Original Article Oxygen for Newborns: How Much Is Too Much?

Original Article Oxygen for Newborns: How Much is Too Much?

Ola Didrik Saugstad, MD, PhD World Health Organization, the estimated global burden of birth asphyxia is between 4 and 9 million newborns out of 130 million births. It is estimated that the number of infants with asphyxia- related neurologic disability is around 1 million, and the number International guidelines for newborn resuscitation recommend the use of of neonatal deaths due to birth asphyxia is also in the range of 100% oxygen. However, high concentrations of oxygen after asphyxiation 1 million.4 In addition, there are approximately 1.6 million activate reactive oxygen species that may contribute to a number of intrapartum stillbirths and many of them are related to morbidities. Animal models have been useful in describing their intrauterine asphyxia. mechanisms, but only large-scale clinical trials can provide evidence that In 1999, an advisory statement from the Pediatric Working may be used to alter clinical practice. It has been demonstrated that Group of the International Liaison Committee on Resuscitation neonates recover faster when resuscitated with room air as opposed to pure said that 100% oxygen should be used for newborn resuscitation.5 oxygen and neonatal mortality rates are improved. Increases in saturation This statement is, in my opinion, outdated and no longer accurate. are equal with oxygen and room air resuscitation. Studies of normal oxygen There is a building body of evidence that suggests that room air is saturation immediately after birth suggest that clinicians may unnecessarily sufficient for neonatal resuscitation and that the use of pure be rushing to high saturations. In the first weeks of life, lower saturation oxygen generates oxygen radicals and may be toxic. However, large targets in preterm infants reduce retinopathy of prematurity and pulmonary clinical trials that satisfy the requirements of evidence-based complications and may improve growth. The neonatologist would be well medicine are still needed in this field. served to think of oxygen as a medication, and use it sparingly. Journal of Perinatology (2005) 25, S45–S49. doi:10.1038/sj.jp.7211321 Sources of Free Radicals Hypoxanthine accumulates during hypoxia, and during reoxygenation, superoxide radicals are produced.2,3 As more oxygen INTRODUCTION is given during resuscitation, more superoxide radicals are produced, leading to cell injury.2 Oxygen radicals are produced by Scheele and Priestley are credited with the discovery of oxygen, but a number of sources, not only by the hypoxanthine–xanthine 170 years before their publications, a Polish alchemist, Michael oxidase system. The most important source may be the Sendivogius, described a gas that could be obtained by heating 1 mitochondria in the electron transfer chain. Oxygen radicals are potassium nitrate. He called the gas ‘‘the elixir of life.’’ Ironically, also produced by activated macrophages and oxidation of in neonates suffering asphyxia, dispensing the ‘‘elixir of life’’ in its catecholamines and arachidonic acid. The hydroxyl radical is pure form activates toxic substances, ultimately resulting in cell 2,3 produced through Fenton chemistry, which might induce further death. It is this paradoxical role of oxygen in full-term and oxidation, protein oxidation, DNA damage and may also affect premature neonates that will be reviewed to better understand its signal transduction.6 application in resuscitation after asphyxia, the optimal saturation the first weeks of life, and considerations related to its overuse. There is a need for a greater respect for oxygen as an entity that An In Vivo Animal Model possesses both therapeutic and toxic characteristics, similar to A model of hypoxia/reoxygenation in pigs has been developed.7–10 prescribed medications. Piglets of 1 to 2 days old are exposed to 8% oxygen until they are Asphyxia is a devastating condition that causes morbidity and close to collapse. They are then resuscitated with either room air, mortality in neonates in a number of forms. According to the 100% oxygen, or any other oxygen concentration of interest to be studied. The system features ability to measure systemic and local hemodynamics and by applying microdialysis local metabolic

Department of Pediatric Research, Rikshospitalet University Hospital, University of Oslo, Oslo, changes in the brain as well. The microcirculation in the brain Norway. may be measured by laser Doppler, and brain oxygenation Proprietary interest/disclosure: None. measured by oxygen probe. A nitric oxide sensor is also used in

