Assessment of Right Ventricular Afterload in COPD

Abstract

Background: We aimed to study whether pulmonary hypertension (PH) and elevated pulmonary vascular resistance (PVR) could be predicted by conventional echo Doppler and novel tissue Doppler imaging (TDI) in a population of chronic obstructive pulmonary disease (COPD) free of LV disease and co-morbidities.Methods: Echocardiography and right heart catheterization was performed in 100 outpatients with COPD. By echocardiography the time-integral of the TDI index, right ventricular systolic velocity (RVSmVTI) and pulmonary acceleration-time (PAAcT) were measured and adjusted for heart rate. The COPD patients were randomly divided in a derivation (n = 50) and a validation cohort (n = 50). Results: PH (mean pulmonary artery pressure (mPAP) ≥ 25mmHg) and elevated PVR ≥ 2Wood unit (WU) were predicted by satisfactory area under the curve for RVSmVTI of 0.93 and 0.93 and for PAAcT of 0.96 and 0.96, respectively. Both echo indices were 100% feasible, contrasting 84% feasibility for parameters relying on contrast enhanced tricuspid-regurgitation. RVSmVTI and PAAcT showed best correlations to invasive measured mPAP, but less so to PVR. PAAcT was accurate in 90- and 78% and RVSmVTI in 90- and 84% in the calculation of mPAP and PVR, respectively. Conclusions: Heart rate adjusted-PAAcT and RVSmVTI are simple and reproducible methods that correlate well with pulmonary artery pressure and PVR and showed high accuracy in detecting PH and increased PVR in patients with COPD. Taken into account the high feasibility of these two echo indices, they should be considered in the echocardiographic assessment of COPD patients.

Keywords:

• COPD
• pulmonary pressure
• pulmonary vascular resistanceright heart catheterization
• right ventricular afterload
• Tissue Doppler imaging

Introduction

Measurement of mean pulmonary artery pressure (mPAP) and pulmonary vascular resistance (PVR) by right heart catheterization are currently gold standards and reference methods when it comes to definition of pulmonary hypertension (PH) and PVR (1). In clinical practice, systolic pulmonary artery (PA) pressure and its derived mPAP, is calculated by echocardiographic tricuspid regurgitation peak velocity by the Bernoulli equation (2). This method, however, could be hampered by low feasibility and poor accuracy. Moreover, measures of PVR for routine use has not been recommended (3), although several simple non-invasive methods of estimating PVR have been proposed in different patient's populations (4–8).

Previous studies, measuring afterload noninvasively by echocardiography, have mainly been performed in patients with idiopathic PH and to a lesser extent in the largest group of patients with PH; chronic obstructive pulmonary disease (COPD). From 2006 to 2010 we recruited a well-defined group of COPD patients, which previously have been thoroughly described (9–11). Moreover, validation of echo indices against right heart ­catheterization regarding clinical accuracy, in particular in COPD, has hardly been performed.

The main objective of the present study was therefore to investigate the diagnostic ability and accuracy, precision and feasibility of echocardiography in predicting PH and PVR in a cohort of stable COPD patients by established methods, such as pulmonary acceleration time (PAAcT) (12) and by novel tissue Doppler measures. We hypothesized that the TDI indices and a novel heart rate adjusted PAAcT would provide better estimations and be more feasible than the estimation by standard tricuspid jet.

Methods

Study population

The present study is an analysis of a previously described cohort of outpatients enrolled in a cross-sectional study at a secondary referral centre (Oslo University Hospital-Aker) for diagnosis and evaluation of COPD associated pulmonary hypertension, and who prospectively underwent right heart catheterization for the purpose of this study. The cohort was in advance screened with regard to established left sided heart disease and comorbidities, and excluded from participation if such conditions were present. Patients with a history of congenital, rheumatic, valvular and ischemic heart disease, treated arterial hypertension with blood pressure (BP) >160/90 mmHg, arrhythmias (including atrial fibrillation), cardiac surgery, other acute or chronic pulmonary disease, sleep apnea syndrome, malignancy, hyperthyroidism, systemic inflammatory diseases and renal failure, were excluded, as well as the use of beta-blockers, warfarin or clopidogrel.

