Pulmonary pulse wave transit time is associated with right ventricular–pulmonary artery coupling in pulmonary arterial hypertension

Abstract Abstract

Pulmonary pulse wave transit time (pPTT), defined as the time for the systolic pressure pulse wave to travel from the pulmonary valve to the pulmonary veins, has been reported to be reduced in pulmonary arterial hypertension (PAH); however, the underlying mechanism of reduced pPTT is unknown. Here, we investigate the hypothesis that abbreviated pPTT in PAH results from impaired right ventricular–pulmonary artery (RV-PA) coupling. We quantified pPTT using pulsed-wave Doppler ultrasound from 10 healthy age- and sex-matched controls and 36 patients with PAH. pPTT was reduced in patients with PAH compared with controls. Univariate analysis revealed the following significant predictors of reduced pPTT: age, right ventricular fractional area change (RV FAC), tricuspid annular plane excursion (TAPSE), pulmonary arterial pressures (PAP), diastolic pulmonary gradient, transpulmonary gradient, pulmonary vascular resistance, and RV-PA coupling (defined as RV FAC/mean PAP or TAPSE/mean PAP). Although the correlations between pPTT and invasive markers of pulmonary vascular disease were modest, RV FAC (r = 0.64, P < 0.0001), TAPSE (r = 0.67, P < 0.0001), and RV-PA coupling (RV FAC/mean PAP: r = 0.72, P < 0.0001; TAPSE/mean PAP: r = 0.74, P < 0.0001) had the strongest relationships with pPTT. On multivariable analysis, only RV FAC, TAPSE, and RV-PA coupling were independent predictors of pPTT. We conclude that shortening of pPTT in patients with PAH results from altered RV-PA coupling, probably occurring as a result of reduced pulmonary arterial compliance. Thus, pPTT allows noninvasive determination of the status of both the pulmonary vasculature and the response of the RV in patients with PAH, thereby allowing monitoring of disease progression and regression.

Pulmonary arterial hypertension (PAH) is a lethal disease, characterized by pathological pulmonary vascular remodeling, that leads to elevated pulmonary arterial pressure (PAP) and pulmonary vascular resistance (PVR), decreased pulmonary arterial compliance, and eventual right ventricular failure and death.^1,2 Growing evidence suggests that pulmonary arterial compliance plays a critical role in the pathogenesis of PAH and right ventricular failure.³ Decreased pulmonary arterial compliance causes premature reflection of waves from the distal pulmonary vasculature, leading to increased pulsatile right ventricular afterload and altered right ventricular–pulmonary artery (RV-PA) coupling.^3-5 Both loss of pulmonary arterial compliance and impaired RV-PA coupling are clinically important, because they are associated with increased mortality in patients with pulmonary hypertension.^6-11

Pulse wave velocity, defined as the velocity of pressure waves traveling through the arterial system, correlates inversely with arterial compliance in the systemic circulation.¹² With decreasing compliance, pulse wave velocity increases, and the time taken by the pressure wave to travel between 2 vascular points (pulmonary pulse wave transit time [pPTT]) shortens.¹³ Both pulse wave velocity and pulse transit time have been identified as clinically important surrogates for assessing arterial compliance in the systemic circulation.¹² However, the relationship between pulse wave velocity and arterial compliance in the pulmonary circulation has not been examined in detail, probably due to the technical difficulties of measuring pulse wave velocity in the pulmonary circulation.^14,15

Recently, Wibmer et al.¹⁶ showed that pPTT, defined as the delay between onset of ventricular electrical activity (on electrocardiogram) and the arrival of the pulse wave in the pulmonary vein (as determined by Doppler echocardiography of the pulmonary vein), was reduced in 6 patients with World Health Organization (WHO) group 1 pulmonary hypertension and 6 patients with WHO group 3 pulmonary hypertension. Wibmer et al.¹⁶ postulated that pPTT was shortened because of decreased pulmonary artery compliance and increased pulse wave velocity, and they proposed pPTT as a potential index of pulmonary arterial compliance; however, a definitive mechanism for reduction in pPTT was not demonstrated. Here, we sought to identify the determinants of pPTT in PAH. We tested the hypothesis that the shortening of the pPTT in patients with PAH results from reduced RV-PA coupling, likely occurring as a result of reduced pulmonary arterial compliance.^3-5

Methods

Study cohort

We studied patients in the Minnesota Pulmonary Hypertension Repository, which is a customized patient database created to collect specific variables for every patient treated at the University of Minnesota Pulmonary Hypertension Clinic (from March 2014 to the present). Data are collected by chart review and entered using an internet-based electronic data-capture system. Patients who received a diagnosis before March 2014 were entered retrospectively. All patients gave informed consent for participation in the repository. The University of Minnesota institutional review boards approved the Minnesota Pulmonary Hypertension Repository.