Address correspondence and reprint requests to Ola Didrik Saugstad, MD, PhD, Department of some studies to measure local nitric oxide concentrations. In this Pediatric Research, Rikshospitale, 0027 Oslo, Norway. model, pulmonary arterial blood pressure follows a biphasic pattern

11 during the hypoxemic phase. During resuscitation, there is a very SaO2 in 1 min Apgar < 4 rapid and high increase in the pulmonary arterial pressure, which 100 11 Room air is followed by a gradual normalization. 100% Oxygen Hydrogen peroxide from the sagittal sinus in these newborn 75 12 piglets has also been studied. Hydrogen peroxide concentration % 2 increases during resuscitation with 100% oxygen, and there is no 50 P= 0.003 increase in animals resuscitated with room air. These data agree SaO with a study by Kondo et al., in which the authors used a newborn 25 apneic piglet model, measuring the concentration of free radicals 13 0 on the lung surface for 6 minutes. Animals resuscitated with 1 3 room air had no increase in free radical production. Animal Min after birth models, therefore, consistently demonstrate that resuscitation of Figure 1. SaO2 at 1 and 3 minutes of life in asphyxiated neonates hypoxic neonates with 100% oxygen results in the production of with 1 minute Apgar score <4 resuscitated with room air and 100% free radicals, whereas resuscitation with room air does not induce oxygen. Data from the RESAIR 2 database. A small but signiﬁcant higher SaO was noted in room air babies. free radical production. 2

Table 1 Time to First Cry in Minutes in Neonates Resuscitated with Room Air and 100% Oxygen; Results from Two Clinical Trials Clinical Trials: RESAIR 2 Saugstad et al.14 Ramji et al.15 The RESAIR 2 study enrolled approximately 600 infants in six countries in a multicenter clinical study of asphyxiated newborn 21% Oxygen 1.6 2.0 infants comparing their resuscitation with either room air or 100% 100% Oxygen 2.0 3.0 oxygen.14 Heart rate, one of the outcome parameters, was recorded p 0.005 0.008 for the first 30 minutes of life. There was no difference in heart rate between neonates resuscitated with 100% oxygen and neonates resuscitated with room air.14 More surprising, there was no This OR represents a 40% reduction in neonatal mortality difference in oxygen saturation between neonates resuscitated with associated with resuscitation with room air instead of 100% oxygen. 100% oxygen and room air. Oxygen saturation (SaO2) was Examining the cohort from Spain separately, neonatal mortality measured in a subset of infants and was approximately 60% at 1 was reduced from 3.5 to 0.5%. It is possible that this level of minute, and 80% at 3 minutes, and gradually increases up to 90% mortality is more representative of industrialized countries and at 10 minutes. Neonates resuscitated with room air had rather therefore reflects the situation in North America and Western similar oxygen saturation as those given 100% oxygen. Figure 1 Europe. Preterm neonates were also analyzed separately. Also, for shows SaO2 values in infants with 1 minute Apgar score <4. this group, neonatal mortality decreased significantly when infants Further, resuscitation with 100% oxygen delayed the time to first were resuscitated with room air (OR 0.51, 95% CI 0.28 to 0.90). cry14 (Table 1). The delay was in median 24 seconds in the Applying these rates of reductions in mortality to the occurrence RESAIR 2 study, and 60 seconds in a study by Ramji et al.15 of birth asphyxia, a 5% reduction in 4 million neonates worldwide Regarding neonatal mortality, neonates resuscitated with room suffering from birth asphyxia corresponds to 200,000 lives. If the air had a rate of 14% as compared to 19% in the group resuscitated 3% reduction from Spain is representative of North America and with 100% oxygen. One possible explanation of the high mortality Western Europe, both of which have 200,000 neonates annually rate in these infants is that most of these babies were enrolled from that require resuscitation, these figures imply the rescue of 6000 developing countries. The odds ratio (OR) of 0.69 (0.44 to 1.06) lives every year. was not significant, but there is a strong tendency in favor of Although the impact on outcomes seems to be evident, the room air.14 optimal oxygen concentration for resuscitation remains to be To date, five studies have been published including 1737 infants defined. In order to define the optimal level, the normal time- randomly assigned (or pseudorandomly assigned) to room air or course of oxygen saturation during first minutes postpartum must oxygen resuscitation. In a meta-analysis, we examined neonatal be known. Rao and Ramji17 published a study suggesting that mortality as the primary outcome parameter, and secondary normal newborn babies have a median or a mean oxygen parameters of heart rate, APGAR score and time to first breath.16 saturation of 70% at 1 minute of age, and that there is a large The meta-analysis found a reduction in neonatal mortality from variation in normal levels which include saturations as low as 40% 13% in the oxygen group to 8% in the room air group, giving ORs in normal healthy neonates. At 3 minutes, average saturation favoring resuscitation with room air of 0.57 (95% CI 0.4 to 0.8). increases to 75%, reaches 80% at 5 minutes and begins to plateau

