Chinese Medical Journal 2010;123(20):2938-2942
Oxygen administration in the care of neonates: a double-edged

Phyllis A. Dennery

Phyllis A. Dennery (University of Pennsylvania School of Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA 19072, USA)

Correspondence to:Phyllis A. Dennery,University of Pennsylvania School of Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA 19072, USA (Tel: 215-590-1653. Fax:267-426-5632. E-mail:dennery@ email.chop.edu)
Keywords
supplemental oxygen; prematurity; retinopathy of prematurity; bronchopulmonary dysplasia
Abstract

Objective  To evaluate the use of oxygen in neonates in the past, present and future       

Data sources  The data are mainly from Pubmed with relevant published articles from the 1940s to the present with some information gathered from web searches.

Study selection  Studies evaluating the use of oxygen in premature and term infants through history with original milestone articles included.

Results  There are still many unknowns about the proper use of oxygen in preterm and term infants but many studies suggest that both liberal use (resulting in a blood oxygen saturation of greater than 94%) as well as restrictive use (resulting in a blood oxygen saturation of less than 80%–85%) are detrimental and have long term consequences on the infant. 

Conclusions  Definitive studies evaluating the appropriate concentration and duration of supplemental oxygen are ongoing and will help in the management of term and preterm infants requiring oxygen.

This review will serve to describe the history of oxygen use in newborns, the benefits and toxicity of oxygen, as well as focus on clinical uses and misuses as well as, management strategies to prevent oxygen related injuries in newborns.

HISTORICAL BACKGROUND

In 1604 the Polish alchemist, Michal Sedziwoj, (Sendivogious) discovered oxygen and stated, “this is the elixir of life which no mortal can live without”. He did not publish his findings and the discovery did not receive much fanfare. Therefore, oxygen had to be “rediscovered”, as it was by Scheele in 1772, followed by Priestly in 1774.1,2 Because Karl Wilhelm Scheele, never published his observations, the credit for the discovery ultimately went to Priestly who wrote a seminal paper on “dephlogisticated air” in 1774.1 In the late 1700s Antoine Lavoisier, renamed this compound oxygène.2 In 1780, another Frenchman, Francois Chaussier, used oxygen for the revival of “near-dead” infants,3 but it was not until the 1930’s that oxygen began to be used liberally in neonates. Doctor J.H. Hess described an oxygen unit for premature and very young infants in the 1930s.4 This was followed by another publication by C.C. Chapple in 1938, on an incubator for infants.5 In this unit, 100% oxygen gas was delivered at 4 liters per minute, which gave an FIO2 of approximately 46% in the isolette.

The adverse consequences of this high oxygen delivery only became clear several years later. Between 1942 and 1945, Dr. Terry collected 117 examples of a novel type of blindness in infants born premature, characterized by a thick fibrotic membrane in the retrolental space, whose etiology was not yet elucidated.6 It was not until 1951 that Dr. Campbell first suspected a role for supplemental oxygen in the etiology of the new blindness entitled retrolental fibroplasia.7 Of note, one of the most famous victims of this condition is the American musical prodigy Stevie Wonder, who performs to this day despite his blindness.8 To confirm the role of supplemental oxygen in the etiology of this new disease, in 1952, Dr. Patz9 performed an NIH-funded randomized trial of 78 infants who were either subjected to high (100% oxygen to the isolette) or restricted oxygen (supplemental oxygen only when showing evidence of desaturation). In this study where 11 infants were excluded, of the 29 exposed to high oxygen, 17 (61%) developed retrolental fibroplasia vs. 6 out of 37 (16%) in the restricted oxygen group. This strongly suggested that oxygen was the culprit in the disease. A prospective study including 719 premature infants10 revealed that the most significant risk factors for the development of retrolental fibroplasia were birth weight <1200 g and length of exposure to supplemental oxygen. The concentration of oxygen, while a risk factor, did not correlate as much. Many of the studies cited were hampered by the lack of availability of accurate and continuous oxygen saturation monitoring.

Early clinical studies had also suggested that restricted oxygen use resulted in fewer deaths, decreased respiratory failure and decreased lung disease compared to liberal oxygen use.11 The use of supplemental oxygen for most premature infants irrespective of their clinical need in the 1940s and 1950s, combined with the newfound awareness of the role of prolonged supplemental oxygen use in the etiology of retrolental fibroplasia led to practitioners allowing babies to be severely desaturated and, unfortunately, an increased incidence of cerebral palsy. In fact, in a multicenter study of 1080 infants weighing <1800 g, a median duration of supplemental oxygen exposure of <2 days was associated with a 17.4% incidence of cerebral palsy and an 8.7% incidence of retrolental fibroplasia whereas a median supplemental oxygen exposure of >10 days was associated with a 5.8% incidence of spastic diplegia but a 21.7 % incidence of retrolental fibroplasia, illustrating the delicate balance between too little and too much supplemental oxygen exposure in premature infants.11 Overall, these early studies suggested that oxygen is a necessary evil that must be used in moderation.

