The Prognostic Utility of Invasive Right Heart Catheterization in Patients Prior to Transcatheter Aortic Valve Replacement

Introduction. Invasive right heart catheterization (RHC) provides valuable prognostic information in cases of severe aortic stenosis, but is not done routinely prior to transcatheter aortic valve replacement (TAVR). Therefore, we sought to investigate the prognostic utility of pre-TAVR RHC for assessing outcomes post TAVR. Methods. This is a single-center, retrospective study of patients who underwent RHC prior to TAVR between June 2011 and March 2019. We evaluated abnormalities in the following variables as predictors of poor outcomes post TAVR: pulmonary capillary wedge pressure (PCWP), systolic pulmonary artery pressure (PASP), mixed venous oxygen saturation (MVO₂), right ventricular stroke work index (RVSWI), and right atrial pressure (RAP). Cox proportional hazard regression models were used to assess the primary composite outcome of all-cause mortality and hospitalization for heart failure. Results. A total of 522 patients (mean age, 83.5 ± 4.5 years; 49.4% women) with complete invasive hemodynamic data were included. At a median follow-up of 529.5 days, there were 127 deaths and 59 heart failure hospitalizations. On multivariate analysis, PCWP ≥15 mm Hg (hazard ratio [HR], 1.48; 95% confidence interval [CI], 1.02-2.07), PASP ≥50 mm Hg (HR, 1.66; 95% CI, 1.17-2.36), MVO₂ ≤60% (HR, 1.42; 95% CI, 1.01-1.98), RVSWI ≥12 g•m/m²/beat (HR, 1.40; 95% CI, 1.004-1.94), and RAP ≥10 mm Hg (HR, 1.66; 95% CI, 1.09-2.51) were independent predictors of death or heart failure hospitalization. Conclusions. Preprocedural invasive RHC provides useful prognostic information. A comprehensive invasive hemodynamic assessment should be considered for risk stratification in patients undergoing TAVR.

J INVASIVE CARDIOL 2022;34(1):E14-E23. Epub 2021 December 5.

Key words: right heart catheterization, risk stratification, transcatheter aortic valve replacement

The use of transcatheter aortic valve replacement (TAVR) for the treatment of symptomatic severe aortic stenosis (AS) is on the rise and will only continue to increase with the aging population, approval of TAVR for low-risk surgical candidates, and ongoing trials evaluating TAVR in asymptomatic patients and patients with moderate AS.^1-4 While TAVR has the potential to improve mortality and provide patients with significant symptomatic relief, the overall mortality rate for patients who undergo TAVR remains high.^5,6 As our population grows older and the criteria for patient eligibility for TAVR continues to expand, the need to be able to accurately identify individuals at greatest risk for poor outcomes post TAVR is becoming even more important.

Early investigations into structural abnormalities as predictors of poor outcomes post TAVR focused primarily on left-sided cardiac structures, particularly the left ventricle; however, subsequent investigations have shifted attention to structural changes in the right heart as being prognostically significant.^6-11 Most of these prior studies investigating the prognostic utility of right heart measurements have relied on transthoracic echocardiogram (TTE) for patient assessment and have focused primarily on pulmonary hypertension (PH) as a risk factor.^7,8,11 While TTE is an exceptionally useful non-invasive tool in the assessment of cardiac structure and function, there is evidence that it is limited in terms of accuracy and reproducibility in the assessment of the right heart.^12-14 To the best of our knowledge, there has never been a study demonstrating the prognostic utility of the easily obtained measurements from invasive right heart catheterization (RHC), other than pulmonary artery pressure (PAP) as it relates to PH. Hence, we sought to evaluate the utility of measurements obtained from pre-TAVR invasive RHC in identifying poor outcomes post TAVR.

Patient population. We performed a retrospective study of patients undergoing TAVR between October 2011 and March 2019 at the Regional Cardiac Catheterization Lab at Kaiser Permanente Los Angeles Medical Center in Los Angeles, California. This is a tertiary referral center for Kaiser Permanente medical centers across the Southern California region. Patients referred for TAVR are evaluated utilizing a heart team approach that includes a cardiothoracic surgeon and interventional cardiologist who both deem TAVR to be the optimal treatment option. Type of prosthetic valve used (ie, self-expanding vs balloon expandable), and other procedural technical considerations were chosen at the discretion of the operator. Inclusion criteria for this study were: (1) ≥18 years of age; and (2) complete RHC data available. A total of 522 patients met the inclusion criteria and were included in the analysis. This study was approved by the institutional review board of Kaiser Permanente Southern California and received the proper ethical oversight.

