The assessment of pulmonary pressure is important for the diagnosis and management of patients with pulmonary hypertension. Mean pulmonary artery pressure (MPAP) has been used in the current definition of pulmonary hypertension. However, invasive derivation by Doppler echocardiography provides the peak pulmonary artery systolic pressure (PASP). The aim of this study was to derive a method to predict MPAP from PASP.
Invasive hemodynamic pressures in 307 patients who underwent right heart catheterization were examined. Simple regression techniques were used to determine the relationship between MPAP and PASP in a derivation cohort ( n = 198) and a validation sample ( n = 109). Bland-Altman analysis was performed to examine predicted versus observed values of MPAP.
MPAP and PASP at catheterization were strongly related over a range of pressures ( R 2 = 0.89, n = 198; SE, 4.04; P < .0001). The relation of MPAP to PASP in the derivation cohort (MPAP = 0.61 × PASP + 1.95 mm Hg) was validated in the test sample, with an R 2 value of 0.94 for predicted versus observed MPAP (SE, 2.87; P < .0001). The relationship of predicted versus observed MPAP was constant across different degrees of pressure elevation, as well as different etiologies of pulmonary hypertension. Applying the equation to Doppler-derived pulmonary pressures, there was excellent correlation of predicted MPAP from echocardiography and invasively measured MPAP ( R 2 = 0.78, P < .0001).
MPAP can be accurately predicted from PASP over a wide pressure range for different etiologies of pulmonary hypertension. This finding may help define MPAP noninvasively.
Pulmonary hypertension is a common disorder that leads to progressive vascular remodeling and eventually right heart failure. A variety of etiologies can lead to the development of pulmonary hypertension, including left heart disorders and those due to intrinsic pulmonary vascular or parenchymal disease. Timely recognition and interventional pharmacotherapy are essential to the management of these patients, because right ventricular failure frequently ensues, leading to substantial morbidity and a poor prognosis.
Echocardiography has a fundamental role in the diagnostic assessment of pulmonary hypertension in suspected patients. Right-sided hemodynamics can be accurately evaluated using established Doppler-derived methods, primarily using the peak tricuspid regurgitation velocity to estimate the peak pulmonary artery systolic pressure (PASP). However, an important parameter in the assessment and management of pulmonary hypertension is mean pulmonary pressure, which can be difficult to derive with current noninvasive methods. Pulmonary hypertension is defined by a mean pulmonary artery pressure (MPAP) ≥25 mm Hg at rest or ≥30 mm Hg during physical activity, with current guidelines emphasizing the need for documentation with invasive right heart catheterization. A response to vasodilators is defined by a decrease in MPAP of >10 mm Hg to an absolute MPAP <40 mm Hg.
To derive an echocardiographic method for the calculation of MPAP, in the present study we examined the relation between phasic variables of pulmonary pressure obtained from direct invasive hemodynamic assessment.
The Mayo Clinic Institutional Review Board approved this study. To form a derivation cohort, 198 patients who underwent right heart catheterization between December 2002 and January 2004 for evaluation of pulmonary pressures in the Mayo Clinic Cardiac Catheterization Laboratory (Rochester, MN) were randomly selected and enrolled in the study. A separate group of 109 randomly selected patients who underwent right heart catheterization between January 2011 and May 2011 were retrospectively enrolled to form a validation cohort. No exclusion criteria were applied to enrollment for either cohort. All patients provided informed consent for review of their medical record for research purposes in accordance with Minnesota state law.
Patients were brought to the cardiac catheterization laboratory in the fasting state. All procedures were performed with conscious sedation. After placement of 7-Fr or 8-Fr vascular sheaths, intracardiac pressures (right atrial, right ventricular, pulmonary artery, and pulmonary capillary wedge) were obtained in standard fashion using 7-Fr or 7.5-Fr balloon-tipped fluid-filled catheters (Arrow International, Asheboro, NC, or Edwards Lifesciences, Irvine, CA). Mean pulmonary capillary wedge pressure (PCWP) was confirmed with verification of oxygen saturation in all patients. Cardiac output was measured in standard fashion using the thermodilution method, except when there was evidence of intracardiac shunting or moderate or severe right-sided valvular regurgitation. In these latter instances, the Fick method was used, with direct measurement of myocardial oxygen consumption. Pulmonary resistance index was calculated as (MPAP−mean PCWP)/cardiac index. Hemodynamic recordings were obtained before any pharmacologic infusion or saline infusion in all patients. Digital acquisition of pressures at 3-msec to 5-msec intervals was performed, with recordings available for subsequent offline analysis.
Phasic Pressure Analysis
Pulmonary pressures were analyzed using visual analysis with digital quantification using dedicated software (CathCoding; Mayo Foundation, Rochester, MN) by reviewers who were blinded to clinical data. Variables of peak PASP, diastolic pulmonary artery pressure, and mean PCWP were ascertained. In all analyses, careful attention was given to excluding catheter and pressure-damping artifacts (e.g., whip). MPAP was defined as the area under the systolic pressure curve divided by the systolic pulse interval and was calculated from the hemodynamic record using Mac-Lab software (GE Healthcare, Milwaukee, WI). All pressures were examined in steady state at end-expiration in triplicate.
Standard M-mode, two-dimensional, color Doppler imaging and continuous-wave Doppler echocardiographic examinations were performed by registered diagnostic cardiac sonographers using standardized instruments and protocols and were interpreted by echocardiologists blinded to invasive data. Left ventricular ejection fraction and cardiac output were derived by standard methods. PASP was estimated by Doppler echocardiography from the systolic right ventricular–to–right atrial pressure gradient using the modified Bernoulli equation (4 × [peak tricuspid regurgitant velocity] ). Right atrial pressure assessments were performed in accordance with previously described methods. No patients had significant right ventricular outflow tract obstructions. All parameters were measured in triplicate at steady state and averaged.
