San Francisco, March 10–12, 2016
Address correspondence to Dr. Jeffrey Fineman, UCSF Benioff Children’s Hospital, 513 Parnassus Avenue, HSE-1418, Box 1346, San Francisco, California 94143-1346, USA (jeff.fineman@ucsf.edu).
. 2016 Sep;6(3):389–396. doi: 10.1086/687256
Pulm Circ. 2016;6(3):389–396.
Real-time three-dimensional echocardiography (RT3DE) evaluation of the right ventricle (RV) decreases geometric assumptions from conventional echocardiography. RV strain has been demonstrated to be more sensitive in evaluation of RV function than conventional echocardiography. TomTec 4D RV-Function 2 software is vendor neutral and can generate RV size and functional parameters. We compared 4D RV-Function 2 data between controls and pediatric pulmonary hypertension (PH) patients. We also compared tricuspid annular plane excursion (TAPSE) and fractional area change (FAC) measured by 4D RV-Function 2 and by traditional two-dimensional echocardiography. We enrolled 145 children with PH with 161 visits (mean age: 11.5 ± 6.9 years; 47% male) and 13 controls (mean age: 8.6 ± 2.5 years; 66% male) who underwent RT3DE evaluation of the RV in 2014–2015. Offline analysis of the 4D RV-Function 2 data of these patients generated RV indices: 3D ejection fraction (EF), RV septal and free-wall longitudinal strain (RVLS), TAPSE, and FAC. Differences in RV indices between PH patients and controls were assessed with linear mixed models, adjusting for age and accounting for repeated measures. A t test was performed on the TAPSE and FAC from traditional two-dimensional echocardiography and compared to the TAPSE and FAC generated from the 4D RV-Function 2 software in pediatric PH patients. TAPSE is measured in 4D RV-Function 2 by taking the difference between the length from the apex to the tricuspid valve annulus in systole and that in diastole. Intraclass correlations (ICCs) between two users were tested. There were significant differences in RV indices between PH patients and controls, with the exception of TAPSE. There were also significant differences in TAPSE and FAC between the two-dimensional echocardiography and 4D RV-Function 2. The ICCs showed good agreement for EF and RV free-wall longitudinal strain (0.92 and 0.82, respectively). The ICCs showed moderate agreement in RV septal longitudinal strain, TASPE, and FAC (0.62, 0.66, and 0.64, respectively). Pediatric PH patients have impaired RV function compared to normal children. The 4D RV-Function 2 software is feasible and generates RV functional parameters from one RT3DE acquisition, making it easier for clinical use. Differences between the values of TAPSE and FAC from the two methods will need to be evaluated further to define the normal values generated from 4D RV-Function 2, as this method measures TAPSE differently than the traditional M-mode measurement of TAPSE.
Pulm Circ. 2016;6(3):389–396.
Bronchopulmonary dysplasia–associated pulmonary hypertension (BPD-PH) is a prominent source of morbidity and mortality in the premature infant population. Both hypoxia and hyperoxia have been implicated as potential contributors to the pathogenesis of BPD-PH. The optimal oxygen saturation target range to minimize risk for BPD-PH is in dispute. A large, single-center neonatal intensive care unit has changed its pulse oximetry targets twice in the past 7 years for reasons unrelated to PH risk. We hypothesized that exposure to increased pulse oximetry targets would result in lower rates of BPD-PH. We retrospectively compared rates of PH and use of PH medications (sildenafil, inhaled nitric oxide [iNO], and bosentan) in at-risk infants exposed to varying saturation targets. Infants with birth weight of <1,500 grams and gestational age of 23–32 weeks were divided into 3 cohorts: group 1, May 1, 2009–April 30, 2011, goal saturation: 85%–94%; group 2, May 1, 2011–May 31, 2013, goal saturation: 88%–94%; group 3, June 1, 2013–May 31, 2015, goal saturation: 90%–95%. There was a significant reduction in the incidence of PH with higher saturation targets (86/463 [15.7%] in group 1, 30/494 [5.7%] in group 2, 28/403 [6.5%] in group 3; P < 0.001). Infants in group 2 (OR: 0.33 [95% CI: 0.21–0.51], P < 0.001) and group 3 (OR: 0.37 [95% CI: 0.24–0.59], P < 0.001) had significantly reduced odds of PH compared to infants in group 1. Use of iNO was also reduced in the higher-saturation-target groups (13/536 in group 1, 9/515 in group 2, 8/423 in group 3, P = 0.004), while bosentan and sildenafil use were not statistically different. Increasing the goal saturation range for premature infants at risk for BPD-PH is associated with reduced rates of PH diagnosis and reduced use of iNO. Era bias may have contributed to the findings. A more comprehensive analysis to account for potential confounders is underway.
Pulm Circ. 2016;6(3):389–396.
In children destined for single-ventricle physiology, increased PVR may halt surgical progression or result in failing Fontan. The use of pulmonary vasodilators, even at pressures of <25 mmHg, may seem reasonable. However, knowledge on indication, dosage, and effect is limited. All (n = 30) children with single-ventricle physiology treated with vasodilators by the UK Pulmonary Hypertension Service for Children in 2002–2014 were retrospectively reviewed. Mean follow-up time was 4.0 years (range: 0.5–10.9). There was no mortality and no significant ventricular dysfunction or severe AV valve regurgitation. Therapy was initiated after stage 1 (n = 11), Glenn (n = 7), or TCPC (n = 12). In stage 1 or Glenn, treatment indications were mainly increased mPAP. In TCPC, indications were PLE, ascites, plastic bronchitis, or cyanosis/reduced exercise tolerance. Average dose of phosphodiesterase inhibitor was 2.0 mg/kg/day and that of endothelin receptor antagonist (ERA) 3.4 mg/kg/day; 60% of the children received dual therapy. Single therapy reduced mPAP from 18.5 to 13.9 mmHg (n = 8, P = 0.004), and dual therapy from 20.2 to 14 mmHg (n = 6, P = 0.002). Five of 11 patients progressed from stage 1 to Glenn or Kawashima, and 2 of 7 from Glenn to TCPC. Primary therapy with either one or two vasodilators yielded a significant increase in saturation, from a mean of 79% to 85% (P = 0.001). Sequential addition of a second vasodilator was effective only in the Glenn group. The only severe side effect observed was a single case of reversible renal failure secondary to systemic hypotension. ERA was discontinued in 9 of 19 patients because of insufficient effect (n = 6), side effects (n = 1), or surgical progression (n = 2), and saturations remained stable. Pulmonary vasodilator therapy appears to be safe and associated with improvements in oxygenation and mPAP in children with single-ventricle physiology. It is, however, also safe to attempt weaning when no clear benefit is observed.