Clinical characteristics and outcomes. International Classification of Disease (ICD-9 and ICD-10) codes were used to obtain patient demographics and comorbidities, and were adjudicated via manual review of electronic medical records. Echocardiogram reports were manually reviewed for left ventricular ejection fraction, valvular regurgitation, peak aortic velocity, aortic gradient, and aortic valve area. TAVR procedural reports were manually reviewed for valve type and size. The outcomes evaluated were all-cause mortality and heart failure hospitalization after TAVR. All-cause mortality was obtained from the California State Death Master Files and Kaiser Permanente Healthcare Plan database. Heart failure hospitalization was defined as any hospitalization or emergency room visit with the primary encounter diagnosis of heart failure as identified by ICD-9 and ICD-10 codes.

RHC data. All patients undergoing TAVR with complete RHC data available were included in the study. RHC was routinely performed during initial TAVR evaluation. The following RHC data were manually obtained from records: right atrial pressure (RAP), pulmonary artery systolic pressure (PASP), pulmonary capillary wedge pressure (PCWP), mixed venous oxygen saturation (MVO₂), cardiac output and cardiac index, and right ventricular end-diastolic pressure (RVEDP). Right ventricular stroke work index (RVSWI) was used as an invasive surrogate for right ventricular function and was calculated as: RVSWI = (stroke volume index) x (mean PAP – mean right atrial pressure) x 0.0136. The cut-offs for hemodynamic measurements were chosen according to prior investigations.^15-19

Statistical analysis. Comparisons were made between groups stratified by RHC variable. The Shapiro-Wilk test was used to evaluate for normality. Independent Student’s t-test, expressed as mean ± standard deviation, and Mann-Whitney- U test, expressed as median (interquartile range [IQR]), were used where appropriate for continuous variables. The Chi-square test (expressed as number [%]) was used for categorical variables. Cox proportional hazard regression analysis and the Kaplan-Meier method were used to evaluate the association between predictor variables and outcomes after TAVR. For heart failure hospitalization outcome, patients were censored on the first event. A 2-sided P-value of <.05 was used as the cut-off for statistical significance. All statistical analysis was performed using R, version 3.6.3.

Between June 2011 and March 2019, a total of 683 patients underwent TAVR for severe AS. Of these, a total of 522 patients (mean age, 83.5 ± 4.5 years; 50.6% men) were included in the analysis (Table 1, Part 1 and Table 1, Part 2). Median time to follow-up was 529 days (IQR, 231-1043 days). There were 127 total deaths (24%) among the patients analyzed and 59 total hospitalizations (11%) due to heart failure.

PCWP ≥15 mm Hg. There were 245 patients (47%) with PCWP ≥15 mm Hg. These patients were more likely to have a higher body mass index (27.2 kg/m² vs 25.8 kg/m²), chronic obstructive pulmonary disease (42% vs 33.2%), atrial fibrillation/flutter (50.6% vs 28.9%), more beta-blocker use (66.9% vs 58.1%), higher brain natiuretic peptide (457.5 pg/mL vs 258 pg/mL), lower ejection fraction (60% vs 65%), and at least moderate mitral regurgitation (16.7% vs 8.9%) vs patients with PCWP <15 mm Hg, respectively (all P<.05). At long-term follow-up adjusting for age, sex, body mass index, peripheral vascular disease, atrial fibrillation/flutter, and glomerular filtration rate, PCWP ≥15 mm Hg was an independent predictor of the composite outcome of all-cause mortality and heart failure hospitalization (hazard ratio [HR], 1.48; 95% confidence interval [CI], 1.02-2.07; log rank P=.01 for time-to-event analysis) (Table 3 and Figure 1). When evaluating the outcomes separately, PCWP ≥15 mm Hg was independently associated with heart failure hospitalization (HR, 2.12; 95% CI, 1.21-3.40), but not all-cause mortality (HR, 1.36; 95% CI, 0.94-1.95) after TAVR. At 30-day and 1-year follow-up, PCWP ≥15 mm Hg was not associated with worse outcomes, ie, composite outcome (30 days: HR, 1.27; 95% CI, 0.60-2.69; 1 year: HR, 1.32; 95% CI, 0.83-2.10), all-cause mortality (30 days: HR, 1.54; 95% CI, 0.57-4.14; 1 year: HR, 1.40; 95% CI, 0.81-2.44), or heart failure hospitalization (30 days: HR, 1.03; 95% CI, 0.33-3.28; 1 year: HR, 1.69; 95% CI, 0.83-3.44) on multivariate analysis.