The present report describes 100 patients with stable, smoke-induced COPD (with ≥ 10–pack-year history of smoking) in age of 40–75 years with different airway severity according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) in stages I to IV (13). Baseline, pulmonary function and arterial blood gas characteristics, invasive- resting and exercise hemodynamic data and selected echocardiographic data on RV myocardial function according to COPD with and without PH has been described in previous reports (9–11). In contrast to previous articles, the present article includes GOLD stage I participants. To identify clinical relevant cutoff values for different echocardiographic markers by receiver operating characteristic (ROC) curve analyses, the COPD cohort was randomly computer divided into equal sized groups; i) a derivation cohort (n = 50), and ii) a validation sample (n = 50). The study complies with the Declaration of Helsinki and was approved by the Local Research Ethics Committee. Written informed consent was obtained from all the participants.

Right heart catheterization

Standard right heart catheterization was performed after transthoracic echocardiography within 120 minutes as previously described (9). Volume loading or sedation was not administered to patients prior to invasive measurements. A Swan–Ganz catheter was used for flow measurements and pressures, which were zeroed in the supine position at the mid-axillary line at the right atrial level. Automatic calibration was done prior to pressure measurements that were performed during temporary breath-hold at end-expiration, verified by flat respiration curve. Cardiac output (L/min) was measured by thermodilution technique. PVR was calculated by equation = (mPAP-pulmonary wedge pressure (PWP))/cardiac output) in Wood unit (WU). A PVR ≥ 2.0 WU was considered abnormal (14).

Doppler-echocardiography

All echocardiograms were obtained by an experienced cardiologist (JMH) using standard techniques and a commercially available system (Vivid Seven, General Electric, Norway). Three consecutive measurements were performed and averaged. Echocardiograms were stored in digital format, and post-processing software (EchoPAC, GE Healthcare, Norway) was used. Echocardiograms were analysed by the same observer (JMH) who was blinded to the clinical status.

The tricuspid valve was interrogated from multiple different views to obtain the most complete Doppler envelope. Agitated saline was administered in all subjects as ordinary bolus (9 ml 0.9% saline and 1 ml air) intravenously to enhance tricuspid regurgitation jet signals, which were used in case of technically limitations (15).

Estimation of systolic PA pressure was based on tricuspid regurgitation peak velocity and calculated by the simplified Bernoulli equation; 4(tricuspid regurgitation peak velocity) ²+ right atrial pressure (RAP). Estimation of RAP was done on the basis of the inferior vena cava diameter and collapse index (3, 16).

According to current guidelines (3), mPAP was estimated by pulsed Doppler, using PA flow acceleration time (PAAcT), defined as time interval between onset of systolic PA flow and peak velocity. Heart rate affects PA acceleration time (17). To avoid false high values due to bradycardia, adjustment was performed: measured PA acceleration time/√RR interval (s) (18) if heart rate < 60, denoted as heart rate-adjusted PA acceleration time (HR-adjusted PAAcT). In addition, mPAP was estimated by a systolic PA pressure estimate: (mPAP = 0.61·systolic PA pressure+2 mmHg) (19) from continuous wave Doppler recordings.

The ratio of peak early transmitral velocity (E) by pulsed Doppler and peak early diastolic mitral annular velocity (Ea) at the medial mitral annulus by pulsed-Doppler TDI was used as a numeric substitute for pulmonary wedge pressure (20).

PVR (WU) was estimated (4) as:

PVR (WU) = Tricuspid regurgitation peak velocity/[(RV outflow tract pulsed Doppler VTI × 10)+0.16]

TAPSE was measured in the apical four-chamber view by aligning the M-mode cursor along the movement of the lateral tricuspid annulus (3).