Pulm Circ. 2016;6(3):389–396.
The presence of pulmonary artery hypertension (PAH) in Fontan patients is associated with increased morbidity and mortality. Our main objective is to correlate pulmonary vascular resistance (PVR) and transpulmonary gradient (TPG) with cardiac index in Fontan patients. We suspect that patients with pulmonary artery hypertension would have lower cardiac index and worse hemodynamics. We conducted a retrospective analysis of cardiac catheterization hemodynamics of 110 Fontan patients. The median age at Fontan operation was 2.98 years, and the median age of catheterization study was 9.3 years; 31% of the patients were females, and 29% had hypoplastic left heart syndrome. The TPG (R = 0.22; P = 0.02) and PVR (R = −0.31; P = 0.0008) correlated with cardiac index. Thirty-two patients (29%) had TPG ≥ 6 (high-TPG group), while 10 patients (10%) had both TPG ≥ 6 and PVRi ≥ 3 (PAH group). The high-TPG group had a cardiac index similar to that of those with normal TPG (P = 0.13). The PAH group was found to have a lower cardiac index (P = 0.052), lower pulmonary blood flow (P = 0.03), lower pulmonary artery (PA) saturation (P = 0.036), and a lower PVR/SVR ratio (P = 0.0002), compared to the rest of the cohort. They also had a higher age at Fontan (P = 0.046). On multivariate regression analysis, the presence of PAH correlated with lower cardiac index (P = 0.006; OR: 0.32) and lower systemic artery saturation (P = 0.005; OR: 0.76). Sixteen patients (14.5%) had elevated PVR. This high-PVR group even showed significantly lower cardiac index, pulmonary blood flow, PA saturation, and aortic saturation as well as higher TPG, compared to the rest. PVR inversely correlated with cardiac index. Using PVRi ≥ 3 to define PAH will likely identify a more hemodynamically significant pulmonary vascular disease in Fontan patients. This supports the role of selective pulmonary vasodilators in improving cardiac performance in this high-risk group.
Pulm Circ. 2016;6(3):389–396.
The association between pulmonary hypertension (PH) and obstructive sleep apnea (OSA) in children is poorly understood. This study aims to assess characteristics of pediatric patients with PH and OSA. Retrospective data were collected from January 2003 to March 2013 from patients up to 21 years of age seen at one center with ICD-9 diagnoses of primary PH, secondary PH, or congestive heart failure due to PH. Patients were deemed to have PH from echocardiographic and right heart catheterization (RHC) reports describing (1) elevated estimated pulmonary artery pressures (PAP) by tricuspid regurgitant jet velocity or indirect evidence of elevated PAP on echocardiography and (2) increased mean PAP (mPAP) of ≥25 mmHg on RHC. Data were then extracted for patients with OSA diagnosed by polysomnography but without heart disease. OSA was a rare cause of PH (0.68%), with mild to moderate severity. PH was reversible in 3 of 3 children who had OSA treated and underwent repeat echocardiography. One patient did well with noninvasive ventilation. All children with PH and OSA had comorbidities. Of these, trisomy 21 is associated with PH. Prader-Willi Syndrome (PWS), seen in 2 children, has no reported association with PH. The prevalence of PH in children with OSA was low. Children with OSA and trisomy 21 or PWS should have a higher index of suspicion for PH.
This work was presented as a poster at the 29th annual meeting of the Associated Professional Sleep Societies, Seattle, June 6–10, 2015.
Pulm Circ. 2016;6(3):389–396.
Results of acute vasodilator testing (AVT) in patients with segmental pulmonary vascular disease (SPVD) have not been described. We sought to determine whether oxygen (O2), O2 with nitric oxide (NO), and diltiazem cause a similar decrease in mean pulmonary arterial pressure (MPAP) in patients with pulmonary hypertension associated with SPVD. Retrospective review of all subjects with MPAP > 25 mmHg between January 2000 and December 2015 was performed. Subjects with SPVD involving at least two major lung segments due to pulmonary venous or arterial obstruction were included. Baseline hemodynamic measurements were performed with patients breathing 21%–30% O2 (interval 1). AVT was performed using 100% O2 (interval 2). Selected patients were also evaluated using 100% O2 with 20 ppm nitric oxide (O2 with NO; interval 3), 21%–30% O2 (interval 4), and 21%–30% Fio2 with intravenous diltiazem at a dose of 30–60 μg/kg/min (interval 5). A t test was used to compare the change in MPAP between intervals (significant at P < 0.05). Eight of 16 patients were deemed acutely responsive during O2 or O2 with NO, based on a decrease in MPAP of ≥10 mmHg, to a value of <40 mmHg. Two patients were deemed acutely responsive to diltiazem. The change in MPAP with O2 or O2 with NO (19 ± 12 mmHg) was significantly greater than the change with diltiazem (7 ± 5 mmHg). O2, O2 with NO, and diltiazem decrease MPAP in some patients with SPVD. However, the acute response to diltiazem suggests that an optimal response to therapy will not be achieved with a calcium channel blocker alone.
Pulm Circ. 2016;6(3):389–396.
