Background
Pulmonary hypertension (PH) increases right ventricular (RV) pressure, resulting in septal shift and RV dilation. Few echocardiographic measures have been used to evaluate severity and outcomes in children with PH. The aims of this study were to compare the RV to left ventricular (LV) diameter ratio at end-systole (RV/LV ratio) in normal controls and patients with PH, to correlate the RV/LV ratio with invasive hemodynamic measures, and to evaluate its association with outcomes in children with PH.
Methods
The RV/LV ratio was compared retrospectively between 80 matched normal controls and 84 PH patients without shunts. Of the patients with PH, 49 children underwent 94 echocardiographic studies and cardiac catheterizations within 48 hours (13 patients had simultaneous measurements). The RV/LV ratio was correlated against hemodynamic measures. Kaplan-Meier curves and a Cox proportional-hazards regression model were used to assess relationships between RV/LV ratio and time until an adverse clinical event (initiation of intravenous prostacyclin therapy, atrial septostomy, death, or transplantation).
Results
RV/LV ratios were lower in controls compared with patients with PH (mean, 0.51 [95% confidence interval, 0.48–0.54] vs 1.47 [95% confidence interval, 1.25–1.70], P < .01). The RV/LV ratio correlated significantly with mean pulmonary artery pressure, systolic pulmonary artery pressure, systolic pulmonary artery pressure as a percentage of systemic pressure, and pulmonary vascular resistance index ( r = 0.65 [ P < .01], r = 0.6 [ P < .01], r = 0.49 [ P < .01], and r = 0.43 [ P < .01], respectively). Twenty-two patients with PH with RV/LV ratios > 1 had adverse events within a median of 1.1 years from their earliest echocardiographic studies. Increasing RV/LV ratio was associated with an increasing hazard for a clinical event (hazard ratio, 2.49; 95% confidence interval, 1.92–3.24).
Conclusions
The RV/LV end-systolic diameter ratio can easily be obtained noninvasively in the clinical setting and can be used in the management of patients with PH. The RV/LV ratio incorporates both pathologic septal shift and RV dilation in children with PH and correlates with invasive measures of PH. An RV/LV ratio > 1 is associated with adverse clinical events.
Pulmonary hypertension (PH) is defined as mean pulmonary artery pressure ≥ 25 mm Hg at rest. PH is a progressive disease associated with elevations of pulmonary artery pressure and pulmonary vascular resistance, resulting in right ventricular (RV) dilation and dysfunction. The effect of RV dilation on the interventricular septum leads to a septal shift toward the left ventricle in systole. This abnormal interventricular septal shift has been previously described in patients with PH. In those with severe PH, progressive RV dilation and dysfunction result in RV failure and death. Hemodynamic data obtained by cardiac catheterization are used as a gold standard to assess patients with PH and to determine the severity of PH, but cardiac catheterization is an invasive procedure. The inherent risks of cardiac catheterization limit its use as a surveillance tool for children with PH.
Transthoracic echocardiography is a noninvasive method used to evaluate pulmonary hemodynamics and right heart function. It is cost effective, accessible, and safe. Several echocardiographic indices have been described to evaluate PH and its association with clinical outcomes. Tricuspid regurgitation velocity is the most feasible noninvasive technique to estimate systolic pulmonary artery pressure (sPAP) in patients with PH; however, its precision has been debatable, and it cannot be assessed in all patients. We sought to evaluate a new index, the RV to left ventricular (LV) diameter ratio at end-systole (RV/LV ratio), with the following objectives: (1) evaluate the feasibility of this index, (2) compare normal controls and patients with PH, (3) correlate with invasively obtained hemodynamic measures in PH, and (4) evaluate the association between RV/LV ratio and clinical outcomes in children with PH.
Methods
Study Population
Normal Controls
The University of Colorado Children’s Hospital Colorado pediatric normal echocardiographic database (per an institutional review board–approved protocol) was used to retrospectively identify 80 normal controls with similar age and gender distributions as the PH cohort. All 80 normal children were evaluated for heart murmurs and had normal results on echocardiography. RV/LV ratios were obtained in normal controls and compared with RV/LV ratios in patients with PH.
Patients with PH
The University of Colorado Children’s Hospital Colorado pediatric PH database was used to retrospectively identify 84 patients with PH who underwent echocardiographic evaluation, cardiac catheterization, and had outcomes between March 2006 and November 2012. Inclusion criteria were adequate echocardiographic images in parasternal short-axis views at the level of the papillary muscles and patients who could have any category of PH. Exclusion criteria were inadequate-quality echocardiographic images in which the right ventricle was cut off in the parasternal short-axis view, presence of intracardiac shunts, or presence of RV outflow tract obstruction. PH patients’ clinical statuses were obtained from their medical charts. This study was approved by the institutional review board at the University of Colorado.
