Elevated systolic pulmonary artery pressure is a substantial predictor of increased mortality after transcatheter aortic valve replacement in males, not in females

The study population included 303 patients with indication of TAVR at Paracelsus Medical University Hospital, Salzburg and Kepler University Hospital, Linz in the period from 2016 to 2018. Indication for TAVR was made by the interdisciplinary heart team consisting of experienced cardiac surgeons and interventional cardiologists. Inclusion of patients was consecutive and retrospective. Patients with acute cardiac decompensation at the time of transthoracic echocardiography (TTE) or at the time of TAVR, with bicuspid valve or complex congenital heart diseases or, with a history, that might indicate a pre-capillary component of PH (chronic obstructive pulmonary disease (COPD) GOLD 4, idiopathic pulmonary arterial hypertension, chronic thromboembolic PH, interstitial lung disease, or underlying rheumatologic diseases with pulmonary involvement such as scleroderma, lupus erythematosus, etc.) were excluded from the study in advance.

The study protocol was approved by the local ethics committees of Paracelsus Medical University Salzburg (415-E/1969/5–2016) and Johannes Kepler University Linz (E-41–16) and conducted in accordance to principles of the Declaration of Helsinki and Good Clinical Practice. Written informed consent was available from all patients before study inclusion.

TTE was performed routinely, on average 1–4 weeks before TAVR using either an iE33 or Epiq 5 (Philips Healthcare, Hamburg, Germany) ultrasound device. Examinations were conducted by experienced clinicians with more than 4 years of training in echocardiography. Severe AS was classified according to current guidelines of the European Society for Cardiology (ESC). Left ventricular ejection fraction (LVEF) was calculated using Simpson’s method. To graduate mitral, aortic and tricuspid valve regurgitation in minimal, mild (I), moderate (II) and severe (III) spectral and color-Doppler images were used. TRVmax was obtained by continuous wave Doppler over the tricuspid valve. Pulmonary artery pressure (PAP), right atrial pressure (RAP) and at least sPAP was calculated as described previously [6]. As part of an extensive literature search and a self-authored review on the topic of non-invasive ways of determining PH in severe AS, the most commonly used sPAP cut-off values of 40 and 50 mmHg were used [1, 7, 8]. To also assess the severity of echocardiographically determined PH, patients were further subdivided into no PH by sPAP < 35 mmHg, mild PH by sPAP 35–50 mmHg, moderate PH by sPAP 51–70 mmHg and severe PH by sPAP > 70 mmHg [9]. This classification is based on the recommendations of the American Society for Echocardiography [10, 11].

All 303 patients in this study underwent TAVR using transfemoral access with second (CoreValve™ Evolut™ R; Medtronic Inc., Minneapolis, MN, USA) or third (CoreValve™ Evolut™ Pro; Medtronic Inc., Minneapolis, MN, USA) generation devices. The procedure was performed as previously described [12].

Statistical analysis and graphical representation was performed using SPSS (Version 25.0, SPSSS Inc., USA). Kolmogorov–Smirnov–Lilliefors test was carried out to test variables for normal distribution. Normally distributed metric data were expressed as mean ± standard deviation (SD) and analysed using an unpaired student’s t test. Not-normally distributed metric data were expressed as median and interquartile range (IQR); Mann–Whitney U test was applied for statistical analysis here. Frequencies/percentages were used for categorical data and compared using the chi-squared test. First of all, a Kaplan–Meier curve with corresponding log-rank tests and numbers at risk was generated to determine whether there were differences in 1- to 5-year survival between male and female gender. To also graphically illustrate the impact of echocardiographically measured PH on patient survival, Kaplan–Meier curves with different sPAP values were charted. Univariate Cox proportional hazard regression models in dependence of different sPAP values (≥ 40 mmHg and ≥ 50 mmHg) were used to calculate hazard ratios (HR) and 95% CI for several influencing factors associated with 1-, 3- and 5-year-mortality. For better comparability, a z-transformation was performed for metric data. Afterwards, multivariate Cox regression was performed to assess independent predictors of mortality. Therefore, covariates associated with mortality in the univariate analysis (p < 0.100) were entered and a backward variable elimination was performed. Subsequently, the Kaplan–Meier curves were again calculated using the above-mentioned sPAP values and this time strictly separated by gender. A separate univariate and multivariate Cox proportional hazard regression regarding 1-, 3- and 5-year survival as a function of gender was also carried out. Gender-specific area under the receiver operator characteristics (AUROC) curves with area under the curve (AUC) and separate analysis of Youden Index (YI) were performed using different time periods of survival (1, 2, 3, 4 and 5 years) to determine the respective sPAP cut-off value. Ultimately, the categorization of PH severity was determined based on the prevailing guidelines established by the American Society for Echocardiography, classifying it as absent (sPAP < 35 mmHg), mild (sPAP 35–50 mmHg), moderate/severe (sPAP > 50 mmHg). Subsequently, Kaplan–Meier curves were generated for the complete participant group, as well as for the distinct subsets of male and female patients. A p value < 0.05 was considered statistically significant.