Elevated pulmonary vascular resistance (PVR), elevated transpulmonary gradient, and pulmonary vein stenosis are typically contraindications to cavopulmonary anastomosis. Few data exist regarding the use of pulmonary vasodilator therapy as a bridge to second-stage single-ventricle palliation. The study includes a case report and review of the literature. A 10-month-old ex-36-week female diagnosed prenatally with hypoplastic left heart syndrome (HLHS), initially determined to be ineligible for second-stage palliation, underwent both catheterization-based and medical therapies to address her pulmonary hypertension. Palliative bilateral pulmonary artery (PA) banding was performed within the first week of life. Postoperative respiratory failure prompted early catheterization, which revealed left lower pulmonary vein stenosis. Findings from catheterization precluded the infant from further palliative surgery and cardiac transplant. After prostaglandin therapy was discontinued, she was discharged home with comfort care, including furosemide, oxygen, and enteral feeds. At 7 months of age, she presented to our institution, growing and meeting developmental milestones despite worsening oxygen saturation. Repeat catheterization revealed a large PDA, severely elevated PVR, restrictive atrial septum, and no evidence of pulmonary vein stenosis. Subcutaneous (SQ) treprostinil and bosentan were started. Treprostinil was titrated to a maximum dose of 113 ng/kg/min. Atrial septal stent placement, subsequent balloon dilation, and pulmonary vasodilator therapy resulted in significantly improved PA pressures and PVR. At 10 months, she underwent Damus-Kaye-Stansel anastomosis, cavopulmonary anastomosis, and atrial septectomy. Treprostinil was transitioned from SQ to intravenous (IV) perioperatively. Her postoperative course was uncomplicated, and she was discharged home on SQ treprostinil, bosentan, and sildenafil. Cardiac catheterization and atrial stent placement reduced left atrial pressure, facilitating initiation and uptitration of treprostinil to achieve favorable hemodynamics before second-stage palliation in this patient with high-risk HLHS. Aggressive pulmonary vasodilator therapy, including high-dose treprostinil, effectively reduced PA pressure and PVR in this patient, who would not otherwise have been a candidate for cavopulmonary anastomosis. Rapid titration of SQ treprostinil and perioperative transition to IV treprostinil were feasible and well tolerated in this case.
Pulm Circ. 2016;6(3):389–396.
Premature birth disrupts pulmonary vascular growth and development. The long-term consequences of altered pulmonary vasculature on pulmonary hemodynamics beyond the neonatal period are not well known. Doppler echocardiography–derived pulmonary artery acceleration time (PAAT), which inversely correlates with pulmonary artery pressure and resistance, is a reliable quantitative index of pulmonary hemodynamics. The objective of this study was to assess and compare pulmonary hemodynamics by PAAT at 1 year of age in preterm-born and term-born infants. In a prospective study of 80 preterm infants (<29 weeks gestational age at birth) enrolled through the Premature and Respiratory Outcomes Program (NCT01435187), PAAT was measured at 1 year corrected age and compared to those of 100 age- and weight-matched term-born healthy infants. PAAT was derived from pulse Doppler interrogation across the pulmonary valve in the parasternal short-axis view and calculated as the time interval (ms) from onset of systolic pulmonary arterial flow to peak flow velocity. PAAT was adjusted for heart rate against right ventricular ejection time (RVET) using the ratio PAAT∶RVET. Chronic lung disease (CLD) was diagnosed in 48 infants (60%), according to the 2001 NIH workshop definition. All preterm infants had significantly lower PAAT and PAAT∶RVET than term-born infants at 1 year corrected age (73 ± 8 vs. 98 ± 5 ms and 0.31 ± 0.04 vs. 0.40 ± 0.02; P < 0.001). Preterm-born infants with CLD had lower PAAT and PAAT∶RVET than those without CLD (69 ± 5 vs. 79 ± 4 ms and 0.29 ± 0.02 vs. 0.33 ± 0.06; P < 0.01) and significantly lower values than normal-term-born infants at 1 year of age (69 ± 5 vs. 98 ± 5 ms and 0.29 ± 0.02 vs. 0.40 ± 0.02; P = 0.002). Data analysis included adjustment for gestational age at birth, sex, heart rate, and presence of significant left-to-right shunt. Preterm-born infants exhibit abnormal pulmonary hemodynamics at 1 year of age irrespective of neonatal lung disease status. This suggests persistence of pulmonary vascular dysfunction in preterm-born infants beyond infancy that warrants long-term follow-up. PAAT measurements may offer a reliable noninvasive tool for longitudinal monitoring of pulmonary hemodynamics in infants.
Pulm Circ. 2016;6(3):389–396.
Vascular reactivity determines responsiveness to vasodilators and can alter prognosis in pulmonary hypertension (PH). We hypothesized that vascular reactivity is impaired in the Sugen-hypoxia model of PH. Adult male Sprague-Dawley rats injected with Sugen (20 mg/kg sc) or vehicle were exposed to hypoxia (Hx; 9% O2) for 3 weeks followed by 4 weeks in normoxia (experimental groups: SU/Hx, Veh/Hx, and normoxic controls [Norm]). We performed hemodynamic measurements and isolated the pulmonary artery, aorta, and mesenteric arteries for vascular-function studies. We also assessed right ventricular hypertrophy (Fulton’s index). Statistical analysis was done with ANOVA, and mean ± SEM values are presented. Right ventricular systolic pressure was higher in SU/Hx than in Veh/Hx and Norm animals (47.18 ± 2.86 vs. 26.10 ± 0.89 vs. 22.35 ± 1.24 mmHg, P < 0.05). Fulton’s index was higher in SU/Hx than in Veh/Hx and Norm animals (0.56 ± 0.02 vs. 0.31 ± 0.01 and 0.26 ± 0.01, P < 0.05). In isolated pulmonary artery rings, phenylephrine (Phe)-induced maximal contraction was reduced in SU/Hx compared to Veh/Hx and Norm animals (0.11 ± 0.02 vs. 0.17 ± 0.02 vs. 0.20 ± 0.00 g/mg tissue). Pulmonary artery contraction to high KCl (96 mM) was less in Su/Hx than in Veh/Hyx and Norm animals (0.14 ± 0.04 vs. 0.18 ± 0.03 vs. 0.23 ± 0.02 g/mg tissue). Maximal acetylcholine (ACh)-induced relaxation was also less in Su/Hx than in Veh/Hx and Norm animals (33% ± 5% vs. 43% ± 8% vs. 62% ± 3%), suggesting reduced endothelium-dependent vasodilation. ACh relaxation was inhibited via NOS and guanylate cyclase inhibition in all groups, suggesting a role of the NO-cGMP pathway. Nitrate/nitrite production in response to ACh was less in Su/Hx than in Veh/Hx and Norm animals (maximum: 104 ± 18 vs. 423 ± 58 vs. 895 ± 224 pmol/mg tissue), supporting reduced endothelium-derived NO production. Sodium nitroprusside (10−8 M) caused less relaxation in Su/Hx than in Veh/Hx and Norm animals (24% ± 3% vs. 58% ± 3% vs. 59% ± 5%), suggesting decreased responsiveness of vascular smooth muscle (VSM) to vasodilators. Neither contraction nor relaxation differed in the aorta or mesenteric arteries of all groups. There is reduced responsiveness to vasoconstrictors and dilators in the pulmonary, but not the systemic, vessels of SU/Hx-treated rats. This may be due to a phenotypic switch of pulmonary VSM cells to a synthetic and less reactive phenotype and reduction of endothelial NO-cGMP relaxation pathways in this model of severe PH.