Transthoracic Echocardiography
Echocardiography was performed on all normal controls and patients with PH using an iE33 (Philips Ultrasound, Bothell, WA) or a Vivid 7 (GE Medical Systems, Milwaukee, WI) ultrasound machine. Echocardiograms were digitally acquired using a standard protocol with appropriately sized transducers for patient size in all patients. RV/LV ratios were obtained offline from the parasternal short-axis two-dimensional view at the level of the papillary muscles with the RV free wall in view. RV and LV diameters were measured from the endocardial to endocardial surfaces at end-systole. RV/LV ratios were calculated ( Figure 1 ).
Echocardiographic images were analyzed offline using the Agfa Heartlab system (Mortsel, Belgium). Feasibility of RV/LV ratios was assessed in normal controls and patients with PH. RV/LV ratios were compared between normal controls and patients with PH. Forty-nine of 84 patients with PH underwent echocardiography within 48 hours of cardiac catheterization. In these 49 patients, 94 echocardiographic studies and cardiac catheterizations were performed between March 2006 and November 2012. Echocardiographic images were analyzed within 48 hours of cardiac catheterization, and a small subset of patients (13 patients) had simultaneous image acquisition in the cardiac catheterization laboratory. Echocardiographic studies from all 84 patients with PH were analyzed for clinical outcomes.
Cardiac Catheterization
Forty-nine patients with PH underwent 94 standard right-sided heart catheterizations within 48 hours of their echocardiographic assessments. Of these 49 patients, 13 underwent simultaneous echocardiographic studies and cardiac catheterizations. General anesthesia was used in 62 of 94 studies (65%), and the rest were performed under conscious sedation. Eighty-three of 94 studies (88%) were done with the patients on room air. Three studies were performed with the patients on 100% oxygen and 8 studies with the patients on 30% oxygen. Data collected from the catheterization procedure included sPAP, sPAP as a percentage of systemic pressure (%PAP), mean pulmonary artery pressure, pulmonary vascular resistance index, pulmonary capillary wedge pressure, right atrial pressure, and cardiac index. Cardiac output was obtained using thermodilution, and cardiac index was calculated. Pulmonary vascular resistance index was calculated as (mean pulmonary artery pressure − mean pulmonary wedge pressure)/cardiac index.
Clinical Outcomes
Clinical outcomes were analyzed in all 84 patients with PH who underwent echocardiography between March 2006 and November 2012. Clinical outcomes were retrospectively obtained from medical charts in patients with PH. An adverse clinical event was defined as (1) disease progression requiring the initiation of intravenous prostacyclin, (2) clinically indicated creation of an atrial septostomy, (3) death, or (4) transplantation. The earliest echocardiographic studies available (defined as the earliest echocardiograms from 2006 onward for each patient) and the most recent echocardiographic studies (or those performed just before adverse events) were analyzed for outcome analysis. A subgroup analysis of earliest echocardiographic studies was also done to evaluate clinical outcomes of patients who underwent cardiac catheterization within 48 hours and patients who did not.
Statistical Analysis
Descriptive statistics were calculated using percentages and means and standard deviations for categorical and continuous variables, respectively. Intraobserver and interobserver variability for the RV/LV ratio was assessed in 10 randomly selected studies. Intraobserver variability was based on measurements by the same observer (P.-N.J.) at different times (6 months apart). Interobserver variability was evaluated by comparing the results obtained by two independent observers (P.-N.J. and J.H.). Both were blinded to the clip numbers and images of the echocardiograms. Intraobserver and interobserver variability was evaluated using mean percentage error, defined by the absolute difference between observations divided by the mean of the observations. The mean RV/LV ratio values across control subjects and patients with PH were compared using a two-sample t test. Pearson’s correlation coefficients were used to assess cross-sectional correlations between matched echocardiographic and cardiac catheterization measurements. Bivariate mixed models were used to estimate the correlations between RV/LV ratios and catheterization measurements for all paired observations while adjusting for repeated measures. Kaplan-Meier curves were used to assess the cross-sectional relationships between RV/LV ratio, categorized at 1, and time until a clinical event. Kaplan-Meier curves were also used to assess the cross-sectional relationship between patients who did and did not undergo cardiac catheterization within 48 hours of their echocardiographic studies and time until a clinical event. Cox proportional-hazards regression model was fit using all RV/LV ratio measurements as time varying explanatory variables to estimate the association with clinical events. P values < .05 were considered statistically significant. All statistical analyses were performed using SAS version 9.3 (SAS Institute Inc, Cary, NC).