303 patients, 151 men and 152 women, were enrolled at Paracelsus Medical University Hospital, Salzburg and Kepler University Hospital, Linz. An overview of the baseline characteristics is provided in Table 1. On average, the subjects were 82.6 ± 4.8 years old. sPAP was ≥ 40 mmHg in 46.9% of patients. Regarding baseline characteristics in males vs. females, our collective was reasonably comparable other than for expected gender differences in height, body weight, hematocrit and hemoglobin. Immediate pre-intervention creatinine was higher in men compared to women (1.1 ± 0.40 mg/dl vs. 0.90 ± 0.40 mg/dl; p = 0.003); however, the estimated glomerular filtration rate (eGFR) determined using the CKD-EPI formula and thus corrected for gender and age showed comparable values for renal function. However, men, in contrast to women, had a slightly reduced LVEF (53.5 ± 11.5% vs. 56.6 ± 8.7%; p = 0.009) and STS score (2.1 ± 1.3 vs. 3.4 ± 2.2; p < 0.001). Men also had an increased rate of post-TAVR pacemaker implantation (29.8% vs. 15.8%; p = 0.004) and atrial fibrillation (39.7% vs. 28.9%; p = 0.048).

To identify gender-specific differences in long-term survival after TAVR, Kaplan–Meier curves were generated up to 5 years after TAVR with corresponding log-rank tests and numbers at risk calculated annually (Fig. 1). There was a significantly higher mortality in males in the first 4 years after intervention. Only in the 5th year after TAVR, the mortality men and women were similar in mortality (log-rank test 5-year survival: p = 0.124).

To assess the sex-independent effect of PH on survival in the present collective, Kaplan–Meier curves were generated with different sPAP cut-off values (≥ 40 mmHg and ≥ 50 mmHg). Overall, an elevated sPAP, regardless of the cut-off selected, significantly increased mortality rates throughout the 5-year recording period in our cohort, although we would argue this is driven by the elevated mortality in males. Detailed statistical data analysis with log-rank tests and numbers at risk are shown in the corresponding figures (Fig. 2A: sPAP ≥ 40 mmHg; Fig. 2B: sPAP ≥ 50 mmHg).

To investigate influencing factors concerning mortality after TAVR in dependence of different sPAP values (≥ 40 mmHg and ≥ 50 mmHg), a univariate and multivariate Cox proportional hazard regression was performed (Table 2: 1-year mortality; Table 3: 3-year mortality; Table 4: 5-year mortality). Male sex was consistently a significant, and in some cases highly significant, factor for premature death in the univariate analysis, regardless of the observation period and regardless of the sPAP cut-off value. After inclusion of further clinical characteristics with a p < 0.100 in a multivariate analysis, male sex remained an independent risk factor for estimation of mortality after 1, 3 and 5 years in all calculations.

Kaplan–Meier curves with the selected cut-off values were finally further separated according to male and female gender (Fig. 3A: sPAP ≥ 40 mmHg male, Fig. 3B: sPAP ≥ 40 mmHg female; Fig. 4A: sPAP ≥ 50 mmHg male, Fig. 4B: sPAP ≥ 50 mmHg female). At all of the analysed sPAP levels (≥ 40 mmHg and ≥ 50 mmHg), the log-rank test was highly significant at every time point (1–5 year survival) for premature death, for males only. In contrast, there was no statistically significant differences in long-term survival in female study participants, regardless of sPAP cut-off values based on echocardiography at any timepoint.

In addition to different sPAP values, univariate and multivariate Cox hazard regression analysis depending on gender were also analysed (Table 5). In males, the sPAP value was an independent, highly significant risk factor for 1- and 3-year mortality after TAVR (p ≤ 0.001). After 5 years, sPAP still showed a similar trend, although not statistically significant (p = 0.093). In females, creatine kinase (CK) was an independent factor for earlier death after TAVR in all analysed periods (1, 3 and 5 years).