Pulm Circ. 2016;6(3):389–396.
Pulmonary veno-occlusive disease (PVOD) is a rare, severe form of pulmonary hypertension. Lung transplantation is recommended as therapy. We hypothesized that pediatric patients with PVOD might be an appropriate indication for lung transplantation. The study population consisted of 6 patients with PVOD; age range was 6.4–18.0 years. Three patients were female. Initial clinical diagnoses were asthma (4), idiopathic pulmonary arterial hypertension (1), and an asymptomatic sibling of a patient with PVOD (1). Genetic testing identified mutations in 2 siblings in the EIF2AK4 gene. One other patient had positive family history. At referral, 2 patients were receiving pharmacotherapy, and 1 had iloprost added to his regimen while in critical condition. Six-minute walk tests in 4 patients showed distances below predicted values and hypoxemia (lowest Sao2 range: 78%–85%). Echocardiography showed RV dysfunction in 5 patients, with moderately to severely depressed function in 3. Pulmonary vascular resistance ranged from 5 to 39 WU/m2 (median: 20 WU/m2). Pulmonary angiography in 1 patient showed capillary blush that suggesting pulmonary capillary hemangiomatosis. Chest CT scanning was distinctively abnormal, with ground-glass densities in a centrilobular distribution, mediastinal adenopathy in 3 patients, and septal lines in 5 (absent in 1). Five patients were listed for transplantation; 1 died of right ventricular failure. Four waited 3–18 days for donor organs—3 critically ill in ICU, 1 on ECMO—before bilateral lung transplantation. All survived surgery and are alive 7–50 months after surgery. Echocardiography normalized in all surviving patients. We believe that PVOD should be recognized as a distinct, potentially malignant form of PH in children. Referral for lung transplantation should be an urgent consideration. Mid-term outcomes after lung transplantation seem reasonable.
Pulm Circ. 2016;6(3):389–396.
Pulmonary veno-occlusive disease (PVOD) is a rare form of pulmonary hypertension. Characteristic features that distinguish PVOD from idiopathic pulmonary arterial hypertension relate to histopathology and radiographic findings. We hypothesized that a pediatric PVOD cohort might be distinctive. Six patients were referred to us for lung transplantation with a diagnosis of PVOD. Histopathology was available for 5 and chest CT scans for 6. The clinical features of the 6 patients were presented in the previous abstract. Four patients underwent lung transplantation; explant histopathology was available for review. One patient had an autopsy after death on the wait list. Patient 3 showed diffuse sickling of RBCs in the alveolar spaces. He was known to have sickle cell trait before transplantation, and, retrospectively after transplant, a diagnosis of sickle cell trait with alpha thalassemia—a benign combination according to hematology consultants—was made. CTs showed diffuse centrilobular ground-glass opacities. Mediastinal adenopathy was noted in 3 patients. Septal lines were prominent in 3, mild in 2, and absent in the asymptomatic sibling. During cardiac catheterization, prominent capillary blush was noted in 1 patient. In the patient with sickle-alpha thalassemia, the CT findings showed more paucity of vascular markings and more diffuse ground-glass opacities. The histopathology and CT findings in this cohort of pediatric patients with PVOD are similar to those in an adult series. Important interindividual differences exist. CT scan alone is diagnostic.
Pulm Circ. 2016;6(3):389–396.
Pulmonary hypertension (PH) is difficult to recognize clinically. Simultaneous collection of heart sounds with digital stethoscopes offers an opportunity to reevaluate the auscultatory features that may lead to a diagnosis of PH. We hypothesized that the synchrony of simultaneously collected heart sounds would differ between children with PH and those without PH. We recorded heart sounds simultaneously from the 4 traditional auscultatory areas on the chest, using a digital stethoscope, in 39 subjects (20 males) with a median age of 53 years (range: 5 months–85 years) undergoing cardiac catheterization. Seventeen subjects had a mean pulmonary artery pressure (mPAp) of <25 mmHg (9–22 mmHg). Twenty-two subjects had an mPAp of ≥25 mmHg (25–56 mmHg). We applied the synchrony measures with leave-one-out cross validation to the heart sounds to examine the frequency domain through full-frequency directed transfer function (ffDTF). The significance of the results was determined with a rank-sum test. The ffDTF increased significantly in subjects with mPAp ≥ 25 mmHg versus those with mPAp < 25 mmHg, with a P value of <0.005. The increased ffDTF suggests that the heart sounds of PH patients demonstrate greater synchronicity than those of subjects with a normal PA pressure undergoing cardiac catheterization within the optimized frequency band 13–30 Hz. This finding may provide new insights into the cardiovascular physiology of PH and alternative methods of diagnosis using heart sound recordings.
Pulm Circ. 2016;6(3):389–396.
We sought to determine whether N-terminal pro brain natriuretic peptide (NT-proBNP) measurements in children with pulmonary hypertension (PAH) predicted decreased right ventricular (RV) function. We analyzed retrospectively the health records of all children with PAH who had NT-proBNP levels measured at cardiac catheterization and on follow-up from 2008 to 2015. We recorded Panama functional class (PFC), tricuspid annular plane systolic excursion z score (TAPSE z score), RV fractional area change (RVFAC), hemodynamic data, and need for hospitalization. Hospitalization due to worsening PAH was defined as need for additional or up-titration of PAH-specific therapy or death. Fifty-three patients (median age: 46 months [range: 2–204], 30 males) were identified. At baseline, mean pressures were as follows: right atrial: 5 ± 2 mmHg; aortic: 68 ± 15 mmHg; pulmonary artery: 48 ± 20 mmHg; and wedge: 7 ± 3 mmHg. Pulmonary vascular resistance index was 12 ± 8 Wood units·m2. Patients with an NT-proBNP value of >40 pmol/L were more likely to be in PFC IIIa, IIIb/IV (n = 17 vs. 4, P = 0.004) and to have a decreased TAPSE z score (mean −2.1 ± 2.09 vs. −0.5 ± 2.08, P = 0.008) and RVFAC (mean RVFAC: 27% ± 11.3% vs. 37.2% ± 6%, P = 0.0006). Hospitalization occurred in 18 patients (32%) during mean follow-up of 12 (2–98) months. Patients with PAH-related admissions (n = 9) had a median RVFAC of 21% (14%–44%) versus 33% (16%–56%), P = 0.02; a median TAPSE z score of −4.4 (−6 to 3.5) versus −1 (−1.7 to 0.2), P = 0.04; and increased NT-proBNP levels (median: 451 [29–2,632] vs. 20 [8–117] pmol/L, P < 0.02). NT-proBNP levels of >40 pmol/L during subsequent hospitalization predicted worsening PAH with 95% sensitivity and 85% specificity. A baseline NT-proBNP level of >40 pmol/L was associated with higher PFC and decreased RV function. An NT-proBNP level of >40 pmol/L during hospitalization predicted worsening PAH and RVFAC.
