Pulmonary Edema and Acute Lung Injury Research

KENNETH L. BRIGHAM and NORMAN C. STAUB

Vanderbilt University, Nashville, Tennessee; and the University of California San Francisco, San Francisco, California

	INTRODUCTION

TOP INTRODUCTION CONCLUSION REFERENCES

We review here briefly and with admitted biases (sometimes conflicting between the authors) the evolution of our understanding of the mechanisms and clinical consequences of pulmonary edema and acute diffuse lung injury, and we comment on the contributions of the National Institutes of Health (NIH) (in recent years through the National Heart, Lung, and Blood Institute [NHLBI] and its Division of Lung Diseases [DLD]) to that history. The NHLBI and the DLD have played major roles in setting the direction and pace of research in this area, both by providing the major source of research and training support and by introducing funding mechanisms that influenced the direction and organization of the research. We attempt here to interweave developments at the NIH with progress in understanding pulmonary edema and acute lung injury and to give an opinion about relationships between the two.

	IDIOPATHIC ANASARCA OF THE LUNGS

More than a century before President Herbert Hoover signed legislation in 1930, which transformed the National Hygienic Laboratory into the NIH, Rene Laennec, in his classic, Treatise on Diseases of the Chest, published in 1821, described the gross pathology of the heart and lungs in an entity he termed "idiopathic anasarca of the lungs." Laennec inferred that the pathophysiology of the disorder was pulmonary edema without heart failure (1). Here is an excerpt from that description:

Oedema of the lungs is rarely a primary and idiopathic disease. It comes on, most commonly, with other dropsical affections, in cachectic subjects, towards the fatal termination of long-continued fevers, or organic affections, especially those of the heart. Peripneumony that has terminated by resolution, appears also to leave a great predisposition to it; and the most extensive and severe cases that I have met with, occurred during a temporary convalescence from severe attacks of this disease. . . . Although this disease commonly is merely consequent on other affections, and often takes place merely a few hours before death, nevertheless, in some cases, it has certainly lasted several weeks, and even months; and, in a few of these, it even seems to have been idiopathic. The suffocative orthopnea, which sometimes carries off children after attacks of measles, is probably idiopathic anasarca of the lungs. (1)

This is almost certainly a description of the same entity that has become a major focus of contemporary research in pulmonary edema, and that has been called a number of other things, as discussed below. Laennec's "descriptive" research would not compete very well for funds in the current climate, but careful observations clearly communicated are essential to defining a research problem and formulating questions that can be answered. The quote from Laennec also illustrates that mechanisms can be accurately inferred from correlations of structure with the clinical picture.

	PULMONARY EDEMA

In 1941, the NIH awarded 12 research grants, totaling $78,000 in direct costs. Although pulmonary edema was not recognized as a medical subject heading by the National Library of Medicine until about 1950, some research related to the topic did occur in those early projects. Professor Arthur Guyton writes:

The first grant that I received from NIH . . . was H-337 in 1949 for $12,852 total costs . . . was to study multiple aspects of physiology related to weakness of the heart or overloading of the heart by excess fluid. Somewhere in the middle of this grant were a few studies on pulmonary edema. (A. Guyton, personal communication [letters of March and June, 1997])

Those few studies, as Dr. Guyton so modestly puts it, included Guyton and Lindsay's landmark paper (2) about the effect of hydrostatic pressure and protein osmotic pressure on pulmonary edema formation (Figure 1). That grant was funded for more than 20 yr before becoming a Program Project Grant (PPG) in the early 1970s. Just before he retired in 1990, Dr. Guyton writes, his grant was supporting research in the laboratories of eight senior investigators and a number of beginning investigators. The direct costs were about $400,000 per year (A. Guyton, personal communication, 1997).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Relationship between lung water content and left atrial pressure in dogs with normal (A) or decreased (B) plasma oncotic pressure. (Reprinted with permission from reference 2.)

