Abstract 555 Underestimation of right-ventricular function: right-ventricular to pulmonary artery coupling predicts outcomes in TAVI patients independent of flow type

Background

Right-ventricular dysfunction in patients undergoing transcather aortic valve implantation (TAVI) for aortic stenosis (AS) has long been disregarded. Right-ventricular to pulmonary artery coupling (RV/PAc) is a parameter of right-ventricular function that can easily be obtained in echocardiography. Its use for mortality prediction in AS patients warrants further evaluation, data for different flow types of AS are lacking.

Methods

All patients undergoing TAVI for AS at our centre between 2018 and 2020 were assessed. RV/PAc was defined as the ratio of tricuspid annular plane systolic excursion (TAPSE) and estimated systolic pulmonary artery pressure (sPAP). Patients with no tricuspid regurgitation were excluded since right-ventricular/right-atrial pressure gradients necessary for sPAP estimation cannot be obtained in echocardiography.

A receiver operator curve (ROC) analysis was performed to dichotomize the cohort into groups: physiological RV/PAc and pathological RV/PAc.

Primary endpoint was 2-year mortality. Subanalyses were performed for 4 different flow types of AS: high gradient (HG, dPmean ≥40 mmHg), classical low-flow low-gradient (classical LFLG, dPmean <40 mmHg and left-ventricular ejection fraction (LVEF) <50%), paradoxical low-flow low-gradient (paradoxical LFLG, dPmean <40 mmHg, LVEF ≥50%, stroke volume index, SVi, ≤35 ml/m²), and normal-flow low-gradient (NFLG, dPmean <40 mmHg, LVEF ≥50%, SVi >35 ml/m²).

Results

A total of 862 patients were included in the analysis. The cut-off for pathological RV/PAc derived from ROC analysis was 0.512 mm/mmHg, splitting the cohort into 429 patients with physiological and 433 patients with pathological RV/PAc. Median RV/PAc was 0.783 [0.512-3.100] mm/mmHg in the physiological and 0.348 [0.103-0.512] mm/mmHg in the pathological group respectively.

Patients in the pathological group were more often male (54.0% vs. 47.1%, p=0.04), were numerically older (83.0 [79-86.9] vs. 81.0 [77.3-85.4] years, p=0.09), and had higher rates of atrial fibrillation, chronic kidney disease and coronary artery disease. They also had lower LVEF (50 [40-55] % vs. 55 [53-58] %, p<0.01) and higher rates of relevant aortic, mitral and tricuspid regurgitation.

Estimated 2-year mortality was significantly higher in the pathological group (37.3 [32.4-41.8] % vs. 16.5 [12.8-20.0] %, p<0.01, Figure, panel A). Differences in long-term mortality were not driven by differences in procedural outcomes as underlined by the two composite VARC-3 endpoints technical success (p=0.62) and device success (p=0.45). Results were consistent in a multivariate model adjusted for baseline differences.

Median RV/PAc values differed significantly between flow type groups: HG 0.64, classical LFLG 0.42, paradoxical LFLG 0.52, and NFLG 0.61 mmHg, p<0.01. In mortality subanalyses, pathological RV/PAc determined 2-year mortality among HG patients, paradoxical LFLG patients and classical LFLG patients (Figure, panel B).

Conclusions

Right-ventricular dysfunction exemplified by low RV/PAc is a strong predictor for long-term mortality in TAVI patients. In the subanalysis of flow types, classical LFLG and paradoxical LFLG patients appear to have worse RV/PAc than HG patients. RV/PAc retains its strong predictive value in these subgroups of AS.