Pulm Circ. 2016;6(3):389–396.
Right ventricular–pulmonary vascular (RV-PV) dysfunction may exist in extremely preterm infants with chronic lung disease (CLD). Reliable, noninvasive quantitative markers for early recognition of CLD that longitudinally track RV-PV hemodynamics are lacking in preterm infants. The aim of this study was to evaluate novel quantitative echocardiographic measures of RV-PV hemodynamics as early predictive markers in infants at risk for development of CLD and compare them to conventional echocardiographic measures. A prospective study was conducted in 115 preterm infants (<29 weeks gestational age at birth) enrolled through the Prematurity and Respiratory Outcomes Program (NCT01435187). Echocardiographic measures of RV-PV hemodynamics were measured at 5 time points: 1 and 3 days (n = 30), 32 and 36 weeks postmenstrual age (PMA; n = 115), and 1 year corrected age (n = 80). RV function was assessed with fractional area of change (FAC), tricuspid annular plane systolic excursion (TAPSE), and global and free-wall longitudinal strain. RV morphology was evaluated with RV dimensions and areas. Pulmonary hemodynamics were assessed by pulmonary artery acceleration time (PAAT). Conventional measures (tricuspid regurgitation velocity and septal flattening) were evaluated for comparison. CLD was classified by the NIH workshop definition. RV longitudinal global and free-wall strain, FAC, TAPSE, and PAAT were significantly reduced, and RV morphometrics were significantly larger at 32 and 36 weeks PMA in infants who developed CLD. These findings persisted to 1 year corrected age. Conventional measures were present in less than 20% of patients, and only septal flattening was significant at 32 weeks PMA in the CLD group. Quantitative echocardiographic measures of RV-PV function and hemodynamics are reliable early predictive markers of subsequent development of CLD in preterm infants. These measures are superior to commonly used conventional measures and can be useful tools for the early identification and longitudinal tracking of cardiopulmonary function in preterm infants at high risk for CLD.
Pulm Circ. 2016;6(3):389–396.
Respiratory syncytial virus (RSV) infection is the most frequent cause of acute lower respiratory tract disease in infants. Reports have suggested that infants with congenital heart disease (CHD), chronic lung disease (CLD), prematurity, Down syndrome, or pulmonary hypertension (PH) are associated with increased severity of RSV infection. We sought to determine the incidence of PH and assess the severity of illness associated with PH in RSV-infected infants. Echocardiograms of 155 children admitted with RSV bronchiolitis were retrospectively reviewed. Indirect measures of PH were used to categorize subjects with PH, including peak tricuspid regurgitation gradient, interventricular septal curvature geometry, early and end-pulmonary insufficiency gradient, and the left ventricle–to–right ventricle ratio in systole in short-axis view. Geographic data were collected. The association with PH and clinical outcomes, including mortality, ICU admission, prolonged ICU stay (>10 days), tracheal intubation, and need of high-frequency oscillator ventilation (HFOV), were evaluated. Echocardiography detected PH in 29 patients (18.7%). PH was observed more frequently in patients with CHD (11/33), CLD (12/25), prematurity (<38 wGA; 17/59), and Down syndrome (4/10). However, even in patients without these risk factors (n = 68), 5 patients had PH, and PH was associated with ICU admission (P = 0.003) and HFOV ventilation (P = 0.013). Multivariate logistic analysis demonstrated that PH is a risk factor for mortality (P = 0.019, OR: 5.63 [95% CI: 1.33–23.8]), ICU admission (P = 0.024, OR: 3.32 [95% CI: 1.16–9.44]), prolonged ICU stay (P = 0.019, OR: 3.01 [95% CI: 1.19–7.59]), and HFOV (P = 0.002, OR: 6.2 [95% CI: 1.91–20.2]). Presence of PH during RSV bronchiolitis is a risk factor for morbidity and mortality. PH was associated with increased severity of RSV illness even after exclusion of the above risk factors.
Pulm Circ. 2016;6(3):389–396.
Our objectives were to determine prevalence and risk factors for pulmonary hypertension (PH) in preterm infants with bronchopulmonary dysplasia (BPD) by prospective echocardiographic screening and to evaluate the role of biochemical markers in screening for PH. The study was a prospective institutional echocardiographic PH screening at 36–38 weeks corrected gestational age (GA) for infants born at <32 weeks GA who were diagnosed with BPD. A subgroup of 20 infants with <750-g birth weight (BW) and/or <27 weeks GA was tested for quantitative plasma amino acids and NT-proBNP. In total, 159 infants were screened (GA: 25.9 ± 2 weeks, BW: 831 ± 286 g). The PH prevalence was 28% (44/159). BW and GA were lower in infants with PH than in those with no PH. After correction for BW and GA, necrotizing enterocolitis (NEC), severe intraventricular hemorrhage, atrial septal defect, and mortality were independently associated with PH. Screening only infants with NEC and those with BPD who had a BW of <840 g would have yielded an 86% (38/44) sensitivity for detecting PH and reduced the number of screening echocardiograms performed by 31%. In the subgroup of 20 infants who had additional biochemical screening, 25% (5/20) had evidence of PH by echocardiography. NT-proBNP levels of >1,000 pg/mL and citrulline levels of <29 μmol/L predicted PH with 100% sensitivity. In our institution, PH in preterm infants with BPD is associated with lower GA and BW, as well as NEC and severe IVH after correction for GA and BW. If these results are validated in a larger study, NT-proBNP may replace echocardiography as a PH screening tool in infants at lower risk for PH. The lower citrulline levels seen in infants with PH may have therapeutic utility in the future.
