Introduction
Elevated pulmonary vascular pressures are typically observed in fetal life and partially decrease in response to oxygen after birth. The presence of persistent pulmonary hypertension indicates a failure in this transition. Reference Gentle, Abman and Ambalavanan1 Pulmonary hypertension has been reported in 23–37% of premature infants. Reference Slaughter, Pakrashi, Jones, South and Shah2,Reference Ali, Schmidt, Dodd and Jeppesen3 Those with persistent pulmonary hypertension beyond the first few months of life had a high mortality rate. Reference Abman4 When pulmonary hypertension was severe, 47% of them died by two years of age. Reference Abman4,Reference Khemani, McElhinney and Rhein5
Increased pulmonary vascular resistance results from pulmonary vascular remodelling, leading to right ventricular hypertrophy and failure. Right ventricular dysfunction due to increased afterload is a primary driver of disease severity in pulmonary hypertension patients. Reference Abman, Hansmann and Archer6 Cardiac catheterisation is the gold standard for haemodynamic assessment of pulmonary vascular resistance or pulmonary artery pressure, but it is an invasive procedure. Reference Abman, Hansmann and Archer6 Echocardiography is a non-invasive and more available tool for the evaluation of pulmonary hypertension in very low birth weight infants. Tricuspid regurgitation jet velocity is a widely utilised echocardiographic parameter to evaluate pulmonary hypertension, quantitating pulmonary artery pressure. Tricuspid regurgitation jet velocity correlated well with the invasive haemodynamic measurement when performed with a good window view and small insonation angle. Reference Amsallem, Sternbach and Adigopula7,Reference Mourani, Sontag, Younoszai, Ivy and Abman8 However, a tricuspid regurgitation jet was present in only 14–80% of paediatric echocardiograms for pulmonary hypertension. Reference Nagiub, Lee and Guglani9,Reference Burkett, Patel, Mertens, Friedberg and Ivy10 Thus, tricuspid regurgitation jet velocity may not be feasible for treatment effect monitoring or longitudinal analysis of the disease. Qualitative septal flattening is also commonly used in newborn patients. Reference Mourani, Sontag, Younoszai, Ivy and Abman8,Reference Arjaans, Zwart, Roofthooft, Kooi, Bos and Berger11–Reference Mourani, Sontag and Younoszai13 When the right to left ventricular pressure ratio increases, the septum becomes more flattened with reversed curvature in severe right ventricular hypertension. Reference Burkett, Patel, Mertens, Friedberg and Ivy10 However, septal flattening is a subjective assessment with significant interobserver variability. Reference Arjaans, Zwart, Roofthooft, Kooi, Bos and Berger11,Reference Abraham and Weismann14 Left ventricle eccentricity index quantifies septal flattening and is reproducible with less intra- and interobserver variability Reference Ehrmann, Mourani and Abman.15–Reference Averin, Michelfelder, Sticka, Cash and Hirsch17 Utilisation of left ventricle eccentricity index, pulmonary artery acceleration time, and right/left ventricle dimension ratio to quantify pulmonary hypertension is established in adults and has been increasingly studied in children. Reference Amsallem, Sternbach and Adigopula7,Reference Burkett, Patel, Mertens, Friedberg and Ivy10,Reference Abraham and Weismann14 Still, data are limited in preterm infants. Reference Abraham and Weismann14,Reference McCrary, Malowitz and Hornick16
