Abstract
Infants with perinatal asphyxia and moderate-to-severe hypoxic ischemic encephalopathy (HIE) are currently treated with therapeutic hypothermia (TH) as part of a brain protective strategy. However, perinatal asphyxia is a risk factor for development of persistent pulmonary hypertension (PPHN). As such, the aim of this study was to quantify the risk of PPHN in infants undergoing TH and assess short-term outcomes in infants developing PPHN. All N = 59 infants undergoing TH for moderate-to-severe HIE over a period of 3 years (January 2020–December 2022) at a single center were included. PPHN was diagnosed in N = 10 (17%), with this deemed to have been exacerbated by TH in n = 6 (10%). Only 50% (5/10) with PPHN required inhaled nitric oxide, and none of the infants received extracorporeal membrane oxygenation. PPHN was not found to be significantly associated with short-term outcomes, including the extent of HIE on brain magnetic resonance imagings, in-hospital mortality or requirement for nasogastric feeding at discharge. In conclusion, TH appears to be a safe and effective treatment for moderate-to-severe HIE with or without PPHN.
Introduction
Hypoxic ischemic encephalopathy (HIE) is a serious neurological complication of newborns, with an incidence of approximately 1–3 per 1000 live births in developed countries (Kurinczuk et al., 2010; Lai and Yang, 2010). HIE is associated with high morbidity and mortality of infants, and is a major burden for the patient, their family, and society (Lai and Yang, 2010). Currently, therapeutic hypothermia (TH) is the standard treatment for infants with moderate-to-severe HIE (Lemyre and Chau, 2018).
In some newborns, pulmonary vascular resistance (PVR) remains elevated after birth, resulting in shunting of blood away from the lungs, causing severe hypoxemia (Lazar et al., 2012). Perinatal asphyxia is a risk factor for development of persistent pulmonary hypertension (PPHN) (Delaney and Cornfield, 2012). Other risk factors for PPHN include hypothermia, sepsis, meconium aspiration syndrome, severe respiratory distress syndrome, structural lung and heart diseases, and antenatal drug exposure (including selective serotonin reuptake inhibitors [SSRIs], nonsteroidal anti-inflammatory drugs, and cigarette smoking) (Puthiyachirakkal and Mhanna, 2013). In animal studies, hypothermia has been shown to be associated with an increase in PVR, with a 1°C reduction in temperature increasing PVR by 1–2% (Rubini, 2004) However, a 2013 Cochrane review of four major clinical trials involving 614 infants did not demonstrate an increased risk of PPHN with TH in infants with HIE, compared with infants receiving normothermic treatment (Jacobs et al., 2013).
Two more recent studies have assessed outcomes in infants with PPHN undergoing TH. Lakshminrusimha et al. combined pooled data from two randomized control trials, and found PPHN to be common in infants with moderate-to-severe HIE (22%), with a similar prevalence in the TH and normothermic arms (Lakshminrusimha et al., 2018). Length of hospital stay and mortality rates were found to be significantly increased in infants developing PPHN. However, these analyses pooled data from the TH and normothermic treatment arms, hence these findings may not be directly applicable to infants undergoing TH. Yum et al. analyzed short-term outcomes of infants with HIE who had developed PPHN prior to commencement of TH, and reported PPHN to be associated with significantly longer hospital length of stay, but found no significant association with electroencephalography or magnetic resonance imaging (MRI) findings (Yum et al., 2017). However, since this study classified PPHN prior TH treatment, these findings may not be applicable to infants developing PPHN as a complication of TH.
The main purpose of the present study was to investigate the association between TH for HIE and PPHN, and to measure short term outcomes in infants with PPHN, including brain MRI changes and type of feeding at discharge. A better understanding of the potential impact of TH on PPHN is essential for prompt identification and management of this complication.
Methods
Setting
The study was based at Birmingham Heartlands Hospital (BHH), which is part of the University Hospitals Birmingham NHS Trust. BHH is a regional cooling center in the West Midlands region of United Kingdom, to which infants from other centers in the region are referred for TH treatment. TH was achieved by active whole-body cooling, to maintain rectal temperature of 33–34°C for 72 hours.
