Introduction Transcatheter aortic valve implantation (TAVI) is a therapeutic option for high surgical risk patients with symptomatic aortic stenosis. While improvements in pulmonary artery systolic pressure (PASP) following TAVI have been reported, data regarding the effect of changes in PASP on patient functional status are limited.

Methods We performed a retrospective analysis of all consecutive TAVI procedures recorded on the UK TAVI registry from our institution between January 2007 and January 2015. Functional status was defined by the New York Heart Association (NYHA) classification and PASP assessed by transthoracic echocardiography. Repeat measures of PASP and NYHA were performed prior to TAVI and at 1-, 6- and 12-– months following TAVI. Linear mixed model for repeated measures analysis was used to detect changes in PASP and NYHA over the three time points following TAVI and to measure their association. Adjusting for traditional risk factors, we subsequently evaluated whether PASP and NYHA predicted 1-year mortality following TAVI.

Results Over eight years, 299 patients (48% male, age 84 ± 8 years) underwent TAVI. During a mean follow-up of 775 days, 114 patients died. One month following TAVI, there was a significant reduction in PASP (37.3 ± 1.7 to 32.4 ± 1.7 mm Hg, p < 0.0001), which remained at 6-months (33.7 ± 1.9 mm Hg, p = 0.03), yet at 12-months the pressure had returned to baseline levels (35 ± 2.2 mm Hg). There was a significant decrease in NYHA at 1-month following TAVI (2.3 ± 0.1 to 1.5 ± 0.1, p < 0.0001). The improvement in NYHA persisted both at 6-months (1.5 ± 0.1, p < 0.0001) and at 12-months (1.6 ± 0.1, p < 0.0001) following TAVI. In linear mixed model analysis, after adjusting for left ventricular ejection fraction (LVEF), we observed an association between changes in PASP and NYHA following TAVI (coefficient 0.030 ± 0.01, p < 0.0001). An improvement in NYHA by one functional class was therefore associated with a reduction in PASP by 42 mm Hg.

In univariate Cox regression analyses, PASP, NYHA and LVEF, at 1-month following TAVI, all predicted 1-year mortality (p < 0.05). In multivariate analysis, only NYHA at 1-month following TAVI was independently related to 1-year mortality (hazard ratio 1.80, 95% confidence interval 1.21 to 2.69, p = 0.004).

Conclusion These data provide evidence that the reduction in PASP observed following TAVI is closely correlated with an improvement in NYHA functional class along with mortality at 1-year. This study will help enable cardiovascular clinicians to identify those patients likely to have a favourable symptomatic response to TAVI based on the echocardiographic PASP estimate post-procedure.