Is 100% oxygen a sticking plaster for sore neonatal ventilation skills?

The use of supplementary oxygen was generally accepted and used in the delivery room resuscitation of neonates from ~1780, and a century later its use was extended to neonatal special care units 1. The assumption was that reoxygenation with 100% oxygen was beneficial for patients with perinatal asphyxia, due to faster replenishment of the oxygen debt 2, and that the treatment was not toxic 3. However, this practice is now being questioned more than 200 years after clinicians started to resuscitate asphyxiated newborn infants with 100% oxygen 1. Over the past four decades, there has been a lot of debate about the role of the oxygen molecule in sick newborn infants, especially those who are immature and, or, have suffered from hypoxia-ischaemia. In 1980, the Norwegian neonatologist Ola Didrik Saugstad proposed that posthypoxic reoxygenation in newborn infants resulted in a harmful burst of oxygen-free radical production, namely an increase in so-called oxidative stress 4. Since then hyperoxia has been proved to be detrimental in asphyxiated infants. This resulted in changes to the International Liaison Committee on Resuscitation neonatal resuscitation guidelines in 2010. The revised guidelines stated that compromised term or near term infants should be initially ventilated with 21% oxygen and not 100% 5. This led to a change of mindset in professionals caring for newborn infants. To understand the mechanisms of any resuscitative intervention, including assisted ventilation, the complex physiological changes that occur during the immediate transition at birth need to be understood. Any infant needs to establish pulmonary gas exchange to enable its lungs to take over the function of the placenta in supplying oxygen to their body. Lung aeration is critical to initiating the sequence of interdependent physiological changes necessary for a successful transition to extrauterine life 6. Hooper et al. 6 described a three-phase physiological process. In the first phase, the airways are filled with liquid, and no pulmonary gas exchange can occur, which means that the transitional support should focus on clearing the liquid from the lungs. During the second phase, the gas exchanging regions of the lungs are mostly clear of liquid, allowing pulmonary gas exchange. However, in this phase, the liquid that was cleared from the airways is still present in the perialveolar interstitial tissue and there is a risk that it will re-enter the airways. Therefore, respiratory support should focus on minimising the re-entry of alveolar fluid during expiration. The third phase occurs when the liquid has been cleared from the lung tissue 6. Based on the transitional physiology described above, mastering the skill of positive pressure ventilation (PPV) with a bag or T-piece and mask is essential to facilitate the transition in infants who fail to do so spontaneously. Despite that, inadequate PPV during neonatal resuscitation remains a common problem, often resulting in either persistent bradycardia or asystole and the initiation of chest compressions. The delayed restoration of oxygenation and circulation, in addition to providing chest compressions and administering adrenaline, might be harmful. When we consider these aspects of oxidative stress and the importance of lung aeration, it is an intriguing, but also provoking idea, that providing brief exposure to 100% oxygen might mitigate the disadvantage of providing inadequate PPV in the delivery room during cardiopulmonary resuscitation. In this issue of Acta Paediatrica, Linner et al. 7 report the results of a study they performed on asphyxiated piglets who were 12–34 hours of age, to examine the effects of very limited 100% oxygen exposure on the time to return of spontaneous circulation. This model of severe hypoventilation mimicked a blocked airway during cardiopulmonary resuscitation with minimal air entry during assisted ventilation. When the piglets received 100% oxygen, they experienced an improved return of spontaneous circulation with less need for chest compressions compared to inadequate ventilation with 21% oxygen. Asphyxiated piglets with bradycardia and hypotension were exposed to 10 minutes of assisted ventilation with a total of 75 ml/kg of either 21% or 100% oxygen. Thus, the piglets were exposed to only very small amounts of 100% oxygen and, reassuringly, the findings by Linner et al. do not defy the paradigm that we should be concerned about postasphyxial hyperoxia. The authors concluded that no signs of hyperoxia could be measured in this animal model after such a brief exposure to 100% oxygen. They noted the first time points when regional cerebral oxygen saturation exceeded, and then remained, above 30% and the time when the partial pressure of oxygen in brain tissue (PbtO2) had increased by 0.1 kPa from its nadir and remained above that level. In the piglets that received 100% oxygen, PbtO2 had increased by 0.1 kPa from its nadir by two minutes after initiation of resuscitation and remained above that level. The peak value was 2.8 kPa (range: 0.5–5.7 kPa). In all animals ventilated with air, PbtO2 increased by less than 0.1 kPa from its nadir. The median regional cerebral oxygen saturation varied in one-minute cycles, with nadirs of 15% and peaks of 26%–49% in piglets ventilated with 100% oxygen. In the piglets ventilated with air, regional cerebral oxygen saturation remained at the minimum value of 15%. Despite significant differences in the PbtO2 and regional cerebral oxygen saturation between the piglets receiving 21% and 100% oxygen, the authors concluded that there was no brain hyperoxia in the 100% oxygen group. This was supported by the maximum PaO2 of 9.6 kPa and PbtO2 of 5.7 kPa observed in the piglets receiving 100% oxygen. These are both normal values for oxygen tension. The infrequent and unexpected need for resuscitation, coupled with the challenges of obtaining informed consent for such studies, pose barriers to the design and completion of rigorous randomised delivery room resuscitation trials 8. Thus, animal models have been extensively used in neonatal resuscitation research, with piglets being the most frequently used large animal model 9. Animal studies have the advantage of standardised protocols, including timing and rigorously performed interventions. In general, the disadvantages include biological variations in large animals like piglets and sheep and large animal models also tend to suffer from a lack of power because the experiments are labour-intensive and expensive. Also, like most other piglet models of neonatal resuscitation, the study by Linner et al. 7 was performed in post-transitional piglets. The study is an example of a rigorously performed animal study of good quality. However, a weakness was that the model lacked certain important characteristics of a human infant during perinatal transition. The piglets even underwent lung recruitment manoeuvres to ensure open lungs with minimal atelectasis during the experiment. The question about whether 100% oxygen administration can facilitate transition during failed PPV of fluid-filled lungs remains unanswered. Lung expansion is critical for the complex cardio-respiratory changes necessary for the successful transition to extrauterine life, and it would be of interest to investigate whether the findings of Linner et al. could be reproduced in a transitional animal model. In summary, the Linner et al. 7 study provides a valuable contribution to our knowledge about optimal initial stabilisation and resuscitation in the delivery room and informs the on-going discussions on that subject. It is intriguing, yet thought provoking, that clinicians’ shortcomings in providing effective PPV can be overcome by increasing the fraction of inspired oxygen, potentially eliminating the need for chest compressions due to persistent bradycardia. The authors have no conflict of interests to declare.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.