For this study, we identified all adult patients (≥18 years of age) with PAH in the Minnesota Pulmonary Hypertension Repository. In brief, the diagnosis of PAH required the following: (1) a mean PAP (mPAP) ≥25 mmHg at rest, with a pulmonary capillary wedge pressure of <15 mmHg, and a PVR >3 Wood units; and (2) the exclusion of other WHO categories of pulmonary hypertension by clinical evaluation and objective tests, including pulmonary function tests and ventilation-perfusion scan. Patients were excluded if they had obstructive lung disease diagnosed by reduced expiratory flow rates (forced expiratory volume in one second [FEV1]/forced vital capacity [FVC] <75% predicted); more than mild interstitial lung disease diagnosed by reduced total lung capacity <60%; chronic pulmonary thromboembolic disease diagnosed by ventilation perfusion scan (high or intermediate probability), contrast-enhanced chest computed tomography, or pulmonary angiography, if necessary.

We enrolled patients who had both a right heart catheterization (RHC) and a transthoracic echocardiogram with interpretable pulse wave Doppler of the right ventricular outflow tract (RVOT) and the pulmonary vein within a 6-month time interval (median interval in our cohort was 17.5 days). We analyzed the following variables for all patients identified (n = 36): baseline demographic characteristics, clinical characteristics, laboratory values, hemodynamic variables, and selected echocardiographic variables. We also included an age- and sex-matched control group (n = 10) of individuals who had an echocardiogram for clinical concerns but were found to have no evidence of pulmonary hypertension or structural heart disease. Their echocardiograms were defined as having normal findings on the basis of the presence of normal left ventricular and right ventricular size, thickness, and function with no significant valvular abnormalities and normal estimated right ventricular systolic pressures (Table 1).

Table 1.

Characteristics of control patients

Variable	Value (N = 10)
Age, years	58 ± 21
Female sex	6 (60)
Hypertension	4 (40)
Diabetes	3 (30)
Creatinine level, mg/dL	0.92 ± 0.3

^NoteData are presented as mean ± standard deviation or no. (%).

pPTT

Echocardiography was performed using a commercially available Philips system (iE33, Philips Ultrasound, Bothell, WA) with a 3.5-MHz multiphase array probe. pPTT was defined as the time taken by the pressure wave to travel from the pulmonic valve/RVOT to the pulmonary vein (Table S1).¹⁶ RVOT flow was studied using the pulsed-wave Doppler with the Doppler sample volume placed in the main pulmonary artery just distal to the pulmonic valve in the left parasternal, short-axis view, and the pulmonary vein flow was studied by the pulsed-wave Doppler with the Doppler sample volume placed in the right inferior pulmonary vein in the apical 4-chamber view, according to guidelines of the American Society of Echocardiography (Fig. 1).¹⁷ All the Doppler recordings were made at a sweep speed of 75 mm/s with a simultaneous superimposed electrocardiogram. We measured the time interval between the R-wave in the electrocardiogram and the corresponding onset of RVOT pulse Doppler flow velocity (R-RVOT interval; Fig. 1A) for 2 to 3 consecutive cardiac cycles and calculated the mean. We then measured the time interval between the R-wave in the electrocardiogram and the corresponding peak late-systolic pulmonary vein flow velocity (R-PVs2 interval; Fig. 1A) for 2 to 3 consecutive cardiac cycles and calculated the mean. This late-systolic pulmonary vein flow pulse wave has been shown to be closely related to the forward-traveling pressure wave originating from the right ventricle.¹⁸ In 2 patients, these time intervals were available for only 1 cardiac cycle. We then calculated pPTT as the difference between R-PVs2 and R-RVOT intervals, normalized to cardiac cycle length (Fig. S1). We normalized pPTT to cardiac cycle length, because shorter cardiac cycles permit less time between pulse waves. Thus, although pPTT is a measure of time, it is unitless when normalized to cardiac cycle length (Fig. S1). Twenty studies were evaluated by 2 independent reviewers to assess intra- and interobserver variability.