Tissue Doppler

The tricuspid annular peak systolic myocardial velocity (Sm) and its velocity time integral (SmVTI) were acquired in apical four chamber view with a 2-mm sample volume at the junction of the free wall and the anterior leaflet of the tricuspid valve using pulsed-wave TDI. TDI (21) waveforms were accepted when uniform and consistent with little or no beat-to-beat variation of the peak velocity was observed. The resulting velocities and integrals were analysed at sweep speed of 100 mm/s. RVSmVTI was traced exactly in its outer contour from start to cessation of S-wave (Figure 1), and heart rate adjustment was done, in the same manner, as for PAAcT. Gain was optimized to allow a clear signal with minimal background noise. The time from the beginning of isovolumic contraction to the first systolic peak (Sm) was measured and referred to as time to peak Sm (Tpeak). RV diameter was measured in diastole as the internal annular mediolateral dimension of the RV (RVD). RVD/Tpeak (cm/sec) was calculated (22).

Figure 1. Tracing (dotted line) of right ventricular velocity time integral (RVSmVTI) from tissue Doppler obtained from the lateral tricuspid annulus.

Statistical analysis

Data are expressed as mean ± standard deviation (SD) for normally distributed data and median and interquartile range for not-normally distributed data. Paired Student's t-test was used to compare continuous data. Statistical significance was inferred at p < .05. In the first analyses the regression equations were calculated between mPAP to echo markers and of PVR to echo markers in the 50 patients in the derivation group. In the next step, the algorithm derived from the regression line was used to calculate mPAP and PVR from the echo variables of 50 patients in the validation cohort. Finally, the agreement between the calculated mPAP and PVR values, and the invasive measures from RHC were evaluated using Bland-Altman plots.

Accuracy was predefined as 95% limits of agreement within ±5 mmHg for mPAP and ±1 Wu for PVR. Receiver operating characteristic (ROC) curves were constructed to evaluate the ability of echocardiographic parameters, HR-adjusted PAAcT and RVSmVTI, in addition to RVD/Tpeak, to predict PH. Furthermore, HR-adjusted PAAcT and SmVTI were used to predict elevated PVR. Comparison to clinically established parameters was done, and comparative analyses of area under curve (AUC) for different variables were performed.

ROC curve analysis was performed in the derivation group to establish cutoff values for the detection of i) PH (mPAP ≥ 25 mmHg) and ii) increased PVR (PVR ≥ 2 WU) (1, 14). The cutoff values formed by the derivation group were applied to the test group to evaluate the diagnostic performance of the ROC cutoff values. Diagnostic utility was reported by balanced sensitivity- and specificity, by positive- and negative predictive values and diagnostic odds ratio, calculated as [((sensitivity × specificity)/(1-sensitivity) × (1-specificity)]. Reproducibility data were tested by intraclass correlation coefficient for RVSmVTI, RVSm, Tpeak and RVD1 in a randomly chosen sample of 10 study participants. The measurements were performed by two of the authors (KS and JMH). Statistical analyses were performed using SPSS v.20 and SigmaPlot v.12.

Results

Clinical, hemodynamic, and echocardiographic characteristics are presented in Tables 1 and 2. COPD patients partitioned in derivation and validation group did not differ with regard to clinical, hemodynamic or echocardiographic characteristics. Pulmonary stenosis was not identified, and only trivial and mild pulmonary regurgitation was observed. Significant linear relationships (p < 0.01) were observed between invasive systolic PA pressure vs. mPAP (r = 0.92), and between mPAP and transpulmonary pressure gradient (r = 0.81). A difference of> 10 mmHg in systolic systemic blood pressure and of > 10 beats/min in heart rate between the invasive- and non-invasive data for any given patients were observed only in one patient.

