Background
Right ventricular (RV) function is an important determinant of mortality in patients with idiopathic pulmonary arterial hypertension (iPAH). The aim of this study was to serially evaluate global and regional RV two-dimensional strain and their relation to transplantation-free survival in children with iPAH.
Methods
RV regional and global longitudinal strain was retrospectively assessed in children with iPAH. Serial echocardiograms at 3 to 6 months from presentation and then at yearly intervals were analyzed. Results were compared with those from controls and between iPAH survivors (group 1) and those who died or needed transplantation (group 2). Survival stratified by RV global longitudinal strain at presentation was analyzed.
Results
Seventeen patients with iPAH (mean age, 8.4 ± 4.8 years; seven male patients), of whom 11 were alive (group 1) and six had died or undergone transplantation (group 2), and 17 age-matched controls were studied. The median follow-up period was 1.5 years (range, 0.04–7.8 years). RV global longitudinal strain was significantly reduced in patients with iPAH compared with controls (−13.5 ± 5.9% vs −24.4 ± 3.9%, P < .001) and in group 2 compared with group 1 at presentation (−9 ± 2.8% vs −16 ± 5.7%, P < .05) and throughout follow-up. During follow-up, RV global and regional longitudinal strain worsened in group 2, especially in RV apical segments (−6.3 ± 5% vs −1.9 ± 1.6% at presentation compared with the last echocardiographic assessment in group 2, P < .05), but was unchanged in group 1. RV global longitudinal strain > −14% predicted transplantation-free survival with 100% sensitivity and 54.5% specificity.
Conclusions
RV strain imaging may be useful for serial follow-up and prognostication in children with iPAH.
Pulmonary arterial hypertension (PAH) is a progressive disease that carries important morbidity and mortality. In children, idiopathic PAH (iPAH) carries worse prognosis than that associated with congenital heart disease. Right ventricular (RV) function is a key determinant of outcomes in patients with iPAH, and therefore, assessment of RV function is important. In a previous study, we examined the prognostic value of two-dimensional (2D) and M-mode echocardiographic parameters in children with iPAH or that associated with congenital heart disease. We found that mortality occurred only in the group with iPAH and that 2D and M-mode echocardiographic parameters of RV function may be useful to predict death in these patients. In that study, most parameters worsened on the last study in nonsurvivors and improved or remained unchanged in survivors. However, values at baseline were not very different between survivors and nonsurvivors, and in that study, we did not evaluate RV strain in relation to outcomes, which has been reported in the adult population as being a more sensitive indicator of function that predicts outcomes and allows more direct evaluation of myocardial performance.
Few studies have evaluated the usefulness of strain imaging to assess RV function in pediatric iPAH, and few studies have investigated its relationship to clinical outcomes in this population. Furthermore, to date, there has been no serial investigation of RV myocardial performance in children with iPAH.
Accordingly, the objective of this study was to investigate RV global and regional longitudinal systolic strain and its relation to death or lung transplantation in children with iPAH. We investigated the same group of patients with iPAH previously studied in relation to 2D and M-mode parameters, hypothesizing that RV function, as assessed by 2D speckle-tracking strain, is related to transplantation-free survival in children with iPAH.
Methods
Echocardiograms and patient records were retrospectively analyzed in children 0 to 18 years with iPAH, who were evaluated at our institution between 2004 and 2012. Patients were included if they met published criteria for diagnosis of PAH and did not have RV outflow obstruction, pulmonary stenosis, or left-sided obstructive lesions. Twelve patients were able to perform 6-min walk tests according to a standardized protocol. To calculate pulmonary vascular resistance (PVR), we performed right heart catheterization at presentation to our institution. PVR was calculated as the ratio of the difference between pulmonary arterial pressure and left atrial pressure or pulmonary arterial wedge pressure to mean cardiac output, expressed as units times body surface area (BSA) in square meters. As performed in our previous study, children with iPAH were divided into two groups: group 1 included children alive without lung transplantation at last follow-up, and group 2 included children who had died or undergone lung transplantation at the time of the study. We analyzed serial echocardiograms in all patients: at presentation to our institution, at 3 to 6 months after diagnosis, and at every 1-year follow-up visit and before death or transplantation in group 2. All indices were remeasured by a single investigator (K.O.) for the purposes of this study (SyngoDynamics version 5.1; Siemens Medical Solutions, Erlangen, Germany).
Echocardiography was performed on various ultrasound platforms used in our laboratory over the study period (HDI 300/5000 [Philips Medical Systems, Bothell, WA], Vivid 7 and Vivid E9 [GE Medical Systems, Milwaukee, WI], and iE33 [Philips Medical Systems, Andover, MA]). Compression and gain were adjusted to optimize the images. Sweep speeds were set at 50 mm/sec. Images were acquired during quiet respiration. RV systolic pressure was estimated from tricuspid regurgitation Doppler using the modified Bernoulli equation. We added an assumed right atrial pressure of 5 mm Hg to the RV–right atrial pressure gradient measured from Doppler. Tricuspid annular plane systolic excursion (TAPSE) was measured by M-mode echocardiography or from tricuspid annular maximal displacement in the four-chamber view when M-mode TAPSE was unavailable. RV end-diastolic and end-systolic areas, indexed to BSA, were measured from the apical four-chamber view. The RV fractional area change percentage was calculated as (end-diastolic area − end-systolic area)/end-diastolic area. The myocardial performance index was measured as described previously. The 6-min walk distance closest in time to the first echocardiographic examination was recorded. Clinical data at time of echocardiography was collected from the medical record.