Assessment of the right ventricular function is the key to determine disease severity and prognosis. Reference Rosenzweig, Abman and Adatia18 Tricuspid annular plane systolic excursion and fractional area change are parameters to assess right ventricular function in preterm infants. Reference Murase and Ishida19,Reference Koestenberger, Nagel and Ravekes20 Tissue Doppler imaging is a technique currently used to evaluate ventricular function by measurement of myocardial motion velocities during systole and diastole in infants with bronchopulmonary dysplasia. Reference Patel, Mills and Cheung21,Reference Sehgal, Malikiwi, Paul, Tan and Menahem22 However, its application for pulmonary hypertension screening and monitoring has not been well characterised and is still challenging in the very premature heart. Reference Borges, Knebel and Eddicks23 This study aimed to determine the feasibility of quantitative echocardiography parameters for pulmonary hypertension screening and analyse longitudinal changes up to 36 weeks postmenstrual age in pulmonary pressure and right ventricular function in very low birth weight infants. Secondary objectives were to determine the predictive factors associated with newly developed or persistent pulmonary hypertension at 36 weeks postmenstrual age. Standardising screening and follow-up will allow for the early detection, objective monitoring, risk stratification, and effective management of infants with pulmonary hypertension. Reference Arjaans, Zwart, Roofthooft, Kooi, Bos and Berger11
Materials and methods
Study population
This study was a secondary analysis of a prospective cohort single-centre study. The primary analysis of this prospective cohort single-centre study is the feasibility of pulmonary hypertension screening in very low birth weight infants at risk of bronchopulmonary dysplasia by using tricuspid regurgitation gradient and a cardiac biomarker. It was conducted in the neonatal intensive care unit between August 1, 2019, and September 30, 2020, and was approved by the Institutional Ethics Committee. Preterm infants with a birth weight of 400 g–1,500 g or gestational age 230/7–316/7 weeks admitted within 24 hours of age were enrolled. Exclusion criteria included congenital heart diseases [except patent ductus arteriosus, non-significant atrial septal defect, or non-significant ventricular septal defect], chromosomal abnormalities, multiple congenital anomalies, hydrops fetalis, cardiomyopathy, diaphragmatic hernia, or administration of a fluid bolus ≥10 ml/kg intravenously in the past 24 hours to avoid volume load dependence that might interfere with evaluating ventricular function. After obtaining informed consent from parents, the infants received pulmonary hypertension screening according to Chiang Mai University protocol. Preterm infants with gestational age <28 weeks and body weight <1,000 g underwent echocardiographic screenings at 7 and 28 days of age and at 36 weeks postmenstrual age. Preterm infants with gestational age 280/7–316/7 and body weight 1,000 g–1,500 g underwent echocardiographic screening at 28 days of age and at 36 weeks postmenstrual age (Supplemental Figure 1).
Definitions
-
Criteria for pulmonary hypertension: left ventricular eccentricity index ≥1.2 Reference Amsallem, Sternbach and Adigopula7, Reference D’Alto, Bossone, Opotowsky, Ghio, Rudski and Naeije24
-
Very low birth weight: birth weight of less than 1500 g
-
Extremely low birth weight: birth weight of less than 1000 g
-
Small for gestational age: birth weight of less than 10th percentile for gestational age
-
Birth asphyxia: Activity, Pulse, Grimace, Appearance, Respiration (APGAR) score of <7 at 5 minutes
-
Haemodynamically significant patent ductus arteriosus: patent ductus arteriosus size ≥1.5 mm and left atrial-to-aortic root ratio ≥1.5
-
High-flow nasal cannula: the gas flow is operated between 3–8 L/min
Clinical data
Demographic data, including sex, gestational age, birth weight, history of asphyxia, route of delivery, extensive resuscitation at birth, intrauterine growth restriction, small for gestational age, and respiratory support, were collected.