Table 1. Baseline characteristics of study participants

Variables (unit)	COPD (n = 100)
Demographics
Age (years)	63 ± 7
Sex Female/Male (n)	51/49
Body mass index (kg/m^2)	25 ± 5
Body surface area (m^2)	1.8 ± 0.3
Caucasian (%)	100
GOLD stages I-IV (n)	2/37/31/30
Smoking habits^§ (n)	69/31/0
Pack-years of smoking	40 ± 19
Sinus rhythm (%)	100
Atrial arrhythmia (%)	0
Pulmonary function
FEV1/FVC (%)	48 ± 12
FEV1% predicted	43 ± 16
RV % predicted	207 ± 63
DLCO % predicted	53 ± 22
Arterial blood gases
PaO2 (kPa)	9.4 ± 1.5
PaCO2 (kPa)	5.4 ± 0.7

Values are mean ± SD. COPD: Chronic obstructive pulmonary disease; GOLD: Global Initiative for Chronic Obstructive Lung Disease; FEV₁: Forced expiratory volume in the first second; RV: Residual volume: DLCO: Diffusion capacity for carbon monoxide of the lungs; PaO₂: Arterial oxygen tension; PaCO₂: Arterial carbon dioxide tension; 3DE: Three-dimensional echocardiography; RV/ LVEF; Right and Left ventricle ejection fraction.^§ Former, current and never smoker, respectively.

Table 2. Hemodynamic and echocardiographic characteristics in COPD

Variables (unit)	COPD (n = 100)
RHC data
sPAP (mmHg)	31 ± 8
dPAP (mmHg)	11 ± 6
mPAP (mmHg)	21 ± 6
mPWP (mmHg)	9 ± 4
mRAP (mmHg)	6 ± 3
TPG (mmHg)	12 ± 5
CI (l/min/m^2)	2.9 ± 0.5
SVI (ml/m^2)	39 ± 7
PVR (WU)	2.3 ± 1.2
Heart rate at RHC (beats/min)	77 ± 13
Systemic systolic blood pressure (mmHg)	139 ± 21
Proportion of mPAP ≥ 25 mmHg (%)	26
Proportion of PWP > 15 mmHg (%)	0
Proportion of PVR > 2 and > 3 WU (%)	35/19
Echocardiographic data
TAPSE (cm)	2.3 ± 0.4
Systemic systolic blood pressure (mmHg)	138 ± 18
Heart rate at echo (beats/min)	75 ± 14
Proportion of pulmonary regurgitation (%)	4
Left ventricular E/Ea	9 ± 2
Left atrial volume (ml/m^2)	23 ± 5
Left ventricular mass index (g/m^2)	77 ± 21

Values are mean ± SD. Abbreviations: PAP = pulmonary artery pressure; s, d, and m: systolic, diastolic and mean; PWP: Pulmonary wedge pressure; RAP: Right atrial pressure; TPG: Transpulmonary pressure gradient; CI:cardiac index; SVI: indexed stroke volume; PVR: Pulmonary vascular resistance; RHC: Right heart catheterization; TAPSE: Tricuspid annular plane systolic excursion.

Agreement between right heart catheterization and echocardiography

Feasibility of HR-adjusted PAAcT and TDI derived variables were 100% compared to 84% using tricuspid regurgitation peak velocity signals. Only mild-to-moderate tricuspid regurgitation jets were observed.

The derivation group revealed the following linear regression equations between invasive mPAP and the echo variables PAAcT, RVSmVTI, and RVD/Tpeak: y = -0.25x + 49.24, y = -8.67x + 43.47, and y = 0.48x +9.77, respectively. When applied to the validation group, strong correlation and satisfactory agreement with narrow limits of agreement by Bland-Altman plots were observed between echocardiography and right heart catheterization, in particular for PAAcT and RVSmVTI (Figure 2). Difference between observed and predicted mPAP by PAAcT and RVSmVTI in the validation cohort were: -1.86 (95% CI: -2.72, -0.97) and 0.2 (95% CI: -0.66, 1.07) mmHg, respectively (Figure 2, panel B and D, respectively). Using the predefined definition of accuracy, both PAAcT and RVSmVTI were accurate in 90%, and RVD/Tpeak in 80%. Comparable value for the established clinical method for predicted mPAP by 0.61 × sPAP+2 mmHg versus observed mPAP is shown in Figure 2G. A strong correlation between predicted and invasive mPAP was also observed (Figure 2), but weaker than validation indices, HR-adjusted PAAcT and HR-adjusted RVSmVTI (Figure 2). Further, the established method showed a larger bias (+3.42 mmHg) and wider limits of agreement (Figure 2H), compared to HR-adjusted PAAcT and RVSmVTI (Figures 2B and 2D).

