The clinicoradiologjcal symptoms of the early phase of the respiratory distress syndrome (RDS) and its anatomical substratum (hyaline membranes) have been known for many years. However, the existence of a close relationship between RDS and lung maturation and surfactant production is a more recent concept.
It is now generally accepted that during the disease there is a vicious circle- collapse of the lung, hypoxia, acidosis, diminished blood flow, damage of alveolar cells and capillary endothelia, facilitating transudation of fluids into alveoli. The primary cause of the disease and the relative part that each of these factors plays are still being questioned.
Within a short period of time after birth, the infant develops the pathognomonic clinical and radiological signs of RDS with low PO2 and acidosis. This development lasts from a few hours until five or six days, at which time the baby usually either recovers or dies.
If death occurs during the acute phase, the lungs at autopsy are reddish-purple, solid and liverlike in consistency, with no significant increase in weight. The characteristic histopathologic features affect all lobes of both lungs.
Early hyaline membrane formations, collapsed alveoli and overdistended air ducts.
Within a few hours after birth, an eosinophilic deposit of fibrin with necrotic debris of lung epithelium and serum protein line the inner surface of the few expanded alveoli, respiratory ducts and bronchioles. Most of the alveoli are collapsed, and the ducts proximal to the atelectasis are often overdistended (Fig. 1). Interstitial emphysema is common.
Both optic and electron microscopy have disclosed epithelial necrosis, disruption of the basement membrane and thickening of the pulmonary capillary endothelium, plus an increase in the reticular network.
Researchers have described active cellular response from about 24 hours on 220.127.116.11 Tne hyaline membranes are fragmented and engulfed by "membranophages;" other types of histiocytes and fatty macrophages are seen later (Fig. 2).
The resolution of the membrane is accompanied by an epithelial repair. Large epithelial cells, lining alveoli and terminal air ducts, are hypertrophic and hyperplasic with mitosis; these cells protrude into the lumen as rackets and are often detached into it (Fig. 3).
An identical cellular reaction and epithelial regeneration has been described after injection of various substances into the lungs of rabbits18 and after giving toxic levels of oxygen to monkeys.10,11
Hyaline membranes engulfed by "membranophages" and progressively resorbed.
Normal reparative phase: Type II epithelial cells, bulging or detached into the lumen.
These type II epithelial cells or granular pneumocytes have an extremely rapid turnover, and they may form a festoon around the air cavities; this new epithelium will soon be revascularized. According to Kaplan and Kapanci, during the regenerative phase, these cells make up 95 per cent of the alveolar epithelium in monkeys, as compared to 15 per cent in controls. On the other hand, the membranous pneumocytes are almost completely destroyed.
The granular pneumocytes are characterized by the densely osmiophilic lamellated granules and numerous organelles in the cytoplasm. These are the sites of surfactant production. During this period of repair in the lung, the appearance of osmiophilic granules and surfactant (detected on the Wilhelmy balance) has been demonstrated.8,9
In the cases that recover, clinical status and radiological appearance of both lung fields are back to normal within five to ten days of age. In the severe cases, death occurs within a few days.
Transition phase: Intense cellular proliferation occluding terminal air ways. Capillary network of the open alveoli is quite rich.
SEQUELAE OF HYALINE MEMBRANE DISEASE*
Mechanical ventilation has helped to cure and/or to prolong survival of many infants who would have died otherwise.
A new clinicoradiological and anatomical form of the disease was first noticed in 1964 by Shepard et al., as "long term residue of RDS/' after mechanical ventilation. On their long-term follow-up, 48 per cent of survivors had significant x-ray changes compatible with pulmonary fibrosis.
During the same year, Robertson et al. 17 brought to light the anatomical evolution of the lung lesions in three infants who died at 13, 21 and 24 days, respectively, and in the case of a lung biopsy performed on an eightweek-old infant. Because of the initial severity of the RDS, the Engström respirator had to be used on all of these cases.
Transition Phase: Coalescent alveoli and thickened septa; myxoid aspect due to very young fibrosis not yet stained by Masson's trichome.
Pseudoglandular parenchyma due to rapid granular pneumocyte proliferation; yet. poor capillary network.