Echocardiographic data
All echocardiograms were performed using a Philips CX 50 and Phillips Epiq 7 with 8,12 MHz probes (Philips Medical Systems) according to patient size. The data were collected and analysed by a single observer (YP). Non-significant intracardiac shunts such as atrial septal defect, patent foramen ovale, and ventricular septal defect were examined. Patent ductus arteriosus was identified by 2D, colour Doppler, and shunt gradient. Left atrial-to-aortic root ratio was measured to determine left heart volume loading. Reference Levy, Tissot and Horsberg Eriksen25
Echocardiographic determination of pulmonary hypertension and pulmonary vascular resistance
Pulmonary artery pressure was estimated using tricuspid regurgitation jet velocity if tricuspid regurgitation jet was present. Tricuspid regurgitation jet velocity was measured using continuous-wave Doppler from an apical four-chamber view. Reference Skinner, Boys, Hunter and Hey26 Left ventricle eccentricity index was measured as the diameter in the parasternal short-axis view at the midpapillary muscle level during end-systole. Reference Averin, Michelfelder, Sticka, Cash and Hirsch17 Left ventricular eccentricity index was calculated as follows: left ventricle eccentricity index = left ventricular diameter parallel to the interventricular septum/left ventricular diameter perpendicular to the interventricular septum. Reference Amsallem, Sternbach and Adigopula7,Reference Averin, Michelfelder, Sticka, Cash and Hirsch17 The pulmonary artery velocity wave Doppler was used to calculate the pulmonary artery acceleration time and right ventricular ejection time. Reference Murase and Ishida19
Echocardiographic determination of right ventricular function
Tricuspid annular plane systolic excursion was measured by M-mode recording in the apical four-chamber view with the cursor placed at the free wall of the tricuspid valve annulus. Reference Koestenberger, Nagel and Ravekes20 Fractional area change was measured by manual tracing of the right ventricular endocardial border from the lateral tricuspid annulus along the free wall to the apex and back along the interventricular septum to the medial tricuspid valve annulus at end-diastole and end-systole. Reference Augustine, Coates-Bradshaw and Willis27 For tissue Doppler imaging, the right ventricular myocardial velocities were measured with pulse-wave tissue Doppler imaging at the lateral tricuspid annulus from an apical four-chamber view. Reference de Boode, Singh and Molnar28 The insonation angle should be less than 15°. The mean of 3−5 consecutive cardiac cycles was obtained for each echocardiographic parameter. Peak isovolumic systolic velocity and peak systolic ejection velocity (S) were used to quantify systolic velocities, which corresponded to early systolic isovolumic contraction and later systolic contraction ejection phases, respectively. Reference Lindqvist, Waldenström, Wikström and Kazzam29 Diastolic velocities measured were the early diastolic E’ velocity corresponding to early diastolic relaxation and later diastolic A’ velocity, corresponding to atrial contraction. Reference Oki, Tabata and Yamada30 Right ventricular myocardial performance index was obtained by the sum of isovolumic contraction and relaxation time divided by ejection time derived from tissue Doppler imaging. Reference de Boode, Singh and Molnar28
Echocardiographic determination left ventricular function
Mitral annular plane systolic excursion was measured by M-mode recording in the apical four-chamber view with the cursor placed at the free wall of the mitral valve annulus. Left ventricular ejection fraction was measured by M-mode in the left ventricular short-axis view. Reference Gaspar and Morhy31
Statistical analysis
All statistical analyses were performed using Stata 17 (StataCorp, College Station, Texas, USA). We described categorical variables with frequency and percentage. For continuous variables, mean and standard deviation or median and interquartile range were used for data description depending on the underlying distribution.
An independent t-test or Mann–Whitney U test was used to compare the difference of continuous variables between groups, whereas Fisher’s exact probability test was used for categorical variables. We also calculated the standardised difference and an area under the receiver operating characteristics curve for each predictor. Multivariable analysis was not performed as a means to identify significant predictors of pulmonary hypertension at postmenstrual age 36 weeks, owing to significant limitations in terms of study size. P-value < 0.05 was defined as significant statistical testing. In this study, we defined significant predictors of pulmonary hypertension based on three criteria: (1) statistical testing at P-value ≤ 0.1, (2) a lower bound of 95% confidence interval of an area under the receiver operating characteristic curve ≥0.5, and (3) a standardised difference ≥0.5 (moderate effect size according to Cohen). Reference Sullivan and Feinn32
Individual profile plots were illustrated to visualise the changes in echocardiographic parameters over time. We modelled the trend of these values and performed an interaction test using multi-level Gaussian regression. Pearson’s correlation coefficient and correlation plots were used to quantify the strength of a linear relation between left ventricular eccentricity index and tricuspid regurgitation jet velocity.