Figure 2. The panels A, C, E and G, show the relationship between heart rate adjusted pulmonary artery acceleration time (HR-adjusted PAAcT), the tissue Doppler derived right ventricle (RV) systolic myocardial velocity time integral (RVSmVTI), RV dimension to time to peak (RVD/Tpeak) and mPAP derived from Doppler echocardiography by the standard equation: [(0.61 × systolic artery pressure (sPAP)) + 2 mmHg] versus invasive mean pulmonary artery pressure (mPAP), respectively. The horizontal dashed lines represent definition of pulmonary hypertension (mPAP ≥ 25 mm Hg) by invasive right heart catheterization, while the vertical dashed lines reflect elevated mPAP by the PH cutoff value of the echo indices. The number within each quadrant of true-positive, false-negative, false-positive, and true-negative results are shown in the right plots. The red and green dots show misclassified and correctly classified patients, respectively. In panel B, D, F and H corresponding Bland-Altman plots are shown for observed minus estimated mPAP, based on regression equations from the derivation group, versus the average of predicted minus observed values of mPAP. The horizontal lines correspond to the mean difference and 95% limits of agreement. Less accuracy of Doppler echocardiography in estimating mPAP is apparent at high mPAP.

The derivation group showed the two linear regression equations between invasive PVR and the echo indices, HR-adjusted PAAcT and RVSmVTI, y = –0.04x + 6.87 and -1.34 + 5.85, respectively. Applied to the validation group, strong correlations (Figures 3A and 3C) and acceptable limits of agreement were observed ­(Figures 3B and 3D). Accuracy according to predefinition of PVR for PAAcT and RVSmVTI was observed in 78% and 84%, respectively. There was also good agreement between PVR predicted as proposed by Abbas et al. and calculated PVR from RHC. Bland-Altman plot showed a mean difference -0.01 (95% CI: -0.30 -0.28). PAAcT and RVSmVTI were comparable to Abbas et al. (4) with minor bias in mean differences of -0.23 (95% CI: -0.50, 0.04) and 0.09 (95% CI: -0.18, 0.36), respectively, and by comparable limits of agreements compared to PVR by Abbas (Figures 3B and D).

Figure 3. Scatterplots showing the relation between invasive pulmonary vascular resistance (PVR) versus heart rate adjusted pulmonary artery acceleration time (HR-adjusted PAAct) (panel A), right ventricular systolic velocity time integral (RVSmVTI) (panel C) and the recommended method by the Tricuspid regurgitation peak velocity/[(RV outflow tract pulsed Doppler VTI × 10)+0.16] (panel E). Correctly and misclassified subjects are illustrated by green and red dots, respectively. Bland-Altman plots of predicted minus observed PVR (vertical axis) against mean of observed and predicted PVR (horizontal axis) for PAAcT (panel B), RVSmVTI (panel D) and the recommended echo method (panel F). Good agreement with negligible mean difference and narrow limits of agreement for all three indices were observed; however, a tendency for greater negative differences with greater PVR, seen in both panel B, D and F.

Prediction of pulmonary hypertension and elevated PVR

Test indices areas under curve (derivation group) for RVD/Tpeak, RVSmVTI, and HR-adjusted PAAct to predict mPAP ≥ 25 mmHg ranged from 0.87 to 0.96 (Figure 4A), and were not statistical inferior to standard approaches by sPAP (Figure 4A). RVSmVTI had a large area under curve, 0.93 (95% CI 0.86–0.99), and a cutoff value of 2.53 provided the best balanced sensitivity, 85% (95% CI: 55–98%), and specificity, 92% (95% CI: 78–98%) with high test performance (diagnostic odds ratio of 77) to determine mPAP ≥ 25 mmHg (Table 3). A cutoff value of 103 ms for HR–adjusted PAAcT provided the best balanced sensitivity, 98% (95% CI: 75–100%) and specificity, 89% (95% CI: 75–97%) with high diagnostic performance odds ratio of 62.

