A simple echocardiographic prediction rule for hemodynamics in pulmonary hypertension

Abstract

Background: Pulmonary hypertension (PH) has diverse causes with heterogeneous physiology compelling distinct management. Differentiating patients with primarily elevated pulmonary vascular resistance (PVR) from those with PH predominantly because of elevated left-sided filling pressure is critical.

Methods and results: We reviewed hemodynamics, echocardiography, and clinical data for 108 patients seen at a referral PH clinic with transthoracic echocardiogram and right heart catheterization within 1 year. We derived a simple echocardiographic prediction rule to allow hemodynamic differentiation of PH attributed to pulmonary vascular disease (PH(PVD), defined as pulmonary artery wedge pressure [PAWP]≤15 mm Hg and PVR>3 WU). Age averaged 61.3±14.8 years, μPAWP and PVR were 16.4±7.1 mm Hg and 6.3±4.0 WU, respectively, and 52 (48.1%) patients fulfilled PH(PVD) hemodynamic criteria. The derived prediction rule ranged from -2 to +2 with higher scores suggesting higher probability of PH(PVD): +1 point for left atrial anterior-posterior dimension <3.2 cm; +1 for presence of a mid systolic notch or acceleration time <80 ms; -1 for lateral mitral E:e'>10; -1 for left atrial anterior-posterior dimension >4.2 cm. PVR increased stepwise with score (for -2, 0, and +2, μPVR were 2.5, 4.5, and 8.1 WU, respectively), whereas the inverse was true for pulmonary artery wedge pressure (corresponding μPAWP were 21.5, 16.5, and 10.4 mm Hg). Among subjects with complete data, the score had an area under the curve (AUC) of 0.921 for PH(PVD). A score ≥0 had 100% sensitivity and 69.3% positive predictive value for PH(PVD), with 62.3% specificity. No patients with a negative score had PH(PVD). Patients with a negative score and acceleration time >100 ms had normal PVR (μPVR=1.8 WU, range=0.7-3.2 WU).

Conclusions: We present a simple echocardiographic prediction rule that accurately defines PH hemodynamics, facilitates improved screening and focused clinical investigation for PH diagnosis and management.

Figure 1

Representative echocardiographic images of LA size, transmitral flow, tissue Doppler of the lateral mitral annulus, and RVOT Doppler showing score calculation for 2 patients. Column A (top to bottom) shows LA enlargement, ↓E/e′, and normal RVOT Doppler profile(score=−2). Invasive hemodynamics: mPAP=40mmHg, PAWP=29mmHg, PVR=2.0mmHg/l/min. Column B demonstrates normal LA size, normal E/e′, and abnormal RVOT Doppler profile (score=+2). Invasive hemodynamics: mPAP=45mmHg, PAWP=10mmHg, PVR=8.8mmHg/l/min. LA=left atrial; mPAP=mean pulmonary artery pressure; PAWP= pulmonary arterial wedge pressure; PVR= pulmonary vascular resistance; RVOT= right ventricular outflow tract.

Figure 2

Tukey box plot of hemodynamics by score. A:pulmonary artery wedge pressure(PAWP), B:transpulmonary gradient(TPG), C: mean PA pressure(mPAP) and D: pulmonary vascular resistance(PVR). Plus sign (+) signifies mean.

Figure 3

Subject hemodynamic profiles (PVR and PAWP) by score. Horizontal and vertical dotted lines denote PVR=3WU and PAWP=15mmHg, respectively. PAWP<5 and PAWP>30 plotted as PAWP=5mmHg and =30mmHg respectively; PVR>18 plotted as PVR=18 WU.

Figure 4

Clinical and physiologic PH classifications predicted by the model (3A, 3C, 3E, 3G) versus estimated PASP (3B, 3D, 3F, 3H). 3A/B show WHO Group I versus II PH. 3C/D show elevated versus normal PVR. 3E/F do the same for normal versus elevated (>15mmHg) PAWP. 3G/H show classification of PH_PVD (PVR>3WU, PAWP ≤ 15mmHg)

Figure 5

ROC curves for prediction of PH_PVD for estimated PASP (A) and the echo score (B).

Figure 6

Classification and Regression Tree analysis. LA=left atrial dimension; AccT=acceleration time.