Based on large randomized controlled trials, therapeutic hypothermia (TH) is known to improve neurodevelopmental outcomes in infants with moderate to severe neonatal encephalopathy (NE)1. However, to our knowledge, there are currently no trials or recommendations about optimal ventilation management (mechanical ventilation (MV) versus spontaneous ventilation) with moderate NE. By analogy with children with head trauma,2 with a neuroprotective hypothesis, the objective of this study was to compare onset of SBI (secondary brain injuries) in infants with moderate NE treated by TH, according to mechanical versus spontaneous ventilation in the first 72 hours of life.
For this study, all infants ≥34 weeks of gestational age (GA) with moderate NE (Sarnat 2), treated by TH and included in the French LyTONEPAL (Long-Term Outcome of Neonatal EncePhALopathy) cohort were studied. The LyTONEPAL study was approved by all the ethics committee.3
Infants were classified in the MV versus SV groups according to conventional or high-frequency oscillation ventilation versus non-invasive ventilation or room air at 24h of life.
The primary endpoints were each SBI identified at least once between 12h and 72h of life. SBI was defined according to the literature (Table 1). As sedation could be a confounding factor not dissociable from MV or TH, we studied sedation at 24h of life and type of medication (analgesic, sedative, or both) as a secondary endpoint.
From September 2015 to March 2017, 337 (42.5%) of the 794 infants were included and 457 (57.5%) excluded: 383 with NE grade 1, 3 or unknown, 46 without TH, and 28 with an unknown ventilation status at 24h of life. The MV and SV groups comprised 173 (51.3%) and 164 infants (48.7%) respectively: 165 infants (95.3%) had conventional ventilation, 8 (4.7%) high-frequency ventilation, 160 (97.5%) no ventilatory support, and 4 (2.5%) non-invasive ventilation. The rate of MV rate varied between centers between 0 and 100% (p<0.001) (Figure 1).
At admission to neonatal intensive care unit, all perinatal characteristics and birth circumstances were similar between MV and SV groups. In contrast, during TH, pulmonary hypertension was more frequent in the MV group (n=16, 10.8% versus n=5, 3.1% in the SV group, p=0.006), as was myocardial injury (n=30, 17.9% in the MV group and n=15, 9.3% in SV group).
Concerning SBI and sedation data are presented in Table 1.
In the present French cohort, 50% of infants with moderate NE treated by TH were intubated. In this population, MV did not provide better control of SBI during TH in infants with grade 2 NE. There was even a higher frequency of hypotension in the MV group (68.8% versus 50% in the SV group, p=0.023), but there was no difference for the other SBI. Almost 90% of infants with moderate NE and TH were under sedative or analgesic drugs, or both. Only use of hypnotics differed: higher (10.1%) in the MV group, versus 2.1% in the SV group (p=0.012). Use of morphine alone or combination of the 2 molecules was comparable. All infants with hypotension during the 72h after birth received sedation.
The proportion of ventilated infants under TH varies in the literature but is quite consistent with the present rate of 50%. The Vermont Oxford Neonatal Encephalopathy Registry reported 65% ventilation in infants with moderate to severe encephalopathy; Lopez Laporte reported 70% in 2017.4,5
The higher frequency of hypotension may be due to greater use of hypnotics, with their side effects, or higher frequency of myocardial failure and pulmonary hypertension in the MV group. It is known that hypotension is the most unfavorable prognostic factor, along with hypoxia, in children with traumatic brain injury and must be controlled to limit secondary brain injury.6,7 It should be noted that 100% of infants with hypotension were sedated, which prevented more precise analysis.
There was no significant difference in capnia between the 2 groups. In a retrospective study of 198 infants, the authors found that invasive ventilation was associated with a lower pCO2 level in infants under TH; however, the 2 groups were not comparable in terms of severity, with more severe NE, more severe brain injury and more deaths in the intubated group.5 Hypocapnia is in fact often due to compensatory effects of initial metabolic acidosis and reduced metabolic rate by TH, which leads to a decrease in CO2 production.8
Regarding sedation, 11.3% of our cohort received none. By analogy with head trauma, infants under TH should be appropriate sedated, to limit SBO by reduced cerebral metabolic rate and adverse physiologic responses such as stress, which can decrease the effectiveness of TH and increase the intracranial pressure.9 Moreover, neonatal comfort is considered to be an important aspect of neonatal practice, and sedation must be adequate and sufficient.10 On the other hand, sedation can lead to adverse effects. Currently there are no guidelines for adequate sedation in infants under TH. A systematic review was conducted by Bäcke et al. in 2022, but the authors were unable to reach any conclusion, due to the heterogeneity of the studies and lack of randomized controlled studies.11
The present study has certain limitations: missing data on SBI made it difficult to use a score and cross-reference relevant data. Concerning blood pressure missing information, we were able to use other data (fluid bolus and amine use). Moreover, we did not have more precise data concerning the other drugs used -such as paracetamol- or concerning dose or pain scales.
This study also has some strengths. The data were from a prospective national population-based cohort. To our knowledge this was the first national-level French study to evaluate SBI and MV in the context of NE with TH.2 To go further, work is underway on ventilation patterns and outcomes in the same cohort.
To conclude, MV did not seem to improve the management of SBI during the first 72h of life in infants with moderate NE and TH. Although indications for initiation of TH are now well defined, more studies on ventilation mode and sedation in TH are needed to enhance the comfort and neurodevelopment of these infants.