Since men on a clinicoradiological and/or anatomical basis, this condition was described by Northway et al., who coined the name, bronchopulmonary dysplasia;14 by Pusey et al., who called it pulmonary fibroplasia;15 by Larroche and Nessman, calling it sequelae of HMD;13 and by other groups of researchers. 1,3,5,9,23 All these descriptions were of infants who required mechanical ventilation.
Schematically, there are three phases of the disease: transition, sequelae and complete recovery.
In our experience the transition form occurred in infants who died within a week. The lungs at autopsy are generally overexpanded, pink with patchy pale zones and, on microscopic sections, consolidated lobules are associated with emphysematous areas.
Microscopically, cellular reaction resembles that which is observed in the normal repair of HMD. However, its intensity is unusual. The parenchyma is dense; alveoli ducts and bronchioles are occluded by "membranophages," histiocytes, macrophages and abundant mucous secretions (Fig. 4). After a week, typical hyaline membranes have usually been resorbed.
During this transition phase, the collapsed alveoli and coalescent septa form a loose meshwork of a myxoid appearance (Fig. 5), when faintly stained with Masson's trichome. The necrosed alveolar epithelium is replaced by an amorphous, eosinophilic border. Elsewhere the festoon-like epithelium of regeneration seen in normal repair is forming; the parenchyma has a fetal, pseudoglandular appearance (Fig. 6). These zones, as well as the necrosed ones, have no capillary beds.
The lymphatic channels are enlarged, the bronchi may undergo MaJpighien metaplasia (Fig. 7) and the mucous glands are hyperplasic.
The transition phase appears within a week and will exist simultaneously later on with the sequelae phase, which occurs when fibrosis is evident.
At autopsy during the sequelae phase, the lungs are overexpanded, the surface mottled (Fig. 8) with retracted lobules>on microscopic sections firm and pale nodules coexist with emphysematous zones, and fibrosis is the most striking pathological change. Fibrosis lines the air spaces, thickens the septa or proceeds from necrosed and occluded bronchi to form scattered nodules.
Cartilage and/or satellite arteries are the sole landmark of the bronchial tree (Fig. 9). When bronchi are not necrosed, Malpighien metaplasia occurs, sometimes with a hypertrophic muscular ring. We have even observed very unusual strands of muscular fiber, which have no relationship to normal bronchovascular musculature. Mucous secretions and fibroblastic reaction are widespread and prominent.
In contrast with dense, hypercelhilar and fibrous areas, exist zones of overdistention with parenchymal and interstitial emphysema. Pseudoglandular neoepithelium undergoing revascularization may be seen in a few lobules.
Epithelial metaplasia of a bronchus. Note the thick muscular ring.
Phase of sequelae: Gross appearance of the iung. The surface is mottled due to fibrous retraction.
Two nodules of fibrosis, extending from necrosed and occluded bronchi. Small satellite arteries are visible. Note the pseudoglandular aspect of the parenchyma.
Recently, Bomsel et al.5 reviewed a series of 408 cases of HMD. Onehundred-and-thirty-nine infants were ventilated with an intermittent positive pressure respirator. Repeated chest films permitted us to follow the evolution of the disease.
In infants not treated by ventilation, the chest x-ray was normal within five to ten days of age. They represented cases which recovered.
In infants treated by ventilation, the initial reticulogranular pattern on xray turned into blurred nodular opacities between the fourth and seventh day of life. These opacities were scattered over both lung fields; they were located in different areas within the same day after physiotherapy, postural drainage and suctioning. In survivors after extubation, the opacities generally disappeared within three or four weeks of life. However, in 27 out of 209 survivors, the opacities persisted and chronic respiratory distress followed the acute episode. The nodular opacities were more distinct. Areas of hyperaeration were mixed with the opacities. The connective pulmonary network thickened and later on the radiolucencies of the lower lobes became confluent. The thorax was overexpanded, the ribs were thin and poorly calcified. These symptoms generally occurred around the third week of life. Then and only then was the diagnosis of HMD sequelae made while, on an anatomical basis, the diagnosis of HMD sequelae i.e., fibrosis, could be made already after a week of survival.