With better tools to monitor oxygen saturation in the blood, investigators verified that in infants weighing between 500 g and 1300 g, the longer the hours where the transcutaneous oxygen concentration was >80 mmHg, the greater the incidence and severity of retinopathy of prematurity, especially when this occurred in the 2nd to 4th week of life.12 Other studies began to focus on targeting oxygen saturations to a particular level in preterm infants. The STOP-ROP study demonstrated that supplemental oxygen to keep saturations in the 94%–98% range did not prevent the progression of pre-threshold disease but increased the risk of adverse pulmonary events including pneumonia, chronic lung disease, diuretics, and hospitalization at 3 months of corrected age.13 Askie et al14 in a multicenter trial, confirmed that targeting oxygen saturations between 95%–98% did not confer any benefits in terms of growth and development of premature infants as compared to targeting oxygen saturations to a lower range of 91%–94%. Hua et al15 analyzed the risk factors for ROP in Chinese infants and confirmed that the lower the birth weight, the longer the duration of exposure and the higher the concentration of oxygen, the higher the incidence of ROP. In this analysis, infants exposed to supplemental oxygen for >8 days and in infants who inhaled >80% oxygen, the incidence of ROP was 38.8% and 52.6% respectively.15 Most recently, a study defined the ranges of oxygen saturation in normal infants in the delivery room and concluded that at 1, 2 and 5 minutes of life, 50% of infants have oxygen saturations of only 66%, 73% and 89% respectively. It took a median of almost 8 minutes to reach saturations above 90%.16 This clearly suggests that normal infants take some time to transition and that, perhaps we are too aggressive with delivering supplemental oxygen to these infants without proof that this is beneficial.

MOLECULAR MECHANISMS OF OXYGEN TOXICITY

Oxygen is an important molecule in the production of ATP. It serves as the final acceptor for electrons passed along the hydrogen ion gradient of the respiratory chain in the inner mitochondrial membrane. Although this process is necessary for cell survival, an excess amount of oxygen results in the accumulation of reactive oxygen species (ROS), which are generated proportionally to the increasing amount of oxygen in mitochondria. With the generation of ROS, increased gene expression is observed through the up-regulation of various transcription factors including nuclear factor kappa B (NF-kB),17,18 STAT,19 activator protein (AP)-1,20 p5321 and hypoxia inducible factor (HIF)-1.22 This results in increased expression of proteins involved in cellular proliferation as well as cell death.23 Because of their capacity to influence both cell proliferation and cell death, ROS may have beneficial or harmful effects depending on cell type, tissue and susceptibility.

Hyperoxia can result in the formation of lipid peroxides, ROS resulting in mitochondrial DNA damage with mitochondrial dysfunction, loss of sodium/potassium ATPase function. Conversely lipid peroxides can attack the inner membrane of the mitochondria leading to cytochrome C release and apoptosis. Thereafter, ATP hydrolysis, necrosis, cellular dysfunction and ultimately death occur.24

Although excess oxygen can be toxic, too little oxygen is also a problem. In hypoxia there is activation of xanthines oxidase and ATP oxidase, which leads to increased ROS. Furthermore, hypoxia leads to the activation of HIF-1α, which results in increased activation of iNOS and increased production of nitric oxide. Lastly, hypoxia can also result in energy depletion and down-regulation of a variety of protective enzymes including superoxide dismutase, catalase, glutathione peroxidase, NADPH oxidase and sodium potassium ATPase.25

ROLE OF OXYGEN DURING DEVELOPMENT

In embryonic development, ROS can be beneficial. For example during sperm capacitation, the superoxide anion and hydrogen peroxide are required to allow the sperm to enter the zona pellucida of the ovum.26,27 Interestingly, excess ROS can lead to decreased fertility but lack of ROS can also lead the same.28 Later in development, during the formation of the early blastocyst, hydrogen peroxide is needed to allow for apoptosis of the inner cell mass and the formation of the central cavity of the embryo required for gastrulation.29 Increased concentrations of ROS are also needed to guide apopototic events required for digit individualization of developing limbs.30

