Background. Traditionally, asphyxiated newborn infants have been ventilated using 100% oxygen. However, a recent multinational trial has shown that the use of room air was just as efficient as pure oxygen in securing the survival of severely asphyxiated newborn infants. Oxidative stress markers in moderately asphyxiated term newborn infants resuscitated with either 100% oxygen or room air have been studied for the first time in this work.
Methods. Eligible term neonates with perinatal asphyxia were randomly resuscitated with either room air or 100% oxygen. The clinical parameters recorded were those of the Apgar score at 1, 5, and 10 minutes, the time of onset of the first cry, and the time of onset of the sustained pattern of respiration. In addition, reduced and oxidized glutathione concentrations and antioxidant enzyme activities (superoxide dismutase, catalase, and glutathione peroxidase) were determined in blood from the umbilical artery during delivery and in peripheral blood at 72 hours and at 4 weeks' postnatal age.
Results. Our results show that the room-air resuscitated (RAR) group needed significantly less time to first cry than the group resuscitated with 100% oxygen (1.2 ± 0.6 minutes vs 1.7 ± 0.5). Moreover, the RAR group needed less time undergoing ventilation to achieve a sustained respiratory pattern than the group resuscitated with pure oxygen (4.6 ± 0.7 vs 7.5 ± 1.8 minutes). The reduced-to-oxidized-glutathione ratio, which is an accurate index of oxidative stress, of the RAR group (53 ± 9) at 28 days of postnatal life showed no differences with the control nonasphyxiated group (50 ± 12). However, the reduced-to-oxidized-glutathione ratio of the 100% oxygen-resuscitated group (OxR) (15 ± 5) was significantly lower and revealed protracted oxidative stress. Furthermore, the activities of superoxide dismutase and catalase in erythrocytes were 69% and 78% higher, respectively, in the OxR group than in the control group at 28 days of postnatal life. Thus, this shows that these antioxidant enzymes, although higher than in controls, could not cope with the ongoing generation of free radicals in the OxR group. However, there were no differences in antioxidant enzyme activities between the RAR group and the control group at this stage.
Conclusions. There are no apparent clinical disadvantages in using room air for ventilation of asphyxiated neonates rather than 100% oxygen. Furthermore, RAR infants recover more quickly as assessed by Apgar scores, time to the first cry, and the sustained pattern of respiration. In addition, neonates resuscitated with 100% oxygen exhibit biochemical findings reflecting prolonged oxidative stress present even after 4 weeks of postnatal life, which do not appear in the RAR group. Thus, the current accepted recommendations for using 100% oxygen in the resuscitation of asphyxiated newborn infants should be further discussed and investigated.
- HIE =
- hypoxic ischemic encephalopathy •
- RAR =
- room-air resuscitated •
- OxR =
- 100% oxygen resuscitated •
- GSH =
- reduced glutathione •
- GSSG =
- oxidized glutathione •
- GPx =
- glutathione peroxidase •
- SOD =
- superoxide dismutase •
- CAT =
- catalase •
- Hb =
- hemoglobin •
- ROS =
- reactive oxygen species
The transition from fetal to neonatal life at birth implies acute and complex physiologic changes, which are not always successfully accomplished. In approximately 6% to 10% of deliveries, a number of adverse maternal and/or fetal conditions interfere with the infant's ability to make this adaptation adequately.1Hypoxic-ischemic encephalopathy (HIE), after-birth asphyxia, and the development of major neurodevelopmental handicaps such as cerebral palsy, mental retardation, and epilepsy, constitute the long-term sequelae of this pathologic event.2
Hypoxanthine, a purine metabolite accumulated during hypoxia, is oxidized in the presence of xanthine oxidase to uric acid during reoxygenation.3 This generates a burst of oxygen free radicals, rendering the natural defense systems incapable of their neutralization. As a result, damage to cellular structures is caused.4,5 We have shown that measuring the glutathione redox ratio can be an accurate procedure to monitor oxidative damage.6,7 Accordingly, we have described the oxidative stress accompanying the fetal-to-neonatal transition in animal experiments under physiologic conditions,8 and in human neonates breathing an oxygen-enriched atmosphere.9
