Abstract

Introduction

Assessment of right ventricular (RV) and pulmonary arterial (PA) function is critical for monitoring disease progression and identifying successful treatment strategies in patients with pulmonary arterial hypertension (PAH). In particular, metrics of RV-PA hemodynamic coupling and PA compliance have demonstrated better prognostic value than traditional hemodynamic measurements used in the clinical setting, such as mean pulmonary arterial pressure (mPAP) and pulmonary vascular resistance.[1–4] Furthermore, the response of the RV and pulmonary vasculature to exercise may be even more revealing.[5, 6] However, changes in hemodynamic coupling efficiency and PA compliance with exercise have been under-investigated, particularly in subjects with PAH.

In healthy subjects performing physical exercise, vascular resistance typically stays constant or decreases due to pulmonary arterial distension and recruitment of normally collapsed arteries.[7] The effect of exercise on resistance in patients with PAH have been inconsistent.[8–10] Furthermore, it is unclear how PA compliance changes during exercise.[11, 12] The resistance-compliance characteristic time (RC time), which is constant under most conditions, decreases during exercise in healthy subjects, while the results for PAH patients have not been definitive.[12] A more comprehensive measure of cardiopulmonary status is ventricular-vascular coupling efficiency.[13] Coupling efficiency at rest is decreased in PAH patients,[14, 15] and recent findings indicate that uncoupling occurs during exercise.[16] However, the associations of coupling efficiency at rest with vascular stiffness and maximum exercise level have not been investigated.

Here, a novel methodology was used to investigate the effect of exercise on RV afterload, including proximal PA stiffness (by characteristic impedance Z_C), proximal and distal PA compliance (by C_PA), and total pulmonary vascular resistance (by Z₀), in subjects with PAH. Simultaneous PA pressure and flow measured with echocardiography during right heart catheterization (RHC) and exercise were analyzed. The hypothesis tested was that hemodynamic coupling efficiency at rest associates with proximal PA stiffness and maximum exercise level tolerated.

Methods

Study population and design

Subjects with known or suspected PAH who were referred for clinical RHC at Northwestern Memorial Hospital or University of Wisconsin-Hospital were invited to participate. Inclusion criteria were subjects aged 18–80 with diagnosed or suspected PAH, including idiopathic PAH (IPAH), PAH associated with systemic sclerosis (SSc-PAH) or chronic thromboembolic pulmonary hypertension (CTEPH). Exclusion criteria included World Health Organization (WHO) functional class IV patients, as well as a contraindication to either MRI or physical exercise, or recent syncope.

The study included simultaneous RHC and echocardiography at rest as well as during one or more stages of incremental exercise. MRI at rest was performed within 2 weeks of the RHC. All study procedures are summarized in Table 1, which includes the number of subjects who completed each test and the primary endpoints for each test. All patients gave written informed consent. The study was HIPAA-compliant and approved by the Institutional Review Boards of both institutions.

Table 1

Summary of the tests performed in this study.

Test	n	Primary endpoints
Clinical RHC
Rest (0 W)	26	Baseline hemodynamics
Research RHC+echo
Rest (0 W)	22	RV pressure, PA pressure, PA flow, LAP
15 W	22	PA pressure, PA flow, LAP
30 W	18	PA pressure, PA flow, LAP
45 W	15	PA pressure, PA flow, LAP
MRI
Rest (0 W)	18	RV volume, EDV, ESV, SV, EF, SV/ESV

Open in a separate window

RV pressure, PA pressure, PA flow, and RV volume data were collected as time-dependent waveforms, averaged over multiple heartbeats.

RHC = right heart catheterization; MRI = magnetic resonance imaging; RV = right ventricle; PA = pulmonary artery; LAP = left atrial pressure; EDV = end-diastolic volume; ESV = end-systolic volume; SV = stroke volume; EF = ejection fraction.

Analysis of pulmonary vascular characteristics

Figure 1A summarizes data collection and analytical methods used to assess pulmonary vascular function at rest and exercise. Synchronized PA pressure (from RHC) and flow (from echocardiography) were used to calculate Z₀, which is a measure of tPVR, and Z_C, which is a measure of proximal PA stiffness.