From the large series of 50 infants with HMD sequelae, 27 of whom survived, we tried to analyze several factors: 1) role of the initial disease and its severity, 2) degree of immaturity of the lung, 3) role of the respirator and O2 as iatrogenic factors and 4) associated visceral and/or brain damage.
Our conclusions can be summarized as follows:
1. There is an increased rate of sequelae from grade I to grade ?? (according to Bomsel's classification of 1970).
2. Hyaline membranes sequelae seem to occur more frequently under 34 weeks of gestational age.
3. Statistical analysis shows that the severity of the disease plays a greater role than prematurity.
4. The minimum duration of ventilation was five days, 11 hours. The minimal time with O2 above 80 per cent concentration was four days, four hours.
5. In survivors with HMD sequelae, the mean duration time with O2 above 80 per cent concentration was greater than in cases without sequelae. This difference was even more significant when 35 per cent or less concentration of O2 was utilized (22 days in survivors with fibrosis, versus eight days in normal survivors).
6. In 23 deceased infants, the severity of the lung disease versus length of survival was difficult to establish (all were ventilated until death with 80 per cent to 100 per cent O2). The associated visceral or cerebral pathology deserves a few comments. Thirteen died during the second week of life, six during the third and three during the fourth and fifth week.
Of these 23 cases, 20 had cerebral damage (hemorrhage and/or necrosis) with or without significant visceral pathology (lobar bronchopneumonia, lobar pulmonary hemorrhage and above all vascular thrombosis due to indwelling catheter).12
In addition we should mention the high frequency of abnormally large patent ductus arteriosus. The interior circumference has been compared to control cases previously published.7
Only two had neither significant brain nor visceral damage; we then assumed that the lung disease was the direct cause of death.
In survivors on a long-term followup, the clinical and radiological symptoms disappear between three and six months of life in the mild cases and between 18 months to two years in the severe ones.
In infants who have died within two months, the lung pathology has always confirmed the x-ray abnormalities. In survivors, from a few lung biopsies,20 a personal communication21 and one of our own cases who died at three-and-one-half months with a quite normal lung, it was possible to follow the parallellism of the radiological pattern improvements and the lung tissue healing.
Thus, in spite of the severity of the initial anatomical lesions of the lungs, it seems reasonable to conceive the possibility for an immature lung to undergo a complete repair.
The etiology and the mechanisms involved in these pulmonary sequelae of severe hyaline membrane disease remain debated. Does the mechanical ventilation permit the natural history of repair to take place in the damaged lung or do the iatrogenic factors (endotracheal tube, intermittent positive pressure, O2) recently introduced play an important role in the outcome of the disease? Although these questions are unanswered, there is little doubt that the subsequent clinicoradiological and anatomical appearance of the lung in HMD is related to:
1. The severity of the disease in young premature infants.
2. The high concentration of O2 as well as the length of exposure even at low concentrations.
3. The possible feeding aspiration and the mucosal and cellular obstruction of the small necrotic airways. Thus, repeated suctions along with early and careful physiotherapy should help prevent part of the sequelae.
A high proportion of infants (almost 50 per cent) die within the first two months; 90 per cent exhibit severe visceral and/or cerebral associated damage, probably the direct cause of death. In survivors a large number of total radiological and clinical recoveries are observed by the third year.
The respiratory distress syndrome remains the major cause of death of prematurely born infants even though the recent utilization of mechanical ventilation has indeed decreased the mortality and morbidity rate. Meanwhile a new form of the disease has been described on a clinicoradiological and anatomical basis.
The chronic clinical respiratory distress is accompanied by abnormal chest films. The fixity of the nodular opacities and the subsequent overexpanded lungs indicate the disease.
When death occurs, the severe pathological changes of the lungs and the fibrosis confirm the hyaline membrane sequelae. In most of these cases, severe visceral and/or brain lesions are associated and probably the main cause of death. In survivors improvement and complete recovery are observed.
1. Barnes, N.; Hull. D.; Glover. W. and Milner, A. Effects of prolonged positive pressure ventilation in infancy. Lancet 22 (1969). 1096.