PASP ≥50 mm Hg. PH (defined as mean PAP ≥25 mm Hg) was noted in a total of 284 patients (54%) and 154 of these patients (29.5% overall) had a PASP ≥50 mm Hg; 121 of the 154 patients (79%) had a transpulmonary gradient of >12 mm Hg. The 154 patients with PASP >50 mm Hg were slightly younger (mean age, 82 years vs 84 years), predominantly male (57.8% vs 47.6%), and were more likely to have a higher body mass index (27.5 kg/m² vs 25.9 kg/m²), diabetes (60.4% vs 47%), atrial fibrillation/flutter (51.9% vs 33.7%), anemia (33.8% vs 24.2%), lower glomerular filtration rate (54.5 mL/min vs 60.5 mL/min), higher brain natriuretic peptide (575.5 pg/mL vs 278 pg/mL), lower ejection fraction (60% vs 65%), at least moderate tricuspid regurgitation (20.3% vs 12.6%), and higher left ventricular end-diastolic pressure (LVEDP) prior to valve deployment (19 mm Hg vs 16 mm Hg) vs patients with PASP <50 mm Hg, respectively (all P<.05). At long-term follow-up on multivariate analysis, PASP ≥50 mm Hg was independently associated with the composite outcome of death and heart failure hospitalization (HR, 1.66; 95% CI, 1.17-2.36; log rank P<.001 for time-to-event analysis) (Table 3 and Figure 1). When evaluating the outcomes separately, PASP ≥50 mm Hg was independently associated with all-cause mortality (HR, 1.60; 95% CI, 1.10-2.33) and heart failure hospitalization (HR, 1.81; 95% CI, 1.05-3.11) after TAVR. At 30-day follow-up, PASP ≥50 mm Hg was independently associated with the composite outcome (HR, 2.61; 95% CI, 1.22-5.59) and all-cause mortality (HR, 3.10; 95% CI, 1.15-8.35), but not heart failure hospitalization (HR, 2.04; 95% CI, 0.63-6.60). At 1-year follow-up, PASP ≥50 mm Hg was independently associated with all-cause mortality (HR, 1.81; 95% CI, 1.04-3.17), but not the composite outcome (HR, 1.56; 95% CI, 0.97-2.52) or heart failure hospitalization (HR, 1.59; 95% CI, 0.78-3.21).

MVO₂ ≤60%. There were a total of 167 patients (32%) with MVO₂ ≤60%. These patients were more likely to have diabetes (58.1% vs 47.6%), coronary artery disease (83.2% vs 73%) with prior myocardial infarction (30.5% vs 22.5%), atrial fibrillation/flutter (47.9% vs 34.9%), anemia (34.1% vs 23.7%), higher BNP (521 pg/mL vs 286 pg/mL), more beta-blocker use (70.1% vs 58.6%), lower ejection fraction (60% vs 65%), and more cases of at least moderate tricuspid regurgitation (22.5% vs 11.3%) vs patients with MVO₂ >60%, respectively (all P<.05). At long-term follow-up on multivariate analysis, MVO₂ ≤60% was a significant predictor of the primary composite outcome of death and heart failure hospitalization (HR, 1.42; 95% CI, 1.01-1.98; log rank P=.02 for time-to-event analysis) (Table 3 and Figure 1). When evaluating the outcomes separately, MVO₂ ≤60% was independently associated with all-cause mortality (HR, 1.47; 95% CI, 1.02-2.11) after TAVR. At 30-day follow-up, MVO₂ ≤60% was not associated with worse outcomes, ie, composite outcome (HR, 1.30; 95% CI, 0.61-2.78), all-cause mortality (HR, 2.01; 95% CI, 0.77-5.25), and heart failure hospitalization (HR, 0.68; 95% CI, 0.18-2.54). At 1-year follow-up, MVO₂ ≤60% was independently associated with the composite outcome (HR, 1.68; 95% CI, 1.06-2.65) and heart failure hospitalization (HR, 2.14; 95% CI, 1.09-4.19), but not all-cause mortality (HR, 1.41; 95% CI, 0.82-2.44) on multivariate analysis.