Open in a separate window

Figure 1

Data analysis flowchart

A) Pulmonary vascular impedance analysis was performed using PA pressure data from RHC and PA flow data from echocardiography, at rest and each exercise level. Not every subject was able to exercise up to 45 W, so the group size decreased with increasing exercise intensity. A frequency-domain analysis of the pressure-flow data was performed. Endpoints included various metrics of impedance, pulmonary arterial compliance, and RC time. B) Pressure-volume loop analysis was performed at rest using RV pressure data from RHC and RV volume data from MRI. A single-beat method was used for this analysis. Endpoints included E_es, E_a and η.

Pulmonary vascular impedance was calculated using frequency-domain analysis as the ratio of PA pressure to flow moduli,[18] and normalized by BSA. Z₀ was taken as the zero-frequency value of the pulmonary vascular impedance modulus, whereas Z_C was computed averaging the 4^th through 10^th harmonic. C_PA was assessed as the ratio of RV-SVI (estimated from echocardiography) to PAPP. The RC time was assessed as the product of C_PA and Z₀. Left atrial pressure (LAP) was estimated from echocardiography as

LAP = 1.9 + 1.24 E/e^′

(1)

as reported previously [19] and recently confirmed to be moderately accurate in patients with PAH.[20]

Results

Data were obtained from a study group of 26 patients. After accounting for patients who either declined MRI or had incomplete hemodynamic data, 22 patient datasets were usable for pulmonary vascular analysis at exercise, and 18 patient datasets were usable for cardiac contractility and RV-PA coupling analysis at rest. The demographics for the 26 patients included in the study are summarized in Table 2. One of the subjects included in this study had severe tricuspid regurgitation.

Table 2

Demographic data of the 26 patients included in the study group.

Variable	n	All subjects Mean ± SD	All subjects Range
Age (years)	26	55.7 ± 10.6	36 – 78
Female (%)	15	58
BSA (m2)	26	1.90 ± 0.28	1.46 – 2.70
Diagnosis (%)
IPAH	7	27
SSc-PAH	13	50
CTEPH	6	23
WHO class
I	1	4
II	11	42
III	14	54
PAH medications (%)
Sildenafil	11	42
Ambrisentan	4	15
Treprostinil	6	23
Tadalafil	3	12
Bosentan	1	4
Flolan	1	4
No treatment*	11	42

Open in a separate window

^*Patients who were not on pulmonary vasodilator therapy presented for RHC for the initial diagnosis of PAH.

BSA = body surface area; IPAH = idiopathic pulmonary arterial hypertension; SSc-PAH = pulmonary arterial hypertension associated with systemic sclerosis; CTEPH = chronic thromboembolic pulmonary hypertension; WHO = World Health Organization.

Effect of exercise on hemodynamics

Hemodynamic data obtained at rest and during exercise are summarized in Table 3. All measured and calculated parameters increased with exercise except LAP, tPVR, tPVRI, and RV-SVI. During RHC, patients were able to perform exercise up to a maximum workload of 46.7 ± 3.5 W, on average.

Table 3

Hemodynamic parameters measured at rest and maximum exercise level.

