This study examined maternal, infant, clinical and echocardiographic markers in a recent cohort of infants and showed that infants with PPHN had significantly higher fraction of inspired oxygen, oxygenation index, received more inotropes and had lower Apgar scores at 1 and 5 minutes. TR Vmax and EI significantly categorised cases from invasively ventilated and non-invasively ventilated controls.
Gestational diabetes is a known risk factor for the development of PPHN.[16, 17] This finding was confirmed in our small-to-moderate cohort, as offspring of mothers with gestational diabetes more frequently developed PPHN. A similar proportion of ventilated cases and controls had a diagnosis of hypoxic ischaemic encephalopathy (HIE), however grade 3 HIE was seen only in cases of PPHN. This observation may have contributed to severity in symptoms seen in PPHN cases, as evidenced by higher inotrope usage and a significantly longer period of invasive ventilation. Levels of inspired oxygen, OI, Apgar score at 1 and 5 minutes and days of invasive ventilation were significantly different between the groups. OI, a useful clinical parameter, was higher in PPHN cases similar to findings from other studies.[18, 19]
Most echocardiographic markers studied were significantly different between cases and non-invasively ventilated controls. TR Vmax, S/D TR, PAAT were significantly different between cases and non-ventilated controls. Previous studies examining S/D TR have shown that a value > 1.3 was associated with death/ need for ECMO.[20] This has been replicated in our study thus confirming the predictive ability of S/D TR to predict death. PAAT, which has previously been demonstrated to have good reproducibility and was validated with pulmonary pressures on catheterisation,[21] was lowest in cases in our cohort. TAPSE was negatively correlated with Apgar score at 1 min, possibly reflecting the effects of hypoxic ischaemic insult on myocardial function. Furthermore, TAPSE was lower in cases in comparison to controls. This finding is important, since TAPSE values below 4mm have been shown previously to predict adverse outcomes in PPHN.[22] Stroke volume measured was lowest in cases and highest in non-invasively ventilated controls reflecting the close relationship between the cardio-respiratory system and its usability as a marker of sickness in infants with PPHN. A low SV observed in cases could be due to reduced preload secondary to right-to-left shunting causing reduced pulmonary venous return and the effects of septal flattening.[9, 23, 24] However, a compensatory rise in heart rate in in infants with PPHN preserved the LVO.[25] Interestingly, the LV EDD, a surrogate of preload, when related to SV showed a trend exhibiting the properties of the Frank-Starling curve.[26] Infants with PPHN had a lower response when compared to controls, reflecting the reduced preload[23] and suboptimal myocardial function often seen in PPHN.[27]
Among the several echocardiographic markers explored in this study, TR Vmax and eccentricity index (both end systolic and end diastolic) were the only markers that discriminated between cases, invasively and non-invasively ventilated controls. TR Vmax, when interpretable, is a useful parameter to determine the pulmonary pressures, however, this may not be always possible.[28] In this study, 15/22 (68%) of cases had an interpretable TR Vmax envelope, which is similar to findings from other studies.[24, 29, 30] The S/D TR can give meaningful results even if the CW Doppler envelope is incomplete as shown in our cohort where we could assess S/D TR in all infants.
In the absence of an interpretable TR Vmax, sEI ,a quantitative measurement, can provide important prognostic information as it has been shown to predict adverse outcomes in PPHN.[20] Though the regression model identified sEI as the only echocardiographic parameter predicting PPHN in this small to moderate sample size, sEI was found to have an excellent AUC of 0.94 and good sensitivity. This is comparable with other studies that have examined sEI in PPHN.[30] sEI has been shown to have excellent inter-rater variability[30] and intra-rater variability, as also demonstrated in our cohort. Furthermore, it showed good sensitivity and specificity, making this a useful parameter in aiding diagnosis and possibly monitoring progress to treatment. sEI has several advantages such as the ability to assess if from subcostal view which is free from lung artefact, angle of insonation does not affect measurements and finally it is independent from reliance on a complete Doppler envelope.
The study limitations include the retrospective nature and the small number of infants but being studied in a single centre, this reduced the variation in practice frequently seen with studies involving multiple centres. Secondly, this study only provides a snapshot of the physiology and clinical characteristics during a single echocardiography measurement. This also includes the administration of inotropic support, which would have changed after performing the echocardiography. While some echocardiographic markers depend on loading conditions, we did not include data on the amount of fluid resuscitation and exact inotropic dosage. Lastly, the echocardiographic parameters performed for invasive ventilated controls could be different if performed at other times, however the clinical parameters such as oxygenation index and inotropic support were significantly different in this group when compared with infants with PPHN. Despite these limitations, the strengths of this study are manifold. Firstly, this study was performed in a single unit routinely caring for infants with PPHN. Furthermore, this cohort included infants with sepsis and hypoxic ischaemic encephalopathy, thus capturing the effect of these conditions on the cardiac physiology. Lastly, echocardiographic markers such as sEI were examined offline several months after their initial recording using an anonymised database by a single operator thus reducing variability and therefore providing less biased results.