S46 Journal of Perinatology 2005; 25:S45–S49 Oxygen for Newborns Saugstad

near 90% at 10 minutes. More studies are needed, but these pilot occurrence of threshold ROP, fewer days on oxygen, fewer days on data suggest that normal healthy neonates may not be pink until 5 artificial ventilation and fewer neonates with weights below the to 10 minutes after birth. Clinically, this calls into question the third percentile at discharge.18 utility of APGAR scores. Since color is one of five variables in the The BOOST (Benefits Of Oxygen Saturation Targeting) trial APGAR score, clinicians may strive to increase clinical scores in from Australia compared neonates with oxygen saturations of 91 to their patients, leading them to hasten increased oxygen saturation 94% with neonates with saturations of 95 to 98%.20 While the beyond the levels that Rao and Ramji found to be normal. authors found no difference in long-term development, there was an increase in the duration of oxygen therapy in the high saturation group, an increase in the occurrence of home oxygen Oxygen Saturation After Birth therapy and more frequent chronic lung disease in the high What, then, is the best protocol for saturation in the first weeks of oxygen saturation group. life? Tin et al.18 analyzed data from NICUs in Northern England Five studies have recently addressed the question of the optimal arterial saturation the first weeks after birth of extremely low and identified four oxygen policies in neonates according to oxygen 18–21 saturation limits that were set at: (i) 70 to 90%, (ii) 84 to 94%, (iii) weight birth infants. These are summarized in Table 2. It 85 to 95%, or (iv) 88 to 98%. The occurrence of retinopathy of seems that a high oxygen saturation of 93% or more increased the prematurity requiring cryotherapy (threshold ROP) was four times risk for severe ROP, and also the risk for lung complications. There higher in the high saturation group compared to the low was no difference in mortality or neurodevelopment. Whether growth is inhibited by a high saturation is not completely clarified. saturation group with saturation limits between 70 and 90%. This 21 finding was confirmed in a recent study, which found that The study by Chow et al. emphasizes the importance of avoiding neonates nursed in oxygen saturations greater than 92% have more peaks in SaO2 and constant teaching of the staff in keeping the severe ROP than babies nursed in saturations less than 93%.19 Tin saturation within strictly defined limits. also found that days on ventilation in the high saturation group were twice as many compared to the 70 to 90% group. There was also evidence of growth retardation. The percentage of neonates Recent Experimental Data below the third percentile of weight at discharge was significantly Why is a brief exposure of oxygen immediately after birth so toxic higher in the high saturation group. At follow-up, infant mortality to the newborn infant? Data from Vento et al.22 have shown that and neuro-development at 18 months were not different between newborn infants in need of resuscitation already in cord blood have these regimes. In summary, neonates in the lowest range, when an increased oxidative stress when this is assessed by the ratio compared with the highest range, presented with significantly lower between oxidized to reduced glutathione (GSSG/GSH ratio) in

Table 2 A Summary of Prospective Clinical Trials Comparing Outcome Parameters in Infants Dichotomized into Higher or Lower Oxygen Saturation Groups (Adapted from Tin26)

Study Subject population SaO2 groups Survival CLD ROP 3-4 ROP therapy

Tin (2001) <27 weeks Low 70 to 90% 53% 18% 6%

High 88 to 98% 52% 46%z 27% yy Sun (2002) r1500 g Low r92% 83% 27% 10% 4% High >95% 76% 53%z 29%z 12%w Anderson (2004) r1500 g Low r92% 5.7% 1.4% >2 weeks High >92% 2.5%w 3.3%z Chow (2003) 500 to 1500 g Low 85 to 93% 88% 2.5% 0% to 1.3% High 90 to 98% 81% 12.5%* 4.4%w Askie (2003)y <30 weeks Standard 91 to 94% 97% 46% 16% 11% w r32 weeks High 95 to 98% 95% 64% 12% 6% *p<0.01. wp<0.001. zp<0.0001. yRandomization after 32 weeks. Survival after randomization. yySun, Pediatr Res 2002;51:350A.