Results
Thirty-three infants were enrolled in the study, but three were excluded due to missing echocardiograms. The final enrolled number was 30, with no patient death up to 36 weeks postmenstrual age. Fourteen infants (46.67%) were male. A total of 79 echocardiograms were analysed. No patients received a pharmacological pulmonary vasodilator during the echocardiographic study. Nineteen infants (63.33%) had extremely low birth weight (gestational age 27.07 ± 2.51 weeks and weight 1181.43 ± 507.89 g). Eleven infants (36.67%) had very low birth weight (gestational age 29.63 ± 0.94 weeks and mean weight 1861.59 ± 462.95 g). Of the entire cohort, 18 (22.8%) events were diagnosed as pulmonary hypertension. The prevalence of pulmonary hypertension was 31.6% (6/19) on day 7 of life and 20% (6/30) on day 28 of life. The final six infants (6/30, 20%) had persistent or progressive pulmonary hypertension at 36 weeks postmenstrual age. Three patients presented with pulmonary hypertension at the last echocardiography. Three patients who presented with at least one prior echocardiography showed evidence of pulmonary hypertension before the last echocardiography.
No significant differences in clinical data were observed at baseline, day 7 of life, and 36 weeks postmenstrual age, with the exception that male infants were more likely to be in the pulmonary hypertension group than the non-pulmonary hypertension group (Tables 1, 2, and 3).
Baseline fixed factors
APGAR: Activity, Pulse, Grimace, Appearance, Respiration; Asphyxia: APGAR at 5 min score <7; ETT = endotracheal tube; IUGR = intrauterine growth restriction; nCPAP = nasal continuous positive airway pressure; PMA = postmenstrual age; PPV = positive pressure ventilation; ROC = receiver operating characteristic curve; SGA = small for gestational age; *standardised difference ≥ 0.5; **statistical testing at P-value ≤ 0.1; ***lower bound of 95% confidence interval of an area under ROC curve ≥ 0.5.
Dynamic factors at 7 days in preterm infants with BW <1,000 g (N = 19)
A’ = late diastolic velocity; E’ = early diastolic velocity; S’ = peak systolic ejection velocity; AT = acceleration time; CO = cardiac output; EF = ejection fraction; FAC = fractional area change; IVV = peak isovolumetric systolic velocity; LA:Ao = left atrial-to-aortic root ratio; LV = left ventricular; LVEI = left ventricular eccentric index; MAPSE = mitral annular plane systolic excursion; MPI = myocardial performance index; PA = pulmonary artery; PDA = patent ductus arteriosus; PMA = postmenstrual age; ROC = receiver operating characteristic curve; RV = right ventricular; TAPSE = tricuspid annular plane systolic excursion; TR = tricuspid regurgitation.
Haemodynamic significant PDA: PDA ≥1.5 mm and left atrial-to-aortic root ratio ≥1.5.
*standardised difference ≥0.5; ** statistical testing at P-value ≤ 0.1; *** lower bound of 95% confidence interval of an area under ROC curve ≥0.5.
At day 28, infants with late pulmonary hypertension had significantly lower current weight and postmenstrual age than the non-pulmonary hypertension group (975 ± 204.45 g. vs. 1300.95 ± 402.68 g., P = 0.067, receiver operating characteristic curve = 0.73) and (30.48 ± 1.32 wk vs. 32.64 ± 2.51 wk, P = 0.052, receiver operating characteristic curve = 0.78), respectively (Table 3). Infants with late pulmonary hypertension also required high-flow oxygen therapy at day 28 more than those without pulmonary hypertension. No differences were noted in patent ductus arteriosus to body weight ratio, size, and left atrium to aorta ratio between the pulmonary hypertension group and the non-pulmonary hypertension group.