Figure 1.

Reduced pulmonary pulse wave transit time (pPTT) in patients with pulmonary arterial hypertension (PAH) determined using echocardiography. Shown are examples of determination of pPTT from a control patient (A) and a patient with PAH (B). Pulse wave Doppler interrogation of right ventricular outflow track (RVOT; upper boxes) and pulmonary veins (PVs; lower boxes). C, Quantification of pPTT (mean ± standard error of the mean) from 10 control patients and 36 patients with PAH. Asterisk indicates P < 0.0001. R-PVs2: time interval between the R-wave in the electrocardiogram and the corresponding peak late-systolic PV flow velocity; R-RVOT: time interval between the R-wave in the electrocardiogram and the corresponding onset of RVOT pulse Doppler flow velocity.

RHC

Patients underwent RHC, as clinically indicated, in the University of Minnesota cardiac catheterization laboratory. Hemodynamic characteristics were obtained using a 7 Fr, balloon-tipped, flow-directed catheter placed into either the internal jugular vein or the common femoral vein. The following hemodynamic variables were recorded at the end of expiration: right atrial pressure, right ventricular systolic and end-diastolic pressures, systolic PAP, diastolic PAP, mPAP, and pulmonary capillary wedge pressure. Cardiac output was determined as the mean of 3 measurements with the thermodilution method or indirect Fick method on the basis of total body oxygen consumption estimated via the formula of LaFarge and Miettinen.¹⁹ Transpulmonary gradient was calculated as the difference between mPAP and pulmonary capillary wedge pressure (Table S1). PVR was calculated in Wood units as the difference between mPAP and pulmonary capillary wedge pressure divided by the cardiac output (Table S1). Diastolic pulmonary gradient (DPG) was calculated as the difference between diastolic PAP and pulmonary capillary wedge pressure (Table S1). Pulmonary arterial compliance (mL/mmHg) was calculated as the ratio of stroke volume to the pulmonary artery pulse pressure (Table S1), as described elsewhere.⁶

Right ventricular function and RV-PA coupling

Right ventricular size was semiquantitatively described as normal size (two-thirds or less of the left ventricular size) or as mildly (right ventricle similar to the left ventricle size), moderately (right ventricle larger than the left ventricle), or severely (right ventricle much larger than the left ventricle) enlarged, as described elsewhere.²⁰ We also quantified right ventricular function by calculating right ventricular fractional area change (RV FAC) and tricuspid annular plane systolic excursion (TAPSE).¹⁷ Right ventricular function was categorized as reduced if RV FAC was <35%.²¹ If patients did not have m-mode of the tricuspid annulus recorded, we measured TAPSE by postprocessing of 2-D 4-chamber images using a Java-based imaging software program (Image J, National Institutes of Health) as described elsewhere.²²

We estimated RV-PA coupling in all patients by two methods: RV FAC divided by invasively measured mPAP (RV FAC/mPAP) and TAPSE divided by invasively measured mPAP (TAPSE/mPAP) as described elsewhere.^23-26 In a subgroup of patients (n = 8) who had cardiac magnetic resonance imaging (MRI), we also estimated RV-PA coupling using cardiac MRI-based volume method defined as end-systolic volume divided by stroke volume.^27,28Table S1 lists the 3 formulae used to estimate RV-PA coupling.

Statistical analysis

Categorical data are expressed as frequency and proportions, whereas continuous data are presented as mean ± standard deviation unless otherwise stated. Unpaired t tests were used to compare means for normally distributed continuous variables between 2 groups, whereas χ² or Fisher exact test were performed to compare proportions for categorical variables. To better understand the relationship between pPTT and the various demographic, clinical, laboratory, and hemodynamic variables, we performed univariate and multivariable linear regression analyses with pPTT as the dependent variable. From the univariate analysis of variables associated with pPTT at P < 0.1, we determined that the following covariables should be entered into our multivariate model: age, diabetes, RV FAC, TAPSE, mPAP, transpulmonary gradient, DPG, PVR, RV FAC/mPAP, and TAPSE/mPAP. Because RV FAC, TAPSE, RV FAC/mPAP, and TAPSE/mPAP were highly correlated, we created 4 separate multivariable models. Model 1 included age, diabetes, RV-PA coupling defined as RV FAC/mPAP, mPAP, transpulmonary gradient, DPG, and PVR. Models 2, 3, and 4 included the same variables, except that the RV FAC/mPAP in model 1 was replaced with RV FAC, TAPSE, or TAPSE/mPAP, respectively. Right ventricular systolic pressure, systolic PAP, and diastolic PAP were highly correlated with mPAP; therefore, only mPAP was retained in the multivariate models to avoid multicollinearity. Pearson correlation was used to compare pPTT with invasively measured hemodynamic parameters. All statistical analyses were performed using Stata software, version 13 (Stata, College Station, TX).