Resuscitation of the asphyxiated newborn should be performed rapidly and in an orderly manner to ensure the best possible outcome. Traditionally, asphyxiated newborn infants have been ventilated with bag and mask using 100% oxygen as gas source at a flow rate of 6 to 8 L/min.10 However, in a recent multinational trial, the use of room air instead of 100% oxygen was shown to be just as efficient in securing the survival of severely asphyxiated newborn infants. Moreover, those infants resuscitated with room air seemed to recover more rapidly (as assessed by Apgar scores, time to first cry, and time required to establish a sustained respiratory pattern) than those receiving 100% oxygen.11 Recent reviews of perinatal resuscitation have questioned traditional recommendations and have started to consider the possibility of using room air under certain conditions.12–14
We hypothesize that the use of room air in the resuscitation of asphyxiated newborn infants might reduce the generation of oxygen free radicals and thus oxidative stress, allowing a more rapid and harmless recovery. Oxidative stress markers, as well as postnatal clinical parameters in moderately asphyxiated term newborn infants resuscitated with either 100% oxygen or room air, have been studied for the first time in this work.
This was a randomized, blind, monocentric clinical study performed at our hospital (Hospital Virgen del Consuelo, Valencia, Spain). Eligible patients were recruited for a 24-month period among term neonates (37–40 weeks' gestational age as deduced from serial ultrasound evaluations). Gestations were controlled in our Obstetric Outpatient Clinic and deliveries were conducted in the obstetric ward of our hospital. Attending obstetricians were informed of the ongoing clinical trial. At admission, parents were informed of the characteristics of the ongoing clinical trial and the previous published experience. Informed consent was required in all cases for participation in the study. Eligible neonates showed clinical and biochemical signs of asphyxia. Thus, they exhibited hypotonia and apnea, which were nonresponsive to external stimuli, pale skin and mucous color, and bradycardia (< 80 bpm). The biochemical signs of asphyxia were hypoxia (Po 2 <70 mm Hg), hypercarbia (Pco 2 >60 mm Hg), and acidosis (pH ≤ 7.15) in umbilical blood. The Apgar score at 1 minute in both groups ranged from 3 to 5. Gas sources were located at the wall and connected to an oxygen blender that was not visible for the resuscitation team. The flow was set at 6 L/min. The nurse in charge switched from 21% to 100% oxygen after an aleatoric number in a sequential manner, which corresponded to room air or 100% oxygen in each asphyxiated infant. Nurses provided the neonatologists with a bag and mask for resuscitation that were connected to the corresponding gas mixture (room air or 100% oxygen). Thus, the members of the resuscitating team were unaware of the type of gas they where using on each different infant. In addition, the maximal pressure ventilation, gas flow, and the ventilatory rate were limited (as described in “Clinical Proceedings”). Asphyxiated infants who were not blindly resuscitated were excluded from the study. None of the resuscitated infants had to switch from room air to 100% oxygen or vice versa because of failure in the process. As the study was monocentric, homogeneous obstetric and perinatal care were assured, thus avoiding external circumstances that could have biased the results.
The Ethical Committee of the Hospital Virgen del Consuelo (Valencia, Spain) approved the study protocol. Parents' written consent was obtained in each case when admitted to the obstetric ward before delivery.
From a total of 245 eligible patients, 40 newborn infants were enrolled in the study and placed in 2 experimental groups. Patients who did not fulfill all the requirements (ie, informed consent, aleatoric number assignment, complete blood testing, and follow-up clinical examination) were not admitted to the study. The RAR (room-air resuscitated) group consisted of 19 moderately asphyxiated neonates with a median (5–95 percentiles) Apgar score at 1 minute of 4 (2–6). The 100% oxygen resuscitated (OxR) group consisted of 21 moderately asphyxiated neonates with a median Apgar score at 1 minute of 4 (1–6) resuscitated with 100% oxygen. Controls consisted of 26 nonasphyxiated term neonates with a median Apgar score at 1 minute of 8 (7–9). All infants were born by vaginal delivery under epidural analgesia.
