論文ID: CJ-15-1080
Background: Mean pulmonary arterial pressure (MPAP) is an important pulmonary hemodynamic parameter used in the management of patients with chronic thromboembolic pulmonary hypertension (CTEPH). We compared echocardiography-derived estimates of MPAP with right heart catheterization (RHC) to identify reliable noninvasive methods of estimating MPAP-derived RHC (MPAPRHC) in these patients.
Methods and Results: Echocardiography and RHC were performed in 56 patients with CTEPH (60.5±12.0 years; 44 females). We measured the tricuspid regurgitation (TR) pressure gradient (TRPG) using echocardiography. The mean systolic right ventricular (RV)-right atrial (RA) gradient was calculated by tracing the TR time velocity flow. Systolic and mean pulmonary artery pressures (SPAPTR and MPAPTR) estimated from TRPG and mean systolic RV-RA gradient were calculated by adding RA pressure based on the inferior vena cava. MPAPChemla was calculated using Chemla’s formula: 0.61×SPAPTR+2 mmHg. MPAPRHC and pulmonary vascular resistance were 35.9±11.3 mmHg and 6.6±3.6 Wood units, respectively. The mean difference from MPAPRHC and limits of agreement were −1.5 mmHg and −19.6 to 16.5 mmHg for MPAPTR, and −4.6 mmHg and −24.5 to 15.2 mmHg for MPAPChemla. Accuracy within 10 mmHg and 5 mmHg of MPAPRHC was 80.4% and 46.4% for MPAPTR, and 71.4% and 48.2% for MPAPChemla, respectively.
Conclusions: MPAPTR and MPAPChemla are reliable estimates for MPAPRHC in patients with CTEPH.
Chronic thromboembolic pulmonary hypertension (CTEPH) is a form of pulmonary hypertension caused by non-resolving thromboembolism of the pulmonary arteries and pulmonary vascular remodeling.1 Hemodynamic evaluation is important in the management of these patients, so pulmonary arterial pressure (PAP) and pulmonary vascular resistance (PVR) must be accurately assessed. Mean PAP (MPAP) is an important parameter in the diagnosis, assessment of severity, and response to therapy in patients with PH.2
Although right heart catheterization (RHC) is the gold standard for assessing pulmonary hemodynamics in patients with PH, it is invasive and costly.3 Systolic PAP (SPAP) is estimated from the peak flow velocity of tricuspid regurgitation (TR) on echocardiography and is used for patient screening and follow-up. However, MPAP is not routinely estimated on echocardiography. Although MPAP can also be estimated from the velocity of pulmonary regurgitation (PR) at the start of diastole (MPAPPR), MPAPPR is difficult to measure and is therefore not routinely used.4,5 The measurement of MPAPPR becomes more difficult with the anatomical modifications of right ventricular (RV) structure in patients with severe PH. Chemla et al reported that MPAP can be reliably calculated from SPAP estimated from the peak flow velocity of TR (SPAPTR) using the formula: MPAP=0.61×SPAPTR+2 mmHg.6,7 Steckelberg et al also reported that a similar formula, MPAP=0.61×SPAPTR+1.95 mmHg, was useful over a wide range of pressures for different etiologies of PH.8 Conversely, Aduen et al reported that MPAP calculated from adding the mean systolic RV-right atrial (RA) gradient derived by tracing the TR velocity flow to the RA pressure (MPAPTR) had an accuracy and precision similar to that of MPAP as estimated by the Chemla method (MPAPChemla).2,9 However, the number of patients with CTEPH included in those previous reports was insufficient.2,7,10 Furthermore, proximal obstruction in CTEPH may cause high pulse pressure (PP: systolic-diastolic PAP) with low MPAP compared with pulmonary arterial hypertension (PAH), but there have been no reports that evaluated the usefulness of MPAPChemla and MPAPTR in patients with CTEPH. Therefore, the aim of the present study was to identify whether MPAPChemla and MPAPTR were reliable estimates of MPAP in patients with CTEPH.