RVSWI ≥12 g•m/m²/beat. There was a total of 213 patients (40.8%) who had RVSWI ≥12 g•m/m²/beat (Table 2, Part 1 and Table 2, Part 2). These patients were more likely to have chronic obstructive pulmonary disease (43.7% vs 33%), anemia (33.3% vs 22.7%), and higher brain natriuretic peptide (453.5 pg/mL vs 276 pg/mL) vs patients with RVSWI <12 g•m/m²/beat, respectively (all P<.05). On multivariate analysis, RVSWI ≥12 g•m/m²/beat was a significant predictor of the primary composite outcome of death and heart failure hospitalization (HR, 1.40; 95% CI, 1.004-1.94; log rank P=.02 for time-to-event analysis) (Table 3 and Figure 1). At 30-day and 1-year follow-up, there were no differences (RVSWI ≥12 g•m/m²/beat vs RVSWI <12 g•m/m²/beat) in the composite outcome  (30 days: HR, 1.16; 95% CI, 0.55-2.43; 1 year: HR, 1.02; 95% CI, 0.65-1.62), all-cause mortality (30 days: HR, 1.52; 95% CI, 0.58-4.02; 1 year: HR, 1.13; 95% CI, 0.66-1.94), and heart failure hospitalization (30 days: HR, 0.82; 95% CI, 0.24-2.77; 1 year: HR, 0.89; 95% CI, 0.45-1.77) on multivariate analysis.

RAP ≥10 mm Hg. There were a total of 92 patients (17.6%) with RAP ≥10 mm Hg. These patients were slightly younger (mean age, 81 years vs 85 years), and were more likely to have a higher body mass index (29.4 kg/m² vs 25.9 kg/m²), atrial fibrillation/flutter (51.1% vs 36.5%), liver disease (8.7% vs 3.7%), higher brain natriuretic peptide (490 pg/mL vs 320 pg/mL), and lower ejection fraction (60% vs 65%) vs patients with RAP <10 mm Hg, respectively (all P<.05). On multivariate analysis, RAP ≥10 mm Hg was a significant predictor of the primary composite outcome of death and heart failure hospitalization (HR, 1.66; 95% CI, 1.09-2.51; log rank P=.02 for time-to-event analysis) (Table 3 and Figure 1). When evaluating the outcomes separately, RAP ≥10 mm Hg was independently associated with heart failure hospitalization (HR, 2.20; 95% CI, 1.19-4.06), but not all-cause mortality (HR, 1.38; 95% CI, 0.86-2.20) after TAVR. At 30 days, RAP ≥10 mm Hg was not associated with the composite outcome (HR, 2.02; 95% CI, 0.85-4.84), all-cause mortality (HR, 2.14; 95% CI, 0.65-6.98), and heart failure hospitalization (HR, 1.86; 95% CI, 0.52-6.74). At 1-year follow-up, RAP ≥10 mm Hg was independently associated with the composite outcome (HR, 1.95; 95% CI, 1.13-3.34) and all-cause mortality (HR, 1.97; 95% CI, 1.03-3.77), but not heart failure hospitalization (HR, 1.99; 95% CI, 0.91-4.34).

Additional analyses. As part of our statistical analysis, we performed multiple separate multivariate assessments and used inverse probability treatment weighting to further account for possible confounding. The propensity score for each respective variable (PCWP, PASP, MVO₂, RVSWI, and RAP) was estimated by a logistical regression model including 21 covariates that covered demographics, clinical characteristics, and baseline medication use. Assessment of balance between groups was determined by independent t-test and Chi-square test for continuous and categorical variables, respectively, and each variable was determined to be very well balanced between groups. We performed these assessments for the 5 main variables described above, as well as for RVEDP (Table 3). We also went on to compare the unadjusted HRs for all of the adverse predictors of our primary composite outcome of mortality and heart failure hospitalization (Figure 2).

Most prior research attempting to identify hemodynamic predictors of poor outcomes post TAVR have relied on TTE and have focused primarily on PH.^{8,9,11,15,16,20-23} To the best of our knowledge, our study is the first to demonstrate that pre-TAVR invasive RHC provides additional measurements that can be used to predict poor outcomes post TAVR. These findings support consideration for the use of RHC as a part of pre-TAVR patient assessment.

There have only been 5 prior studies evaluating the predictive value of hemodynamic measurements obtained from pre-TAVR RHC and clinical outcomes post TAVR.^{9,15,16,20,23} These studies focused on pre-TAVR PH as a predictor of poor outcomes and their results suggest that baseline PH alone is not a reliable predictor of post-TAVR adverse events and that other factors need to be considered when evaluating patient risk prior to TAVR.^{9,15,16,20,23}

In 2015, O’Sullivan et al evaluated 433 patients for PH (defined as mean PAP ≥25 mm HG on pre-TAVR RHC) as a predictor of mortality post TAVR. Patients with PH were separated into precapillary (LVEDP or PCWP ≤15 mm Hg) and postcapillary (LVEDP or PCWP >15 mm Hg) PH groups. The postcapillary group was further divided into isolated postcapillary PH vs combined pre/postcapillary PH based on diastolic pressure gradient. Worse mortality was seen in patients with unspecified PH, precapillary PH, and combined pre/postcapillary PH, but was not seen in patients with isolated postcapillary PH compared with patients without baseline PH.¹⁵