Figure 4. Comparative receiver operating characteristic curve analyses are presented in 50 patients (derivation group) for; heart rate adjusted pulmonary artery (HR-adjusted PAAcT), right ventricular (RV) systolic velocity time integral (HR-adjusted RVSmVTI), RV diameter (RVD)/time to peak Sm (RVD/Tpeak), estimation of mPAP from established formula (0.61·systolic PAP+2 mmHg, whereby ^*sPAP is obtained by continuous wave Doppler by tricuspid regurgitation (TR) peak velocity by Bernoulli + right atrial pressure), and echo pulmonary vascular resistance Tricuspid regurgitation peak velocity/[(RV outflow tract pulsed Doppler VTI × 10)+0.16]. The best discriminators for pulmonary hypertension (≥ 25 mmHg) and elevated pulmonary vascular resistance (>2WU) are shown in panel A and B, respectively. Area under the curve superscript (1–8): significant difference (p < 0.05) vs. numbered area under the curve.

Table 3. Diagnostic performance of echocardiographic indices in predicting pulmonary hypertension and elevated PVR in the validation group (n = 50) with cutoff values from the derivation cohort (n = 50)

Echocardiographic test variables	Cutoff	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	DOR
Predicting PH (mPAP ≥ 25 mmHg)
HR-adjusted PAAcT (ms)	<103	85	92	79	94	62
HR-adjusted RVSmVTI (cm)	<2.53	92	86	71	97	77
RVD/Tpeak (cm/s)	>22.5	77	84	63	91	17
TR peak by Bernoulli (mmHg)^*	>39	67	65	62	99	11
0.61× sPAP + 2 (mmHg)^*	≥25	83	88	71	94	36
Predicting PVR (PVR>2WU)
HR-adjusted PAAcT (ms)	<116	64	86	86	66	11
HR-adjusted RVSmVTI (cm)	<2.7	75	82	84	72	14
PVR Abbas (WU)^*	>2.1	68	60	68	60	3

Abbreviations; PPV: positive predictive value; NPV: negative predictive value; DOR: diagnostic odd ratio; HR: heart rate; PAAcT: pulmonary artery acceleration time; RVSmVTI: right ventricle systolic myocardial velocity time integral; RVD/Tpeak: right ventricular dimension to time to peak; TR: tricuspid regurgitation; PVR: pulmonary vascular resistance. ^* Patients (n = 42).

The ability of PAAcT and RVSmVTI to predict elevated PVR > 2WU were also tested, and areas under curve of 0.80 and 0.77 were noted, respectively (Figure 4B). PAAcT had statistical significant better ­predictive ability compared to PVR by Abbas. Cutoff values of 116 ms for PAAcT provided the best balanced sensitivity, 99% (95% CI: 75–100%) and specificity, 68% (95% CI: 50–82%) to determine PVR>2WU, and of 2.7 cm for RVSmVTI with sensitivity of 92% (95% CI: 64–99%) and specificity of 76% (95% CI: 59–88%).

Diagnostic performance by sensitivity, specificity and diagnostic odd ratio for all test variables is shown in Table 3 for the validation group. Mean PAP by HR-adjusted PAAct and RVSmVTI showed superior diagnostic odd ratio compared to established methods (tricuspid regurgitation peak jet regurgitation by Bernoulli and mPAP by 0.61 × sPAP+2 (mmHg) (Table 3).

