Multiparametric vs. Inferior Vena Cava–Based Estimation of Right Atrial Pressure

Background: Right atrial pressure (RAP) can be estimated by echocardiography from inferior vena cava diameter and collapsibility (eRAPIVC), tricuspid E/e′ ratio (eRAPE/e′), or hepatic vein flow (eRAPHV). The mean of these estimates (eRAPmean) might be more accurate than single assessments. Methods and Results: eRAPIVC, eRAPE/e′, eRAPHV (categorized in 5, 10, 15, or 20 mmHg), eRAPmean (continuous values) and invasive RAP (iRAP) were obtained in 43 consecutive patients undergoing right heart catheterization [median age 69 (58–75) years, 49% males]. There was a positive correlation between eRAPmean and iRAP (Spearman test r = 0.66, P < 0.001), with Bland–Altman test showing the best agreement for values <10 mmHg. There was also a trend for decreased concordance between eRAPIVC, eRAPE/e′, eRAPHV, and iRAP across the 5- to 20-mmHg categories, and iRAP was significantly different from eRAPE/e′ and eRAPHV for the 20-mmHg category (Wilcoxon signed-rank test P = 0.02 and P < 0.001, respectively). The areas under the curve in predicting iRAP were nonsignificantly better for eRAPmean than for eRAPIVC at both 5-mmHg [0.64, 95% confidence interval (CI) 0.49–0.80 vs. 0.70, 95% CI 0.53–0.87; Wald test P = 0.41] and 10-mmHg (0.76, 95% CI 0.60–0.92 vs. 0.81, 95% CI 0.67–0.96; P = 0.43) thresholds. Conclusions: Our data suggest that multiparametric eRAPmean does not provide advantage over eRAPIVC, despite being more complex and time-consuming.


INTRODUCTION
Right atrial pressure (RAP) is an important prognostic factor in pulmonary hypertension (PH), regardless of whether this latter is due to pulmonary vascular disease, especially pulmonary arterial hypertension (PAH), or heart failure (HF) (1)(2)(3).
The scope of this study was to investigate the correlation between eRAP mean and its components, including eRAP IVC , and invasively measured RAP (iRAP) in a cohort of subjects undergoing right heart catheterization (RHC) for different reasons.

Study Population
In this prospective, observational, single-center study, we consecutively enrolled the patients who underwent RHC between September 2018 and January 2020 and had at least two components of eRAP mean measurements. For subjects undergoing multiple RHC during the study period, only the first one was considered. As per institutional policy on admission, all patients signed an informed consent to the use of their anonymized clinical data for research purposes. The study protocol was conducted in accordance with the ethical guidelines of the 1975 Declaration of Helsinki.

Echocardiography
A two-dimensional transthoracic echocardiogram was performed by two cardiologists (M.T. and S.G.) blinded to the results of RHC, on the same day of the hemodynamic assessment.
Standard images were acquired with the patient in the lateral decubitus position. Left ventricular (LV) dimensions and function were evaluated in the parasternal long-axis and apical four-chamber views. Mitral and aortic regurgitations were evaluated using color Doppler and continuous-wave Doppler in the apical four-and five-chamber views. LV diastolic function was examined through PW Doppler of the transmitral flow (Ewave and A-wave peak velocities, E/A ratio, deceleration time of the E-wave) and pulsed-tissue Doppler-derived e ′ velocity of the septal mitral annulus. Right ventricular (RV) end-diastolic basal diameter, tricuspid annular plane systolic excursion (TAPSE), Tissue Doppler S ′ peak velocity, fractional area change, and tricuspid regurgitation peak velocity (TRV) were assessed in the RV-focused apical four chamber view (4,9,15,16). RV systolic pressure was computed from TRV with the simplified Bernoulli equation (4,9).
IVC diameter was measured in the subcostal view just proximal to the junction of the HV, at end-expiration and then end-inspiration to determine the respiratory variation (4)(5)(6)9). HV flow was evaluated by PW Doppler in the subcostal view. Peak systolic and diastolic wave velocities (Vs and Vd, respectively) and the relevant velocity-time intervals (VTIs and VTId) were measured, and then the HV systolic filling fraction (HVFF) was calculated as VTIs/(VTIs + VTId) (4,7,10). Tricuspid E/e ′ ratio was derived by the tricuspid inflow E wave velocity (as determined by PW Doppler, with the sample volume at the tips of the leaflets during the RV-focused apical fourchamber view) and tricuspid lateral annulus e ′ wave velocity (with tissue Doppler imaging) (4,8,9). As tricuspid inflow and HV flow are highly sensitive to the respiratory phase, measurements from multiple beats were averaged. eRAP IVC , eRAP E/e ′ , and eRAP HV were given a value between 5 and 20 mmHg on a 5-mmHg scale as summarized in Table 1 and exemplified in Figure 1. eRAP mean was calculated as (eRAP IVC + eRAP E/e ′ + eRAP HV )/3 and thereby consisted of continuous values.