This study was a single-center retrospective investigation of consecutive patients with a high clinical suspicion of CTEPH who underwent echocardiography and RHC from October 2012 to April 2015. This study was approved by the ethics committee of Chiba University (approval date: June 1, 2009; approval number: 826), and written informed consent was given by each patient before echocardiography and RHC. Patients with complications, including left heart disease and atrial fibrillation, and patients who had undergone pulmonary endarterectomy within 1 year of the study, were excluded.
EchocardiographyWithin 2 days of RHC, Doppler echocardiography using an AplioTM 300 ultrasound (Toshiba Medical, Tochigi, Japan) with a PST-25BT transducer (2.5 MHz; Toshiba Medical) was performed on all of the patients at the end of expiration. The recordings were obtained from the left parasternal long axis, left parasternal short axis, and the apical 4-chamber and 5-chamber views. All results were the mean of 3 measurements and the analyses were performed without knowledge of the patients’ clinical status. There were no changes in medication or oxygen therapy between RHC and echocardiography. The TR velocity was obtained using continuous wave Doppler from an appropriate view where TR could be clearly visualized, such as the apical 4-chamber, parasternal, and subcostal views, and the highest peak value (TRV) was recorded. The TR pressure gradient (TRPG) was determined from the TRV using a simplified Bernoulli equation: TRPG=4×TRV2. The mean systolic RV-RA pressure gradient was calculated by tracing the TR velocity flow to obtain the mean value from the area under the curve (Figure 1).2,9 The RA pressure was estimated by the combination of diameter and respiratory variation of the inferior vena cava (IVC) as follows: (1) IVC diameter ≤2.1 cm that collapses >50% with a sniff was defined RA pressure of 3 mmHg, (2) IVC diameter >2.1 cm that collapses <50% with a sniff was defined a high RA pressure of 15 mmHg, (3) indeterminate cases in which the IVC diameter and collapse do not fit this paradigm was defined an intermediate value of 8 mmHg.11 SPAPTR and MPAPTR were calculated by adding the estimated RA pressure to TRPG and the mean RV-RA pressure gradient, respectively. MPAPChemla was calculated from the following formula: MPAPChemla=0.61×SPAPTR+2 mmHg.7 The severity of TR was assessed from the regurgitation jet area using standard definitions.12
Mean pulmonary artery pressure estimated from tricuspid regurgitation (TR) flow on echocardiography (MPAPTR) in a 49-year-old woman. The mean systolic right ventricular-right atrial (RA) pressure gradient was measured by tracing the TR velocity flow to obtain the mean value from the area under the curve (38.5 mmHg). The mean RA pressure was estimated to be 8 mmHg from a combination of the diameter and respiratory variation of the inferior vena cava. MPAPTR=38.5+8=46.5 mmHg. MPAPTR, mean pulmonary artery pressure estimated from tricuspid regurgitation flow with echocardiography.
A 7.5Fr Swan-Ganz thermodilution catheter (Edwards LifeSciences, Irvine, CA, USA) was positioned via a jugular approach. At end-expiration, pressure measurements from the RA, RV, and main pulmonary artery, as well as the pulmonary arterial wedge pressure (PAWP), were recorded. The zero point was defined as mid-thoracic. MPAPRHC was calculated from SPAP and the diastolic PAP (DPAP) using the following formula: MPAPRHC=DPAP+(SPAP−DPAP)/3. Cardiac output (CO) was determined using the thermodilution method by averaging at least 3 measurements. The PVR was calculated as follows: (MPAP−PAWP)/CO in Wood units. Possible left-to-right shunting was excluded by oximetry.
Statistical AnalysisAll results are expressed as mean±standard deviation unless otherwise indicated. Spearman’s correlation was used to assess the correlations between MPAPRHC and the echocardiography-derived estimates MPAPTR and MPAPChelma. No ordinal categorical data were analyzed using the χ2 test. Finally, a Bland-Altman analysis was performed to determine the limits of the agreements between MPAP estimated by echocardiography and MPAPRHC. P<0.05 was considered statistically significant. All of the statistical analyses were performed using JMP 9.0 software (SAS Institute, Cary, NC, USA).