Dynamic factors at 36 weeks PMA in all preterm infants with BW <1,500 g (N = 30)
A’ = late diastolic velocity; E’ = early diastolic velocity; S’ = peak systolic ejection velocity; AT = acceleration time; CO = cardiac output; EF = ejection fraction; FAC = fractional area change; IVV = peak isovolumetric systolic velocity; LA:Ao = left atrial-to-aortic root ratio; LV = left ventricular; LVEI = left ventricular eccentric index; MAPSE = mitral annular plane systolic excursion; MPI = myocardial performance index; PA = pulmonary artery; PMA = postmenstrual age; ROC = receiver operating characteristic curve; RV = right ventricular; TAPSE = tricuspid annular plane systolic excursion; TR = tricuspid regurgitation.
*Standardised difference ≥0.5; **statistical testing at P-value ≤ 0.1; ***lower bound of 95% confidence interval of an area under ROC curve ≥0.5.
Echocardiography data
Tricuspid regurgitation was presented and interpretable in 35 of 79 (44.30%) echocardiograms. Tricuspid regurgitation jet velocity correlated with left ventricular eccentricity index (r = 0.77, P = 0.005) (Figure 1). Left ventricular eccentricity index was performed successfully in all examinations. All infants with tricuspid regurgitation jet velocity >25 mmHg (2.5 m/sec) also had left ventricular eccentricity index >1.2. Tissue Doppler imaging at the lateral tricuspid annulus (right ventricular peak isovolumic systolic velocity, S’, E,’ A’ and myocardial performance index) was available for interpretation in 72 of 79 (91%) echocardiograms.
Relation between tricuspid valve jet velocity (TRJV) and left ventricular eccentric index (LVEI) postmenstrual age at 36 weeks; PMA: postmenstrual age.
The longitudinal changes in echocardiography parameters (Fig. 2)
Left ventricular eccentricity index showed a significantly different trend, increasing over time in the pulmonary hypertension group and decreasing over time in the non-pulmonary hypertension group (P = 0.007). The mean change per day in the pulmonary hypertension group was +0.004 (P = 0.090). The mean change in the non-pulmonary hypertension group was −0.001 (P = 0.041). Right ventricular peak isovolumic systolic velocity trend showed a progressive worsening of right ventricular function in the pulmonary hypertension group, which was the opposite of the improving function found in the non-pulmonary hypertension group (P = 0.197).
Longitudinal changes of PHT group versus non-PHT group: a ) longitudinal change in LVEI, b ) longitudinal change in RV IVV, c ) longitudinal change in RV MPI, d ) longitudinal change in current weight. The thick blue and red lines represent the overall summary trend of the group of patients between PHT and no PHT at 36 weeks PMA. The thin lines represent the individual trends of each patient. LVEI = left ventricular eccentric index; PHT = pulmonary hypertension; RV IVV = right ventricular isovolumic systolic velocity; RV MPI = right ventricular myocardial performance index.
Echocardiography parameters at 7 days of age (Table 2)
Compared to the non-pulmonary hypertension group, there was no significant difference between the two groups regarding the size of patent ductus arteriosus and patent ductus arteriosus by body weight ratio. Although the number of infants having tricuspid regurgitation jet velocity was significantly higher in the non-pulmonary hypertension group, tricuspid regurgitation jet velocity was in the normal range (4.5 mmHg [IQR 0,10.5]). Two infants in the late pulmonary hypertension group had early pulmonary hypertension at day 7 of life. The ratio of early pulmonary hypertension was higher in late pulmonary hypertension (33.3%) than in non-pulmonary hypertension infants (12.7%) but was not statistically significant.
Echocardiography parameters at 28 days of age (Supplemental Table1)
Two infants in the late pulmonary hypertension group met pulmonary hypertension criteria at day 28 of life. There was no difference in echocardiographic parameters between the late pulmonary hypertension group and the non-pulmonary hypertension group.