Intra- and interobserver variability

The intraclass coefficient of intra- and interobserver (JMH vs. JMH -and JMH vs. KS) variability for post-processing analyses for RVSmVTI, RVSm, Tpeak and RVD1 were tested in 10 participants. The results for intraobserver (JMH) analyses with 95% CI were 0.99 (0.97–1.0), 0.97 (0.86–0.99), 0.99 (0.96–1.0), 0.99 (0.95–1.0) and for interobserver (JMH vs. KS) were 0.96 (0.83–0.99), 0.90 (0.61–0.98), 0.99 (0.96–0.99) and 0.91 (0.64–0.98), respectively. All parameters were significant (p < 0.001).

Discussion

In a COPD population with and without PH where post-capillary contribution to elevated mPAP has been excluded, we have shown that PH can be faithfully predicted by Doppler echocardiography and novel TDI with a considerable degree of sensitivity and specificity. TDI indices and PAACT were highly feasible, and showed good test accuracy both for mPAP and PVR estimations.

Non-invasive estimation of PH and PVR

Calculation of mPAP can be estimated by simple standard Doppler echocardiogram by the inverse relationship between PA acceleration time and mPAP, which also has been shown in the present study (23, 24). We therefore consider that our data support the use of PA acceleration time as a predictor of PH in COPD. However, there are currently two empirical regression formulas available in determining mPAP by PA acceleration time depending on its duration; >120 or <120 ms (3). Applying these formulas; patients with PA acceleration times of 119 and 120 ms would give estimated mPAP values of 16 and 25 mmHg, respectively, which creates an artificial difference. Our proposed regression equation, however, will avoid this problem. HR-adjusted PAAcT demonstrates satisfactory prediction of PH as shown in the ROC curve analysis (Figure 4A) and was significantly different in a statistical perspective compared to standard approaches based on tricuspid regurgitation, and hence, provided better discriminative ability.

ROC curve analyses showed diagnostic usefulness in predicting PH in our COPD population in particular by RVSmVTI (Figure 4) similar to PAAcT. Melek et al. studied a comparable COPD population by echocardiography and found a similar relationship between echo-estimated systolic PA pressure by peak tricuspid regurgitation, and the TDI index as presented in our study to PA pressure (21). To the best our knowledge, we are the first to report strong inverse correlation of RVSm and its velocity time integral to invasively measured mPAP. The afterload sensitivity of RVSmVTI can be utilized in the prediction of PH, as the vast majority of RV dysfunction in the setting of pulmonary vascular disease is afterload dependent. We clearly show that this measure will track with changes in RV afterload and also with good accuracy.

RVD/Tpeak was tested as a new index of systolic PA pressure in a recent study by McLean et al. (22). This index was strongly correlated to invasively measured systolic PA pressure (r = 0.87) in nine patients, and a cutoff of 22 cm/sec predicted the presence of PH with 80% sensitivity and 83% specificity. Our study, comprising 100 COPD patients without and with mild to moderate PH, documents data that are consistent with the study of McLean et al. (22). The numbers of misclassified subjects were 9 of 50 participants with slightly lower sensitivity and specificity than PAAcT and RVSmVTI (Figures 2 and 4). However, both sensitivity and specificity were still high and not inferior to standard approaches by tricuspid regurgitation based estimate.

Two studies have demonstrated good correlations between peak tricuspid regurgitation jet velocity and invasive measured systolic PA pressure (25, 26). This method, however, is limited in particular in COPD by reduced quality of the Doppler signals due to hyper inflated lungs, expansion of the thoracic cage and alteration in the position of the heart. The success rate in estimating systolic PA pressure from unenhanced tricuspid regurgitation jet velocity ranges from 30 to 66% (12, 27). We improved the feasibility of the tricuspid regurgitation velocity method by applying agitated saline to enhance the jet signals, and succeeded in 84% of COPD patients, which is similar to Gartman et al. (28). Using agitated saline will imply intravenous access, which is time consuming and uncomfortable for the patients.