Eligibility for treatment with TH for moderate-to-severe HIE was as per regional guidelines. Specifically, infants were eligible for TH if they had gestational age >35 weeks, birth weight of >1.8 kg, were aged <6 hours, and met criteria A, B, and C below:
Criteria A: At least one of the following: (a) pH of <7; (b) base deficit ≥16 mmol/L in cord blood or any blood gas sample within 60 minutes of birth; (c) Appearance, Pulse, Grimace, Activity, and Respiration of <5 at 10 minutes; (d) Assisted ventilation for >10 minutes after birth.
Criteria B: Moderate-to-severe encephalopathy consisting of altered state of consciousness and at least one of the following: (a) hypotonia; (b) abnormal reflexes including oculomotor or pupillary abnormalities; (c) absent or weak sucking; (d) clinical seizures.
Criteria C: Infants meeting criteria A and B should be assessed for at least 30 minutes of cerebral function monitoring (CFM). Where the referring center did not have the ability to perform CFM, only criteria A and B were used to assess the eligibility for TH.
Infants with major congenital abnormality, metabolic diseases, or massive hemorrhage were subsequently excluded for treatment with TH.
The diagnosis of PPHN was based on either clinical signs in the first 24 hours of life (pre- and post-ductal saturation difference >10%, and a high oxygen requirement, with or without systemic hypotension), and/or echocardiography findings of raised pulmonary pressures (tricuspid regurgitation, right-to-left shunting at patent foraman ovale (PFO) and/or patent ductus arteriosus (PDA), right ventricular dysfunction, and D-shaped left ventricle in short-axis view). In infants diagnosed with PPHN, this was deemed to have been exacerbated by TH if the oxygen requirement was seen to increase significantly (more than 20% increase in FiO2) during or following a drop in rectal temperature within the first 24 hours after birth. All infants underwent a brain MRI between day 5–15 of life, as per regional guidelines; where evidence of HIE was observed, this was classified as mild, moderate, or severe by a consultant radiologist (Trivedi et al., 2017). During the period of TH, CFM was performed on all infants for at least 72 hours continuously, unless the infant died before completing 72 hours of TH. Any evidence of electrical seizure activity on CFM was recorded, as was any clinical evidence of seizures during this period.
Data collection
All infants undergoing TH for moderate-to-severe HIE at BHH between 1st January 2020, and 31st December 2022, were included in the study. Infants were identified retrospectively, and data were collected from the electronic patient record system (“Badger”). This included maternal and neonatal characteristics, as well as treatment and infant outcomes, specifically the extent of HIE on brain MRI, feeding method at discharge, seizures (identified either on CFM or clinically), duration of invasive respiratory support, length of hospital stay, and treatment with inhaled nitric oxide (iNO) and inotropes. For infants with PPHN, data for all routine observations recorded in the first 24 hours after birth (i.e., oxygen saturation, oxygen requirement [FiO2], temperature, heart rate, and blood pressure) were respectively reviewed, and used to determine whether the PPHN was exacerbated by TH, as previously defined.
The study was registered as a clinical audit on the local clinical audit and registration management system (audit ID: CARMS-20984). The need for informed consent was waived due to the retrospective nature of the study, and the fact that the data were collected in a pseudonymized format. Infants were excluded if they had opted out of their data being used for research, as part of the NHS National Data Opt-Out.
Statistical methods
Comparisons between the PPHN and non-PPHN groups were performed using independent samples t-tests for continuous variables that were approximately normally distributed; Mann–Whitney U tests for non-normal or ordinal variables; or Fisher’s exact tests for nominal variables. Continuous variables were reported as mean ± standard deviation where approximately normally distributed, or as median (interquartile range; IQR) otherwise. All analyses were performed using IBM SPSS v29 (IBM Corp. Armonk, NY), with p < 0.05 deemed to be indicative of statistical significance throughout.
Results
Cohort characteristics
A total of N = 59 infants underwent TH for moderate-to-severe HIE during the study period, of whom 17% (N = 10) developed PPHN, as diagnosed by clinical and/or echocardiographic findings. In those developing PPHN, this was deemed to be exacerbated by the TH in 60% (6/10) of cases (i.e., 10% [6/59] of the overall cohort). No significant differences in maternal or neonatal characteristics were detected between the PPHN and non-PPHN groups (Table 1).