There is considerable speculation that lack of ROS due to increased expression of superoxide dismutase results in the phenotypic characteristics of trisomy 21 or Down syndrome.31 These individuals have an additional copy of the Cu, Zn superoxide dismutase, which should provide more protection against oxidative stress from the superoxide radical. Paradoxically, they have abnormal brain development, accelerated aging, abnormal pulmonary vessels and altered immune function.31 Laboratory investigations substantiate that decreased expression of Cu, Zn SOD is associated with increased DNA damage, aberrant mitochondrial protein function and apoptosis.32

DISEASES OF OXYGEN IN THE NEONATE

In the normal state, oxygen radicals and antioxidants defenses are in balance. When oxygen radicals overwhelm the cell, there is induction of antioxidants defenses resulting in tolerance to hyperoxia. In preparation for birth, the expression of various antioxidants enzymes is up-regulated in the latter third of gestation.33 This allows for newborns accustomed to the relatively hypoxic environment in utero to adapt to the relatively hyperoxic environment outside the womb. These antioxidant defenses are compromised in premature infants where interruption of normal development occurs. Incomplete retinal vascularization, immature lungs and an immature brain in conjunction with a lack of placental transfer of antioxidants (which usually occurs in the third trimester), inadequate nutrition, and increased exposure to oxidative stress create the perfect circumstances for toxicity.34

With exposure to excess oxygen, the developing retinal endothelial cells activate various transcription factors including HIF-1α and vascular endothelial growth factor (VEGF). This leads to abnormal retinal vascular proliferation and the formation of a ridge, which places traction on the retina and increases the risk of detachment, as seen in retinopathy of prematurity.35

The lung is directly exposed to delivered oxygen and therefore is at high risk for injury. In the initiation phase of hyperoxia mediated lung injury, oxygen exposure leads to the inflammatory phase where polymorphonuclear cells and macrophages infiltrate the lung. This promotes interstitial edema. Thereafter, in the destructive phase, injury to endothelial cells occurs first followed by injury to Type I epithelial cells. In the late phase of oxygen toxicity, proliferation and trans-differentiation of Type II epithelial cells occurs so as to compensate for the loss of other cell types.36

The phenotype of neonatal chronic lung disease or bronchopulmonary dysplasia, can be replicated in the mouse model simply by exposing the animals to 80% oxygen for 28 days.37 The lung then becomes simplified with enlarged alveolar spaces and decreased vascular elements. In the newborn, hyperoxia, ischemia and inflammation all lead to the formation of ROS, which then cause increased apoptosis, abnormal fibroblast proliferation, vascular smooth muscle hypertrophy as well as a simplified cuboidal epithelium.37 The result is an underdeveloped lung with lower alveolar numbers and decreased ability to diffuse oxygen due to the reduced proximity of the alveolar and vascular elements. This has consequences for the affected neonate into adulthood.

Hyperoxia can also interfere with control of breathing. When newborn rats were resuscitated with 100% oxygen, this delayed the onset of diaphragmatic excursions required for spontaneous breathing.38 The pups had significantly increased lag-time to onset of spontaneous diaphragmatic movements as measured by EMG, as compared to rats exposed to 21% oxygen. Interestingly, when the animals were exposed to 40% oxygen, they showed a similar increase in lag-time as did the animals exposed to 100%.38 This strongly suggests that supraphysiologic oxygen at any concentration suppresses diaphragmatic function.

In an experiment using endothelial colony forming cells, exposure to hyperoxia resulted in the suppression of endothelial nitric oxide synthase (eNOS) as well as decreased expression of the VEGF receptor.39 This would favor a vasoreactive state. In a neonatal lamb model of persistent pulmonary hypertension (PPHN) obtained by ductal ligation at 126–128 days of gestation, despite the presence of 20 ppm NO, exposure to hyperoxia increased pulmonary vascular resistance compared to animals exposed to room air,40 again suggesting that hyperoxia may lead to enhanced vasoreactivity and worsened PPHN.