Results

We measured pPTT in 10 healthy controls (Table 1) and 36 patients with PAH (Table 2). Median duration between echocardiogram and RHC was 17.5 days (interquartile range: 2.5–61.5 days). There was a significant reduction in pPTT in patients with PAH (0.164 ± 0.01) when compared with control patients (0.301 ± 0.02, P < 0.0001; Fig. 1C). The distribution of pPTT in controls and patients with PAH is depicted in Figure 2.

Table 2.

Baseline demographic, clinical, and hemodynamic characteristics of patients with pulmonary arterial hypertension (PAH)

Variable	Total cohort (N = 36)
Age, years	58 ± 17
Female sex	25 (69)
Etiology of PAH:
Idiopathic	11 (31)
Connective tissue disease	12 (33)
Congenital	7 (19)
Liver disease	6 (17)
Comorbidities:
Hypertension	15 (42)
Diabetes	7 (19)
Coronary artery disease	3 (8 )
Laboratory values:
Serum creatinine, mg/dL	1.1 ± 1.1
Hemoglobin, g/dL	13.1 ± 2.5
NT pro-BNP, pg/mL	5,334 ± 12,112
Echocardiographic analysis:
LV EF, %	58 ± 6
TR maximum pressure gradient, mmHg	59 ± 20
RA enlargement	26 (84)
Mild-moderate RV dilation	16 (62)
Severe RV dilation	6 (23)
RV dysfunction	15 (42)
RV FAC, %	32 ± 12
TAPSE, cm	1.7 ± 0.5
Invasive hemodynamic:
Heart rate, bpm	75 ± 16
Mean RA pressure, mmHg	9 ± 5
RV systolic, mmHg	73 ± 24
RV diastolic, mmHg	9 ± 7
PA systolic, mmHg	75 ± 20
PA diastolic, mmHg	32 ± 12
Mean PA, mmHg	47 ± 13
Mean PCWP, mmHg	12 ± 5
Cardiac output, L/min	4.6 ± 1.6
Stroke volume, mL	65 ± 27
Compliance, mL/mmHg	1.7 ± 0.9
TPG, mmHg	34.5 ± 10.6
DPG, mmHg	19.5 ± 9.1
PVR, Wood units	8.4 ± 4.8
RV FAC/mPAP, %/mmHg	0.8 ± 0.4
TAPSE/mPAP, cm/mmHg	0.04 ± 0.02

^NoteData are presented as mean ± standard deviation or no. (%). BUN: blood urea nitrogen; DPG: diastolic pulmonary gradient; EF: ejection fraction; LV: left ventricle; mPAP: mean pulmonary artery pressure; NT pro-BNP: N-terminal pro-brain natriuretic peptide; PA: pulmonary artery; PCWP: pulmonary capillary wedge pressure; PVR: pulmonary vascular resistance; RA: right atrium; RV: right ventricle; RV FAC: right ventricular fractional area change; TAPSE: tricuspid annular plane systolic excursion; TPG: transpulmonary gradient; TR: tricuspid regurgitation.

Figure 2.

Distribution of pulmonary pulse transit time (pPTT) in controls and patients with pulmonary arterial hypertension (PAH). Most patients with PAH had lower pPTT when compared with the median pPTT of the age- and sex-matched control group. pPTT in the X-axis has no units, because it is normalized to the cardiac cycle length in seconds using the formula (time from R-wave to peak S-wave on pulmonary vein Doppler − time from R-wave to onset of right ventricular outflow track Doppler)/cardiac cycle length.

Next, we performed linear regression analysis to delineate determinants of pPTT. Age, RV FAC, TAPSE, right ventricular systolic pressure, PAP, DPG, transpulmonary gradient, PVR, RV FAC/mPAP, and TAPSE/mPAP were identified as significant predictors of pPTT in the univariate analysis (Table 3).

Table 3.