Immediately after birth, all infants were put under a radiant heating unit (Infant Resuscitator, Draeger Inc, Lübeck, Germany) and resuscitated following the usual procedures of our nursery.10 A specific gas mixture (room air or 100% oxygen) was assigned to each infant at random. To avoid differences between infants, the maximal gas flow was limited to 6 L/min, the inspiratory pressure was limited to 40 mbar, and the ventilation frequency was kept below 30 rpm. Once the initial resuscitating procedures at ∼1 minute of postnatal life had stabilized the patient, nurses placed the probes to monitor the infants' clinical parameters (electrocardiogram, temperature, respiration, and pulseoximetry).
Neonatal nurses determined the Apgar score and clinical parameters at 1-minute intervals beginning immediately after cord clamping. The clinical parameters recorded were as follows: 1) onset of the first cry, defined as the first audible cry spontaneously emitted by the infant; 2) onset of a regular respiratory pattern, defined as the establishment of a spontaneous and sustained respiratory pattern of efficacious respiratory movements, which allowed the neonate to maintain adequate clinical parameters (heart and respiratory rate), and hemoglobin saturation above 90%, thus needing no additional intervention from the resuscitation team.
Blood samples were drawn from the umbilical artery before detachment from the placenta and from a peripheral vein or arterialized capillary thereafter. Determinations were made at delivery, after 72 hours of postnatal life, and at 4 weeks' postnatal age.
Neurologic follow-up, including clinical evaluation, ultrasound, and electroencephalogram, was performed at 28 days of postnatal life in all cases.
Routine biochemical determinations were performed at the clinical laboratory of the hospital following the standard procedures.
Reduced glutathione (GSH) and oxidized glutathione (GSSG) were determined in whole blood as we have previously described.15,16
Glutathione peroxidase (GPx), superoxide dismutase (SOD), and catalase (CAT) activities were determined in erythrocytes as described by Flohé and Güzzler,17 Flohé and Otting,18 and Aebi,19 respectively. Blood was collected into heparinized tubes and then immediately centrifuged for 10 minutes at 500 g and 4°C. Plasma was removed and erythrocytes were washed twice with 0.9% NaCl. The supernatant was aspirated and the cell pellet was hemolyzed with distilled water. GPx, SOD, and CAT activities were assayed in the hemolysate. Enzymatic activities were expressed per gram of hemoglobin content in the hemolysate.
Statistical analysis has been made using nonparametric statistics because the data obtained did not have a normal distribution; thus we used the Mann-Whitney's test for nonpaired samples and the Kruskal-Wallis test for >2 nonpaired (independent) samples. Data have been analyzed using the GB-STAT computer program (Dynamics Microsystems, Inc, Silver Spring, MD).
Table 1 shows that no significant differences between the 3 groups were found for gestational age, birth weight, type of delivery, and analgesia received by the mother confirming the homogeneity of the population enrolled. Fetal bradycardia (<80 bpm), however, was significantly more frequent in both the asphyxiated groups (RAR and OxR).
Table 2 shows that the median of the Apgar score at 1 minute as well as pH and Po 2 were significantly lower in the OxR and RAR groups than in the control group, whereas Pco 2 was significantly higher in the OxR and RAR groups than in the control group. These results reflect perinatal asphyxia, acidosis, and hypoxia in the experimental groups. At 5 minutes of postnatal life, the median Apgar score in nonasphyxiated controls was 9 compared with 8 in the RAR group (not significant) and 7 in the OxR group (P < .05). No significant difference was found between the RAR and OxR groups. However, RAR infants had already reached an Apgar score similar to that of controls in 5 minutes. By contrast, in the OxR group this was not the case until 10 minutes of life.
