Quantitative analysis of pulmonary congestion using computed tomography in patients with severe aortic stenosis and its long-term prognostic impact after TAVR

Background and aims:
Pulmonary congestion (PG) is common in symptomatic severe aortic stenosis (AS). Extent of PG can be analyzed by quantitative computed tomography (QCT). The study aims to investigate PG by QCT in patients with severe AS undergoing transcatheter aortic valve replacement (TAVR) and to analyze its impact on long-term outcomes.

Methods:
Mean lung density (MLD), extravascular lung water volume (EVLWV) and ground glass opacification (GGO) as QCT measures of PG were analyzed in 144 patients with AS undergoing TAVR at baseline. Long-term follow-up (median follow-up time 58 months) was conducted to assess all-cause mortality and a composite endpoint of mortality and/or hospitalisation for heart failure. MLD, EVLWV and GGO were dichotomized by median values as the binary cut-off.

Results:
Patients with MLD > median were significantly older [83 (IQR 80-86) vs. 80 (IQR 77-84), P < 0.05], had higher levels of NT-pro-BNP [2963.0 pg/ml (IQR 1249.5-7296.0) vs. 1302.0 pg/ml (IQR 571.0-2233.0), P < 0.001] and NYHA stages ≥ III were more prevalent (80.8% vs. 66.2%, p < 0.05) compared to patients with MLD < median. 69 patients died during follow-up and 82 reached the combined endpoint. According to Kaplan-Meier analysis, MLD, EVLWV and GGO were significant predictors for mortality and combined endpoint (each P < 0.05). In multivariate cox regression analysis, MLD was an independent predictor for mortality (OR 1.76, P < 0.05) and combined endpoint (OR 1.70, P < 0.05), whereas GGO was only for the combined endpoint (OR 1.78, P < 0.05). EVLWV was no significant predictor for both endpoints (each P = n.s.).

Conclusion:
PG assessed by QCT can predict adverse long-term outcome in patients with AS undergoing TAVR. Especially MLD seems to incorporate a relevant prognostic capacity and should be evaluated further.