Right Heart Catheterization
RHC was performed under local anesthesia in the cardiac catheterization laboratory by other cardiologists (G.C., M.B., I.P., and P.A.), who were unaware of the results of the echocardiography. A balloon-tipped Swan-Ganz catheter was introduced through a sheath inserted into the femoral, antecubital, or jugular vein. The zero reference level was set at the midthoracic level. The catheter was advanced through the right heart chambers to the pulmonary artery, and pressures were measured. Then, the balloon was inflated, and the catheter was pushed forward up to the wedge position to record pulmonary artery wedge pressure. Finally, RV pressures and iRAP (mean over 5 cardiac cycles) were measured during catheter pull-back. Cardiac output was obtained by means of the thermodilution technique or Fick's indirect method (Dehmer formula), with the latter one being preferred in the presence of intracardiac or extracardiac shunts or severe tricuspid regurgitation.

Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics version 25.0. GraphPad Prism was also used to make the Figures.
Normality was assessed with the Kolmogorov-Smirnov test. Continuous variables are presented as mean ± standard deviation or median with interquartile range, as appropriate. Categorical variables are reported as absolute count and percentages. The relationship between iRAP and IVC diameter, tricuspid E/e ′ ratio, HVFF, or eRAP mean was analyzed by Spearman correlation ρ test. The correlation between eRAP mean and iRAP was also visually appraised by the Bland-Altman method. Furthermore, the correspondence between iRAP and eRAP IVC , eRAP E/e ′ , and eRAP HV values was evaluated using the Wilcoxon signed-rank test. The 5-and 10-mmHg eRAP thresholds were tested against the same iRAP thresholds by receiver operating characteristic (ROC) areas under the curve (AUCs), and the eRAP HV , eRAP E/e ′ , and eRAP mean AUC were compared with the eRAP IVC AUC by means of the Wald test.
Because eRAP mean and its components had not been compared before, no expected difference between these assessments was available to set a minimum number of enrolled patients.