Many have evaluated the use of oxygen in the resuscitation of newborns in the delivery room. When 100% oxygen was used to resuscitate newborns with hypoxic ischemic encephalopathy, mean arterial blood pressure and increased cerebral microcirculation returned to baseline more rapidly compared to room air use.41 However, in a multicenter randomized control trial of 609 infants, there were no differences in neonatal death or HIE nor were there differences in heart rate, Apgar scores, duration of resuscitation, time to breath or cry, arterial blood gas and neurologic exam at 4 weeks in infants resuscitated in room air vs. those resuscitated in 100% oxygen. Furthermore there were no differences at 18–20 months as to the development of cerebral palsy, attainment of milestones, hearing and language development.42 Therefore one would suggest that hyperoxic resuscitation is not any more harmful than resuscitation with room air. However, studies do not yet exist to help us understand the long-term sequelae of this intervention. In fact, few of the studies looking at outcomes with resuscitation with 100% oxygen versus room air were blinded. These were mostly focused on short-term outcomes in the first weeks or months of life. Additionally, larger studies did not really show differences in HIE, death or other morbidities between the groups. Several ongoing trials are investigating the effects of restricted oxygen administration with neonatal resuscitation. The recently published SUPPORT trial showed no differences in the rates of BPD between infants randomly assigned to intubation and surfactant treatment within 1 hour after birth compared to infants where CPAP treatment was initiated in the delivery room.43 Other trials such as the COT, BOOST 2, and POST-ROP trials are not yet completed. Result of these investigations will hopefully guide us as to oxygen use for resuscitation of neonates. Of note, the Neonatal Resuscitation Program of the American Academy of Pediatrics is currently revising its guidelines for resuscitation of newborns but still recommends oxygen use in the delivery room.

LONG-TERM CONSEQUENCES OF NEONATAL EXPOSURE TO HYPEROXIA

The association between supplemental oxygen exposure and childhood cancer was evaluated in a group of 27 900 infants.44 With a cumulative oxygen exposure of 0–2 minutes, 2 cases of neonatal cancer were found. When the cumulative oxygen exposure was more than 3 minutes, 11 out 370 000 children developed childhood cancers with a health risk (HR) of 2.87. In the 3 383 000 children never exposed to hyperoxia, the cases of childhood cancer were 35 with an HR of 1.45 These provocative data suggest that there may be an association between minimal exposure to hyperoxia in the delivery room and childhood cancer. This needs to be investigated prospectively with attention to confounding factors.

CONCLUSIONS AND MANAGEMENT GUIDELINES

Use of supplemental oxygen in neonates is not without risk. This could lead to altered cell proliferation, altered gene expression, increased inflammation, decreased cellular function and increased cell death. Judicious use of oxygen within the correct clinical scenario is important. Underuse (not providing oxygen despite a saturation below 80%–85%) could result in a resurgence of cerebral palsy whereas overuse (supplemental oxygen despite a saturation of greater than 90%) could lead to ROS-mediated diseases such as ROP and BPD. In fact, the increased incidence of blindness attributed to ROP in emerging middle income nations (Latin America, India, Vietnam, South Africa, Eastern Europe) is caused by preterm infants surviving at increasing rates and receiving supplemental oxygen for more than a few days,46-48 reminding us of our past mistakes. Despite many advances, we do not know the exact level and duration of oxygen that would be safe in neonates and whether this level differs with gestational age. Administration of supplemental oxygen in the delivery room should be reserved for documented cases of hypoxemia (saturation less than 85% after 5 minutes of life). New guidelines from the American Academy of Pediatrics will be available soon to further define safe ranges for this population. In preterm infants, we must carefully monitor oxygen saturations and keep these at less than 94% to prevent adverse pulmonary consequences such as bronchopulmonary dysplasia and to prevent ROP. Nevertheless, we must prevent desaturations below 80%–85% to prevent adverse consequences to the brain such as cerebral palsy. Even with the most careful practice, direct exposure of the lungs and the retinas at vulnerable periods in development could result in adverse consequences. Hopefully, in the next years, the results of several well-conducted large multi-centered trials will help provide guidelines for the appropriate use of oxygen in neonates. The words of Joseph Priestly ring even more true today: “The air which nature has provided for us is as good as we deserve.”

REFERENCES

1. Severinghaus JW. Fire-air and dephlogistication. Revisionisms of oxygen’s discovery. Adv Exp Med Biol 2003; 543: 7-19.

2. Sternbach GL, Varon J. The discovery and rediscovery of oxygen. J Emerg Med 2005; 28: 221-224.

3. Stofft H. New born’s apparent death (1781–1806) through Francois Chaussier’s Work. Hist Sci Med 1997; 31: 341-349.

4. Raju TN. The Nobel chronicles. 1949: Walter Rudolf Hess (1881–1973); and Antonio Egas Moniz (1874–1955). Lancet 1999; 353: 1281.