Figure 1 depicts the time required for the onset of a regular respiratory pattern. As shown, the nonasphyxiated control group established a regular respiratory pattern significantly faster than the RAR and OxR neonates. Furthermore, differences between both experimental groups, as shown in Fig 1, are also statistically significant. Thus, the RAR group needed significantly less time of ventilation administered by the resuscitation team to achieve a sustained respiratory pattern than the OxR group did.
The OxR group reached a hemoglobin saturation of >90% sooner than the RAR group (see Table 2).
Table 3 shows that GSSG was significantly higher in the umbilical artery in the asphyxiated infants before resuscitation procedures were initiated. After 72 hours of postnatal life, as shown in Table 3, GSSG was still significantly higher in the RAR and OxR groups compared with the nonasphyxiated control group. Furthermore, GSSG was also significantly greater (P < .01) in the OxR group (94.2 μM) than in the RAR group (63.9 μM).Table 3 shows that after 28 days of postnatal life, GSSG in the RAR group and in the nonasphyxiated control group did not show significant differences. However, Table 3 shows that the GSSG concentration was much higher (P < .01) in OxR group (56.7 μM) than in nonasphyxiated controls (19.2 μM) and in the RAR group (18.8 μM).
Figure 2 depicts the evolution of the GSH/GSSG ratio throughout the study. It shows the significant differences between the experimental groups and the nonasphyxiated control group immediately before cord clamping (day 0) and at 3 days of postnatal life (day 3). The GSH/GSSG ratio of the RAR group at 28 days of postnatal life shows no differences compared with the nonasphyxiated control group. However, Fig 2 shows how the GSH/GSSG ratio of the OxR group is significantly different, not only from the nonasphyxiated control group, but also from the RAR group, revealing persistent oxidant stress in this group (Figure 2).
SOD, CAT, and GPx results throughout the study are shown in Table 3. In the first determination in the umbilical artery, a significant increase of SOD and CAT is evidenced in the asphyxiated infants, before initiation of resuscitation, when compared with the nonasphyxiated group. Table 3 shows there are significant differences at 72 hours of postnatal life for SOD, CAT, and GPx between the asphyxiated OxR and RAR groups and nonasphyxiated controls. However, Table 3 shows how the activities of the antioxidant enzymes SOD and CAT in the OxR group are remarkably higher than in the RAR group at 72 hours of postnatal life.Table 3 also shows that in the last determination, at 28 days of postnatal life, there were still higher SOD and CAT activities in the OxR group than in the nonasphyxiated control and the RAR groups, whereas there were no differences between the latter two.
No differences were found in the follow-up evaluation conducted at 28 days of postnatal life regarding the clinical and neurologic condition (including ultrasound and electroencephalogram) between control, RAR, and OxR groups.
In a previous international randomized clinical trial11 we showed that resuscitation with room air instead of 100% oxygen favors a prompter initiation of a sustained spontaneous pattern of respiration in the asphyxiated neonate. Our present study confirms this finding. The use of high oxygen concentrations during resuscitation as generally recommended10 may cause hyperoxia, accounting for various negative effects on the neonate, such as retarded initiation of spontaneous ventilation, increased oxygen consumption, or alterations in cerebral circulation.20–22Hyperoxia itself can affect primarily neonates' ventilation. Thus, research done on lambs20 and human neonates21has shown that ventilation for <1 minute with 100% oxygen leads to a decrease in minute ventilation when compared with RAR. In the Resair 2-trial11 and in our present study, we have found a delay in the initiation of the first cry (Table 2) and in the establishment of a sustained pattern of respiration (Figure 1) when using 100% oxygen rather than room air.