There are limited data available on non-invasive assessment of PVR especially in COPD. In the present study strong correlations between heart rate adjusted PA acceleration time by quadratic regression and SmVTI versus invasive PVR were observed, however, correlation was less robust at markedly elevated PVR (>5Wu) (Figures 3A and 3C), which is in line with ­others (29). However, correlations and limits of agreement were similar to the recommended method by Abbas et al. (Figure 3). Compared to the method by the same scientist, the area under the curve was significantly better for PA acceleration time and statistical similar for RVSmVTI. Both methods, however, showed a similar correlation to PVR by right heart catheterization and were not inferior predictors of PVR (Figure 4). They also showed high feasibility compared to the standard tricuspid regurgitation velocity method, which is in line with previous reports (18, 30). The present study reports on elevated PVR, defined as >2 WU, to identify early remodeling in the pulmonary circulation, contrasting the most used clinical cutoff value, PVR>3WU.

We consider measurements of intraclass coefficient of intra- and interobserver variability in echo recordings to be important according to the relatively large variability in the assessment of different echo indices. In the present study RVSmVTI and RVSm were selected, since they represented relative new TDI indices for the evaluation of pulmonary circulation and RV function, respectively. The two other indices were chosen in this respect according to their ability to measure time in the TDI image (Tpeak) and length in the 2D image (RVD1).

Clinical implications

The present study has demonstrated data on agreement and correlations between PA acceleration time and RVSmVTI versus invasive measured PA pressures and PVR, which are similar or better than the method previously described by Abbas et al. and others (4, 18). Besides, both echo indices have demonstrated 100% feasibility, which we consider to be of significant clinical importance in particular in patients with COPD with their reduced acoustic conditions. We therefore suggest that both methods should be implemented in future echo evaluation in this group of patients.

It is debated whether elevated mPAP in COPD is of clinical relevance. Recent data, however, suggest that even small elevation in mPAP can influence mortality (31), hospitalization (32) and patient outcomes (9, 33, 34). Although targeted treatment for PH in COPD remains to be established, future clinical attention should nevertheless be concentrated on early identification of altered pulmonary circulation and PH, because interventions at an early stage may reverse the process or slow its progression (35). As previously shown by our group (11), patients with clinical suspicion of PH and PaO₂<9.5 kPa should be further examined with regard to PH.

Limitations

The invasive and non-invasive investigations were performed non-simultaneously due to comprehensive echocardiographic examinations. The time interval between examinations introduces a potential source of error due to biologic variation. To minimize these variations, echocardiographic examinations preceded RHC by no more than 120 minutes and were performed under comparable fasting and loading conditions.

The present study comprises a narrowly defined population of stable COPD patients without co-morbidities and clinical established LV disease to study exclusively the impact of pre-capillary pulmonary vascular disease. We therefore cannot extrapolate our findings to all patients with COPD.

Our patients had slightly elevated systemic blood pressure and pulmonary wedge pressure in the upper part of normal limits. Thus, latent LV diastolic heart failure with preserved ejection fraction cannot entirely be excluded. These considerations are thoroughly discussed in a previous paper including the same patients (9).

TDI S velocity and its integral might be influenced by other mechanisms than RV afterload as impaired regional contractility due to myocardial infarction, postoperative state, tachy- or brady arrhythmias, all conditions that should be taken into consideration using this method as a substitute for invasive PAP. Measurements by TDI can be challenging because careful placement of the sample of volume at tricuspid annulus is required. However, this measurement is now being increasingly performed on a routine basis and errors from inaccurate sampling are less likely to occur.

Conclusion

In a COPD population free of LV disease and co-morbidities, mPAP and PVR were determined noninvasively by echocardiography with conventional and novel TDI techniques, Heart rate-adjusted PA acceleration time and RVSmVTI are simple and reproducible methods that correlate well with PA pressure and PVR and show a high degree of accuracy. They are therefore echo indices to be considered in the echocardiographic assessment of COPD patients with suspicion of PH, and may be of particular interest in cases with inadequate tricuspid Doppler signal.

Funding

This work was supported by Eastern Norway Regional Health Authority, 2303 Hamar, Norway and Department of Cardiology, Oslo University Hospital, Aker, Norway.

Declaration of Interest Statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Color versions for one or more of the figures in the article can be found online at www.tandfonline.com/icop.