Baseline Neonatal and Maternal Characteristics
Continuous variables are reported as “mean ± standard deviation” with p-values from independent samples t-tests, or as “median (interquartile range)” with p-values from Mann–Whitney U tests, as appropriate. Categorical variables are reported as “N (column %)” with p-values from Fisher’s exact tests. Bold p-values are significant at p < 0.05. PPHN, persistent pulmonary hypertension.
APGAR, Appearance, Pulse, Grimace, Activity, and Respiration.
Neonatal outcomes
For the N = 52 infants who underwent brain MRIs, the extent of HIE was found to be similar in the PPHN and non-PPHN groups (p = 0.861, Table 2). Seizures were observed in 69% infants who underwent TH for HIE, with similar rates in the PPHN and non-PPHN groups (70% vs. 69%, p = 1.000). No infants in either group required extracorporeal membrane oxygenation (ECMO). Infants with PPHN required significantly longer durations of invasive respiratory support (median: 6 vs. 3 days, p < 0.001), and significant higher rates of inotropic support (100% vs. 41%, p < 0.001) and treatment with iNO (50% vs. 0%). However, in-hospital mortality rates did not differ significantly between the PPHN and non-PPHN groups (10% vs. 8%, p = 1.000), with infants surviving to discharge having similar lengths of stay in hospital (median: 11 vs. 10 days, p = 0.352) and rates of nasogastric feeding at discharge (33% vs. 27%, p = 0.696).
Neonatal Outcomes
Continuous variables are reported as “mean ± standard deviation” with p-values from independent samples t-tests, or as “median (interquartile range)” with p-values from Mann–Whitney U tests, as appropriate. Categorical variables are reported as “N (column %)” with p-values from Fisher’s exact tests, unless stated otherwise. Bold p-values are significant at p < 0.05.
p-value from Mann–Whitney U test, as the factor is ordinal.
The number of infants who underwent seizures, identified either from electrical activity on cerebral function monitoring, or from clinical observations.
Excludes infants who died in hospital. ECMO, extra corporeal membrane oxygenation; HIE, hypoxic ischemic encephalopathy; MRI, magnetic resonance imaging; PPHN, persistent pulmonary hypertension.
Discussion
HIE due to perinatal asphyxia can lead to severe neurological depression, seizures, and long-term neurological sequalae (Shankaran et al., 1991). High admission FiO2 or high oxygen requirement during the early phase of hypothermia is an indicator of the development of PPHN, which requires immediate management. Prompt evaluation of PPHN by echocardiography, followed by initiation of iNO and maintenance of an appropriate blood pressure are vital to prevent these infants sliding into refractive pulmonary hypertensive crisis (Joanna et al., 2021). In the current study of infants undergoing TH for moderate-to-severe HIE, PPHN exacerbated by TH was observed in 10% (6/59) of the cohort. This rate is comparable to the 11% (13/116) reported by Yum et al. (Yum et al., 2017).
HIE can lead to significant brain injury, which can be exacerbated by uncontrolled PPHN and asphyxia, leading to impaired brain oxygenation; this was reported in a case series by Gagnon et al. (Gagnon and Wintermark, 2016). However, the current study found no significant difference in MRI brain abnormalities between the PPHN and non-PPHN groups, which is consistent with the findings by Yum et al. (Yum et al., 2017). This is an encouraging finding since brain MRI is a useful predictor for future neurodevelopmental outcome. The non significant difference in MRI brain changes may be indicative of a positive influence of iNO administration. This is supported by animal studies, which have described a protective effect of iNO in HIE induced rodents in the restoration of cerebral blood flow to ischemic regions of the brain (Charriaut-Marlangue et al., 2012; Terpolilli et al., 2012). The prevalence of seizures was also similar in PPHN and non-PPHN groups, which is consistent with the findings of Yum et al. and Vijverberg et al. (Yum et al., 2017; Joanna et al., 2021). As such, there is currently no evidence to suggest that PPHN increases the risk of seizures in infants undergoing TH.