Strain Echocardiography
Because we analyzed previously acquired 2D images in Digital Imaging and Communications in Medicine format (stored at a default setting of 30 frames/sec), which were acquired on different imaging platforms, we analyzed strain using TomTec Imaging Systems (Unterschleissheim, Germany) software. The software is similar in principle to 2D speckle-tracking–derived strain, tracking B-mode myocardial speckles, as well as the endocardium and other anatomic landmarks. RV strain was derived from a RV-focused four-chamber view. Strain curves of the basal, middle, and apical segments of the right lateral ventricular wall and septum were generated. Strain curves were accepted only after visual assessment confirmed good tracking. Longitudinal systolic strain values were calculated from an average of three cardiac cycles. We also calculated the average of the right lateral ventricular wall segments as average RV free wall strain.
Data Analysis
Variables are expressed as mean ± SD or as medians and ranges. Results were compared between patients with iPAH and controls using unpaired t tests and among iPAH group 1, iPAH group 2, and controls using one-way analysis of variance. RV regional strain values over time in children with iPAH were compared between the first and last echocardiographic examinations by paired t tests. Cutoff sensitivity and specificity values for RV global longitudinal strain from the initial echocardiographic study to predict transplantation-free survival were generated from receiver operating characteristic curves and the area under the curve as well as the hazard ratio calculated. Kaplan-Meier estimates of survival were then analyzed, and transplantation-free survival between iPAH patients above or below the echocardiographic cut points were assessed using Cox regression analysis.
Intraobserver and Interobserver Reliability Analysis
Intraobserver and interobserver reliability analysis of RV global longitudinal strain was assessed in a subset of eight randomly selected patients with iPAH. Measurements were repeated after 4 weeks by the same observer to assess intraobserver reliability. Two independent observers independently generated strain curves from the same cardiac cycle to assess interobserver variability. The coefficient of variation was used to express intraobserver and interobserver reliability.
Statistics were analyzed using SPSS version 19.0 (IBM, Armonk, NY). P values < .05 were considered significant. The study was approved by the institutional research ethics board.
Results
Baseline Characteristics
Seventeen children with iPAH were studied. Eleven patients (65%) were alive without transplantation at time of study (group 1), and six (35%) had died or undergone transplantation (group 2). Median follow-up duration for patients with iPAH was 2.5 years (range, 0.04–6.7 years) for group 1 and 1.3 years (range, 0.7–5.4 years) for group 2 ( P = .40). Demographic data for the study population are shown in Table 1 . There were no significant differences between group 1 and group 2 in PVR indexed to BSA and 6-min walk distance. The mean intervals from the initial echocardiographic examination to 6-min walk test and assessment of PVR indexed to BSA in patients with iPAH were 3.6 ± 5 and 0.95 ± 18 months, respectively. Most patients with iPAH were on two or more pulmonary vasodilators, especially children in iPAH group 2.
| Variable | iPAH (all patients) | iPAH group 1 | iPAH group 2 | P |
|---|---|---|---|---|
| Number | 17 | 11 | 6 | |
| Age (y) | 8.4 ± 4.8 | 8.9 ± 4.8 | 7.6 ± 5.3 | NS |
| Weight (kg) | 28.4 ± 1 | 26.6 ± 13.2 | 31.8 ± 21.3 | NS |
| BSA (m 2 ) | 0.97 ± 0.38 | 0.94 ± 0.34 | 1.04 ± 0.47 | NS |
| Female | 10 (59%) | 6 (55%) | 4 (67%) | NS |
| Heart rate (beats/min) | 96 ± 23 | 90 ± 18 | 107 ± 29 | NS |
| Systolic BP (mmHg) | 92 ± 12 | 91 ± 13 | 93 ± 11 | NS |
| Diastolic BP (mmHg) | 54 ± 8 | 52 ± 8 | 58 ± 7 | NS |
| 6-min walk distance (m) | 390 ± 79 | 384 ± 88 | 414 ± 13 | NS |
| PVRi (WU · m 2 ) | 23 ± 11 | 25 ± 13 | 20 ± 6 | NS |
| Number of medications | ||||
| 0 | 1 (9%) | 0 (0%) | ||
| 1 | 6 (55%) | 3 (50%) | ||
| 2 | 3 (27%) | 0 (0%) | ||
| 3 | 1 (9%) | 2 (33%) | ||
| 4 | 0 (0%) | 1 (17%) |
Seventeen age-matched control subjects were studied. There were no significant differences between controls and patients with iPAH in mean age (8.5 ± 4.8 vs 8.4 ± 4.8 years, control vs iPAH), body weight (28.4 ± 1 vs 26.6 ± 13.2 kg, control vs iPAH), BSA (0.97 ± 0.38 vs 0.94 ± 0.34 m 2 , control vs iPAH), and gender (female 59% vs 59%). The mean heart rate in patients with iPAH was significantly higher than in controls (96 ± 16 vs 79 ± 16 beats/min, respectively, P < .05, control vs iPAH).