RESULTS
Forty-three patients were included in the analysis. Their characteristics are shown in Table 2. The reasons for RHC were PH diagnosis (29 subjects, of whom 6 were found with PAH, 5 with chronic thromboembolic PH, and 4 with left heart disease-associated PH, and 14 did not actually have PH), PAH reassessment (10 subjects), or evaluation of HF eligibility to LVAD or heart transplant (4 subjects). Median age was 69 (58-75) years, and 28 (65%) patients were older than 65 years; male and female genders were equally distributed. Functional class was most often II, and median N-terminal pro-brain natriuretic peptide was 462 (114-2,045) ng/L. At the hemodynamic evaluation, median iRAP was 7 (3-11) mmHg, and 67% of the patients had an iRAP value below the 8-mmHg cutoff that identifies a higher risk of mortality (1,18).
Echocardiographic assessment of IVC was feasible in the entire study population, whereas HV parameters and tricuspid E/e ′ ratio were not determinable in 4 and 2 patients, respectively. Interobserver agreement was very good (weighted k = 0.84, 0.90, and 0.87 for eRAP IVC , eRAP E/e ′ , eRAP HV , respectively). Median eRAP IVC , eRAP E/e ′ , eRAP HV , and eRAP mean were 5 (5-10), 5 (5-20), 10 (5-10), and 6.7 (5-11.7) mmHg, respectively. The parameters from which eRAP IVC , eRAP E/e ′ , and eRAP HV are derived were positively correlated with iRAP: r was 0.47 for IVC diameter (P = 0.002), 0.44 for tricuspid E/e ′ ratio (P = 0.004), and 0.46 for HVFF (P = 0.007). Consistently, there was also a positive correlation between eRAP mean and iRAP (r = 0.66, P < 0.001; Figure 2, left). The Bland-Altman plot showed that eRAP mean was in agreement with iRAP especially when ≤10 mmHg (Figure 2, right). For all eRAP components, 5 mmHg was the most frequent estimate, and the actual iRAP was not significantly different from it (Figure 3). For the 10-mmHg category, the concordance between eRAP components and iRAP was less frequent, particularly for eRAP E/e ′ and eRAP HV , although not to a statistically significant extent. For the 15-mmHg value, it was possible to test only the correlation between iRAP and eRAP IVC (no significant difference), as the number of eRAP E/e ′ and eRAP HV was too low. A statistically significant difference between iRAP and eRAP E/e ′ (P = 0.02) and eRAP HV (P < 0.001) was instead found for the 20-mmHg threshold (Figure 3).
The accuracy in predicting iRAP was numerically highest for eRAP mean for both the 5-and the 10-mmHg categories ( Table 3 and Figure 4). Nonetheless, the AUC of eRAP mean was not significantly different from that of eRAP IVC , nor were the AUC of eRAP HV and eRAP E/e ′ ( Table 3).

