Noninvasive assessment of pulmonary arterial capacitance by pulmonary annular motion velocity in children with ventricular septal defect

Pulmonary vascular hemodynamic assessment has traditionally evaluated pulmonary vascular resistance (PVR) [1‐3]. However, PVR primarily reflects small vessel status and the static component of the pulmonary circulation [4], and it does not reflect the properties of the large and medium pulmonary vessels or account for pulsatile elements of the pulmonary circulation. Pulmonary artery (PA) capacitance reflects the dynamic component of the pulmonary circulation and is thought to be determined largely by the properties of the large proximal capacitance vessels [5]. Because it is the immediate environment encountered by the right ventricle (RV), PA capacitance is a major determinant of RV work [5]. It has recently been shown that lower PA capacitance is strongly associated with worse survival in patients with pulmonary arterial hypertension (PAH) [5‐7]. Furthermore, PA capacitance is also reported to be a strong prognostic indicator in patients with left ventricular dysfunction, even in patients with normal PVR [8]. Consequently, accurate determination and serial follow-up of PA capacitance are important in the management of patients with various cardiac diseases with congestive heart failure. However, despite its potential importance for RV performance and clinical outcomes, PA capacitance is not routinely measured. Noninvasive quantitative assessment of PA capacitance remains challenging.

We have previously reported that the pulmonary annular motion velocity waveform obtained using tissue Doppler imaging (TDI) reflects right ventricular outflow tract (RVOT) performance [9]. In the present study, we hypothesized that the pulmonary annular motion would be affected not only by RVOT function, but also by PA capacitance because the pulmonary annulus is located adjacent to the RVOT and PA. In large arteries, deformation of the arterial wall during systole and diastole occurs in the radial and longitudinal directions (Additional file 1). The present study attempted to assess the relationship between the longitudinal pulmonary arterial deformation and capacitance.

The aim of this study, therefore, was to determine whether tissue Doppler-derived pulmonary annular motion velocity can be used as a noninvasive assessment of PA capacitance in children with ventricular septal defect (VSD).

No subjects were excluded from the subsequent analyses due to suboptimal recording from poor echocardiographic imaging. Accordingly, the study group included 23 healthy children (2.5 ± 1.6 years; range, 0.6–5.0 years) and 35 VSD patients (mean age, 2.3 ± 1.5 years; range, 0.6–5.0 years). Table 1 shows the clinical, echocardiographic, and hemodynamic data of the participants. Age, height, and HR did not differ significantly between the VSD group and the Control group, whereas body weight and BSA were significantly lower in the VSD group than in the Control group. Left ventricular end-diastolic dimension (LVEDD), left ventricular fractional shortening (LVFS), and LVEF were significantly higher in the VSD group than in the Control group.

Figure 1b shows the pulmonary annular velocity curve. The tricuspid annular motion is shown in Fig. 1c for comparison with the pulmonary annular velocity waveform. The systolic wave was monomodal (s’) for the tricuspid annular velocity and bimodal (s1′ and s2′) for the pulmonary annular motion velocity. The systolic waveform peaked earlier in pulmonary annular motion than in the tricuspid annular velocity curve. The shapes of the e’ and a’ waves in diastole were similar. Simultaneous recordings of pulmonary annular motion, RV pressure, and PA pressure in a patient with VSD are shown (Fig. 1d). The pulmonary s1′ wave corresponds to the steep RV pressure elevation during early systole. The s2′ wave coincides with mid to late systolic phase around the peak RV and PA pressure. Measurements of the sPAC and the dPAC are shown with the PA pressure profile (Fig. 1e). The comparison between the sPAC and dPAC obtained from the VSD group is demonstrated in Fig. 1f. The sPAC was significantly higher than the dPAC (4.2 ± 1.9 vs. 1.4 ± 0.7 mL/mmHg · m², p < 0.0001). There was a significant correlation between these parameters (Fig. 1g; r = 0.65, p < 0.0001).

Pulmonary annular motion was compared between the Control and VSD groups (Fig. 2). The peak velocities of the s1′, s2′, and e’ waves were significantly lower in the VSD group than in the Control group (9.9 ± 1.9 vs. 11.5 ± 2.2 cm/s, p = 0.0070; 3.4 ± 0.7 vs. 4.1 ± 1.1 cm/s, p = 0.0107; 8.8 ± 2.3 vs. 13.0 ± 2.7 cm/s, p < 0.0001, respectively). The s1′ duration, s2′ duration, e’ duration, and a’ duration were significantly shorter in the VSD group (p = 0.0008, <0.0001, 0.0392, and 0.0008, respectively), whereas isovolumic contraction time (ICT) and isovolumic relaxation time (IRT) were not significantly different between the two groups.

Next, the relationships between the parameters obtained from TDI-derived pulmonary and tricuspid annular motion and right heart performance were examined in the VSD group. Table 2 summarizes the univariate regression analysis. TDI-derived parameters were significantly correlated with and determined by RV and PA function. Figure 3 shows the relationship between the sPAC and the pulmonary annular motion waveform. The peak velocity of the pulmonary annular s1′ wave was significantly correlated with the sPAC (r = 0.58, p = 0.0002). The s2′ and e’ velocity also showed significant correlations with the sPAC (r = 0.62, p <0.0001; r = 0.66, p <0.0001; respectively). ICT and IRT measured using TDI and the time duration of each wave were also assessed in terms of their correlations with the sPAC. The sPAC had significant correlations with s1′ and e’ and a’ wave duration (r = 0.49, p = 0.0029; r = 0.48, p = 0.0034; r = 0.44, p = 0.0088; respectively), whereas there were no significant correlations with ICT, IRT, and s2′ wave duration. Figure 4 shows the relationship between dPAC and TDI-derived pulmonary annular motion. The peak velocities of pulmonary annular s1′, s2′, and e’ wave were significantly correlated with the dPAC (r = 0.60, p = 0.0002; r = 0.46, p = 0.0051; r = 0.55, p = 0.0006; respectively). In regard to the timing issue, the e’ wave duration was significantly correlated with the dPAC (r = 0.53, p = 0.0011).