Echocardiography parameters at 36 weeks postmenstrual age (Table 3)
Tricuspid regurgitation jet velocity was significantly higher in the pulmonary hypertension group than the non-pulmonary hypertension group (28.5 mmHg [IQR 21.5, 48.5] vs. 13 mmHg [IQR 0, 21], P = 0.012). Right ventricular peak isovolumic systolic velocity was significantly reduced in the pulmonary hypertension group (5.5 ± 2.16 cm/sec vs. 7.69 ± 1.93 cm/sec, P = 0.025), and right ventricular myocardial performance index in the pulmonary hypertension group was significantly prolonged compared to the non-pulmonary hypertension group (0.43 ± 0.11 vs. 0.35 ± 0.05, P = 0.033).
Discussion
Twenty percent (6/30) of preterm low birth weight infants were diagnosed with newly developed or persistent pulmonary hypertension at 36 weeks postmenstrual age in our study. This prevalence is slightly higher than 11.7–14% in previous prospective screening of pulmonary hypertension, Reference Bhat, Salas, Foster, Carlo and Ambalavanan12,Reference Mourani, Sontag and Younoszai13 probably due to different diagnostic criteria of pulmonary hypertension and no pulmonary vasodilator treatment in our study. In our study, small for gestational age was not significantly different between the two groups of patients. A previous review article showed that poor growth in utero may be one of the risk factors for bronchopulmonary dysplasia and may cause pulmonary hypertension. Reference Underwood, Wedgwood, Lakshminrusimha and Steinhorn33 A recent study showed premature infants suffering from bronchopulmonary dysplasia have a tendency to experience pulmonary hypertension. Reference Abman and Lakshminrusimha34 We had a similar finding in that the late pulmonary hypertension group had a higher incidence of bronchopulmonary dysplasia than the other group. In our study, preterm infants who still needed high-flow oxygen therapy at 28 weeks of age were likely to develop pulmonary hypertension at 36 weeks postmenstrual age, consistent with the study of Bhat et al. Reference Bhat, Salas, Foster, Carlo and Ambalavanan12 Prolonged oxygen therapy in infants with abnormal lung development may contribute to bronchopulmonary dysplasia and pulmonary hypertension. Reference Gentle, Abman and Ambalavanan1 The haemodynamic significance of patent ductus arteriosus or patent ductus arteriosus treatment was not related to the late pulmonary hypertension, similar to the findings of Bhat et al. Reference Bhat, Salas, Foster, Carlo and Ambalavanan12 and Mourani et al. Reference Mourani, Sontag and Younoszai13 In our study, there were no significant differences between the two groups of patients in using invasive positive pressure ventilation, non-invasive positive pressure ventilation, or fractional concentration of administered oxygen, which can be indicators of the disease severity. This could be due to the small sample size.
Echocardiographic evaluation of pulmonary hypertension and right ventricular function
To our knowledge, this is the first prospective serial study of quantitative echocardiographic screening of pulmonary hypertension in preterm infants from day seven and day 28 of life until 36 weeks postmenstrual age. In this study, tricuspid regurgitation jet velocity was measurable in 44.30%, whereas left ventricular eccentricity index was performed successfully in all cases. Further, the left ventricular eccentricity index showed a good correlation with tricuspid regurgitation jet velocity. Previous studies in preterm infants at 36–38 weeks postmenstrual age also revealed 15–66% of measurable tricuspid regurgitation jet velocity, Reference Bhat, Salas, Foster, Carlo and Ambalavanan12,Reference Abraham and Weismann14,Reference McCrary, Malowitz and Hornick16 and 86–100% of available left ventricular eccentricity index, Reference Abraham and Weismann14–Reference McCrary, Malowitz and Hornick16 with a good correlation between both parameters Reference Abraham and Weismann14,Reference McCrary, Malowitz and Hornick16 . The measurement of left ventricular eccentricity index is easily obtained in parasternal short-axis view at mid-left ventricle between the two papillary muscles and does not depend on insonation angle, as does the tricuspid regurgitation jet velocity Doppler. Reference Forfia and Vachiéry35 However, care must be taken to keep the 2D plane perpendicular to the left ventricular long axis to avoid oblique rotation leading to artifactual septum flattening. Reference Forfia and Vachiéry35 Too near to the cardiac base may cause septal distortion by the aortic root, and too near to the apex may not detect the abnormality. Reference Ryan, Petrovic, Dillon, Feigenbaum, Conley and Armstrong36