Results
The data consist of 80 echocardiograms in 80 normal controls and 194 echocardiographic measurements in 84 patients with PH, with a median of two observations per patient with PH (range, 1–6). Clinical diagnoses and medications in patients with PH are shown in Table 1 .
| Variable | Patients with PH ( n = 84) | Normal controls ( n = 80) |
|---|---|---|
| Age (y), mean ± SD | 8.7 ± 6.0 | 8.3 ± 5.5 |
| Male/female | 43/41 | 41/39 |
| Events | 22 (26%) | |
| IV prostacyclin | 11 | |
| Atrial septostomy | 6 | |
| Transplant | 1 | |
| Death | 4 | |
| IV medications | ||
| Treprostinil | 8 (9.5%) | |
| Epoprostenol | 3 (3.5%) | |
| Oral medications | ||
| Phosphodiesterase type 5 inhibitor | 41 (49%) | |
| Endothelin receptor antagonists | 20 (25%) | |
| Inhaled prostanoid | 7 (8.8%) | |
| Diagnosis | ||
| Idiopathic pulmonary arterial hypertension | 34 (40%) | |
| ASD s/p repair | 12 (14.2%) | |
| Pulmonary vein stenosis | 5 (6%) | |
| “Absent” LPA or RPA | 4 (5%) | |
| AVSD s/p repair | 4 (5%) | |
| DORV s/p repair | 1 (1.2%) | |
| TGA s/p repair | 1 (1.2%) | |
| Coarctation s/p repair | 1 (1.2%) | |
| PDA s/p closure | 1 (1.2%) | |
| Others (CDH, pulmonary fibrosis, altitude connective tissue disease, sickle cell) | 21 (25%) |
Feasibility of RV/LV Ratio
RV/LV ratios were obtained for all normal controls and all patients with PH. There were 3 of 194 echocardiograms (1%) in patients with PH for which RV/LV ratios were not obtained, whereas tricuspid regurgitation velocity could not be estimated on 29 (18%) echocardiograms. The RV/LV ratio measured in the parasternal short-axis views in end-systole was highly reproducible, with low intraobserver and interobserver variability (3.4% and 5.2%).
Comparison across Normal Controls and Patients with PH
Normal subjects had similar gender and age range as patients with PH ( Table 1 ). Eighty normal echocardiograms were compared with 84 earliest echocardiograms in patients with PH. The mean RV/LV ratio for the matched group of normal subjects (0.51; 95% confidence interval, 0.48–0.54) was significantly lower compared with the mean of earliest RV/LV measurements (1.47; 95% confidence interval, 1.25–1.70) from patients with PH ( P < .01; Figure 2 ).
Correlation between Echocardiographic and Hemodynamic Variables
Forty-nine patients underwent 94 echocardiographic studies and cardiac catheterizations within 48 hours. Cross-sectional correlations between RV/LV ratios and hemodynamic variables were calculated using the most recent echocardiograms for patients obtained within 48 hours outside the catheterization laboratory. The RV/LV ratio correlated significantly with mean pulmonary artery pressure, sPAP, %PAP, and pulmonary vascular resistance index ( Figure 3 , Table 2 ) when evaluated using both cross-sectional and repeated-measures data. In patients with multiple echocardiograms (repeated measurements), there were no significant changes in RV/LV ratios over time on average (slope estimate, −0.01; standard error, 0.02). There was no statistically significant correlation of RV/LV ratio with pulmonary capillary wedge pressure, right atrial pressure, or cardiac index.
| Hemodynamic variable | RV/LV ratio (single measure) | RV/LV ratio (repeated measures) | ||||
|---|---|---|---|---|---|---|
| n | Pearson’s correlation | P | n | Pearson’s correlation | P | |
| mPAP | 49 | 0.65 | <.01 | 94 | 0.67 | <.01 |
| sPAP | 49 | 0.63 | <.01 | 94 | 0.67 | <.01 |
| %PAP | 49 | 0.49 | <.01 | 94 | 0.61 | <.01 |
| PVRi | 49 | 0.48 | <.01 | 94 | 0.70 | <.01 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