The study group comprised 56 consecutive patients (mean age: 60.5±12.0 years; 44 females) with CTEPH confirmed by RHC and pulmonary angiography. Table 1 summarizes the subjects’ baseline clinical characteristics, hemodynamic data, and echocardiographic parameters.
| Parameter | |
|---|---|
| Age (years) | 60.5±12.0 |
| Sex, n (F/M) | 44/12 |
| Body surface area (m2) | 1.53±0.28 |
| Oxygen therapy, n | 41 (0.5–4 L/min) |
| Postpulmonary endarterectomy | 15 |
| Vasodilators | |
| Oral prostaglandin I2, n | 14 |
| Intravenous prostaglandin I2, n | 0 |
| Phosphodiesterase V inhibitor, n | 15 |
| Endothelin antagonist, n | 8 |
| Soluble guanylate cyclase stimulators, n | 2 |
| Pulmonary hemodynamic data | |
| MPAPRHC (mmHg) | 35.9±11.3 |
| SPAPRHC (mmHg) | 64.8±22.1 |
| DPAP (mmHg) | 18.3±7.0 |
| PVR (Wood units) | 6.6±3.6 |
| CO (L/min) | 4.5±1.1 |
| CI (L·min−1·m−2) | 2.9±0.6 |
| PAWP (mmHg) | 8.6±2.6 |
| mRA (mmHg) | 5.6±3.1 |
| Echocardiographic parameters | |
| Heart rate (beats/min) | 67.7±11.3 |
| TR grade | |
| Trivial, n | 8 |
| Mild, n | 33 |
| Moderate, n | 10 |
| Severe, n | 5 |
| TRPG (mmHg) | 56.6±25.1 |
| Mean systolic RV-RA pressure gradient (mmHg) | 30.8±12.7 |
| Inspiratory IVC diameter (cm) | 12.5±3.8 |
| Expiratory IVC diameter (cm) | 7.3±3.7 |
| Estimated RA pressure (mmHg) | 6.6±2.6 |
| SPAPTR (mmHg) | 63.2±25.3 |
| MPAPTR (mmHg) | 37.4±13.1 |
| MPAPChemla (mmHg) | 40.5±15.4 |
Data are presented as mean±standard deviation. CI, cardiac index; CO, cardiac output; CTEPH, chronic thromboembolic pulmonary hypertension; DPAP, diastolic pulmonary arterial pressure; IVC, inferior vena cava; MPAPChemla, mean pulmonary arterial pressure by the method of Chemla et al;7 MPAPRHC, mean pulmonary arterial pressure; MPAPTR, mean pulmonary artery pressure estimated from tricuspid regurgitation flow with echocardiography; mRA, mean right atrium pressure; PAWP, pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; RA, right atrial; RV, right ventricular; SPAPRHC, systolic pulmonary arterial pressure; SPAPTR, systolic pulmonary artery pressure estimated from a peak flow velocity of TR; TR, tricuspid regurgitation; TRPG, tricuspid regurgitation pressure gradient.
Figure 2 shows the linear regression plots for MPAPRHC vs. MPAPTR and MPAPChemla for the 56 patients. MPAPTR and MPAPChemla significantly correlated with MPAPRHC, with good correlations between MPAPTR and MPAPRHC (r=0.737, P<0.001) and between MPAPChemla and MPAPRHC (r=0.799, P<0.001). Figure 3 shows the Bland-Altman analysis used to evaluate the level of agreement between MPAPTR, MPAPChemla, and MPAPRHC across the patient sample. This analysis revealed a mean difference between MPAPTR and MPAPRHC of −1.5 mmHg with limits of agreement from −19.6 to 16.5 mmHg, and a mean difference between MPAPChemla and MPAPRHC of −4.6 mmHg with limits of agreement from −24.5 to 15.2 mmHg. The accuracy within 10 mmHg and 5 mmHg of MPAPRHC was 80.4% and 46.4% for MPAPTR, and 71.4% and 48.2% for MPAPChemla, respectively. However, there were no significant differences in accuracy between MPAPTR and MPAPChelma (10 mmHg, P=0.269; 5 mmHg, P=0.850) (Table 2). MPAPChemla tended to underestimate MPAPRHC by a difference of greater than 10 mmHg compared with MPAPTR, although it did not reach statistical significance (21.4% vs. 10.7%, respectively; P=0.123).