5. Robertson AF, Baker JP. Lessons from the past. Semin Fetal Neonatal Med 2005; 10: 23-30.

6. Terry TL. Retrolental fibroplasia in the premature infant: V. Further studies on fibroplastic overgrowth of the persistent tunica vasculosa lentis. Trans Am Ophthalmol Soc 1944; 42: 383-396.

7. Campbell K. Intensive oxygen therapy as a possible cause of retrolental fibroplasia: a clinical approach. Med J Aust 1951; 2: 48-50.

8. Noguera-Palau JJ. Stevie Wonder. Saginaw (USA), 1950. Arch Soc Esp Oftalmol 2008; 83: 281.

9. Patz A, Hoeck LE, De La Cruz E. Studies on the effect of high oxygen administration in retrolental fibroplasia. I. Nursery observations. Am J Ophthalmol 1952; 35: 1248-1253.

10. Kinsey VE, Arnold HJ, Kalina RE, Stern L, Stahlman M, Odell G, et al. PaO2 levels and retrolental fibroplasia: a report of the cooperative study. Pediatrics 1977; 60: 655-668.

11. Usher RH. Clinical investigation of the respiratory distress syndrome of prematurity. Interim report. N Y State J Med 1961; 61: 1677-1696.

12. Flynn JT, Bancalari E, Snyder ES, Goldberg RN, Feuer W, Cassady J, et al. A cohort study of transcutaneous oxygen tension and the incidence and severity of retinopathy of prematurity. N Engl J Med 1992; 326: 1050-1054.

13. Supplemental Therapeutic Oxygen for Prethreshold Retinopathy Of Prematurity (STOP-ROP), a randomized, controlled trial. I: primary outcomes. Pediatrics 2000; 105: 295-310.

14. Askie LM, Henderson-Smart DJ, Irwig L, Simpson JM. Oxygen-saturation targets and outcomes in extremely preterm infants. N Engl J Med 2003; 349: 959-967.

15. Hua SD, Chen YQ, Dong JY, Kong XY, Feng ZC. Screening and risk factors analysis of retinopathy of prematurity. Chin J Pediatr (Chin) 2009; 47: 757-761.

16. Dawson JA, Kamlin CO, Vento M, Wong C, Cole TJ, Donath SM, et al. Defining the reference range for oxygen saturation for infants after birth. Pediatrics 2010; 125: e1340-e1347.

17. Wright CJ, Zhuang T, La P, Yang G, Dennery PA. Hyperoxia-induced NF-kappaB activation occurs via a maturationally sensitive atypical pathway. Am J Physiol Lung Cell Mol Physiol 2009; 296: L296-L306.

18. Yang G, Abate A, George AG, Weng YH, Dennery PA. Maturational differences in lung NF-kappaB activation and their role in tolerance to hyperoxia. J Clin Invest 2004; 114: 669-678.

19. Lee PJ, Camhi SL, Chin BY, Alam J, Choi AM. AP-1 and STAT mediate hyperoxia-induced gene transcription of heme oxygenase-1. Am J Physiol Lung Cell Mol Physiol 2000; 279: L175-L182.

20. Yang G, Madan A, Dennery PA. Maturational differences in hyperoxic AP-1 activation in rat lung. Am J Physiol Lung Cell Mol Physiol 2000; 278: L393-L938.

21. Maniscalco WM, Watkins RH, Roper JM, Staversky R, O’Reilly MA. Hyperoxic ventilated premature baboons have increased p53, oxidant DNA damage and decreased VEGF expression. Pediatr Res 2005; 58: 549-556.

22. Michiels C, Minet E, Mottet D, Raes M. Regulation of gene expression by oxygen: NF-kappaB and HIF-1, two extremes. Free Radic Biol Med 2002; 33: 1231-1342.

23. Wright CJ, Dennery PA. Manipulation of gene expression by oxygen: a primer from bedside to bench. Pediatr Res 2009; 66: 3-10.

24. Chandel NS, Budinger GR. The cellular basis for diverse responses to oxygen. Free Radic Biol Med 2007; 42: 165-74.

25. Bunn HF, Poyton RO. Oxygen sensing and molecular adaptation to hypoxia. Physiol Rev 1996; 76: 839-85.

26. de Lamirande E, Gagnon C. Reactive oxygen species (ROS) and reproduction. Adv Exp Med Biol 1994; 366: 185-197.

27. Bize I, Santander G, Cabello P, Driscoll D, Sharpe C. Hydrogen peroxide is involved in hamster sperm capacitation in vitro. Biol Reprod 1991; 44: 398-403.