Univariate analysis of determinants of pulmonary pulse wave transit time

	Univariate analysis
Variable	Coefficient (95% CI)	P value
Age	0.002 (0.0005–0.003)	0.009
Sex	0.02 (−0.04 to 0.04)	0.54
BMI	0.002 (−0.002 to 0.006)	0.31
Hypertension	0.03 (−0.03 to 0.08)	0.30
Diabetes	0.06 (−0.004 to 0.12)	0.07
CAD	0.009 (−0.09 to 0.1)	0.85
Serum creatinine level	0.01 (−0.01 to 0.03)	0.40
Hemoglobin level	−0.007 (−0.02 to −0.004)	0.21
LV ejection fraction	0.003 (−0.002 to 0.007)	0.29
TR maximum pressure gradient	0.00003 (−0.001 to 0.001)	0.96
RV enlargement	−0.03 (−0.1 to 0.03)	0.28
RV dysfunction	−0.05 (−0.1 to −0.01)	0.12
RA enlargement	−0.03 (−0.1 to 0.04)	0.39
RV FAC	0.004 (0.002–0.005)	<0.0001
RV FAC/mPAP	0.12 (0.08–0.16)	<0.0001
TAPSE	0.09 (0.06–0.13)	<0.0001
TAPSE/mPAP	2.8 (1.9–3.6)	<0.0001
RA pressure	−0.003 (−0.008 to 0.003)	0.37
RV systolic	−0.001 (−0.002 to 0.00003)	0.04
RV diastolic	−0.001 (−0.005 to 0.002)	0.46
PA systolic	−0.001 (−0.003 to −0.0002)	0.02
PA diastolic	−0.003 (−0.005 to −0.0006)	0.01
Mean PA pressure	−0.003 (−0.005 to −0.0008)	0.006
Mean PCWP	−0.002 (−0.005 to 0.002)	0.35
Cardiac output	0.006 (−0.01 to 0.023)	0.43
Stroke volume	0.6 (−0.7 to 1.9)	0.37
Compliance	20.7 (−14.3 to 55.8)	0.24
PVR	−0.006 (−0.01 to −0.0004)	0.04
TPG	−0.004 (−0.006 to −0.002)	0.001
DPG	−0.005 (−0.008 to −0.003)	0.0002

^NoteBMI: body mass index, defined as weight in kilograms divided by the square of height in meters; CAD: coronary artery disease; CI: confidence interval; DPG: diastolic pulmonary gradient; LV: left ventricle; mPAP: mean pulmonary artery pressure; PA: pulmonary artery; PCWP: pulmonary capillary wedge pressure; PVR: pulmonary vascular resistance; RA: right atrium; RV: right ventricle; RV FAC: right ventricular fractional area change; TAPSE: tricuspid annular plane systolic excursion; TPG: transpulmonary gradient; TR: tricuspid regurgitation.

Then, we conducted a correlational analysis between pPTT and invasive measures of pulmonary vascular disease to determine whether pulmonary vascular disease could explain shortened pPTT in PAH. We observed modest but statistically significant inverse correlations between pPTT and DPG (r = −0.61, P = 0.0002; Fig. 3A), transpulmonary gradient (r = −0.55, P = 0.001; Fig. 3B), PVR (r = −0.38, P = 0.04; Fig. 3C), and mPAP (r = 0.45, P = 0.006; Fig. 3D). Pulmonary arterial compliance (r = 0.24, P = 0.24) was not significantly associated with pPTT (Table 3).

Figure 3.

Moderate correlation between pulmonary pulse transit time (pPTT) and invasively measured markers of pulmonary vascular disease. Scatterplots showing moderate negative correlations between pPTT and diastolic pulmonary gradient (DPG; r = −0.61, P = 0.0002; Fig. 3A), transpulmonary gradient (TPG; r = −0.55, P = 0.001; Fig. 3B), pulmonary vascular resistance (PVR; r = −0.38, P = 0.04; Fig. 3C), and mean pulmonary arterial pressure (mPAP; r = −0.45, P = 0.006; Fig. 3D).

Subsequently, we analyzed how pPTT correlated with markers of RV-PA coupling. There were highly significant and moderate to strong correlations between pPTT and RV FAC (r = 0.64, P < 0.0001; Fig. 4A) and TAPSE (r = 0.67, P < 0.0001; Fig. 4B). Strong associations between RV-PA coupling and pPTT as estimated by RV FAC/mPAP (r = 0.72, P < 0.0001; Fig. 4C) and TAPSE/mPAP (r = 0.76, P < 0.0001; Fig. 4D) were observed. There was a strong and significant correlation between pPTT and MRI-derived RV-PA coupling (r = 0.78, P = 0.023).