The cut-off left ventricular eccentricity index values to identify the presence of pulmonary hypertension among previous studies Reference Amsallem, Sternbach and Adigopula7,Reference Burkett, Patel, Mertens, Friedberg and Ivy10,Reference Abraham and Weismann14,Reference McCrary, Malowitz and Hornick16,Reference Averin, Michelfelder, Sticka, Cash and Hirsch17,Reference Forfia and Vachiéry35 varied from 1.15 to 1.24, probably due to different patient ages, pulmonary hypertension aetiologies, and using reference catheterised data or other echocardiographic parameters to compare. We chose >1.2 as an abnormal ratio to avoid overdiagnosis of pulmonary hypertension. Reference Amsallem, Sternbach and Adigopula7 After birth, pulmonary pressure should be high initially. Over time, there should be a gradual decrease in pulmonary pressure, normalising within two weeks. Therefore, left ventricular eccentricity index should return to normal. In infants facing pulmonary hypertension, the progression of left ventricular eccentricity index recovery may be prolonged. Left ventricular eccentricity index >1.2 could differentiate the trend of progressive pulmonary hypertension in late pulmonary hypertension infants from decreasing pulmonary pressure in non-pulmonary hypertension infants. The persistent or progressive increase of pulmonary pressure may reflect ongoing pulmonary vascular disease Reference Rosenzweig, Abman and Adatia18 and prolonged oxygen therapy in these very low birth weight infants. Reference Gentle, Abman and Ambalavanan1 Interestingly, we observed a significant ongoing resolution of pulmonary pressure after 7 days of life until 36 weeks postmenstrual age with an estimated rate of left ventricular eccentricity index changes of –0.004 per day.
Right ventricular peak isovolumic systolic velocity represents the systolic function of right ventricle and was significantly reduced in the late pulmonary hypertension group in this study. Patel et al. also found reduced right ventricular peak isovolumic systolic velocity in term newborns with pulmonary hypertension than in the control group. Reference Patel, Mills and Cheung21 In another cohort of bronchopulmonary dysplasia infants, Reference Alabed, Sabouni, Al Dakhoul and Bdaiwi37 both right ventricular peak isovolumic systolic velocity and right ventricular peak systolic ejection velocity (S’) were diminished, but only right ventricular peak isovolumic systolic velocity had an association with a longer duration of respiratory support. In our study, serial right ventricular peak isovolumic systolic velocity was likely to differentiate the progressively diminished right ventricular systolic function in late pulmonary hypertension infants from increasing right ventricular systolic function in non-pulmonary hypertension infants. The progressively depressed right ventricular systolic function may result from persistent or progressive increased right ventricular afterload or chronic hypoxia related to impaired pulmonary vascular and alveolar growth and maturation. Reference Rosenzweig, Abman and Adatia18,Reference Patel, Mills and Cheung21,Reference Sehgal, Malikiwi, Paul, Tan and Menahem22 The major advantage of right ventricular peak isovolumic systolic velocity measurement is its simplicity and high reproducibility, Reference Lindqvist, Waldenström, Wikström and Kazzam29 which should encourage implementation in preterm infants. The limitations of right ventricular peak isovolumic systolic velocity are angle dependency and load dependence, which are similar to other conventional tissue Doppler imaging measurements. Reference Forfia and Vachiéry35
Right ventricular myocardial performance index, which reflects combined systolic and diastolic function, was prolonged significantly in the pulmonary hypertension group at 36 weeks postmenstrual age. This result was consistent with prior studies. Reference Sehgal, Malikiwi, Paul, Tan and Menahem22 However, right ventricular myocardial performance index trends could not differentiate between pulmonary hypertension and non-pulmonary hypertension groups. Pseudo-normalization of myocardial performance index, seen in a patient with increased atrial pressure, Reference Forfia and Vachiéry35 may obscure the result.