28. de Lamirande E, Gagnon C. Impact of reactive oxygen species on spermatozoa: a balancing act between beneficial and detrimental effects. Hum Reprod 1995; 10 Suppl 1: 15-21.

29. Pierce GB, Parchment RE, Lewellyn AL. Hydrogen peroxide as a mediator of programmed cell death in the blastocyst. Differentiation 1991; 46: 181-186.

30. Salas-Vidal E, Lomeli H, Castro-Obregon S, Cuervo R, Escalante-Alcalde D, Covarrubias L. Reactive oxygen species participate in the control of mouse embryonic cell death. Exp Cell Res 1998; 238: 136-147.

31. Groner Y, Elroy-Stein O, Avraham KB, Schickler M, Knobler H, Minc-Golomb D, et al. Cell damage by excess CuZnSOD and Down’s syndrome. Biomed Pharmacother 1994; 48: 231-240.

32. Shin JH, London J, Le Pecheur M, Hoger H, Pollak D, Lubec G. Aberrant neuronal and mitochondrial proteins in hippocampus of transgenic mice overexpressing human Cu/Zn superoxide dismutase 1. Free Radic Biol Med 2004; 37: 643-653.

33. Frank L. Developmental aspects of experimental pulmonary oxygen toxicity. Free Radic Biol Med 1991; 11: 463-494.

34. Dennery PA. Role of redox in fetal development and neonatal diseases. Antioxid Redox Signal 2004; 6: 147-153.

35. Chen J, Smith LE. Retinopathy of prematurity. Angiogenesis 2007; 10: 133-140.

36. Crapo JD. Morphologic changes in pulmonary oxygen toxicity. Annu Rev Physiol 1986; 48: 721-31.

37. Warner BB, Stuart LA, Papes RA, Wispe JR. Functional and pathological effects of prolonged hyperoxia in neonatal mice. Am J Physiol 1998; 275(1 Pt 1): L110-L117.

38. Martin RJ, Bookatz GB, Gelfand SL, Sastre J, Arduini A, Aguar M, et al. Consequences of neonatal resuscitation with supplemental oxygen. Semin Perinatol 2008; 32: 355-366.

39. Fujinaga H, Baker CD, Ryan SL, Markham NE, Seedorf GJ, Balasubramaniam V, et al. Hyperoxia disrupts vascular endothelial growth factor-nitric oxide signaling and decreases growth of endothelial colony-forming cells from preterm infants. Am J Physiol Lung Cell Mol Physiol 2009; 297: L1160-L1169.

40. Lakshminrusimha S, Swartz DD, Gugino SF, Ma CX, Wynn KA, Ryan RM, et al. Oxygen concentration and pulmonary hemodynamics in newborn lambs with pulmonary hypertension. Pediatr Res 2009; 66: 539-544.

41. Solas AB, Kalous P, Saugstad OD. Reoxygenation with 100 or 21% oxygen after cerebral hypoxemia-ischemia-hypercapnia in newborn piglets. Biol Neonate 2004; 85: 105-111.

42. Saugstad OD, Ramji S, Irani SF, El-Meneza S, Hernandez EA, Vento M, et al. Resuscitation of newborn infants with 21% or 100% oxygen: follow-up at 18 to 24 months. Pediatrics 2003; 112: 296-300.

43. Finer NN, Carlo WA, Walsh MC, Rich W, Gantz MG, Laptook AR, et al. Early CPAP versus surfactant in extremely preterm infants. N Engl J Med 2010; 362: 1970-1979.

44. Spector LG, Klebanoff MA, Feusner JH, Georgieff MK, Ross JA. Childhood cancer following neonatal oxygen supplementation. J Pediatr 2005; 147: 27-31.

45. Cnattingius S, Zack MM, Ekbom A, Gunnarskog J, Kreuger A, Linet M, et al. Prenatal and neonatal risk factors for childhood lymphatic leukemia. J Natl Cancer Inst 1995; 87: 908-914.

46. Sanchez ME, Andrews BJ, Karr D, Lansingh V, Winthrop KL. The emergence of retinopathy of prematurity in Guatemala. J Pediatr Ophthalmol Strabismus 2009; 8: 1-4.

47. Munoz B, West SK. Blindness and visual impairment in the Americas and the Caribbean. Br J Ophthalmol 2002; 86: 498-504.

48. Gilbert C, Rahi J, Eckstein M, O’Sullivan J, Foster A. Retinopathy of prematurity in middle-income countries. Lancet 1997; 350: 12-14.