Figure 4.

Pulmonary pulse transit time (pPTT) is most strongly associated with markers of right ventricular–pulmonary artery (RV-PA) coupling in pulmonary arterial hypertension (PAH). Scatterplots showing correlation between pPTT and right ventricular fractional area change (RV FAC; r = 0.64, P < 0.0001; Fig. 4A), tricuspid annular plane systolic excursion (TAPSE; r = 0.64, P < 0.0001; Fig. 4B), RV-PA coupling as defined by RV FAC/mean pulmonary arterial pressure (mPAP; r = 0.72, P < 0.0001; Fig. 4C), and RV-PA coupling as defined by TAPSE/mPAP (r = 0.74, P < 0.0001).

To further probe the relationship of pPTT and measures of RV-PA coupling, we performed multivariable analysis. After correcting for age, diabetes, mPAP, transpulmonary gradient, DPG, and PVR, the associations between pPTT and RV FAC, TAPSE, and RV-PA coupling defined by RV FAC/mPAP and TAPSE/mPAP remained significant (Fig. 5).

Figure 5.

Forest plot of multivariate analysis of pulmonary pulse transit time (pPTT) in pulmonary arterial hypertension (PAH). Right ventricular fractional area change (RVFAC; A), tricuspid annular plane systolic excursion (TAPSE; B), and RV-PA coupling defined as RV FAC/mean pulmonary artery pressure (mPAP; C) and TAPSE/mPAP (D) were significantly associated with pPTT after correcting for pulmonary vascular resistance (PVR), diastolic pulmonary gradient (DPG), transpulmonary gradient (TPG), mPAP, presence of diabetes, and age. P values are listed to the right. Data are presented as coefficient with 95% confidence interval.

Intra- and intervariability in pPTT measurements

Reproducibility of pPTT was acceptable with correlation coefficients of 0.96 for intraobserver variability and 0.98 for interobserver variability. Mean intraobserver difference was 0.04 ± 0.12 and the mean interobserver difference was −0.06 ± 0.17 when the reviewers were blinded to previous measurements.

Discussion

In this study, we show that pPTT is reduced in 36 patients with PAH, which is consistent with the data from 6 patients with PAH reported by Wibmer et al.¹⁶ We also demonstrate pPTT has modest correlation with invasively measured markers of pulmonary vascular disease, which suggests pPTT likely reflects changes in the precapillary structures. We show pPTT has moderate to strong correlations with RV FAC, TAPSE, and RV-PA coupling (Fig. 4). Importantly, on multivariable analyses, RV FAC, TAPSE, and RV-PA coupling were the only independent predictors of pPTT. Taken together, these results suggest that the abbreviated pPTT in PAH is due to altered RV-PA coupling and that pPTT is a clinically useful surrogate marker for RV-PA coupling in PAH.

Although pulse wave velocity has been shown to predict systemic vascular compliance,¹² we did not observe a significant relationship between pPTT and pulmonary arterial compliance. The exact reason for this is unclear. One possible reason could be that compliance in the pulmonary circuit is distributed differently than in the systemic circuit. Although the aorta accounts for 80% of compliance in the systemic circulation, the proximal vessels in the pulmonary circuit account for only 15%–20% of compliance.⁴ Therefore, the differential distribution of compliance may not manifest as changes in pPTT in PAH, a syndrome in which much of the structural remodeling is in smaller, resistance-level arteries. Another probable explanation is that, in PAH, compliance has a negative hyperbolic relationship with PVR.^29,30 Thus, pulmonary arterial compliance decreases significantly even when PVR is normal or mildly increased. This concept has been reinforced by recent experimental data.³¹ Most of the patients with PAH in this study had advanced pulmonary vascular disease with significantly elevated PVR (mean PVR ± SD: 8.4 ± 4.8 Wood units) and low pulmonary vascular compliance (1.7 ± 0.9 mL/mmHg). This may have prevented us from detecting a relationship because of the relatively small distribution of compliance in the population we studied. Furthermore, although pulse wave velocity is related to arterial compliance in the systemic circulation, there is a greater ability for the left ventricle to adapt to changes in arterial compliance. However, as the right ventricle is very sensitive to changes in pulmonary arterial compliance, small changes in compliance can result in large changes in right ventricular function. Thus, the relationship between pulmonary arterial compliance and pulse wave velocity may be revealed through RV-PA uncoupling in the pulmonary circulation, and assessing RV-PA coupling can provide a better assessment than compliance alone.