Variable	n	Rest	Maximum exercise	p-value
Workload	22	-	45.7 ± 3.9	-
RHC + echo	22
sPAP (mmHg)		65.6± 4.6	88.0± 5.9 *	< 0.001
dPAP (mmHg)		29.8± 2.8	37.4±3.0 *	0.0012
mPAP (mmHg)		43.6 ± 3.3	58.8 ± 3.8 *	< 0.001
PAPP (mmHg)		35.9 ± 2.9	51.2 ± 4.1 *	< 0.001
LAP (mmHg)		11.1 ± 0.8	12.2 ± 1.3	0.238
HR (bpm)		80.2 ± 2.7	111.8 ± 2.8 *	< 0.001
CO (L/min)		5.86 ± 0.34	8.41 ± 0.80 *	< 0.001
CI (L/min/m2)		3.02 ± 0.16	4.28 ± 0.35 *	< 0.001
tPVR (WU)		7.91 ± 0.74	8.17 ± 0.83	0.668
tPVRI (WU × m2)		15.3 ± 1.4	15.8 ± 1.6	0.689
CPA (mL/mmHg/m2)		1.32 ± 0.17	0.88 ± 0.10 *	0.0019
RC time (s)		1.00 ± 0.07	0.67 ± 0.05 *	< 0.001
RV-SVI (mL/m2)		38.0 ± 1.9	39.1 ± 3.5	0.659
Z1 index (WU × m2)		3.57 ± 0.38	4.63 ± 0.57 *	0.0047
ZC index (WU × m2)		1.33 ± 0.15	2.31 ± 0.38 *	0.026
MRI	18
RV-ESVI (mL/m2)		75.8 ± 8.9	NA1	-
RV-EDVI (mL/m2)		106.1 ± 8.8	NA1	-
RV-EF		0.32 ± 0.03	NA1	-
RHC + MRI	18
RV-SW (mL mmHg)		2398 ± 296	NA1	-
Ees (mmHg/mL)		0.89 ± 0.13	NA1	-
Ea (mmHg/mL)		1.36 ± 0.25	NA1	-
SV/ESV		0.53 ± 0.09	NA1	-
Pmax/mPAP −1		1.46 ± 0.16	NA1	-

Open in a separate window

¹MRI was only performed at rest so MRI-based calculations are only available at rest.

Values are expressed as mean ± SE.

^*p < 0.05.

At increasing levels of exercise, increases in mean pressure (mPAP) and mean flow (CI) were correlated. When data were normalized using Poon’s adjustment, the slope of the mPAP-CI curve was 8.70 mmHg/(L/min/m²), and the intercept was 18.98 mmHg (R² = 0.764, p < 0.05), as reported in Figure 2. While Z₀ did not change with exercise (Figure 3A), Z_C and C_PA changed significantly: from resting conditions to 45 W, Z_C was nearly doubled (Figure 3B), whereas C_PA decreased by about 21% (Figure 3C). The RC time was 1.00 ± 0.07 s at rest, and progressively decreased with exercise (Figure 3D). Online supplementary figure S1 shows a comparison between Z_C from frequency-domain analysis (as reported in this study) and time-domain analysis.

Open in a separate window

Figure 2

Mean PA pressure-flow relationship during incremental exercise

Pooled data include all patients who completed the exercise protocol, and were adjusted for individual variability following Poon’s method.[22] Pearce’s linear correlation (R² = 0.764) is superimposed on the individual data points.

Open in a separate window

Figure 3

Effect of incremental exercise on measures of RV afterload

A) Pulmonary vascular resistance (Z₀) did not change from rest to 45 W exercise, likely reflecting that in PAH patients the pulmonary vasculature is already fully recruited and distended at rest. B) Characteristic impedance (Z_C) increased significantly from rest to 45 W. The exercise-induced increase in Z_C is consistent with arterial strain-stiffening and vasoconstriction mechanisms, also observed in healthy subjects at incremental exercise.[12] C) PA compliance (C_PA, normalized by BSA) progressively decreased with exercise, consistent with the increase in Z_C. D) The RC characteristic time progressively decreased from rest to 45 W exercise. The resting RC time is consistent with observations in healthy controls.[12] The values obtained in this study are generally larger than in other studies in which PVR instead of tPVR was used to calculated the RC time.[23–26]

Data at the four exercise levels were obtained from 22 subjects (0 and 15 W), 18 subjects (30 W) and 15 subjects (45 W). *p < 0.05, compared to rest (0 W); ^†p < 0.05, compared to 15 W.

Discussion

This study sought to investigate the effect of exercise on PA stiffness in patients with PAH as well as the association of hemodynamic coupling efficiency at rest with exercise ability. One major result was that exercise induced stiffening of pulmonary arteries with no change in vascular resistance. Also, significant associations between coupling efficiency at rest and PA stiffness and maximum exercise level tolerated were demonstrated for the first time.

This is the first study to measure pulmonary vascular impedance in PAH patients during exercise. As a consequence, the findings of increased proximal PA stiffness, quantified by Z_C, and decreased proximal and distal compliance, quantified by C_PA, with exercise in PAH patients, are novel. These results are consistent with a recent report that exercise reduces compliance in both controls and CTEPH patients.[12] Since Z₀ was preserved in the current study, meaning that an increase in CO was associated with an increase in mPAP, PA stiffening with exercise may be the result of (1) strain-stiffening of the proximal PAs due to the increase in mPAP, (2) exercise-induced sympathetic vasoconstriction due to the increase in CO, or (3) a combination of these mechanisms.