Next, multiple regression analysis for predictors of the pulmonary annular motion waveform was performed (Table 3). The analysis showed that the sPAC had an effect on the peak velocities of s1′, s2′, and the e’ wave and the time duration of the s1′ and a’ waves (β = 0.41, p = 0.0131; β = 0.62, p <0.0001; β = 0.35, p = 0.0314; β = 0.49, p = 0.0029; and β = 0.27, p = 0.0488, respectively). The dPAC affected the s1′ wave peak velocity (β = 0.34, p = 0.0354). The other right heart performance parameters, including RVEDP, RVSP, and Qp/Qs, also had an impact on pulmonary annular motion. Importantly, although RVEF had been assumed to determine pulmonary annular motion, no significant relationship was demonstrated with pulmonary annular motion in the VSD group.

The present results showed that the pulmonary annular motion waveform is mainly determined by PA capacitance. Assessment of pulmonary annular motion is useful for estimating PA capacitance in children with VSD. Pulmonary annular TDI was found to be a simple, rapid, reproducible, and highly distinctive method for evaluating PA capacitance.

Recent reports demonstrated that PA capacitance has great impact on the prognosis in various cardiac diseases [5‐8]. Patients require lifelong follow-up that includes serial assessment of PA capacitance. Therefore, noninvasive assessment of PA capacitance is considered quite important. However, its assessment is challenging due to the difficulty of measurement and poor reproducibility. Therefore, an invasive method is necessary to accurately assess PA capacitance in clinical practice. Several studies have been conducted to identify noninvasive ways for estimating PA capacitance using echocardiographic parameters [13, 14]. These studies attempted echocardiographic estimation of PA capacitance from the PA pulse pressure and RV stroke volume. Systolic PA pressure was estimated from the tricuspid regurgitation velocity by the modified Bernoulli equation. RV stroke volume was calculated using the pulmonary valve diameter measured in the parasternal short-axis view and the velocity-time integral of the PA Doppler flow [13]. In other reports, the right pulmonary arterial diameter change during the cardiac cycle was used to calculate PA capacitance [14]. However, these methods are complex and time-consuming, errors can easily occur, and they have low reproducibility. Previous reports that attempted echocardiographic estimation of PA capacitance all had limitations due to poor reproducibility and difficulty, because accurate quantitative assessment by two-dimensional echocardiography is hampered by the complex geometry and difficult depiction. Pulmonary arterial wall strain evaluated by speckle tacking method can be a candidate for evaluating PA distensibility and capacitance. However, in our preliminary experimental study, the tracking was insufficient and the reproducibility was quite low. Therefore, nongeometric methods to assess PA compliance and PA deformation should be explored. One such method, TDI, allows the quantitative assessment of longitudinal PA deformation.

To the best of our knowledge, this is the first application of pulmonary annular motion velocity obtained by TDI as a tool for PA capacitance assessment. We postulated that the PA longitudinal deformation reflects stroke volume, and that the pulmonary annular motion waveform shows PA capacitance.

In the present study, PA capacitance was represented by two parameters: sPAC and dPAC. There was a significant correlation between the two parameters, although sPAC was significantly higher than dPAC. The significant difference between the two parameters was assumed to be caused by the time phase difference during the cardiac cycle. The sPAC is the arterial capacitance from end-diastole to systole, whereas the dPAC is the capacitance from end-systole to diastole.

The peak velocities of the pulmonary annular s1′, s2′, e’, and a’ waves were significantly lower in the VSD group than in the Control group (Fig. 2), whereas the peak velocity of the tricuspid s’ wave was significantly higher in the VSD group than in the Control group (Table 1). This discrepancy between pulmonary and tricuspid annular motions might have resulted from the different RV contractility properties between RV inflow and outflow regional properties that originate from the RV geometry, including myocardial fiber orientation [15, 16]. Furthermore, we assumed that PA motion velocity might be affected by PA properties, including PA capacitance. The annular motions are presumed to be affected by RV performance and the adjacent structures [9]. The pulmonary annulus and tricuspid annulus are located adjacent to the PA and right atrium, respectively. Tricuspid annular motion would be affected by right atrial stiffness, contraction, and pressure. Pulmonary annular motion is assumed to be affected by PA properties, including arterial pressure, compliance, and stiffness. These conditions might contribute to the opposite results for the peak velocities of pulmonary and tricuspid annular motions in the VSD group.

The present data demonstrated that pulmonary annular motion parameters can be used to represent PA capacitance. On multiple regression analysis, PA capacitance independently affected the pulmonary annular motion waveform and parameters. Importantly, RVEF was not an independent variable that determined pulmonary annular motion. This is likely due to the fact that PA stroke volume is not correlated with RVEF because VSD shunt flow is directly expelled to the PA during systole in the VSD group. The present data indicate that pulmonary annular motion is determined not by RV contraction, but by PA deformation in children with VSD. Nevertheless, the present study demonstrated the significant negative correlation between the pulmonary annular s1′ wave and RVEDP in the VSD group. Furthermore, RVSP and RVEDP affect the peak velocity of the e’ wave. It would be useful to investigate the relationship between pulmonary annular motion velocity and RV performance or RV overload. Further studies are necessary to determine the various factors affecting pulmonary annular motion.

Pulmonary annular TDI is a promising echocardiographic tool for evaluating PA capacitance in children with VSD. It can be a simple, rapid, reproducible, and highly distinctive method for evaluating PA capacitance.