Simple linear regression analyses were performed to determine the relation between continuous variables. These analyses were used to derive a regression equation with the SE for the relation of MPAP to PASP in the 198 patients from the derivation cohort that was then tested in the 109 patients in the validation cohort. Bland-Altman plots were generated to compare predicted versus observed MPAP values. To examine the regression equation in different clinical states, analyses also were performed according to levels of right atrial pressure (≥10 vs <10 mm Hg), PASP (≥40 vs <40 mm Hg), mean PCWP (≥15 vs <15 mm Hg), and left ventricular ejection fraction (≥50% vs <50%). To examine the utility of the derived equation with noninvasive methods, data from nonsimultaneous echocardiograms obtained within 1 month of cardiac catheterization were reviewed. Right ventricular dysfunction, severity of tricuspid regurgitation, estimation of mean right atrial pressure, and Doppler-derived estimation of right ventricular systolic pressure were recorded using standard definitions. The equation was applied to Doppler echocardiographic data in the validation cohort for the prediction of MPAP. Continuous variables are reported as mean ± SD. Statistical significance was inferred at P < .05.
In the derivation cohort, the mean age of the study patients was 58.3 ± 15.1 years ( Table 1 ). The majority of patients had preserved left ventricular ejection fractions (mean, 54 ± 16%), and congenital heart disease (24%) and right ventricular dysfunction detected on echocardiography (42%) were also common. Histories of valvular heart disease were also present in the majority of patients (72%), with the following types of valvular disease represented: aortic regurgitation (11.1%), aortic stenosis (11.6%), mitral regurgitation (39.9%), mitral stenosis (7.8%), tricuspid regurgitation (45.0%), pulmonary regurgitation (9.6%), pulmonary stenosis (0.5%), and tricuspid stenosis (0.5%). Overall, elevated PASP (>40 mm Hg), MPAP (>25 mm Hg), and PCWP (>15 mm Hg) were present in 67.2%, 67.7%, and 54.0% of patients, respectively. Systemic hypertension (systolic blood pressure ≥120 mm Hg) was present in 60.6% of patients. There were no significant differences in clinical or invasive hemodynamic variables between patients in the derivation and validation cohorts. Preserved left ventricular ejection fraction (mean, 53.8 ± 15.9%), coronary artery disease with >50% stenosis (35.4%), atrial fibrillation (28.8%), prior cardiac surgery (28.8%), and moderate or severe right-sided valvular regurgitation (24.7%) were common. Pulmonary hypertension (PASP >60 mm Hg) was present in 37.8% of patients. In comparison with the derivation cohort, patients in the validation cohort less commonly had atrial fibrillation, hypertension, and coronary artery disease, but had a higher frequency of prior cardiac surgery, moderate or severe right-sided regurgitation, and moderate or severe right ventricular dysfunction. The invasive hemodynamic data for the entire study population, derivation sample, and validation sample is shown in Table 2 .
|Variable||All patients ( n = 307)||Derivation cohort ( n = 198)||Validation cohort ( n = 109)||P|
|Age (y)||58.4 ± 15.1||58.3 ± 15.1||58.5 ± 14.9||.91|
|Men||168 (54.7%)||98 (49.5%)||70 (64.2%)||.01|
|Atrial fibrillation||69 (22.4%)||57 (28.8%)||12 (11.0%)||.0004|
|Hypertension||108 (35.2%)||95 (47.9%)||13 (11.9%)||<.0001|
|Coronary artery disease||93 (30.2%)||70 (35.4%)||23 (21.1%)||.009|
|Congenital heart disease||18 (9.9%)||14 (7.1%)||4 (3.7%)||.22|
|Cardiac surgery||105 (34.2%)||57 (28.8%)||48 (44.0%)||.007|
|Left ejection fraction (%)||53.0 ± 17.2||53.8 ± 15.9||51.8 ± 19.2|
|Moderate or severe right-sided regurgitation||92 (29.9%)||49 (24.7%)||43 (39.4%)||.007|
|Moderate or severe right ventricular dysfunction||50 (16.2%)||21 (10.6%)||29 (26.4%)||<.0001|
|Variable||All patients ( n = 198)||Derivation sample ( n = 100)||Validation sample ( n = 98)|
|Heart rate (beats/min)||75.9 ± 15.4||75.9 ± 14.8||75.9 ± 16.2|
|Systolic blood pressure (mm Hg)||130.3 ± 28.6||126.8 ± 29.57||133.0 ± 27.85|
|Diastolic blood pressure (mm Hg)||68.2 ± 13.2||67.5 ± 14.9||69.6 ± 13.4|
|Right atrial pressure (mm Hg)||12.6 ± 12.2||12.4 ± 11.7||12.7 ± 12.7|
|Right ventricular systolic pressure (mm Hg)||54.0 ± 21.9||53.3 ± 21.1||54.1 ± 23.1|
|PASP (mm Hg)||54.0 ± 22.9||55.0 ± 23.6||53.1 ± 22.2|
|Pulmonary artery diastolic pressure (mm Hg)||22.4 ± 12.3||22.4 ± 12.3||22.3 ± 11.8|
|MPAP (mm Hg)||34.3 ± 13.9||34.2 ± 12.9||34.3 ± 14.9|
|PCWP (mm Hg)||17.8 ± 8.7||18.5 ± 9.4||17.2 ± 8.0|
|Cardiac output (L/min)||6.1 ± 2.5||6.1 ± 3.2||5.7 ± 1.6|
|Pulmonary artery resistance index (Wood units · m 2 )||8.2 ± 9.8||8.5 ± 11.3||8.0 ± 8.4|