The duration of ventilation was found to be significantly longer in infants with PPHN, with a median of 6 days compared with 3 days in non-PPHN group. This was consistent with Agarwal et al., who reported a mean of 9 and 2 days respectively. There was no statistically significant difference in length of hospital stay, with a median of 11 days in PPHN group and 10 days in non-PPHN group. This was in contrast to Agarwal et al., who reported a significantly longer mean hospital stay of 23 days in PPHN group, compared to 12 days in the non-PPHN group (Agarwal et al., 2021). This discrepancy may be a result of the small sample size in the current study yielding insufficient statistical power to detect the difference between groups. Infants in PPHN groups are generally more unwell needing longer duration of ventilation and hospital stay. They may need treatment with iNO and inotropes, which can lead to slow weaning and recovery (Nair and Lakshminrusimha, 2014).
Another finding in the current study was that all infants with PPHN required inotropic support, which was significantly higher than the 41% in the non-PPHN group (p < 0.001). This was consistent with the findings reported by Agarwal et al. and Yum et al. (Agarwal et al., 2021; Yum et al., 2017). Inotropes are commonly used in infants with PPHN to improve the systemic blood pressure, which can help in reversal of shunt at PFO and PDA level from right-left to left-right (Sharma et al., 2015).
A total of 50% of infants in PPHN group in our study required iNO, compared with 70% in Agarwal et al. (Agarwal et al., 2021). iNO is commonly used in PPHN infants who are refractory to initial management of PPHN (intubation and ventilation, correction of acidosis, optimizing blood pressure, etc.) or have a high oxygenation index of >20 (Lakshminrusimha and Keszler, 2015). None of the infants in the current study required ECMO.
There was no statistically significant difference between PPHN and non-PPHN groups in the rates of oral feeding at discharge, which is consistent with Agarwal et al. study (Agarwal et al., 2021). Oral feeding at discharge is an important favorable neurological outcome in infants with HIE. Delay in establishing oral feeding could be due to severity of HIE or prolonged intensive care stay.
It was observed that desaturation during TH is due to pulmonary hypertensive crisis, meaning that rigorous monitoring is required. Optimizing hemodynamics by maintaining adequate blood pressure to prevent right-to-left shunting will help to limit episodes of desaturations (Gagnon and Wintermark, 2016). Along with pre- and post-ductal monitoring for PPHN, there is emerging evidence that NIRS (Near Infra-Red Spectroscopy) monitoring of cerebral oxygenation may be more useful to diagnose PPHN. In PPHN, there will be a reduction in cerebral oxygenation due to right-to-left shunting (Kurinczuk et al., 2010). Thus, aggressive treatment of PPHN is required to prevent brain injury. Close monitoring of cardiovascular, hematological, biochemical, and acid–base balance is vital among infants undergoing TH.
Even though in our study, 10% developed TH exacerbated by PPHN, all were successfully treated with appropriate escalation of management for PPHN. This suggests that infants with PPHN may benefit from the neuroprotective effect of TH similar to those without PPHN (Yum et al., 2017). However, individual cases with severe PPHN must be carefully monitored because deterioration can occur rapidly and can lead to very complicated and fatal course. Further multicenter studies evaluating the incidence of PPHN during TH would provide more insight on outcomes in these infants.
The primary limitation of this study was small sample size, particularly in the group of infants with PPHN. This will have resulted in low statistical power, meaning that only large effect sizes will have been detectable. For example, a post-hoc power calculation for the outcome of in-hospital mortality yielded a minimal detectable relative risk of six (at 80% power). Consequently, comparisons between groups will have had an inflated-false negative rate; hence, nonsignificant differences must be interpreted cautiously. The confounding factors like meconium aspiration, maternal diabetes and SSRIs use, which may have contributed to development of PPHN were not studied. Finally, the retrospective nature of the study meant that it was not possible to assess long-term (i.e., 2-year) neurodevelopmental outcome data.
To conclude, infants who develop PPHN during TH require longer durations of ventilation and inotropic support. However, PPHN was not associated with significant differences in short-term outcomes, including MRI brain changes, feeding modality at discharge and in-hospital mortality. As such, while PPHN during TH in infants with HIE is a perilous complication requiring prompt intervention, appropriate escalation of treatment appears to lead to acceptable short-term outcomes.