Conventional Echocardiographic Data
Table 2 compares conventional echocardiographic parameters between controls and patients with iPAH. There were no significant differences between controls and patients with iPAH in left ventricular ejection fraction at the initial echocardiographic study.
| Variable | Control | iPAH (all patients) | iPAH group 1 | iPAH group 2 |
|---|---|---|---|---|
| LV EF (%) | 67.8 ± 4.2 | 69.1 ± 6.4 | 70.6 ± 6.5 | 66.3 ± 5.5 |
| LV SF (%) | 37.4 ± 3.1 | 38.2 ± 5.7 | 39.5 ± 5.6 | 35.9 ± 5.4 |
| RVSP (mm Hg) | NA | 99 ± 27 | 103 ± 28 | 93 ± 27 |
| Indexed RV end-diastolic area (cm 2 ) | 10.7 ± 2 | 19.9 ± 4.7 † | 19.8 ± 3.7 ‡ | 20 ± 6.6 ‡ |
| Indexed RV end-systolic area (cm 2 ) | 6 ± 1.2 | 16.1 ± 4.9 ‡ | 15.6 ± 3.9 ‡ | 17 ± 6.8 ‡ |
| RV FAC (%) | 44.2 ± 6.7 | 19.7 ± 9.3 ‡ | 21.4 ± 9.9 ‡ | 16.6 ± 7.8 ‡ |
| TAPSE (mm) | 15.9 ± 5.9 | 10.6 ± 4.4 ‡ | 12.2 ± 3.5 | 7.8 ± 4.6 † |
| RV E/A ratio | 1.6 ± 0.6 | 1.1 ± 0.3 ∗ | 1.1 ± 0.3 | 1.1 ± 0.4 |
| RV MPI | 0.12 ± 0.1 | 0.84 ± 0.34 ‡ | 0.72 ± 0.32 ‡ | 1.06 ± 0.27 ‡ |
| RV S/D ratio | 0.75 ± 0.16 | 0.91 ± 0.25 ∗ | 0.92 ± 0.25 | 0.89 ± 0.28 |
∗ P < .05 compared with control.
† P < .01 compared with control.
Most RV indices were significantly worse in patients with iPAH compared with controls at the initial echocardiographic study. Specifically, indexed RV end-diastolic area, indexed RV end-systolic area, RV fractional area change, RV myocardial performance index, and TAPSE were significantly reduced in patients with iPAH compared with controls. The RV E/A ratio and the RV S/D ratio were also significantly different between controls and patients with iPAH. Conversely, at the initial echocardiographic examination, TAPSE was not significantly different between controls and patients with iPAH alive at follow-up (group 1). There were no significant differences in any of the RV conventional echocardiographic parameters between group 1 and group 2 at the initial echocardiographic study.
RV Global Longitudinal Strain Assessment
RV global longitudinal strain at baseline echocardiography was markedly reduced in patients with iPAH compared with controls ( Figure 1 A). Likewise, RV global longitudinal strain in each subgroup of patients with iPAH was significantly reduced compared with controls ( Figure 1 B). Furthermore, RV global longitudinal strain in group 2 was significantly lower at presentation compared with group 1 ( Figure 1 B).
Serial Assessment of RV Global Longitudinal Strain and TAPSE
Serial RV global longitudinal strain over time in patients with iPAH is shown in Figure 2 A. RV global longitudinal strain in group 2 worsened over time, whereas RV global longitudinal strain in group 1 remained largely unchanged during follow-up. Consequently, RV global longitudinal strain in group 2 was significantly reduced compared with that in group 1 at 1- and 5-year follow-up.
Serial TAPSE over time in patients with iPAH is shown in Figure 2 B. Similar to the trend in RV global longitudinal strain, TAPSE in group 2 worsened over time, whereas TAPSE in group 1 improved gradually during follow-up. However, the difference in TAPSE over time between the groups was not as divergent as between RV strain over time.
RV Regional Longitudinal Strain in Patients with iPAH
RV regional longitudinal strain in controls and patients with iPAH was calculated at the first and last echocardiographic examinations ( Table 3 , Figure 3 ). At the first study, RV regional longitudinal strain was significantly reduced in all segments in both group 1 and group 2 patients with iPAH compared with controls ( Table 3 ). There were also significant differences between group 1 and group 2 in RV midlateral, midseptal, and average RV free wall strain. At the last echocardiographic study, all RV regional longitudinal strain values in group 2 were significantly reduced compared with those in group 1 ( Figure 3 ). When comparing the first and last echocardiograms, group 2 patients showed worsening of RV regional longitudinal strain in all segments, but especially at the apex and base, whereas RV regional strain was largely unchanged over time in group 1 patients with iPAH ( Figure 3 ).