The early work related to the pathophysiology of pulmonary edema provided a core of knowledge that defined basic processes of edema formation in the lungs, and that separated increased pressure and increased permeability types of pulmonary edema (2, 3). That concept, based on experimental physiology, carried over into the clinical arena, where pulmonary edema was described as cardiac or noncardiac (one of us [K.B.] tried to popularize Laennec's terminology of primary and secondary pulmonary edema without success). The fact that the clinical syndromes were based on physiologic concepts drove development of therapy as well as basic research for several years.

Congress established the National Heart Institute (NHI) in 1948 and James Shannon was appointed its first Director. Shannon later served as NIH Director from 1955-1967, during the period when the NIH became established as the world's leader in biomedical research (4). By 1970 the NIH was awarding more than 11,000 grants annually valued at more than $600,000,000 (averaging $55,000/yr).

	ADULT RESPIRATORY DISTRESS SYNDROME

In the late 1960s and early 1970s several interrelated events dramatically accelerated basic and clinical research related to pulmonary edema and acute lung injury and consolidated the linkages between laboratory observations and clinical medicine. Those events included perceptive clinical observations, major strides in basic science, and new developments at the NIH.

Ashbaugh and colleagues (5) first described a clinical entity that they called the adult respiratory distress syndrome (ARDS) in an article published in 1967. It appears that the description was based on the intuition that comes from clinical experience and an understanding of and interest in physiology. The fact that the reported patients had radiographic evidence of pulmonary edema without heart failure fit the term "noncardiac edema." However, the severity and refractoriness of the respiratory failure and the histologic findings in patients who died seemed more than could be explained by excess lung water alone. Petty believed, even at this early point, that lung inflammation was a major factor (T. L. Petty, personal communication, 1997) which led him to use high doses of corticosteroids as therapy. Subsequent clinical studies failed to demonstrate efficacy of such therapy (6), but the inference that lung inflammation was important was prescient.

Much of the experimental work related to increased lung vascular permeability up to the late 1960s was done with animal models using chemicals that directly injure the lungs. The goal was to cause an isolated increase in membrane permeability without affecting vascular pressures in order to clearly define the physiology; alloxan and alpha napthyl thiourea were popular chemicals for this purpose (3). The concept that the clinical problem was only too much water in the lungs led to simple therapeutic remedies: aggressive diuresis and volume restriction, and administration of oncotically active solutes like albumin, for example. Clinical use of extracorporeal membrane oxygenation was based on the idea that if oxygenation could be maintained during the most acute phase of respiratory failure, then the edema would resolve and the patient would live.

Gradually, it became clear to most people working in the area that what was called ARDS clinically involved diffuse lung injury and inflammation. Research gravitated toward studies of the roles of neutrophils, oxidant stress, and other inflammatory processes. Laboratory models shifted to inflammation-based models, most notably responses to gram-negative bacteria and endotoxin (7). In summary, basic and clinical research in the area became much broader, both in concept and in the experimental approaches employed.

The DLD was formally established as part of the National Heart and Lung Institute in 1972 ("Blood" was not added to the Institute title until 1973). Claude Lenfant was the first Director of the Division, and later (1986) became Director of the NHLBI. According to documents provided by the Director and Deputy Director of the DLD, when the Division was established, it took over responsibility for the administration of more than 200 existing grants, including R01s, Program Projects, Specialized Centers of Research (SCORs), and numerous special programs, with total direct costs of $15,474,111 (8; S. Hurd, personal communication [letter of March, 1997]).

The advent and evolution of the SCOR funding mechanism impacted basic and clinical research related to pulmonary edema and lung injury significantly. The initial SCORs were solicited in 1970, introducing a funding process that was different than the R01 and PPG mechanisms. The Pulmonary SCOR Program was to "foster a concerted research effort, involving both clinical and basic disciplines, directed toward the prevention, diagnosis and treatment of pulmonary diseases identified by the Division of Lung Diseases." Specific requirements of the proposals included, "The SCOR grant must focus on one of the pulmonary disease categories identified in this announcement" and "Clinical aspects of disease must be the primary emphasis of each grant, but fundamental studies must also be included" [underlining is shown as in original] (9). The SCOR programs were a departure from investigator-initiated proposals, a departure that was not universally lauded.