Journal of Perinatology 2005; 25:S45–S49 S47 Saugstad Oxygen for Newborns

erythrocytes. At day 3, this ratio is still high both in infants References resuscitated with room air and 100% oxygen. However, at 28 days, 1. Nick Lane. Oxygen: The Molecule That Made the World. Oxford: Oxford the ratio has normalized in room air-exposed infants but remains University Press; 2003. high in those exposed to pure oxygen. A brief oxygen exposure 2. Saugstad OD, Aasen AO. Plasma hypoxanthine concentrations in pigs. A immediately after birth, therefore, seems to trigger long-term effects prognostic aid in hypoxia. Eur Surg Res 1980;12:123–9. that may influence cell growth and development. 3. Saugstad OD, Sanderud J. Circulatory effects of oxygen radicals. Biomed Studies in hypoxic newborn piglets by Temesvari et al.23 have Biochim Acta 1989;48:S20–4. 4. World Health Organization. Child Health and Development: Health of the shown a poorer short-term neurologic outcome in those Newborn. Geneva, Switzerland: World Health Organization; 1991. resuscitated with 100% oxygen compared with those resuscitated 5. Kattwinkel J, Niermeyer S, Nadkarni V, et al. An advisory statement from the with 21% oxygen. We have shown that resuscitation with 100% Pediatric Working Group of the International Liaison Committee on increases brain glycerol levels as well as metalloproteinases both in Resuscitation. Pediatrics 1999;103:e56. 24,25 the brain and the myocardium. These data indicate that 6. Wardman P, Candeias LP. Fenton chemistry: an introduction. Radiat Res inflammation and injury even in the brain are augmented if 1996;145:523–31. resuscitation is carried out with 100% oxygen. 7. Poulsen JP, Oyasaeter S, Sanderud J, Rognum TO, Saugstad OD. Hypoxanthine, xanthine, and uric acid concentrations in the cerebrospinal fluid, plasma, and urine of hypoxemic pigs. Pediatr Res 1990;28: 477–81. Conclusion 8. Almaas R, Sundar TB, Rootwelt T, Oyasaeter S, Saugstad OD. Plasma hypoxanthine reacts more abruptly to changes in oxygenation than base Compared to resuscitation with room air, newborn resuscitation deficit and uric acid in newborn piglets. J Perinat Med 1997;25:353–60. with 100% oxygen delays early recovery, increases neonatal 9. Rootwelt T, Odden JP, Hall C, Ganes T, Saugstad OD. Cerebral blood flow mortality, increases oxidative stress at least 4 weeks after birth, and, and evoked potentials during reoxygenation with 21 or 100% O2 in newborn in the animal model, increases brain injury and pigs. J Appl Physiol 1993;75:2054–60. neuroinflammation and affects the heart as well. As little as a few 10. Rootwelt T, Loberg EM, Moen A, Oyasaeter S, Saugstad OD. Hypoxemia and minutes of hypoxic exposure triggers long-term effects that are only reoxygenation with 21% or 100% oxygen in newborn pigs: changes in blood now becoming known. In the first weeks after birth, it seems that pressure, base deficit, and hypoxanthine and brain morphology. Pediatr Res an SaO2>93% increases lung complications and severe ROP 1992;32:107–13. although no effect so far has been found on mortality or 11. Saugstad OD. The role of oxygen in neonatal resuscitation. Clin Perinatol neurodevelopment. 2004;31:431–43. Minimizing exposure to high concentrations of oxygen 12. Kutzsche S, Ilves P, Kirkeby OJ, Saugstad OD. Hydrogen peroxide production reduces the risk of these complications. The neonatologist in leukocytes during cerebral hypoxia and reoxygenation with 100% or 21% would be well served to think of oxygen as a medication and use oxygen in newborn piglets. Pediatr Res 2001;49:834–42. 13. Kondo M, Itoh S, Isobe K, et al. Chemiluminescence because of the it sparingly. Oxygen appears to be more toxic than previously production of reactive oxygen species in the lungs of newborn piglets during believed. resuscitation periods after asphyxiation load. Pediatr Res 2000;47:524–7. At the time being, the optimal oxygen saturation is not known 14. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxiated newborn either during resuscitation or in the first weeks of life in extremely infants with room air or oxygen: an international controlled trial: the Resair low birth weight infants. Based on present data it is, however, 2 study. Pediatrics 1998;102:e1. evident that pure oxygen should be avoided for routine newborn 15. Ramji S, Rasaily R, Mishra PK, et al. Resuscitation of asphyxiated newborns resuscitation. This can very well be started with room air, adding with room air or 100% oxygen at birth: a multicentric clinical trial. Indian oxygen if there is a poor response. In our studies, we have typically Pediatr 2003;40:510–7. used oxygen as a back-up if the response (heart rate, color) was 16. Saugstad OD, Ramji S, Vento M. Resuscitation of depressed newborn infants not satisfactory within 90 seconds. Whether 100% oxygen should be with ambient air or pure oxygen: a meta-analysis. Biol Neonate then given could be questioned. The optimal probably would be to 2005;87:27–34. 17. Rao R, Ramji S. Pulse oximetry in asphyxiated newborns in the delivery achieve normal saturations according to the age. Studies with room. Indian Pediatr 2001;38:762–6. neurodevelopmental follow-up are needed; however, it probably 18. Tin W, Milligan DW, Pennefather P, Hey E. Pulse oximetry, severe would be unethical to include resuscitation with 100% oxygen in retinopathy, and outcome at one year in babies of less than 28 weeks one arm. gestation. Arch Dis Child Fetal Neonatal Ed 2001;84:F106–10. The optimal SaO2 is not known in extremely low birth weight 19. Anderson CG, Benitz WE, Madan A. Retinopathy of prematurity and pulse infants but present data indicate that it should be kept at 93% or oximetry: a national survey of recent practices. J Perinatol 2004;24:164–8. less. Many centers today, therefore, are aiming at saturations 20. Askie LM, Henderson-Smart DJ, Irwig L, Simpson JM. Oxygen-saturation between 88 and 92%. Fluctuations with peaks in SaO2 should be targets and outcomes in extremely preterm infants. N Engl J Med avoided. 2003;349:959–67.