Our data suggest that pPTT is a surrogate measure of RV-PA coupling. In our analysis, measures of RV-PA coupling were the only independent predictors of pPTT, which results largely from increased pulsatile afterload due to reduced pulmonary arterial compliance and occurs later in the course of the disease.⁵ Moreover, pPTT was associated with RV FAC and TAPSE, which were shown to correlate well with invasively measured RV-PA coupling.³² Interestingly, there was a strong negative relationship between RC constant (resistance compliance product) and pPTT. When we separated our cohort by median pPTT, the patients with the shortest pPTT had a higher RC constant (high group: 9.5 ± 2.6; low group: 12.6 ± 3.3; P = 0.013), providing more evidence of a complex relationship between pPTT and pulmonary arterial compliance. This would also suggest that patients with advanced pulmonary vascular disease and RV-PA uncoupling have a shorter pPTT.

Although the only variable that was associated with pPTT on multivariable analysis was RV-PA uncoupling, there was a strong relationship between pPTT and DPG, a flow-independent measure of pulmonary vascular disease. Of note, DPG had the strongest correlation with pPTT when compared with other hemodynamic markers of intrinsic pulmonary vascular disease, including transpulmonary gradient, PVR, and pulmonary vascular compliance. Thus, short pPTT is also a marker of intrinsic pulmonary vascular disease. Because we studied only patients with PAH where the vascular remodeling is restricted to the precapillary portion of the pulmonary circulation, we believe that pPTT predominantly reflects the precapillary compartment of the pulmonary circulation.

At present, both diagnosis and subsequent evaluation of therapeutic efficacy in PAH are primarily assessed invasively via RHC. RHC rightfully remains the gold standard for diagnosing PAH; however, due to the low prevalence of the disease and the cost and risk of invasive catheterization, there is a need for noninvasive screening tools to identify and serially assess patients with PAH and hopefully thereby reduce diagnostic delay and allow monitoring of therapy. Although early studies using echocardiography to estimate the pulmonary arterial systolic pressure were promising, because there were strong correlations between estimated and actual pressures,^33,34 Fisher et al.³⁵ recently showed that Doppler-derived pressures were inaccurate in 48% of patients with pulmonary hypertension. Moreover, estimated pulmonary pressures can also be elevated because of increased flow (i.e., high cardiac output) or elevated wedge pressure in the absence of intrinsic pulmonary vascular disease. Thus, there is a need for better noninvasive measures of intrinsic pulmonary vascular disease, right ventricular function, and RV-PA coupling, which are central to the pathogenesis of PAH.

Our results suggest that pPTT is a useful echocardiographic parameter for assessment of RV-PA coupling. The adequacy of functional adaptation of the right ventricle to the afterload (RV-PA coupling) has been associated with survival in patients with PAH.¹¹ Thus, pPTT could be another way to risk stratifying patients with PAH. Current guidelines recommend classifying both patients with incident PAH and those with prevalent PAH as being at high or low risk on the basis of various clinical, imaging, and hemodynamic parameters to guide therapeutic decisions.³⁶ pPTT, a simple, noninvasive measurement, could be added to this armamentarium of risk stratification in PAH after additional validation in external cohorts.

Furthermore, pPTT is another useful echocardiographic parameter for assessment of pulmonary vascular disease in PAH. The first measurement of pulmonary hypertension shown to reflect pulmonary vascular disease was the pulmonary artery acceleration time (PA-AT). Hattle et al.³⁷ showed PA-AT predicted mean and systolic PAP accurately. PA-AT also correlated strongly with PAP in the monocrotaline rat model of PAH.³⁸ Abbas et al.³⁹ showed PVR could be accurately estimated using the ratio of peak tricuspid velocity to the right ventricular outflow time-velocity integral. However, when the Abbas formula was applied to patients with a PVR >8 Wood units, the relationship was no longer significant.⁴⁰ Recently, Opotowsky et al.⁴¹ showed a very strong correlation between invasively measured PVR and estimated PVR using a model based on the ratio of Doppler-estimated systolic PAP to RVOT velocity time integral and the presence of RVOT Doppler midsystolic notching on echocardiogram. Interestingly, the Opotowsky method had a superior receiver-operating characteristic curve for discrimination of PVR >3 Wood units compared with the Abbas method.⁴¹