2. Barter. R.; Byrne. M. and Carter, R. Pulmonary hyaline membrane: late results in injury to the lung lining. Arch. Dis. Child- 41 (1966), 489.
3. Becker-Bloemkolk, M. Changes in pulmonary structure in neonatal hyaline membrane disease treated with high pressure artificial respiration. 14th annual meeting, Pediatric Pathology Society, Rotterdam, October, 1969. Arch Dis. Child. 45 (1970), 145.
4. Bomsel. F. Contribution à l'étude de la maladie des membranes hyalines. Journal de Radiologie et cfElectrologie 51 (1970). 259.
5. Bomsel, F.; Couchard, M.; Pol je, J. and Larroche, J.-CI. Pulmonary sequelae of the hyaline membrane disease. Iatrogenic factors? Radiological and anatomical study. IX meeting of the European Society of Pediatric Radiology, May, 1972. Paris.
6. Boss. J. and Craig, J. Reparative phenomena in lungs of neonates with hyaline membranes. Pediatrics 29 (1962), 890.
7. DesKgneres, C. and Larroche, J.-CI. Ductus arteriosus. Biol. Neonate 16 (1970), 278.
8. Gandy. G.; Jacobson. N. and Gairdner. 0. Hyaline membrane disease. Cellular changes. Arch. Dis. Child. 45 (1970), 289.
9. Hawker, J.; Reynolds, E. and Taghizadeh, A. Pulmonary surface tension and pathological changes in infants dying after respirator treatment for severe HMD. Lancet 2 (1967), 75.
10. Kapanci. Y.; Weibel. E.; Kaplan. H. and Robinson, F. Pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. II. Ultrastructural and morphometric studies. Lab. Investig. 20 (1969), 101.
11. Kaplan, H.; Robinson. F.; Kapanci, Y. and Weibel. E. pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. I Clinical and light microscopic studies. Lab- Investig. 20 (1969). 94.
12. Larroche. J.-CI. Incidents et accidents au cours du cathétérisme ombilical, artériel et veineux. Etude anatomique. Arch. Franc. Ped. 26 (1969), 1065. Symposium on artificial ventilation. Paris, 1959. Bid. Neonate 16 (1970), 101.
13. Larroche, J.-CI. and Nessmann, C. La maladie des membranes hyalines. Evolution, Cicatrisation, Se.quelles. Journées de Réanimation, Nancy, France, April, 1970. Arch, Fr. Ped. 28 (1971), 113.
14. Northway, W.; Rosan, R. and Porter, D. Pulmonary disease following respirator therapy of hyaline membrane disease. Broncho-pulmonary dysplasia. New Engl. J. Med. 276 (1967), 357.
15. Pusey, V.; MacPherson, R. and Chernick. V. Pulmonary fibroplasia following prolonged artificial ventilation of newborn infants. Canad. Med. Assoc. J. 100 (1969). 457.
16. Robertson, B. Reparative phenomena in "hyaline membrane disease.'1 Acta Path. Microb. Scand. 58 (1963). 398.
17. Robertson, B.; Tunell. R. and Rudhe, V. Late stages of pulmonary hyaline membranes of the newborn. Acta Paediat. 53 (1964). 433.
18. Ross, I. Pulmonary epithelium and proliferative reactions in the lungs. Arch. PaOi. 27 (1939), 478.
19. Shepard. F.; Gray. J. and Stahiman. M. Occurence of pulmonary fibrosis in children who had idiopathic respiratory distress syndrome. Society for Pediatric Research J. Pediat. 65 (1964), 1078.
20. Shepard. F.; Johnston, R.; Klatte, E.; Burko. H. and Stahiman, M. Residual pulmonary findings in clinical hyaline membrane disease. New Engl. J. Med. 279 (1968), 1064.
21. Tran- Van-Due. Personal communication, 1971.
22. Wade-Evans. T. The formation of pulmonary hyaline membranes in the newborn baby. Arch. Dis. Child. 37 (1962), 470.
23. Weisser. K.; Ganz. R.; Oiatsson. A. and Wyler, F. Broncho-pulmonary dysplasia in the premature baby. Abstract, meeting of the European Club for Pediatric Research. Interlaken, Switzerland. Sept.. 1969.