From its inception, Shannon led the NIH to support freewheeling research that gave opportunities for imaginative scientists to identify and pursue the problems they considered important. Shannon and his advisors, including his friend Julius H. Comroe, Jr., believed that bright, imaginative, and productive individual scientists would select important problems, devise experiments that could be done, and bring forth discoveries that would benefit mankind quickly. At first there were no RFPs, RFAs, or even PPGs. During those years, there was also liberal funding of research by the NIH so that by the time one of us (N.C.S.) began his career, more than half of all approved grant applications (and not many were disapproved) were funded. About those early years, Julius Comroe wrote, "Those were the good old days when the Director and the Associate Directors of the NIH said, `What can we do to help you?' instead of, `We regret to inform you . . .' " (10). The SCOR mechanism was a quantum leap toward directed research.

The initial SCOR program related to lung injury was in the category of Pulmonary Vascular Diseases, and this program was announced in 1975. The two areas identified in the announcement were pulmonary hypertension and cor pulmonale, and pulmonary edema (9). The latter category was focused entirely on lung solute and fluid exchange. That focus was probably a consequence of exciting work that was developing in the physiologic investigation of the area. Much of that work was developing at the Cardiovascular Research Institute (CVRI) in San Francisco. It is interesting that one of the early "top-down" NIH initiatives was based on research done in a place with a very different philosophy.

Julius Comroe moved to San Francisco in 1957, giving up his own research program to develop the CVRI. This move was not without risk, since Comroe had an imaginative and productive research program at the University of Pennsylvania, and at the time the University of California San Francisco was a small, provincial medical school. Comroe put into practice two principles of research at the CVRI. The first (the Seaborg principle) was: the larger the group of investigators, the more likely it is to discover useful things. (When Comroe asked Glenn Seaborg, himself a Nobel laureate, why there were so many Nobel prize winners in the Physics Department at UC Berkeley, Seabord replied, "Because it's the biggest.") The second principle (dubbed the Comroe Principle by N.C.S.) was: the best way to discover things is to select intelligent, promising or productive investigators, give them the money and space and let them do research without interference from above. (The benefits of unfettered research were upheld later, when Comroe and Dripps analyzed the basic and applied research that led to the achievement of 10 major advances in cardiovascular medicine; they concluded that 40% of the discoveries were made by people looking for something else [11].)

By 1970, the NIH grant funding rate was down to 35.5% of approved applications. It is difficult to know all of the factors influencing the beginning of the SCOR approach, but one surely must have been an effort to obtain increased funding for important research from Congress. Julius Comroe once said that Congress would never restrict NIH funding; friendly congressmen (Hill, Burton, and colleagues) would see to that. But that was before the defense-induced cutbacks began. With that development, it became more necessary than ever for the NIH to convince Congress of the pay-off on their investment, and a disease-focused program was one way to do that. The fact that one focus was on pulmonary edema was an important factor in advancing the research that was emerging on that topic.

The SCOR mechanism also forced interactions among basic scientists and clinician-investigators. There had been a long-standing perception that the NIH would not support clinical research because it was not considered "good science." Relationships between basic scientists and clinicians required by the SCOR structure provided dollars for clinical research and also forced clinical investigators to defend their science in a critical forum.

The NHLBI-supported program to test the utility of extracorporeal membrane oxygenation as therapy for ARDS was winding down by the mid-1970s. A workshop sponsored by the DLD held in Arlington, Virginia, on October 20-21, 1976 is considered by Dr. Suzanne Hurd, current Director of the DLD, to have been the "end of ECMO and the beginning of the ARDS SCOR Program" (S. Hurd, personal communication [memo of September 10, 1997]). The subjects discussed at that workshop reflect what was learned from the ECMO program and laboratory scientific developments in several areas; the discussions also reflect the rapidly evolving focus on investigation of mechanisms of lung injury and inflammation-related phenomena (12). Participants included physiologists, pharmacologists, biochemists, pathologists, and clinicians. Topics discussed included roles of polymorphonuclear leukocytes, platelets, proteolytic enzymes, mast cells, humoral substances, and intrapulmonary nerves in the process of acute lung injury; also considered were increased permeability and alterations in lung mechanics under the heading, "Response to Injury: Functional Abnormalities." Recommendations from that workshop under the separate headings of "Clinical," "Pathologic," and "Experimental" formed the basis of the Program Announcement for Pulmonary Specialized Centers of Research in Acute Respiratory Failure, which appeared December 15, 1977 (13). Thus, the ARDS program was the "son/daughter of ECMO" and began as a separate program from the SCORs in Pulmonary Vascular Disease, which were heavily weighted toward pulmonary edema research.

