J Clin Pathol: first published as 10.1136/jcp.s3-9.1.75 on 1 January 1975. Downloaded from J. clin. Path., 28, Suppl. (Roy. Coll. Path.), 9, 75-80 Drugs and the lung The effect of oxygen on the lung ALBERT E. CLAIREAUX From the Institute of Child Health, University ofLondon The harmful effect of oxygen in high concentration damage can be delayed by interrupted therapy on the lung has been recognized for many years. procedures. Smith (1899) observed the lethal results following Permanent lung damage, both in the human the exposure of experimental animals to oxygen at subject and experimental animal, takes the form of between 3 and 5 atmospheres pressure over a period gross congestion and fibrinous exudation into the of 24 hours. He considered that death resulted from alveoli. The changes are not dissimilar to those pulmonary inflammation. Karsner (1916) and resulting from oxygen toxicity at atmospheric pres- Binger, Faulkner, and Moore (1927) described sure but are of earlier onset. widespread changes in the lungs of animals exposed to high concentrations of oxygen for 48 hours or Oxygen at Atmospheric Pressure more. Much of the early work was concerned with Pulmonary lesions resulting from prolonged expo- animals but later necropsy studies showed that sure to high levels of oxygen tension are well docu- similar pulmonary lesions occurred in man (Pratt, mented and the term 'respirator lung syndrome' has 1958). It was evident that oxygen could exert a been coined by Nash, Blennerhassett, and Pontop- copyright. harmful effect either over a short period at high pidan (1967). These workers found consistent pressure or over a longer time at atmospheric pulmonary damage in patients who had received pressure. oxygen therapy and artificial ventilation. These patients showed a progressive deterioration of Hyperbaric Oxygen pulmonary function with reduced compliance and lowered vital capacity. They finally died from Oxygen administered at high pressures has been pulmonary insufficiency. The investigations of Nash used for a variety of clinical conditions such as and his colleagues revealed that the lung changes http://jcp.bmj.com/ respiratory failure, cyanotic heart disease, ischaemia, resulted from exposure to oxygen rather than to the cardiovascular surgery, shock, air embolism and coal effects of artificial ventilation. At necropsy the gas poisoning (Ledingham, 1969). Pressures about lungs from their patients were unduly heavy, 4 atmospheres are not used clinically because of the markedly congested and inelastic and of increased dangers of toxicity, which can cause rapid derange- consistency. The cut surface was fleshy in appearance. ment of the central nervous system within a matter The histological findings depended upon the duration of minutes if the oxygen overdose is unduly large or of the oxygen therapy. In the early stages the micro- on September 30, 2021 by guest. Protected the pressure unduly high-over 2 5 atmospheres. scopic changes were mainly exudative in character. Below 2-5 atmospheres the onset of toxic effects is The lungs were very congested and patchy intra- more insidious and the lungs are the main target. alveolar haemorrhage was seen. This was accom- Even at 2 atmospheres pressure, however, some panied by fibrinous exudation into the alveoli and deterioration of lung function may be observed hyaline membrane formation. In patients who died soon after oxygen is administered. The high arterial after more prolonged exposure to oxygen, the oxygen tension reached-about 1500 mmHg-gives changes were proliferative in character. The inter- rise to hyperventilation, possibly as a result of high alveolar septa were thickened and fibroblastic carbon dioxide tension in the brain. Cardiac output proliferation and collagen deposition were observed. falls and peripheral resistance rises. A reduction in A few lymphocytes were seen but no evidence of blood flow to every organ ensues. As far as the lungs infection was seen. At the same time diffuse hyper- are concerned, vital capacity, compliance and plasia of alveolar lining cells was noted, and was diffusing capacity are all reduced. These changes are occasionally so pronounced that a cuboidal epithelial reversible, at least initially, and the onset of lung layer was seen lining groups of alveoli. According 75 J Clin Pathol: first published as 10.1136/jcp.s3-9.1.75 on 1 January 1975. Downloaded from 76 Albert E. Claireaux to these observers the severity of the lesions was Although Banerjee et al found it difficult to distin- entirely dependent upon the concentration of oxygen guish between the effects of prolonged oxygen and the duration of the exposure to this form of therapy and prolonged artificial ventilation, they did therapy. point out that prolonged oxygen