Correlations between 2 echocardiography-based estimates of mean pulmonary arterial pressure (MPAP) and MPAP measured by right heart catheterization (MPAPRHC) (n=56). (Left) Scatter plot of MPAP estimated from tricuspid regurgitation (TR) flow on echocardiography (MPAPTR) vs. MPAPRHC. (Right) Scatter plot of MPAP calculated by the method of Chemla et al (MPAPChemla) vs. MPAPRHC. MAPAChemla=mean pulmonary arterial pressure by the method of Chemla et al; MPAPTR=mean pulmonary artery pressure estimated from tricuspid regurgitation flow with echocardiography; MPAPRHC=mean pulmonary arterial pressure measured by right heart catheterization.
Bland-Altman analysis to evaluate the agreement between 2 echocardiography-based estimates of mean pulmonary arterial pressure (MPAP) and MPAP measured using right heart catheterization (MPAPRHC). (Left) The analysis revealed a mean difference of −1.5 mmHg with limits of agreement from −19.6 to 16.5 mmHg between MPAP estimated from a peak flow velocity of tricuspid regurgitation (MPAPTR) vs. MPAPRHC. (Right) The analysis revealed a mean difference of −4.6 mmHg with limits of agreement from −24.5 to 15.2 mmHg between the MPAP estimated by the method of Chemla et al7 (MPAPChemla) vs. MPAPRHC. MPAPTR, mean pulmonary artery pressure estimated from tricuspid regurgitation flow with echocardiography; MPAPRHC, mean pulmonary arterial pressure measured by right heart catheterization.
| Variable | MPAPTR | MPAPChemla | P value |
|---|---|---|---|
| Accuracy | |||
| Difference in absolute value between MPAPRHC within 5 mmHg, n |
26 (46.4%) | 27 (48.2%) | 0.850 |
| Difference in absolute value between MPAPRHC within 10 mmHg, n |
45 (80.4%) | 40 (71.4%) | 0.269 |
| Difference from MPAPRHC | |||
| >5 mmHg, n | 13 (23.2%) | 7 (12.5%) | 0.393 |
| <−5 mmHg, n | 17 (30.4%) | 22 (39.3%) | 0.321 |
| >10 mmHg, n | 5 (8.9%) | 4 (7.1%) | 0.728 |
| <−10 mmHg, n | 6 (10.7%) | 12 (21.4%) | 0.123 |
Abbreviations as in Table 1.
The model derived from linear regression between MPAPRHC and SPAPTR (MPAPCTEPH) in 56 patients with CTEPH was calculated using the following formula: MPAPCTEPH=0.34×SPAPTR+14.5 mmHg. The mean difference and the limits of agreement between the estimated MPAPCTEPH and MPAPRHC were 1.4 mmHg and from −10.6 to 13.5 mmHg, respectively. The accuracy of the estimated MPAPCTEPH within 10 mmHg and 5 mmHg of MPAPRHC was 82.1% and 57.1%, respectively.
ReproducibilityThe inter- and intraobserver reproducibilities of MPAPTR and MPAPRHC were determined from a Bland-Altman analysis of 56 patients. Absolute intraobserver agreement for MPAPTR and MPAPChelma exhibited a mean difference of 0.1 mmHg and 1.0 mmHg, respectively; the agreement range was −5.7 to 6.1 mmHg and −4.2 to 6.2 mmHg, respectively. The absolute interobserver agreement for MPAPTR and MPAPChelma was found to have a mean difference of 0.5 mmHg and 0.0 mmHg, respectively; the agreement range was −8.0 to 9.0 mmHg and −5.8 to 5.7 mmHg, respectively.
