necessary.