Gender-dependent AUROC analyses (Fig. 5) were calculated to assess sPAP cut-off values in relation to 1- to 5-year survival. In males (Fig. 5A), an sPAP value ≥ 53.50 mmHg was a highly significant, relevant predictor (p < 0.001) of premature death after interventional valve replacement in the first 3 years after TAVR. Later, at 4 and 5 years, sPAP values of ≥ 49.50 mmHg (4 years) and ≥ 45.50 mmHg (5 years) were also still highly significant (p = 0.001). In contrast, in female gender (Fig. 5B), the AUROC results showed no relevant significance.

Ultimately, to incorporate the presently accepted echocardiographic categorization of PH, aligning with the recommendations of the American Society of Echocardiography, Kaplan–Meier curves were computed. These calculations encompassed the aforementioned subgroups within the entire cohort (Fig. 6) and were further stratified based on gender (Fig. 7). It is worth noting that due to the limited count of patients exhibiting sPAP values exceeding 70 mmHg (Table 1), the categories of moderate and severe PH were merged into a singular group. In the complete study population (Fig. 6), the most notable risk of premature mortality subsequent to TAVR was evident among patients exhibiting an sPAP exceeding 50 mmHg. The statistical evaluation using log-rank tests revealed significant disparities across all observed survival intervals. Further examination within the male subgroup (Fig. 7A) consistently demonstrated noteworthy distinctions among the different sPAP categories, wherein the cohort with the highest sPAP levels also displayed the highest fatality rates. Conversely, in the female cohort (Fig. 7B), discernible differences in mortality among the varying sPAP classifications were not observed.

Gender-focused investigations exploring disparities in survival outcomes following TAVR have been plentiful in recent literature. Particularly concerning 1-year survival post-TAVR, our study aligns with findings indicating elevated male mortality at this juncture. Zahn et al. (2013) [13] exhibited a markedly increased 1-year mortality in males (23.6%) compared to females (17.3%) in a cohort of 1391 patients from the German Transcatheter Aortic Valve Interventions Registry (p < 0.001). Denegri et al. [14], examining a 3,821-patient Italian multicenter cohort, reported a 1-year mortality of 15.0% in males versus 11.5% in females (p < 0.001). Similarly, Yousif et al. [15] observed significantly higher 1-year mortality in males versus females (18.7% vs. 11.7%, p = 0.037) within the Swiss TAVI Cohort collective (n = 546). This gender-based disparity in survival was not confined to just 1 year; it extended to 3 and 5 years post-TAVR. Denegri et al. [14] undertook a 3-year mortality analysis in the aforementioned Italian cohort, uncovering earlier mortality in males (19.8% vs. 24.9%; p < 0.001) after adjusting for baseline characteristics. In 2017, Zahn et al. [16] demonstrated that female gender served as a protective, non-modifiable factor (HR 0.66, 95% CI 0.56–0.77; p < 0.001) for 5-year survival after TAVR based on the aforementioned registry's 5-year follow-up of 1444 patients. Our study's determination of significantly elevated 1-year mortality rates (17.2% for males vs. 7.2% for females; p = 0.008) seamlessly aligns with previous findings. Our results also revealed statistically significant disparities favoring female gender in 2-year (25.2% vs. 13.8%; p = 0.011), 3-year (27.8% vs. 17.1%; p = 0.020), and 4-year mortalities (34.4% vs. 23.0%; p = 0.037). Many studies have grappled with explaining the heightened male mortality post-TAVR. A common assertion attributes worse baseline vascular comorbidities to male sex across these studies [17]. However, this did not hold true in our analysis. Furthermore, the broader reduced life expectancy in males alone cannot entirely account for the recurrently documented early male mortality post-TAVR. Instead, our study bolsters the hypothesis proposed by Yousif et al. [15], focusing on the idea that the remodeling processes occurring in severe AS due to chronic pressure loading exhibit gender-specific differences. Female patients appear to tolerate left ventricular concentric hypertrophy secondary to AS for a longer duration, which contrasts with males and their propensity for eccentric, partially dilated cardiomyopathy. This notion is backed by the notably lower LVEF before TAVR in males, with nearly identical intraventricular septal thickness and a larger left ventricular end-diastolic diameter. A higher prevalence of atrial fibrillation in males further supports this idea. Hence, it is plausible that the cardiac remodeling process, transitioning from concentric pressure load to eccentric volume load in severe AS, occurs more swiftly and frequently in males, culminating in escalated risk of cardiac contractility loss, consequential heart failure and untimely demise [18;19]. Lastly, an argument could be posited that men present with a more advanced disease state at the time of TAVR, as evidenced by the lower LVEF prior to the procedure, potentially depleting cardiac adaptive capacity in males.