Breathing, which is intermittent in the fetus, becomes continuous after birth, although the mechanism responsible for this transition is not fully understood.23 Continuous breathing does commence if Pao 2 rises after delivery and/or there is a cord occlusion. This mechanism is independent of Paco 2.24 The inspiratory response in hyperoxic neonates with Pao 2 values of >100 mm Hg is impaired (Figure 1). This phenomenon could be explained if the infants become refractory to chemical stimuli on the peripheral chemoreceptors, hence delaying the onset of a spontaneous self-sustained pattern of respiration.
Perinatal HIE is the single most important problem in neonatal neurology, and some patients with HIE develop cerebral palsy and mental retardation.1,2 Several interrelated mechanisms–including excessive stimulation of excitatory amino acid receptors, intracellular calcium accumulation, and free radical generation–may play a role in the evolution of hypoxic-ischemic injury.25
There is a growing body of evidence supporting the hypothesis that inflammatory mediators play a role in hypoxic-ischemic injury to the immature brain. Several strategies to interrupt the inflammatory cascade have proved effective in decreasing the severity of neonatal hypoxic-ischemic brain injury in experimental animals. Nevertheless, these therapies have not yet reached clinical practice.26
During asphyxia, the primary pathophysiological insult to the neonatal brain is caused by a period of ischemia followed by reperfusion. There is an accumulation of the purine derivative hypoxanthine in the body tissues during the ischemic period. Following reperfusion and reoxygenation, hypoxanthine combines with oxygen in the presence of xanthine oxidase to generate large amounts of oxygen free radicals.12 In fact, premature infants with significantly higher hypoxanthine plasmatic concentrations on admission developed prematurity complications more frequently; this was thought to be mediated by oxygen free radicals (ie, periventricular leukomalacia, bronchopulmonary dysplasia, and retinopathy of prematurity).27 Moreover, a recently published study with asphyxiated newborn piglets showed that resuscitation with 100% oxygen produced significantly larger quantities of oxygen free radicals, as detected by chemiluminescence, when compared with resuscitation with room air.28
Oxygen free radicals are highly reactive chemical species with the potential of reacting with almost every type of molecule in living cells. In the presence of transition metals, especially iron, superoxide radical and hydrogen peroxide form the highly reactive hydroxyl radical. The hydroxyl radical is especially toxic, as it reacts with all biological substances, such as proteins, polysaccharides, nucleic acids, and polyunsaturated fatty acids, being susceptible to alteration of their structure and function. Polyunsaturated fatty acids, which are present in high concentrations in cell membranes, are most susceptible to lipid peroxidation.5 DNA damage not only includes oxidative damage, but also structural alterations such as strand-breaks, deletions, base changes, and even chromosomal aberration.29 Furthermore, reactive oxygen species (ROS) modulate signal transduction pathways that stimulate cell proliferation and an excess ROS production may cause cell death by apoptotic and necrotic mechanisms.30 ROS may also promote the expression of adhesion molecules, leading to the accumulation of activated granulocytes and the amplification of cell damage.31
We have endeavored to diminish the generation of oxygen free radicals by decreasing the oxygen concentration in the resuscitating gas. To evaluate the consequences of our strategy we have determined the GSH/GSSG ratio which accurately reflects intracellular redox status.32 An increase in the oxidized form of glutathione (GSSG) reflects a pro-oxidant situation. We have previously shown that this change is indicative of oxidative cell damage.6,7 The antioxidant response in this situation consists in the scavenging of the free radicals by intracellular or circulating antioxidant molecules and/or by the intervention of antioxidant enzymes.32 Our results indicate that neonates resuscitated with room air or 100% oxygen had an increased GSSG concentration in the umbilical artery and in peripheral blood after three days of postnatal life. Concomitantly, an increase in SOD and CAT reflected an antioxidant reaction to neutralize this pro-oxidant situation. However, after 4 weeks of postnatal life, neonates resuscitated with room air had similar GSH/GSSG ratio values as controls, while neonates resuscitated with 100% oxygen (OxR group) exhibited a lower GSH/GSSG ratio when not only compared with the control group but also with the RAR group. The accompanying increase in the SOD, CAT, and GPx activities reflects that these antioxidant enzymes, although higher than in controls, could not cope with the ongoing generation of free radicals in the OxR group. Furthermore, the induction of these enzyme activities might have been a response to the oxidative stress caused by resuscitation with pure oxygen.