S48 Journal of Perinatology 2005; 25:S45–S49 Oxygen for Newborns Saugstad

21. Chow LC, Wright KW, Sola A. Can changes in clinical practice decrease the 24. Munkeby BH, Borke WB, Bjornland K, et al. Resuscitation with 100%

incidence of severe retinopathy of prematurity in very low birth weight O2 increases cerebral injury in hypoxemic piglets. Pediatr Res 2004;56: infants? Pediatrics 2003;111:339–45. 783–90. 22. Vento M, Asensi M, Sastre J, Garcia-Sala F, Pallardo FV, Vina J. Resuscitation 25. Borke WB, Munkeby BH, Halvorsen B, et al. Increased myocardial with room air instead of 100% oxygen prevents oxidative stress in matrix metalloproteinases in hypoxic newborn pigs during resuscitation: moderately asphyxiated term neonates. Pediatrics 2001;107:642–7. effects of oxygen and carbon dioxide. Eur J Clin Invest 2004;34: 23. Temesvari P, Karg E, Bodi I, et al. Impaired early neurologic outcome in 459–66. newborn piglets reoxygenated with 100% oxygen compared with room air 26. Tin W. Optimal oxygen saturation for preterm babies. Biol Neonat after pneumothorax-induced asphyxia. Pediatr Res 2001;49:812–9. 2004;85:319–25.

Journal of Perinatology 2005; 25:S45–S49 S49