therapy at high The findings of Nash et al (1967) concerned the concentrations always resulted in pulmonary fibro- results of oxygen therapy in 70 adults. They were plasia whereas prolonged ventilation did not. similar to those reported by Cederberg, Hallsten, However, Barnes, Glover, Hull, and Milner (1969) and Miorner (1965). Almost precisely similar results have shown that infants who require artificial were noted in the case reported by Fuson, Saltzman, ventilation over a period of six months or more do Smith, Whalen, Osterhout, and Parker (1965) where have impaired pulmonary function even if the a girl of 17 years of age was treated by hyperbaric oxygen concentration is 50% or less. oxygen at a concentration of 100I% at pressures up The relationships between hyaline membrane to 3 atmospheres. Death occurred after five days disease and the exudative phase of oxygen toxicity and here again pulmonary congestion, intraalveolar is difficult to define in the newborn infant. Infants haemorrhage and hyaline membrane formation with respiratory distress and hyaline membrane were in evidence. The pulmonary lesions are thus formation are among those most likely to require almost identical whether hyperbaric oxygen or oxygen therapy in the neonatal period and the latter oxygen at normal atmospheric pressure is used. is likely to aggravate a lesion already present. In the These reports all referred to adult patients but study reported by Banerjee et al (1972) two infants recently there have been similar accounts of pul- with a gestational age of over 41 weeks had pulmon- monary lesions resulting from oxygen therapy in ary fibroplasia and well formed membranes. This infants and young children. Burrows and Edwards type of lesion is extremely rare in such infants not (1970) recorded seven patients whose ages ranged exposed to oxygen therapy. from a few hours to 3 years at the time of onset of Although most of the studies of pulmonary oxygen therapy. Six of the patients died. The lungs fibroplasia or ventilator lung in both adults and were heavy and solid. On microscopic examination infants have been concerned with necropsy material, hyaline membranes were found together with areas it must be assumed that some patients with similar copyright. of haemorrhage and thickening of alveolar walls. pulmonary lesions do survive but probably have Some alveolar epithelial cell proliferation was also permanently impaired pulmonary function. In seen. Although hyaline membrane formation is a addition it must be suspected that the anatomical common accompaniment of the respiratory distress lesions in infants' lungs will interfere with subsequent syndrome in the newborn infant, none of these lung growth, particularly that associated with children had suffered respiratory distress at the time alveolar development. It is also possible that some of birth and only the youngest patient had become air passages will remain blocked by inspissated ill during the first day of life. A larger study under- secretion (Northway, Rosen, and Porter, 1967) and http://jcp.bmj.com/ taken by Banerjee, Girling, and Wigglesworth this in turn could cause overdistension of those (1972) concerned 81 babies who had received oxygen alveoli still capable of some function and a centri- therapy and who had lived for at least 48 hours. lobular emphysema would develop (Rosan and Twenty-three of these infants showed pulmonary Lauweryns, 1970). lesions at necropsy. The extent of the pulmonary The pulmonary lesions which result from oxygen damage could be correlated with the degree of toxicity have been reproduced in a variety of oxygen concentration and the duration of the experimental animals (Binger et al, 1927)-in on September 30, 2021 by guest. Protected therapy. Pulmonary fibroplasia was present in all rabbits (Karsner, 1916) and more recently in the infants who had received over 60% oxygen for more monkey (Robinson, Harper, Thomas, and Kaplan, than 123 hours or over 80% oxygen for more than 1967). Gerschman, Gilbert, and Caccamise (1958) 105 hours. At necropsy the lungs of the affected exposed mice to oxygen at pressures up to 10 atmos- infants were bulky, and on histological examination pheres. The survival time decreased rapidly with showed squamous metaplasia of the bronchial increase in pressure and was only 19 minutes at 10 epithelium, fibroblastic proliferation and an increase atmospheres. Cedergren, Gyllensten, and Wersall in connective tissue and proliferation of alveolar (1959), also usingmice exposed to high concentrations lining cells. In addition hyaline membrane formation of oxygen, studied the effects produced on the lung was a prominent feature as was an increase in the ultrastructure. These included injuries to the musculature and elastic tissue in the