Pulm Circ. 2016;6(3):389–396.
Inoperable pulmonary venous obstruction is a rare, though serious, medical condition leading to a significant risk of morbidity and mortality in children. Hemoptysis may occur secondary to bronchial venous congestion. Pneumonectomy has resulted in resolution of both bronchial varices and hemoptysis. Here we report the use of a vascular plug to limit flow to a segment of lung in order to alleviate hemoptysis in a 15-year-old with Down syndrome and an occluded left lower pulmonary vein. The patient developed varying degrees of stenosis of all pulmonary veins during infancy. Surgery alleviated the stenosis of the right pulmonary veins. He had left lower pulmonary vein occlusion and a stable degree of severe left upper pulmonary vein stenosis that did not have a lasting response to balloon dilation, based on noninvasive estimates of pulmonary venous pressure. He had an elevated mean pulmonary arterial pressure despite treatment with a calcium channel blocker, a phosphodiesterase V inhibitor, and endothelin receptor antagonists. For a year, the patient had large-volume hemoptysis at least twice a month. Bronchoscopy showed dilated vessels and corrugated mucosa of the left main-stem bronchus. An 8-mm Amplatzer Vascular Plug II was placed in the left lower pulmonary artery. The patient tolerated the procedure well. Bronchoscopy 2 months after the procedure showed a marked reduction in the severity of left bronchial vascular congestion after occlusion of the left lower pulmonary artery. Hemoptysis has not recurred over an interval of 28 months. By occluding the peripheral pulmonary arteries that supply areas of lung with inoperable venous obstruction, distal pulmonary arterial pressure and flow may decrease, bronchial varices may regress, and the frequency and severity of hemoptysis may decrease.
Pulm Circ. 2016;6(3):389–396.
Inhaled treprostinil has been shown to be safe in pediatrics up to a maximum dose of 9 breaths 4 times a day (QID). We describe our use of inhaled treprostinil up to and over 9 breaths QID. Medical charts of established PH patients receiving inhaled treprostinil were reviewed. Clinical, imaging, catheterization, and lab data were analyzed. Ten patients were given inhaled treprostinil (5 with idiopathic PAH, 2 with congenital heart disease–associated disease, 2 with congenital diaphragmatic hernia, and 1 with chronic lung disease). Average age was 10 years (range: 4.8–16.1 years). Mean (median) duration of therapy was 1.7 (1.3) years (range: 0.8–4.5 years). Average WHO class at the start of therapy was 2.3, with improvement in WHO class of 0.4 by 9 breaths/dose. All patients were titrated to 9 breaths/dose. Average catheterization mean pulmonary artery pressure (MPAP) improved by 8.33 mmHg (P = 0.9), and pulmonary resistance improved by 3.42 WU*m2 (P = 0.16). Two patients had MPAP decrease to less than 25 mmHg. Treadmill exercise testing (n = 7) by the modified Bruce protocol showed a mean improvement of 2.8 minutes (P = 0.049). Side effects were transient: cough (n = 4), nosebleeds (n = 3), dizziness, flushing, and headache (n = 1 for each). No patient withdrew for side effects. Two patients transitioned to IV therapy because of clinical worsening. Four patients were increased to 12 and one to 15 breaths QID. No change in WHO class or side effects was noted at greater than 9 breaths. Inhaled treprostinil is safe in pediatric patients showing improvement in exercise with a trend toward improvement in MPAP and can be well tolerated even at high doses, up to 15 breaths QID. Inhaled treprostinil is a good prostacyclin option in pediatric patients.
Pulm Circ. 2016;6(3):389–396.
The diagnosis of pulmonary hypertension (PH) is delayed or missed during the early stages of the disease because the clinical signs are difficult to discern. We sought to investigate whether speech-recognition algorithms could differentiate between the pulmonary heart sounds (P2) in subjects with PH and those in subjects without PH. Furthermore, we hypothesized that an automated, speech-recognition-inspired classification algorithm applied to heart sound recordings would outperform trained clinicians in the diagnosis of PH. Heart sounds, electrocardiograms, and pulmonary artery pressure (PAp) were recorded simultaneously. Digitized heart sound recordings were used to train and test speech-recognition-inspired classification algorithms. We extracted heart sound features with mel-frequency cepstral coefficients. We built Gaussian-mixture models to classify the features as PH or non-PH. Clinicians, blinded to clinical data, listened to randomized digitized phonocardiograms. We defined PH as a mean PAp ≥ 25 mmHg. We studied 164 subjects, 86 with PH (median age: 41 years [range: 0.3–84]; 45 females; mean PAp: 41 ± 13 mmHg) and 78 without PH (median age: 17 years [range: 0.6–86]; 41 females; mean PAp: 17 ± 4 mmHg.) The automated speech-recognition-inspired algorithm diagnosed PH with a true positive rate of 74% (area under the receiver-operating-characteristic curve: 0.74), a false negative rate of 23%, and a false positive rate of 34%. Clinicians diagnosed PH from the heart sound recordings with a true positive rate of 56%, a false negative rate 68%, and a false positive rate of 50% (P < 0.05). We found that digital auscultation with the application of automated speech-recognition-inspired classification algorithms to recorded heart sounds results in a 74% correct diagnosis of PH and performed significantly better than clinicians listening to the same heart sound recordings. Our results suggest that changes in P2 are related directly to an increased PAp and that computer-aided diagnosis could be used as a screening tool in the early diagnosis of PH.
We are grateful for funding from the Cardiovascular Medical Research and Education Fund and Stollery Children’s Hospital Foundation, Edmonton, Alberta, Canada. Presentation of this work at the 9th International Conference on Neonatal and Childhood Pulmonary Vascular Disease was made possible through the generous support of an unrestricted educational grant from Mallinckrodt Pharmaceuticals (Canada).
Pulm Circ. 2016;6(3):389–396.