Finally, Choi et al.⁴² showed that Definity dye transit time from peak opacification of right ventricle to first appearance in the left ventricle could be used to estimate PVR and cardiac output. In that study, the transit time for the contrast to travel from the right ventricle to the left ventricle was directly proportional to PVR and inversely proportional to cardiac output. Importantly, the transit time described by Choi et al.⁴² is a measure of blood flow, whereas pPTT is a surrogate measure of pressure pulse wave velocity. Consistent with this, and unlike the dye transit time, pPTT was inversely proportional to PVR and directly proportional to right ventricular function. Moreover, pPTT was much shorter in duration, with times ranging from approximately 70 ms to 350 ms, whereas the dye transit times in the Choi study were approximately 500 ms to 6 s. Thus, our method and the Choi method are measuring 2 different variables and therefore cannot be directly compared.

Although all of the above-mentioned methods show promise as noninvasive assessments in PAH, more studies are needed to determine their widespread validity both individually and possibly in combination. Moreover, characterization of changes in these noninvasive measures associated with therapy would also be useful to further substantiate their utility. Studies are needed to assess the utility of pPTT as a screening tool in assessing the pulmonary vasculature in high-risk populations, such as scleroderma patients, family members of patients with PAH, those with genetic mutations (e.g., mutations of BMPR2) but without overt PAH, or patients with cirrhosis.

Limitations

Our study was limited by the retrospective and single-center design. Although it is the largest study to date of pPTT in PAH, it is nonetheless a relatively small study. Furthermore, pulse wave Doppler of the pulmonic valve/RVOT and the right inferior pulmonary vein were not collected simultaneously, which may have affected the results. Moreover, echocardiographic examinations and RHCs were not done simultaneously, which could have affected our results. Finally, determination of pPTT was limited by availability and interpretability of pulmonary vein Doppler signals. Pulmonary vein Doppler examination was not routinely performed in all patients with pulmonary hypertension who underwent a diagnostic transthoracic echocardiogram at our center, and in some patients, the pulmonary vein Doppler images was not interpretable.

Conclusions

In conclusion, pPTT was reduced in PAH and strongly correlated with measures of RV-PA coupling. pPTT allows noninvasive determination of both the status of the pulmonary vasculature and the response of the RV in patients with PAH, thereby allowing monitoring of disease progression and regression.

Appendix.

Figure S1.

Method of pulmonary pulse wave transit time (pPTT) measurement. pPTT was calculated as time from R-wave to peak S-wave on pulmonary vein Doppler (R-PVs2) minus time from R-wave to onset of right ventricular outflow tract Doppler (R-RVOT) divided by cardiac cycle length. EKG: electrocardiogram.

Table S1.

List of formulas employed

Variable	Formula
pPTT	(Time from R-wave to peak S-wave on pulmonary vein Doppler − time from R-wave to onset of RVOT Doppler)/cardiac cycle length
RV-PA coupling	RV fractional area change/mPAP TAPSE/mPAP CMRI derived end-systolic volume/stroke volume
PVR	(Mean PA pressure − pulmonary capillary wedge pressure)/cardiac output
DPG	PA diastolic pressure − pulmonary capillary wedge pressure
TPG	Mean PA pressure − pulmonary capillary wedge pressure
PAC	Stroke volume/PA pulse pressure

^NoteCMRI: cardiac magnetic resonance imaging; DPG: diastolic pulmonary gradient; mPAP: mean pulmonary artery pressure; PA: pulmonary artery; PAC: pulmonary arterial compliance; PVR: pulmonary vascular resistance; RV: right ventricle; RVOT: right ventricular outflow track; TAPSE: tricuspid annular plane systolic excursion; TPG: transpulmonary gradient.

Source of Support: KWP was funded by National Institutes of Health (NIH) F32 grant HL129554. TT was funded by American Heart Association Scientist Development grant 15SDG25560048. SA was supported by a Canadian Institutes of Health Research Foundation grant, NIH-RO1-HL071115, 1RC1HL099462, a Tier 1 Canada Research Chair in Mitochondrial Dynamics, and the William J. Henderson Foundation.

Conflict of Interest: None declared.

Supplements

• Appendix ^{(362.7KB, pdf)}