Surfactant Deficiency in Hyaline Membrane Disease
The Story of Discovery

MARY ELLEN AVERY

Harvard Medical School, Boston, Massachusetts

	ARTICLE

TOP ARTICLE

This invitation to describe, in a personal way, my major scientific finding and how and why it took place is at once most tempting, but also anxiety producing. The anxiety arises from awareness of the frailty of human memory. The difficulty is in being objective about one's own impulses and actions, and the knowledge that a published work conforms to a style imposed by scientific tradition rather than the prose of a diarist. Thus, the true sequence of events may be difficult to resurrect. No one set of observations or experiments takes place in isolation, but rather follows slowly from a series of earlier findings, and is aided by new technologies. The setting in which we were able to elucidate the cause of atelectasis in hyaline membrane disease illustrates the interdependence of the investigator and the environment. The discovery of surfactant deficiency in lungs of infants who died of hyaline membrane disease (hence their description of these lungs at autopsy as "liver-like") and my own observation of the absence of foam in the airways after death, followed interactions with many individuals from different disciplines, and a publication by M. E. Avery and J. Mead, "Surface properties in relation to atelectasis and hyaline membrane disease" Am. J. Dis. Child 1959;97:517.

My interest in respiratory physiology dates from July 1952, when one month into an internship in pediatrics at Johns Hopkins, my routine tuberculin test was positive, and chest film showed a small right upper lobe infiltrate consistent with tuberculosis. Although I had no symptoms, and no organism was recovered, I was directed to Trudeau Sanitarium for bed rest for six months, along with streptomycin and PAS. I did not stay there, but started bed rest at my parents' home in Moorestown, New Jersey. During that time I corresponded with Richard Riley, my teacher at Johns Hopkins, about the rationale for bed rest; thus my quest for more knowledge of respiratory physiology began.

My interest in the respiratory difficulties newborn infants sometimes face soon after birth had been aroused by an awareness that the most common finding in the lungs of premature infants born alive, but who died in the first days of life, was atelectasis and hyaline membranes. The pathology had been well described by two individuals, both of whom had been my teachers at Johns HopkinsGeorge Anderson and Peter Gruenwald. Both emphasized the lack of clinical description of the course of the disease. In 1947, Gruenwald reported on the unusual expansion patterns of the lungs of premature infants. He assumed that an unusually high surface tension could account not only for the high pressure necessary to introduce air, but also for the fact that air accumulated in the lungs in a Swiss cheese pattern, as was predictable by the law of LaPlace.

Meanwhile, in England, Richard Pattle was studying the foam of pulmonary edema in the Chemical Defense Establishment in Porton, since some "war" gases such as phosgene induce pulmonary edema, and antifoam agents were being sought. The unusual stability of bubbles expressed from normal lung led him to conclude that they must have originated from alveoli, and the alveoli in turn must be covered with a lining layer of very low surface tension. He suggested that absence of the lining substance of the alveoli may sometimes be one of the difficulties with which a premature baby has to contend. He further noted that "such a defect may possibly play a part in causing some cases of atelectasis. The appearance of a hyaline membrane might possibly be due either to a defective lining layer causing transudation from the blood or to excessive secretion of lining substance." This prediction was largely ignored by clinicians.

In the 1950s, I had completed a pediatric residency at Johns Hopkins, then in 1957 moved to Boston to study respiratory physiology at the Harvard School of Public Health with Jere Mead, and to learn more of the newborn infant under the tutelage of Clement Smith at the Boston Lying-In Hospital. I had consulted with Riley about undertaking a fellowship in lung mechanics, as premature infants with hyaline membrane disease had such difficulty inflating their lungs and maintaining air at end-expiration. Riley said Jere Mead was the most knowledgeable person. So I applied for a fellowship in Mead's laboratory in the Department of Physiology under Dr. James Whittenberger.

One of the interesting observations made in Mead's laboratory was that excised lungs inflated with air had a greater elastic recoil than those inflated to the same volume with saline. The conclusion was that surface forces, which are greater at an air-liquid interface than at a liquid-liquid interface, contributed to elastic recoil, especially at large lung volumes. (It was not realized by these investigators until later that von Neergaard had made the same discovery in 1927.) Edward Radford, also in Mead's laboratory, used these observations and an assumed value for surface tension equal to that of plasma to estimate lung surface area. The result differed greatly from that of morphologists and stimulated Clements and Brown to measure surface tension of material expressed from the lung. Clements, then at the Army Chemical Center in Edgewood, Maryland, tried to reconcile Radford's data with Pattle's observations of stable bubbles expressed from lungs with zero surface tension. Whittenberger from the Harvard School of Public Health, was an advisor to Clements at Edgewood and brought news of Clements' findings to Mead and Avery. Clements had reasoned that a dynamic method of measurement of surface tension would be closer to events in lungs, and designed a modified Wilhelmy surface film balance to study changes in surface tension with area. His striking observation established the important feature of the alveolar lining layer, namely a change in surface tension with area, so that at large lung volumes surface tension is high, and at low lung volumes it approaches zero. He named the material presumed to be at the alveolar air interface "pulmonary surfactant," and commented on its central role as an antiatelectasis factor.