Clin Res Cardiol (2023). https://doi.org/10.1007/s00392-023-02180-w
Underestimation of right-ventricular function: right-ventricular to pulmonary artery coupling predicts outcomes in TAVI patients independent of flow type
J. Steffen¹, M. Lux¹, T. Stocker¹, N. Kneidinger², M. Orban³, K. Löw¹, P. Doldi¹, M. Haum¹, J. Fischer¹, L. Stolz¹, H. D. Theiss¹, K. Rizas¹, D. Braun¹, M. Orban¹, S. Peterß⁴, J. Hausleiter¹, S. Massberg¹, S. Deseive¹
¹Medizinische Klinik und Poliklinik I, LMU Klinikum der Universität München, München; ²Medizinische Klinik und Poliklinik V, LMU Klinikum der Universität München, München; ³LH151, Munich Re, München; ⁴Herzchirurgische Klinik und Poliklinik, LMU Klinikum der Universität München, München;
Background Right-ventricular dysfunction in patients undergoing transcather aortic valve implantation (TAVI) for aortic stenosis (AS) has long been disregarded. Right-ventricular to pulmonary artery coupling (RV/PAc) is a parameter of right-ventricular function that can easily be obtained in echocardiography. Its use for mortality prediction in AS patients warrants further evaluation, data for different flow types of AS are lacking. Methods All patients undergoing TAVI for AS at our centre between 2018 and 2020 were assessed. RV/PAc was defined as the ratio of tricuspid annular plane systolic excursion (TAPSE) and estimated systolic pulmonary artery pressure (sPAP). Patients with no tricuspid regurgitation were excluded since right-ventricular/right-atrial pressure gradients necessary for sPAP estimation cannot be obtained in echocardiography. A receiver operator curve (ROC) analysis was performed to dichotomize the cohort into groups: physiological RV/PAc and pathological RV/PAc. Primary endpoint was 2-year mortality. Subanalyses were performed for 4 different flow types of AS: high gradient (HG, dPmean ≥40 mmHg), classical low-flow low-gradient (classical LFLG, dPmean <40 mmHg and left-ventricular ejection fraction (LVEF) <50%), paradoxical low-flow low-gradient (paradoxical LFLG, dPmean <40 mmHg, LVEF ≥50%, stroke volume index, SVi, ≤35 ml/m²), and normal-flow low-gradient (NFLG, dPmean <40 mmHg, LVEF ≥50%, SVi >35 ml/m²). Results A total of 862 patients were included in the analysis. The cut-off for pathological RV/PAc derived from ROC analysis was 0.512 mm/mmHg, splitting the cohort into 429 patients with physiological and 433 patients with pathological RV/PAc. Median RV/PAc was 0.783 [0.512-3.100] mm/mmHg in the physiological and 0.348 [0.103-0.512] mm/mmHg in the pathological group respectively. Patients in the pathological group were more often male (54.0% vs. 47.1%, p=0.04), were numerically older (83.0 [79-86.9] vs. 81.0 [77.3-85.4] years, p=0.09), and had higher rates of atrial fibrillation, chronic kidney disease and coronary artery disease. They also had lower LVEF (50 [40-55] % vs. 55 [53-58] %, p<0.01) and higher rates of relevant aortic, mitral and tricuspid regurgitation. Estimated 2-year mortality was significantly higher in the pathological group (37.3 [32.4-41.8] % vs. 16.5 [12.8-20.0] %, p<0.01, Figure, panel A). Differences in long-term mortality were not driven by differences in procedural outcomes as underlined by the two composite VARC-3 endpoints technical success (p=0.62) and device success (p=0.45). Results were consistent in a multivariate model adjusted for baseline differences. Median RV/PAc values differed significantly between flow type groups: HG 0.64, classical LFLG 0.42, paradoxical LFLG 0.52, and NFLG 0.61 mmHg, p<0.01. In mortality subanalyses, pathological RV/PAc determined 2-year mortality among HG patients, paradoxical LFLG patients and classical LFLG patients (Figure, panel B). Conclusions Right-ventricular dysfunction exemplified by low RV/PAc is a strong predictor for long-term mortality in TAVI patients. In the subanalysis of flow types, classical LFLG and paradoxical LFLG patients appear to have worse RV/PAc than HG patients. RV/PAc retains its strong predictive value in these subgroups of AS.
https://dgk.org/kongress_programme/jt2023/aV1259.html