Recent publications33–36 suggest that free radicals trigger an inflammatory process responsible for the long-term effects on the different tissues in which they are generated. Our findings suggest that hyperoxia during the first minutes of life may cause a prolonged oxidative stress reflected by a sustained decrease in the GSH/GSSG ratio. Arterial oxygen concentration returned to normal values after several minutes of postnatal life in both experimental groups (RAR and OxR). However, even after 4 weeks of postnatal life, the OxR group still exhibited indices of oxidative stress. This may be explained if hyperoxygenation in the first minutes of life triggers a protracted inflammatory process in the neonates' tissues. The RAR did not exhibit this pro-oxidant status after 72 hours of postnatal life, indicating that most probably an inflammatory process had been triggered in this group.
In previous studies, the suitability of room air for resuscitation of asphyctic neonates has been firmly established.11,37Furthermore, there have been no differences regarding mortality, morbidity, or quality of survival when compared with 100% oxygen resuscitated newborn infants.
We conclude that there are no apparent clinical disadvantages in using room air for artificial ventilation of asphyxiated newborn infants rather than 100% oxygen. Moreover, there may even be some advantages. Thus, RAR infants seem to recover more quickly as assessed by Apgar scores, time to first cry, and onset of a sustained pattern of respiration. In addition, neonates resuscitated with 100% oxygen exhibit biochemical findings reflecting prolonged oxidative stress present even after 4 weeks of postnatal life. Thus, the presently accepted recommendations of using 100% oxygen in the resuscitation of asphyxiated neonates10 should be further discussed and investigated.
This work was supported by funds from the Fondo de Investigaciones Sanitarias de la Seguridad Social Number 98/1462 to J.V., the Comisión Interministerial de Ciencia y Tecnologı́a Number SAF 97–0015 to J.S., and the Annual Research Grant (1998) of the Sociedad Española de Neonatologı́a to M.V.
We thank Marilyn R. Noyes for revising the manuscript.
- Received March 28, 2000.
- Accepted August 9, 2000.
- Address correspondence to Máximo Vento, MD, PhD, Jefe de Servicio, Hospital Virgen del Consuelo, Callosa de Ensarriá, 12, E-46007 Valencia, Spain. E-mail:
- World Health Organization. Child Health and Development: Health of the Newborn. Geneva, Switzerland: World Health Organization; 1991
- Volpe J. Hypoxic-ischemic encephalopathy: clinical aspects. In: Volpe J, ed. Neurology of the Newborn. 3rd ed. Philadelphia, PA: WB Saunders; 1995:279,344
- Vento M,
- Asensi M,
- Garcı́a-Sala F,
- Catalá J,
- Viña J
- American Academy of Pediatrics, American Heart Association.Textbook of Neonatal Resuscitation. Dallas, TX: American Heart Association; 1994
- Saugstad OD, Rootwelt T, Aalen O, and the RESAIR 2 Group. Resuscitation of asphyxiated newborn infants with room air or oxygen: an international controlled trial. The Resair 2 Study.Pediatrics. 1998;102(1). URL:http://www.pediatrics.org/cgi/content/full/102/1/e1
- Vento M,
- Garcı́a-Sala F,
- Saugstad OD,
- Stopfkuchen H
- World Health Organization. Basic Newborn Resuscitation: A Practical Guide. Geneva, Switzerland: World Health Organization; 1998
- Mortola JP,
- Frapell PB,
- Dotta A,
- et al.
- Halliwell B,
- Aruoma OI
- Wiseman H,
- Halliwell B
- Copyright © 2001 American Academy of Pediatrics