I was so impressed by Clements' 1957 paper that I visited him and his associate, Brown, in Edgewood in December, 1957 to see the surface film balance. On return to Boston, Mead proposed a way to modify the method to allow study of minced extracts from lungs of human infants. Samples of lungs were obtained courtesy of Kurt Benirshke, then chief pathologist at the Boston Lying-In Hospital. The absence of foam in the lungs at autopsy was a prominent observation that might have led to the conclusion that these lungs were deficient in surfactant even in the absence of measurements on the surface film balance. The first measurements were made before the Pattle magnum opus appeared late in 1958. His observation that bubbles expressed from lungs of immature guinea pigs were unstable reassured us that we were on the right track in pursuing our studies on lungs of infants who had died of hyaline membrane disease.

The selection of a problem for investigation was not as straightforward as the previous account would indicate. I had moved into a laboratory where mechanics of breathing were of central interest. My predecessor as a fellow with Mead was Charles D. Cook, who had explored the changes in compliance in dogs during induced pulmonary edema. Mead had posed the question "where in the lung did the foam first appear?" That question led to reflections in foam stability and the role of surface forces in lungs. C. V. Boys' "Soap Bubbles," a classic introduction to surface tension, was invaluable to me as I struggled with this new discipline. I found the literature on surfaces at the MIT library where N. K. Adam's classic text was available, as were the physical chemical journals in which Langmuir and colleagues had published their studies on force-area relationships of surface films. My intent was to try to answer the question asked by Mead: "Where do the bubbles originate in the lung in cases of pulmonary edema?" Perhaps the pursuit of the question would have opened new vistas. As far as I know, it remains unanswered. I was diverted to measuring surface tension in lung extracts, in part because Clements had devised a method that was applicable to the study of human lungs. I was interested in infants with respiratory problems, and I noted the absence of foam in them, which made it a reasonable possibility that their problem was surfactant deficiency. I visited him at Edgewood to learn his method of measurement with a Wilhelmy balance. It was over Christmas vacation, December, 1957.

This recollection is being written more than forty years after publication of our 1959 paper that described the lack of ability of lung extracts taken from infants who died of hyaline membrane disease to lower surface tension on film compression. The ensuing events involved so many individuals in multiple disciplines that it would be inappropriate to elaborate the further evolution of ideas here. Comroe described many of the early events in his superb article "Premature Science and Immature Lungs" in his 1977 book Retrospectroscope.

My own addendum to Comroe's history concerns the time lag between our publication of the initial observations and general acceptance of their significance. In 1959 multiple theories of pathogenesis of hyaline membrane disease were put forth, and provocative data were assembled. In the first edition of my book The Lung and its Disorders in the Newborn Infant (1964), I tried to be objective about what was known, and, although I included our observations on surface forces, I also said, "No one knows the precise sequence of events which leads to hyaline membrane disease; the view that each individual holds as most important in the pathogenesis directs his therapeutic approach." I then reviewed the arguments for the primacy of aspiration, asphyxia, heart failure, shock, disturbed autonomic regulation, fibrinolytic enzyme defect, prolonged acid-base derangements, and low serum proteins. Subsequent years have permitted greater clarity on many of these issues. The infants who develop hyaline membrane disease are immature for their gestational age in many ways, but necessarily so with respect to the ability to synthesize and secrete adequate amounts of the complex saturated phospholipids that are an essential component of the pulmonary surfactant. The first time I was willing to come out clearly with the statement that "most students of the subject now agree to the central role of immaturity of the lung. . ." was in the fourth edition of my book published in 1981.

I was fortunate to have been in the right place at the right time. My role was to build a bridge between clinicians concerned with infants and physiologists in pursuit of understanding lung mechanics. I was pediatrician-in-charge of newborn infants at Johns Hopkins Hospital from 1961-69. I conceived of an analogy between the clinical course of physiological jaundice in which the infants had a postnatal rise in bilirubin with subsequent improvement over three to five days as the enzyme induction of the liver's ability to conjugate bilirubin evolved. My question was whether hyaline membrane disease could be a delay in maturation of the alveolar cells ability to synthesize and secrete pulmonary surfactants, i.e. antiatelectasis factors. Previously, the concern was that some postnatal adverse event was the cause of the disorder. The paradigm shift that our studies established was that it represented a delayed postnatal adaptation, hence the predisposition of preterm infants for respiratory distress, and the complete recovery in half of them within a few days after birth.

As in any chapter of ongoing investigation, the perception of the truth is limited by existing findings. One can safely predict that the complex chain of events now being described about the molecular regulation of surfactant production will make those simple first observations on a homemade trough seem a very small step. They would probably have been ignored by many in the scientific community because they were published in a pediatric journal (I knew and admired the editor Richard Day, who asked for the article and accepted it by return mail). Gruenwald's related studies a decade earlier had not stimulated much further work. The difference in this case was that Mead and Clements, well known physiologists, discussed the findings, interpreted them, and asked the next set of questions to challenge another generation to pursue more answers.

Of the essence, however, was the collegiality of the investigators, the willingness to share not only data, but importantly, enthusiasm and friendship.

	Footnotes

Correspondence and requests for reprints should be addressed to Mary Ellen Avery, M.D., Harvard Medical School, Boston, MA 02115. E-Mail: AVERY_M{at}a1.tch.harvard.edu