In the current study the RC time decreased significantly with exercise and the RC time at rest was significantly lower in low-η patients. The RC time is approximately constant in a wide range of conditions, both in healthy controls and PAH subjects.[23–26] Based on the assumption of a constant RC time, earlier reports postulated a limited contribution of proximal PA stiffness to pulmonary vascular compliance.[27] However, considering the current results as well as other reports that the RC time constant indeed may not always be constant,[12, 26, 28, 29] the role of proximal PA stiffening in PAH progression deserves further investigation, in particular under stress conditions.

The other major finding of the current study was the associations of coupling efficiency at rest with maximum exercise level, resting C_PA, and C_PA decrease during exercise. Reduced coupling efficiency has been documented in patients with PAH.[14, 15] The prognostic value of resting coupling efficiency has been demonstrated.[1] The current study further suggests that high resting η predicts the ability to accommodate an increase in cardiac output during exercise. In fact, high-η PAH patients have better (i.e., higher) C_PA at rest than low-η patients, and during exercise they can reduce C_PA to a greater extent. These results suggest that PAH patients with more efficient hemodynamic coupling also have a greater pulmonary vascular reserve, which can be used to hemodynamic advantage when cardiac output increases, resulting in improved tolerance to exercise compared to low-η patients. These findings are consistent with the comparison between exercise-induced C_PA reduction in CTEPH patients and healthy controls.[12] In contrast with basic hemodynamics typically assessed clinically, resting coupling efficiency showed a significant association with exercise tolerance. This result indicates that the main reason why patients with PAH become progressively unable to tolerate exercise is the decline in the efficiency in the hemodynamic interaction between ventricle and vasculature, rather than either ventricular or vascular function impairment alone. Assessing resting η from MRI and RHC data may have an important clinical value. Simultaneous MRI and invasive RV pressure recordings is an attractive approach, although not yet clinically viable.[30]

The separation between high-η and low-η PAH patients revealed a number of significant differences between the two groups, which may be analogous to differences between healthy subjects and PAH patients. For instance, the association between reduced η and limited exercise capacity has been observed previously in a comparison between healthy subjects and PAH patients.[16] Also, the mPAP-CI curve was about 32% steeper for patients with low-η compared to high-η in our study, although this difference was not significant, which is consistent with an increase in mPAP-CI slope in patients with pulmonary vascular disease compared to those without.[10, 12]

The comprehensive approach to evaluation of cardiopulmonary performance used in this study (RHC and echocardiography during exercise) is viable for research, but it is unlikely to be used clinically. The clinical significance of this study is pulmonary vascular hemodynamic reserve is an important contributor to exercise intolerance in PAH, based on the relationships found between PA stiffening and ventricular-vascular coupling efficiency at rest and during exercise. Importantly, prior to this study it was unclear whether proximal PAs stiffen during exercise independent of changes in resistance.

These results may help guide the clinical adoption of exercise testing protocols, which have been increasingly recommended to monitor PAH progression.[5] Exercise testing protocols generally include measures of PA resistance only (PVR, mPAP-CI curves). This study suggests that the response to exercise of proximal PAs may also need to be monitored, ideally non-invasively using stress echocardiography.

Conclusions

In patients with PAH, exercise-induced stiffening of pulmonary arteries aggravates RV afterload. PA stiffening was unrelated to changes in resistance, resulting in a progressive decrease in the RC time with exercise. Patients with less efficient hemodynamic coupling at rest exhibited a reduced pulmonary vascular reserve, which likely limited the maximum exercise level tolerated. The results of this comprehensive approach to evaluating cardiopulmonary performance may shed light on mechanical mechanisms of PAH progression and exercise tolerance, which in turn may lead to better, less invasive prognostic indicators. In addition, therapeutic strategies targeting the efficiency of hemodynamic coupling may be beneficial to exercise tolerance in PAH patients and improve quality of life.

Supplementary Material

Acknowledgments

Footnotes

References