Concomitant post-capillary PH in severe AS is a notable risk factor for heightened post-TAVR mortality, a consensus evident across multiple studies involving invasive right heart catheterization. Schewel et al. [20] observed that patients (n = 1400) devoid of PH exhibited notably lower 1-year and 4-year mortality rates compared to those with PH (1-year mortality: 13.8% vs. 22.4%, p < 0.001; 4-year mortality: 37.2% vs. 51.5%, p < 0.001). In recent times, however, the routine use of invasive right heart catheterization, the gold standard for PH diagnosis, prior to TAVR has diminished. For many TAVR-equipped cardiology centers, a non-invasive assessment using TTE to establish PH presence or absence before intervention suffices. In this regard, sPAP predominantly serves in clinical practice, factoring in the patient's current volume status. Various sPAP cut-off values have been applied in comparative studies to define PH [21, 22]. This study, too, adopts the commonly used thresholds of 40 and 50 mmHg. Past analyses employing the FRANCE-2 registry [23] indicated that patients with severe AS and sPAP ≥ 40 mmHg or ≥ 60 mmHg (No PH: 22%; PH ≥ 40 mmHg and PH ≥ 60 mmHg: 28% each; p = 0.032) experienced significantly curtailed 1-year survival after TAVR. Similarly, Barbash et al. [24] documented a similar trend with sPAP ≥ 50 mmHg, highlighting that patients with moderate/severe PH (sPAP ≥ 50 mmHg) had significantly earlier 1-year mortality (p = 0.020) compared to those with no/mild PH (sPAP < 50 mmHg). Drawing from their multicenter registry, D’Ascenzo et al. [25] identified sPAP ≥ 40 mmHg as an independent risk factor for all-cause mortality. Bishu et al. [26] demonstrated that sPAP ≥ 50 mmHg was an independent predictor of long-term mortality post-TAVR. Our current study further underscores elevated mortality rates in severe AS patients with verified PH through TTE, irrespective of the chosen pre-TAVR sPAP cut-off value. Thus, sPAP, a clinically straightforward measurement, emerges not just as a valuable tool for non-invasive PH assessment, but also as a clinical parameter apt for pre-TAVR risk stratification.

Our study is the first to highlight gender-specific differences in survival risk after TAVR based on echocardiographic evidence of PH prior to intervention. For patients with PH, male gender was shown to be an isolated risk factor for premature death after TAVR. In contrast, in females, concomitant PH had no influence on postinterventional mortality regardless of sPAP cut-off values. While average sPAP was slightly higher in males (38.8 ± 19.1 mmHg) in contrast to females (35.5 ± 18.9 mmHg) in our cohort, this was without statistical significance (p = 0.123). Similarly, severity of PH using sPAP as a proxy, was not statistically significantly different (p = 0.109) although sPAP values ≥ 60 mmHg were more prevalent in percentage terms in males in our cohort (15.2% vs. 9.2%). Further breakdown of baseline characteristics according to corresponding sPAP groupings (data available on request) did not yield significant gender-specific differences for an sPAP ≥ 40 mmHg for atrial fibrillation (male: 47.5% vs. female: 36.8%, p = 0.204) or for mitral regurgitation ≥ II° (male: 26.7% vs. female 38.8%, p = 0.120) as a potential additive cause for severe PH. Thus, the question still remains: why is PH, as determined by sPAP, associated with such significantly increased mortality after TAVR, but, only in males? Here, a review by Rodiguez-Arias et al. [27] provides a potential theory, as the cause is also likely to be found in the more rapid remodeling and poorer adaptation of the right ventricle (RV) to pressure or volume loading in males. This hypothesis is supported primarily by the work of Melenovsky et al. [28] and Ventetuolo et al. [29]. Melenovsky and colleagues [28] demonstrated a relevant association between male sex, PH and RV dysfunction using a heart failure with preserved ejection fraction (HFpEF) collective. Ventetuolo et al. [29], using 15,464 veterans with invasive right heart catheterization data, showed that women, in contrast to men, had better survival rates in the presence of PH, regardless of cause, despite higher pulmonary vascular resistance and increased pulmonary artery pulse pressure values. As a potential explanation, the authors suggested a better RV adaptation in females. The question regarding PH reversibility after interventional valve replacement is still open, because in contrast to comparative studies [30, 31], sPAP changes after TAVR were not included in this study. In both Alushi et al. and Masri et al. TAVR patients were divided into different groups according to postinterventional course of sPAP. 51% [30] and 55% [31] of male subjects showed improvement in PH and thus potential reversibility. Whether and to what extent improved sPAP after TAVR could also positively influence male survival needs to be shown in follow-up studies.