Treprostinil, a prostanoid used to treat pulmonary arterial hypertension (PAH), can be delivered by continuous intravenous (IV) or subcutaneous (SC) infusion, by inhalation, or orally (PO). There are no published reports of SC/IV-to-PO transition in pediatrics, and optimal transition protocols have not been defined. Transition of patients receiving high-dose treprostinil is controversial. This series describes two individualized transition strategies. Patient 1, a 17-year-old male with advanced idiopathic PAH (IPAH) since age 2, was unwilling to continue SC infusion because of site pain. He was transitioned from 153 ng/kg/min SC to 25 mg PO TID over 10 weeks at home. Patient 2, a 19 year-old-female with IPAH since age 17, had severe local reaction to SC infusion. She was transitioned from 40 ng/kg/min to 7.5 mg PO TID in the hospital over 5 days. Both patients had pre- and posttransition studies and were monitored during transition for clinical decompensation and adverse reactions. Both patients tolerated the transition well. Both had improvement in 6-minute walk distance and no significant change in B-type natriuretic peptide, functional class, or echocardiographic parameters. Patient 1 was catheterized before and after transition and had no significant change in hemodynamics. Patient 2 was catheterized before any treatment was started and after transition to PO delivery, with dramatic hemodynamic improvement. We describe two successful but very different SC-to-PO transition strategies for patients with IPAH receiving treprostinil. This can be done safely both at home and in the hospital. Patients receiving high-dose SC therapy with advanced disease can be safely switched to oral therapy without significant deterioration in clinical status in the short term. Setting and speed of transition should be individualized to the patient. In the setting of a very high baseline treprostinil dose, a slow transition at home should be considered.
Pulm Circ. 2016;6(3):389–396.
Tetralogy of Fallot is one of the most common congenital heart diseases, affecting approximately 5 in 10,000 births. Patients who undergo surgical repair of tetralogy of Fallot as children are at risk of late complications, including pulmonary regurgitation, right ventricular enlargement, and sudden cardiac death. A 40-year-old man with tetralogy of Fallot who had a Blalock-Taussig shunt at age 2 years and complete surgical repair at age 4 years presented with several years of progressively worsening fatigue, dyspnea, palpitations, and syncope. A transthoracic echocardiogram and cardiac MR imaging demonstrated severe pulmonary regurgitation with a severely dilated right ventricle and a right ventricular ejection fraction of 42%. The patient’s 24-hour Holter monitor showed frequent premature atrial contractions with a 4.3-second pause. The patient was deemed high risk for sudden cardiac death and received an epicardial pacemaker with AICD, along with cryoablation at the right ventricular outflow patch during pulmonary valve replacement surgery. Common chronic postoperative complications from tetralogy of Fallot repair include pulmonary regurgitation and right ventricular enlargement. Such patients are predisposed to ventricular arrhythmias and sudden cardiac death. Demographic factors, such as older age at surgical repair, and electrophysiologic factors, such as prolonged QRS interval, predispose to ventricular arrhythmias. Additionally, hemodynamic factors secondary to right ventricular dysfunction, pulmonic regurgitation, and right ventricular hypertrophy also contribute to sudden cardiac death risk. Management of high–sudden cardiac death risk patients includes implantation of an ICD, ablation at the right ventricular outflow tract, and administration of antiarrhythmic drugs. Pulmonary regurgitation and other hemodynamic factors from chronic complications of postoperative tetralogy of Fallot repair, as well as demographic and electrophysiologic factors, predispose to sudden cardiac death. Management involves ICD placement, ablation, and antiarrhythmic drug administration.
Pulm Circ. 2016;6(3):389–396.
Reduced wall shear stress (WSS)—associated with endothelial mechanotransduction—in the proximal pulmonary arteries in pulmonary arterial hypertension (PAH) patients has been associated with vascular stiffening. Serial, noninvasive evaluation of the relationship of WSS to vascular stiffening is still unknown. Here we aimed to retrospectively investigate time-dependent changes in maximum WSS in pediatric PAH patients who underwent successive phase-contrast MRI (PC-MRI). Twenty-one pediatric PAH patients with more than one PC-MRI visit were included. Segmented cross-sectional planes for WSS measurements were placed at the mid-level of the main pulmonary artery (MPA) and in the right pulmonary artery (RPA) 3 cm distally from the bifurcation. Vascular stiffness in each subject was assessed by relative area change (RAC: [(Amax − Amin)/Amax] × 100 [%]). Changes in WSS and RAC were evaluated with vector origin plots. Spearman rho was used to test correlations among changes in hemodynamic indices. The average time between serial PC-MRI acquisitions was 941 days. In the MPA, all patients with depressed maximum WSS (n = 12) presented a parallel decrease in RAC, and, conversely, all patients showing increased WSS (n = 9) showed increased RAC. Furthermore, chronological change in RAC correlated linearly with the change in maximum WSS (rho = 0.805, P < 0.001). In the RPA, 7 patients showed parallel decreases in WSS and RAC, 4 patients showed no change in RAC with concomitant decreases in WSS, and 5 patients had parallel increases in WSS and RAC. Positive significant linear correlation existed between the change in maximum WSS in the RPA and corresponding RAC (rho = 0.618, P < 0.01). Changes in maximum WSS measured in the MPA in pediatric PAH patients correspond to alterations in vascular stiffness. The noninvasive evaluation of WSS via PC-MRI may become an important follow-up modality in order to investigate the vascular pathophysiology in the progression of pediatric PAH.
Pulm Circ. 2016;6(3):389–396.