DISCUSSION
eRAP is part of the standard transthoracic echocardiographic examination and provides important information. It is fundamental in the diagnostic workup of PH, as systolic pulmonary artery pressure is calculated as the sum of eRAP and RV systolic pressure (1,4,9). Furthermore, elevated eRAP is associated with worse prognosis in HF (19,20) and PAH (21).
In clinical practice, eRAP is obtained by examining the dimension and respiratory collapsibility of IVC (6). Other methods for eRAP exist, but have not been validated across different populations (11). Hence, eRAP IVC is recommended as the default approach, with other modalities being complementary (4,22). Nonetheless, eRAP IVC is approximate. A semiautomated assessment of IVC collapsibility and pulsatility has recently been proposed to overcome the limitations of eRAP IVC (23). Alternatively, eRAP could be more precise if the estimates attained with different techniques were incorporated into a multiparametric scoring system (13).
On this background, we determined the accuracy of averaging the values of eRAP derived from the evaluation of IVC, HV PW Doppler profiles, and tricuspid E/e ′ ratio. Although eRAP mean did correlate with iRAP, it did not perform significantly better than eRAP IVC in predicting iRAP.
Individual comparisons of eRAP IVC , eRAP E/e ′ , and eRAP HV with iRAP have already been drawn (6)(7)(8)(24)(25)(26)(27). By contrast, to our knowledge, only one recent investigation with LVAD patients considered eRAP IVC , eRAP E/e ′ , and eRAP HV together to compute eRAP mean (14). In this study like in ours, eRAP mean had FIGURE 2 | Correlation between multiparametric estimation and invasive measurement of right atrial pressure. (Left) Positive correlation between eRAP mean and iRAP as assessed by RHC (Spearman correlation test). (Right) Bland-Altman plot showing that estimation of iRAP by eRAP mean was especially good for values <10 mmHg. The blue lines represent the average ± 1 standard deviation of (eRAP mean and iRAP). Note that in both analyses some subjects had the same values, hence the relevant dots overlap in the graphs. eRAP mean , multiparametric estimated RAP; iRAP, invasive RAP; RHC, right heart catheterization. FIGURE 3 | Correlation between single-parameter estimation and invasive measurement of right atrial pressure. The actual values of iRAP obtained during RHC are presented for each 5-mmHg threshold and by eRAP component. Horizontal bars indicate medians and interquartile ranges, *P < 0.05 and **P < 0.001, respectively (Wilcoxon signed-rank test). iRAP, invasive right atrial pressure; eRAP, estimated right atrial pressure; IVC, inferior vena cava; E/e ′ ratio of pulsed wave Doppler tricuspid inflow early E-wave peak velocity and tricuspid lateral annulus tissue Doppler imaging e ′ wave velocity; HV, hepatic veins. the greatest AUC for the detection of iRAP >10 mmHg. However, the authors focused on the value of eRAP mean in combination with several other echocardiographic variables in guiding LVAD management, and no statistical comparison between ROC was performed, precluding any conclusion about the higher accurateness of eRAP mean over eRAP IVC . It is also notable that we included a heterogeneous cohort of subjects, a crucial step to understand the potential clinical value of eRAP mean .
For each eRAP mean component (eRAP IVC , eRAP E/e ′ , and eRAP HV ), echocardiographic and invasive values were more often similar when eRAP was <10 mmHg. As a consequence, the correspondence between eRAP mean and iRAP also appeared to be looser for eRAP mean values >10 mmHg. The highest discordance with iRAP was found for eRAP E/e ′ and eRAP HV >10 mmHg. Consistent with our results, the cutoffs beyond which eRAP E/e ′ was less reliable in previous studies were also <10 mmHg (8,28). Overall, the present work supports the systematic use of eRAP IVC in the clinical arena, as it is the simplest way to estimate RAP. Moreover, an extensive literature indicates that the echocardiographic evaluation of IVC offers diagnostic and prognostic cues per se, regardless of which value is assigned to eRAP IVC . Demonstration of a dilated and/or non-collapsible IVC may be sufficient to identify patients with HF and increased LV filling pressures (29) and has been associated with HF hospitalization and mortality (19,30,31). In addition, a larger IVC size at discharge was related to a higher risk of readmission after a first hospitalization for HF (32,33). An independent prognostic role of IVC dilation and reduced collapsibility has also been shown in PAH (34). However, eRAP E/e ′ and eRAP HV may be more convenient in specific populations. eRAP E/e ′ can be helpful in patients with a poor subcostal ultrasound window (24), and HVFF has specifically been evaluated in mechanically ventilated patients (7).
Until eRAP IVC remains the reference in clinical practice, efforts to improve it are desirable, for instance, by tracking the respirophasic movements of the IVC in echocardiographic videoclips (23).
We acknowledge that the sample we examined was small and mostly made of subjects with a low iRAP. Thus, the data presented here should be viewed as preliminary to bigger studies with a wider range of iRAP. On the other hand, this work is the first one addressing the performance of eRAP mean in a series of consecutive patients with different cardiac disorders. It is also remarkable that eRAP and iRAP were assessed on the same day and, in most cases, few hours apart by reciprocally blinded investigators.

CONCLUSIONS
The optimal approach for eRAP during transthoracic echocardiography is debated; recently, it has been suggested that incorporating the analysis of IVC, tricuspid E/e ′ ratio, and HV is better than relying only on IVC assessment.
In this prospective cohort of patients in whom RAP was invasively measured, however, multiparametric eRAP was not more precise than the estimate based on IVC, tricuspid E/e ′ ratio, or HV.
While awaiting for additional studies, we conclude that, at present, evaluation of IVC diameter and collapsibility is preferable for eRAP.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by IRCCS Ospedale Policlinico San Martino Institutional Review Board. The patients/participants provided their written informed consent to participate in this study.