Right ventricular peak systolic ejection velocity (S’), early diastolic velocity (E’), and tricuspid annular plane systolic excursion seemed to be lower in the pulmonary hypertension group in this study, but there was no statistical significance. Right ventricular fractional area change was not significantly different between the two groups. The above findings emphasise the importance of right ventricular peak isovolumic systolic velocity, which may be an earlier marker of cardiac dysfunction than traditional echocardiographic measurement. Early detection of pulmonary hypertension and right ventricular dysfunction using a suitable quantitative approach could have ramifications for therapeutic options and improve outcomes for pulmonary hypertension patients.
Echocardiographic predictor of late pulmonary hypertension
In this study, late pulmonary hypertension infants had a higher incidence of pulmonary hypertension at day seven or day 28 of life than the non-pulmonary hypertension group, but there were no statistical differences. This result was similar to the study of Mourani et al., which revealed that early pulmonary hypertension at day 7 could predict the late pulmonary hypertension by using septal flattening criteria. Reference Mourani, Sontag and Younoszai13 The result may have arisen from using different pulmonary hypertension criteria and/or the number of pulmonary hypertension infants in our study was too small to reach statistical significance. A single echocardiographic examination may reveal pulmonary hypertension status at that time point. The accurate prediction of late pulmonary hypertension may require serial pulmonary hypertension assessment.
Limitations
The strength of this study is its prospective longitudinal study with well-defined echocardiographic criteria and clinical variables. Furthermore, this study used unbiased screening because the echocardiography was performed by a cardiologist blinded to the patient’s diagnosis. However, there were some limitations. First, the initial echocardiogram was performed per protocol, and we did not have data at seven days of age in preterm infants weighing 1000– <1500 g. Longitudinal data in this subgroup may have been lost. Second, there was the lack of simultaneous haemodynamic data obtained by cardiac catheterisation to verify the presence of pulmonary hypertension and optimum left ventricular eccentricity index and right ventricular peak isovolumic systolic velocity cut-off points for clinical use. Third, in some cases, the tricuspid regurgitation gradient could not be measured in the apical 4 chamber view. Parasternal right atrium-right ventricle view is an alternative view for evaluation of tricuspid regurgitation but was not used routinely in this study. Echocardiography protocols should be easy for general paediatricians, and neonatologists to perform, which is why the apical 4 chamber view is the standard view used in this study to access tricuspid regurgitation by visualisation and for measurement. Fourth, this study is a secondary analysis prospective cohort. Some data may not have been presented in this study, including antenatal steroids, history of maternal chorioamnionitis, pre-eclampsia, and preterm prolonged rupture of the membrane, which may be additional predictors of pulmonary hypertension. Future studies should be prospective and include more predictive factors associated with pulmonary hypertension. And lastly, the sample size is small, and clinically relevant findings could have been missed because of limited statistical power. A larger sample size may also identify predictive factors for pulmonary hypertension beyond 36 weeks postmenstrual age.
Conclusion
Left ventricular eccentricity index was feasible for assessing pulmonary hypertension in very low birth weight infants. Serial left ventricular eccentricity index and right ventricular peak isovolumic systolic velocity may help predict late pulmonary hypertension and early detection of right ventricular dysfunction. A study with a larger sample size and more patients with pulmonary hypertension is required to incorporate these parameters into the evaluation and prediction of pulmonary hypertension in preterm infants. We recommend adding both left ventricular eccentricity index and right ventricular peak isovolumic systolic velocity into the echocardiographic assessment of pulmonary hypertension in all very low birth weight infants.