Accuracy of MPAPTR and MPAPChemlaWe analyzed the factors that affected the accuracy of MPAPTR and MPAPChemla. Multiple regression analysis revealed that PVR and CO correlated with the discrepancy between MPAPTR and MPAPRHC (r=0.307, P=0.021; r=−0.370, P=0.005, respectively) or those between MPAPChemla and MPAPRHC (r=0.435, P<0.001; r=−0.507, P<0.001, respectively). In addition, MPAPRHC, SPAPRHC, CI, PP, SPAPTR, and estimated RA pressure also correlated with their accuracy (r=0.297, P=0.026; r=0.345, P=0.009; r=−0.449, P<0.001; r=0.362, P<0.001; r=0.462, P<0.001; r=0.325, P=0.015, respectively) (Table S1). Because the sample size of this study was small, there was a possibility of bias. Therefore, we divided patients into 2 groups depending on whether the discrepancy between echocardiography-derived MPAP and MPAPRHC was greater than 10 mmHg, and performed a logistic analysis to evaluate the relationship between the discrepancy and each parameter. In the group in which the discrepancy between MPAPTR and MPAPRHC was greater than 10 mmHg, PVR was significantly higher (P=0.028), and CO was significantly lower (P=0.025). In the group in which the discrepancy between MPAPChelma and MPAPRHC was greater than 10 mmHg, PVR, MPAPRHC, SPAPRHC, PP, SPAPTR, and estimated RA pressure were significantly higher (P=0.020, P=0.043, P=0.037, P=0.032, P=0.032, and P=0.001, respectively), and CO was significantly lower (P=0.045) (Table S2).
In the present study, MPAPTR and MPAPChelma were the echocardiographic parameters that most strongly correlated with MPAPRHC. To the best of our knowledge, this is the first study to identify MPAPTR and MPAPChemla as reliable estimates of MPAPRHC in patients with CTEPH. Furthermore, MPAPCTEPH: 0.34×SPAPTR+14.5 mmHg was superior to MPAPChelma as a predictive formula using SPAPTR in patients with CTEPH.
MPAPTR and MPAPChemla demonstrated good correlation with MPAPRHC, and the limits of agreement between MPAPTR and MPAPRHC were smaller than those between MPAPChemla and MPAPRHC. Aduen et al examined several MPAP estimates by echocardiography in 117 patients who underwent simultaneous echocardiography and RHC for various conditions, and reported that the mean difference between MPAPTR and MPAPRHC was −1.6 mmHg, with upper and lower limits of agreement of −16.6 to 13.7 mmHg, and the mean difference between MPAPChemla and MPAPRHC was −3.7 mmHg with upper and lower limits of agreement of −18.4 to 11.0 mmHg, respectively.9 The reliability of MPAPTR and MPAPChelma in the present patients with CTEPH was similar to that reported by Aduen et al, although the lower limit of agreement (−24.5 mmHg) of MPAPChelma was lower in our study. Fisher et al reported that almost 50% of patients showed SPAPTR on echocardiography with difference of greater than 10 mmHg above or below SPAPRHC, and that underestimation frequently led to failure to identify PH.13 MPAPTR was superior to MPAPChemla for accuracy when the discrepancy between MPAP estimated by echocardiography and MPAPRHC was within 10 mmHg (80.4% vs. 71.4%; P=0.269), whereas the 2 methods were similar for accuracy within 5 mmHg (46.4% vs. 48.2%; P=0.850). In addition, MPAPTR and MPAPChemla tended to underestimate MPAPRHC. The probability of underestimation greater than 10 mmHg by MPAPTR was less than by MPAPChemla, although there was no significant difference (10.7% vs. 21.4%; P=0.123). The possibility of misclassification of PH severity or even failure to identify CTEPH may be reduced by measuring both MPAPTR and MPAPChemla. MPAPTR and MPAPChemla can be obtained quickly and simply when measuring TRPG and SPAPTR, both of which are always measured on echocardiography for screening and the follow-up of patients with PH. Therefore, MPAPTR and MPAPChemla are reliable parameters that could potentially be adopted in routine echocardiography screening and follow-up of patients with CTEPH.
