Defining Clinical Heart Failure in Patients with Severe Aortic Stenosis and Low Surgical Risk undergoing Transcatheter Aortic Valve Replacement

Background: Objective assessment of clinical heart failure (HF) in patients with severe aortic stenosis (AS) is often challenging due to comorbidity and reduced mobility. New York Heart Association (NYHA) functional classification poorly correlates with cardiac dysfunction and outcomes in severe AS patients and is often limited by subjectivity/inconsistency. Recent data demonstrate a steep increase in the frequency of HF hospitalizations (HFHs) in the 6 months preceding transcatheter aortic valve replacement (TAVR). Here, we aim to compare the prognostic value of NYHA functional class vs a more objective parameter of pre-TAVR HFH in a cohort of low-risk TAVR patients.

Methods: We evaluated long-term survival in consecutive patients with severe AS and a Society of Thoracic Surgeons (STS) score <4% (mean 2.7±0.7) who underwent transfemoral TAVI from 2009 to 2016. The cohort was classified by NYHA function or, alternatively, by the presence/absence of at least one episode of HFH in the 6 months before TAVR. Procedural complications (i.e. bleeding, vascular complications, stroke, pacemaker, paravalvular leak or 30-day mortality) were determined by the Valve Academic Research Consortium (VARC-2). The prognostic effect of each classification on long-term survival was studied by a Cox regression model.

Results: Out of 853 TAVR patients, we included 536 patients (mean age 80±5 years) with a mean STS score of 2.7±0.7%. Classification of the cohort by NYHA functional class (II=11.0%, III=68.2%, IV=22.2%) was directly associated with peri-procedural major bleeding (II=3.4%, III=5.3%, IV=7.6%, p=0.024). Long-term survival (median follow-up 4.2±1.9 years) did not differ between NYHA functional classes (log-rank p=0.35). Classification of patients based on the presence/absence of HFH 6 months before TAVR showed no difference in none of the procedural complications. However, long-term survival was significantly lower in patients with at least one HFH within the 6 months before TAVR (log-rank p<0.001). On multivariate survival analysis, pre-TAVR HFH, but not NYHA classification, was found to be an independent predictor of long-term mortality (hazard ratio 1.87, 1.28-2.73, p=0.001). Others independent predictors of mortality were atrial fibrillation and chronic obstructive pulmonary disease.

Conclusions: In a group of low-risk AS patients undergoing TAVR, HFH 6 months before TAVR, but not NYHA functional class, independently predicted long-term mortality. Early intervention prior to objective symptomatic progression requiring HFH, might improve survival after TAVR.