There is a paucity of data on treprostinil use in neonates with pulmonary hypertension (PH). The aim of the study was to explore the safety and effect of treprostinil in neonates with PH and describe clinical evolution, echocardiographic changes, and morbidity until discharge or death. Between June 2014 and February 2016, we conducted a prospective quasi-experimental study of a cohort with samples before and after treatment. Patients younger than 3 months with PH and severe hemodynamic conditions on mechanical ventilation (Fio2 > 60% with iNO) were included. Twelve infants were enrolled: 66% with left CDH, 25% with right CDH, and 8% with SALAM; 25% received ECMO. Median initiation age was 16.5 days (range: 6–27). All were critically ill, receiving mechanical ventilation. Median airway pressure was 14 (9–26); oxygenation index (OI): 21 (8–64); inotropics: 83%; B natriuretic peptide (BNP): 6,041 pg/mL (2,141–9,940). Echocardiograms performed weekly showed PH signs in all 12 patients. Intravenous treprostinil was started at 6 ng/kg/min and was progressively increased, depending on tolerance. Median dose was 25 ng/kg/min (21–30). Clinical effects after a week were as follows: OI decreased to 10.5 (2–45), P = 0.0028; BNP decreased to 3,729 pg/mL (926–6,532), P = 0.29; tricuspid anular plain systolic excursion was 9 (6–13.2), P = 0.04. Ten patients (83.3%) were transitioned to subcutaneous treatment when hemodynamically stabilized, at a median of 28 days (17–51). Eleven patients were discontinued with pulmonary pressure less than 50% of systemic pressure. NO and milrinone median times were calculated by Kaplan-Meyer: 3 days (2–7) and 8 days (4–16), respectively. One patient died of PH on day 7. Median treatment time was 70 days (49–90). Median hospitalization time was 123 days (89–154). Adverse events were as follows: hypotension (1/12), subcutaneous infusion site hematoma (1/12), thrombocytopenia (<80,000/mm3; 1/12), and persistent cholestasis (3/12). None required drug discontinuation. Treprostinil use was safe and well tolerated in neonates with satisfactory clinical response. Further studies are required to assess efficacy and identify early-responder subgroups.
Pulm Circ. 2016;6(3):389–396.
Patients with pulmonary hypertension may require parenteral prostacyclin therapy. However, titration may be limited by adverse effects. We hypothesized that the use of iloprost in prostacyclin-naïve children as a transitional medication would improve patient tolerance of intravenous epoprostenol. A retrospective chart review was conducted from January 2014 to January 2016. The need for pharmacologic therapy for nausea/vomiting, diarrhea, and pain was compared to that of historical controls who did not receive iloprost before epoprostenol. Four patients (age: 8–14 years; weight: 31–94 kg) received iloprost via a mesh nebulizer before transition to intravenous epoprostenol. Iloprost was initiated at 2.5–5 μg every 3 hours, up-titrated to 15–20 μg every 3 hours, and down-titrated upon epoprostenol initiation. The control group consisted of 4 patients (age: 11–16 years; weight: 62–82 kg). In a small cohort, we observed a trend toward reduced prevalence of diarrhea and pain with the use of iloprost. A larger study is needed to fully assess this novel approach.
Pulm Circ. 2016;6(3):389–396.
Extremely low gestational age neonates (ELGANs) are at increased risk to develop pulmonary hypertension (PH) associated with arrested vascular growth and alveolarization, with resultant decreased pulmonary capillary surface area. While multiple factors have been implicated, we lack precise understanding of the causative mechanisms driving PH in this vulnerable population. Deeply phenotyped observational birth cohorts of ELGANs with available biochemical metrics provide a unique opportunity to probe predisposing factors. The presence of PH, as defined by echocardiogram at 36 weeks gestational age (GA), is associated with significant early respiratory morbidity, increased biomarkers of oxidant stress, and reduced nitric oxide production. The Improving Prematurity-Related Respiratory Outcomes at Vanderbilt (IMPROV) cohort of ELGANs (<29 weeks GA) was used to blindly evaluate echocardiograms obtained at 36 ± 1 weeks GA. Echocardiographic features of PH were compared with clinical features, including markers of respiratory morbidity. Urine samples available for biochemical study were obtained at birth and at 7, 14, and 28 days. Sixty of 172 infants had an echocardiogram at 36 weeks; 10 of these infants had evidence of PH. There was no difference in demographic features, including GA, birth weight, and sex. While few other differences were noted, infants with PH were intubated for a shorter duration (1 day [range: 1.0–4.8] vs. 8 days [1.0–25]; P = 0.041) and had a trend toward a shorter length of stay (70 days [66–88] vs. 90 days [70–121]; P = 0.096). Isoprostanes and isofuranes were not different between groups at any time point, but infants with PH had significantly higher urinary nitric oxide metabolites at birth than infants without PH. Traditional measures of respiratory morbidity in early infancy may not predict the development of PH at 36 weeks among ELGANs. While oxidant stress did not differ between ELGANs with PH and those without PH, nitric oxide metabolite levels at birth may be a predictive biomarker in need of further study.
Pulm Circ. 2016;6(3):389–396.
Cardiac catheterization informs diagnosis and therapeutic strategy in children with pulmonary hypertension (PH). VO2 is commonly estimated from predictive equations for the calculation of pulmonary blood flow despite published data on inaccuracy. This study compares measured and assumed VO2 in baseline and vasodilator conditions. Data from 118 catheter procedures performed under general anesthesia in 94 patients were reviewed. Mass spectrometry was used to measure VO2 at baseline (21%–33% FiO2); measurements were repeated in 60% FiO2 and 20 ppm nitric oxide in 92 procedures. Assumed values were derived from the LaFarge equation. Agreement was evaluated with the Bland-Altman method. Cohort characteristics were as follows: female: 57% (67/118); median age: 8 years (range: 0.8–18 years); age < 3 years: 19% (22/118); congenital heart disease–associated PH: 59% (69/118); and idiopathic PH: 17% (20/118). Mean measured baseline VO2 was 122 mL/min/m2 (range: 78–174 mL/min/m2), bias was 14% (assumed > measured), with an SD of 19%, and 95% limits of agreement −51% and 23%. Mean measured vasodilator-phase VO2 was 124 mL/min/m2 (range: 78–208 mL/min/m2), bias was 12% (assumed > measured), with an SD of 19%, and 95% limits of agreement were −51% and 27%. In a comparison of measured VO2 in baseline and dilator conditions, bias was −2% (dilator > baseline), with an SD of 10%; 95% limits of agreement were −22% and 19%. The strongest determinants of VO2 were body surface area (R2 = 0.71, P < 0.01), height (R2 = 0.79, P < 0.01), mass (R2 = 0.80, P < 0.01), age (R2 = 0.71, P < 0.01), and HR (R2 = 0.12, P < 0.01). Sex was not a significant determinant (R2 = 0.01, P = 0.29). In conclusion, VO2 is higher at baseline than in the vasodilator phase. The indirect Fick method overestimates VO2 in anesthetized children, in both conditions, leading to an underestimation in PVRI. This underestimation may affect decision making in the assessment of congenital heart disease, where small changes in PVRI change operability.