In patients with severe PH, TRV is elevated with increasing PAP and PVR. Because TRPG is derived from TRV (by the formula TRPG=4×TRV2), the difference between SPAPTR and SPAPRHC tends to be greater when TRV cannot be evaluated correctly. In particular, the difference between MPAPChelma and MPAPRHC tends to increase, as well as SPAPTR, when the peak of TR velocity flow is not clearly visualized. In contrast, because MPAPTR is derived by tracing the TR velocity flow, MPAPTR includes the information for the time and change in TR velocity. This may result in a decreased difference between MPAPTR and MPAPRHC. Therefore, MPAPTR is a better estimate of MPAPRHC than MPAPChemla for patients with high PVR, PAP, and SPAPTR. In addition, proximal obstruction in CTEPH may cause high PP with low MPAP compared with PAH.14,15 Indeed, we compared 56 CTEPH patients with 30 PAH patients who had undergone RHC in the same period of this study. PP/MPAPRHC was significantly higher in patients with CTEPH, but there was no significant difference in other pulmonary hemodynamic data between patients with CTEPH and PAH. Although the difference in pulmonary hemodynamics between CTEPH and other types of PH, including PAH, have not been completely clarified, there is a possibility that errors may occur when MPAPChemla is directly used for patients with CTEPH. We examined MPAPCTEPH derived from the linear regression between MPAPRHC and SPAPTR in the present study. MPAPCTEPH was calculated using the following formula: MPAPCTEPH=0.34×SPAPTR+14.5 mmHg. The coefficient for SPAPTR in this model was smaller, and the intercept was larger than in the formula for MPAPChemla. The limits of the agreement between the estimated MPAPCTEPH and MPAPRHC were smaller than MPAPChelma (−10.6 to 13.5 mmHg vs. −24.5 to 15.2 mmHg). The accuracy of the estimated MPAPCTEPH within 10 mmHg and 5 mmHg of MPAPRHC was also superior to MPAPChelma (10 mmHg: 82.1% vs. 71.4%, P=0.179; 5 mmHg: 57.1% vs. 48.2%, P=0.344, respectively). Although there may be differences between institutions that cannot be ignored, the formula for MPAP-derived SPAPTR may vary according to the type of PH. Therefore, it is necessary to study this aspect further for each disease.
Study LimitationsFirst, this was a single-center retrospective study involving a small number of patients; therefore, prospective multicenter studies involving larger patient populations and various types of PH are required to confirm the results. Second, echocardiography and RHC were performed up to 2 days apart, although the clinical condition of the patients was stable and the therapy was the same during each examination. Third, given that we aimed to predict MPAPRHC by echocardiography as a screening or follow-up tool, our study population also included patients treated with vasodilators and pulmonary endarterectomy. However, such treatments may have influenced the echocardiographic parameters. In future research, subgroup analyses will be necessary among therapy-naïve, vasodilator-treated, and postpulmonary endarterectomy patients to assess any possible changes in the correlation between echocardiographic estimates of MPAP and MPAPRHC.
MPAPTR and MPAPChemla were reliable estimates of MPAPRHC in patients with CTEPH, and their reliability was comparable. Furthermore, MPAPCTEPH: 0.34×SPAPTR+14.5 mmHg was superior to MPAPChelma as a predictive formula using SPAPTR in patients with CTEPH.
K.T. received a grant from (1) the Respiratory Failure Research Group from the Ministry of Health, Labor and Welfare of Japan and (2) the Pulmonary Hypertension Research Group from Japan Agency for Medical Research and Development, AMED, and honoraria from Actelion Pharmaceuticals. N.T. received a research grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan (No. 25461148), and belongs to the endowed department sponsored by Actelion Pharmaceuticals. The study sponsors had no roles in the design of the study, the collection, analysis, or interpretation of data, or in writing the manuscript.
Supplementary File 1
Table S1. Multiple regression analysis of the factors influencing the difference between MPAPTR, MPAPChemla, and MPAPRHC in 56 patients with CTEPH
Table S2. Logistic regression analysis of the factors influencing the difference between MPAPTR, MPAPChemla, and MPAPRHC in 56 patients with CTEPH
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-15-1080