In 2015, Lindman et al evaluated 2180 patients from the PARTNER 1 registry and demonstrated an increase in all-cause mortality in patients with moderate-to-severe PH (mean PAP ≥35 mm Hg on pre-TAVR RHC), but this difference was due entirely to an increased mortality in female patients. Among males, there was no increased hazard of death with any degree of PH.²⁰ In 2015, Schewel et al analyzed 439 patients who underwent RHC immediately pre and post TAVR. They found PH (mean PAP ≥25 mm Hg) to be a significant predictor of mortality post TAVR, with all-cause mortality twice as high in patients with PH. These patients also had increased major vascular complications and rates of acute kidney injury.²³

In 2017, Parikh et al analyzed 93 patients for PH prior to TAVR and showed an increased mortality, length of hospitalization, incidence of acute kidney injury, and new atrioventricular block among patients with PASP ≥50 mm Hg on RHC.¹⁶ In 2018, Masri et al analyzed 407 patients who underwent RHC prior to TAVR and assessed for improvement in PH based on post-TAVR TTE as a predictor of all-cause mortality. This study demonstrated that, among patients with PH prior to TAVR (mean PAP ≥25 mm Hg on RHC), baseline PH alone was not a significant predictor of mortality and only patients with persistent PH on post-TAVR TTE had higher mortality.⁹

In our current large observational study of 522 patients, we found that pre-TAVR PH was highly prevalent in severe AS (54% based on mean PAP ≥25 mm Hg), similar to prior investigations.^{9,15,16,20,23} Our findings are in agreement with O’Sullivan et al, Schewel et al, and Parikh et al in that the presence of baseline PH appears to be a significant risk factor for poor outcomes post TAVR (increasing the risk for death or heart failure hospitalization by 66%).^15,16,23 Additionally, we demonstrate that other measurements obtained during pre-TAVR RHC also have predictive value post procedure, including PCWP (HR, 1.48; 95% CI, 1.02-2.07), MVO₂ (HR, 1.42; 95% CI, 1.01-1.98), RVSWI (HR, 1.40; 95% CI, 1.004-1.94), and RAP (HR, 1.66; 95% CI, 1.09-2.51) for our primary outcome. To the best of our knowledge, these measurements have not been shown in previous studies to have prognostic value.

The adverse patient outcomes we describe in this paper are most likely a reflection of the underlying pressure overload and pathologic cardiac remodeling that progressively involves right-sided cardiac structures as a result of severe AS.^7,8 Pathologic changes involving the right heart appear to portend a poor prognosis, and a thorough assessment of right-sided hemodynamics is a valuable component of pre-TAVR assessment.^7,8 Given the high prevalence of baseline PH in patients with severe AS, and research showing that undifferentiated PH alone is not a reliable predictor of post-TAVR outcomes, a comprehensive invasive hemodynamic assessment can provide the necessary additional information needed to be able to accurately identify patients likely to have poor outcomes post TAVR. While TTE is a powerful tool for non-invasive hemodynamic assessment, its utility in assessing the right heart is limited by accuracy and reproducibility.^12-14 RHC remains the gold standard in assessing patients for PH and, as our research demonstrates, can provide additional measurements that can be used in pre-TAVR risk stratification that TTE cannot.¹²

Study limitations. Our study is not without limitations. This is a retrospective study, which may not be able to fully account for confounding variables. This is a single-center study in that all patients underwent RHC and TAVR at the same medical center, but the patients themselves were referred from a large catchment area throughout Southern California. Additionally, all of the patients in this study were insured and had relatively easy access to primary and subspecialty care, potentially limiting the generalizability of these results to other populations.

Over half of the patients in this study had PH prior to TAVR. Beyond PA pressure alone, multiple hemodynamic measurements obtained by RHC may be useful for the risk-stratification of patients prior to TAVR. Our findings suggest that a thorough invasive hemodynamic assessment can help to predict poor outcomes post TAVR and should be routinely considered.

• Bleeding Complications Related to Right Heart Catheterization in the Setting of Elevated INR
• Right Heart Catheterization: Right From the Anatomical Snuffbox
• Real-Time Ultrasound-Guided Venous Access of the Arm for Right Heart Catheterization
• Left Ventricular Pressure Ratio Predicts In-Hospital Outcomes in Hospitalized Heart Failure With Reduced Ejection Fraction