The development of increased rigor in clinical investigations in the area of acute lung injury (and in critically ill patients in general) correlates in time with the advent of the SCOR program in acute respiratory failure and its successor programs. Attribution of cause and effect is, however, not so easy. There were probably many factors that influenced programmatic decisions at DLD and many factors that influenced the development of clinical research in the critically ill; undoubtedly, some of those factors are shared. Whether cause and effect or coincidence, the opportunity to establish multidisciplinary programs, which included of necessity clinical investigators and basic scientists, and the organization of a national network of such programs through the SCOR mechanism were major contributions to the milieu.

	ACUTE LUNG INJURY AND BEYOND

Basic physiologic investigation related to lung water and solute exchange and the clinical studies of the syndrome resulting from diffuse injury to the lungs are different parts of the same elephant. That could also be said of basic investigations of inflammatory cell behavior, endothelial, epithelial, and macrophage cell and molecular biology, and lung mechanics. Progress toward less provincial concepts of the problem has resulted from the accumulation of important clinical databases as well as basic laboratory investigation in many fields of science, and progress has been driven by the need to identify effective therapeutic interventions in a highly lethal clinical disorder.

It has always been a practice of the DLD to conduct a review of the SCOR programs with each funding cycle. This was formerly done through the now defunct Pulmonary Diseases Advisory Committee (PDAC), and such was the case in 1990 when the ARDS and Pulmonary Vascular Disease SCOR programs were reviewed. The committee of non-NIH scientists conducting that review recommended that the Pulmonary Vascular Disease SCORs be eliminated and that the title Acute Lung Injury be substituted for Acute Respiratory Failure, creating a single category of programs with a focus on lung injury, with all of its basic science and clinical implications. (One of us [K.B.] was Chairman of the PDAC at the time and had the dubious distinction of recommending dissolution of a SCOR on which he was the Principal Investigator.) This change resulted in coordinating the SCOR-supported work related to lung injury and paved the way for later initiatives by the NHLBI.

Over the subsequent years, several changes have occurred. The conceptualization of the clinical problem has changed as more coordinated clinical data related to cause, clinical course, and outcomes have accumulated. Such data have made it clear that the clinical syndrome of acute lung injury (or ARDS) is, again, but one part of the elephant. Acute lung injury occurs most often in the setting of sepsis, and sepsis is associated with failure of organs other than the lungs. Also, it has become clear that sepsis is an identifiable clinical syndrome, not always directly attributable to gram-negative bacterial infection. So the terminology changed to multi-organ failure syndrome (MOFS), to sepsis syndrome, and more recently to systemic inflammatory response syndrome (SIRS), a term that offends the sensibilities of even the most determined lumper. These changes in definition are driven both by a broadening pathogenetic concept and by a need to characterize patient groups in which therapeutic interventions can be tested systematically.

A recent initiative by the DLD deserves comment. The establishment of an ARDS network of clinical investigators, which serves a triage function for identifying promising new interventions and as a resource for developing experimental protocols and providing patient populations for clinical studies, is an important contribution to clinical research in the area. Such research has relied heavily on support from industry in recent years. While that is not necessarily evil, the profit motive and the desire for good clinical science are not always compatible.

	RESEARCH TRAINING

A careful review of the scientific advances in any field, and pulmonary physiology is no exception, finds that about 80% of the significant advances are made by 20% of the scientists. As a general statement of all human activity, this phenomenon is called the Pareto Principle. The principle should be kept in mind, but not dwelt on, in designing research training programs for young scientists.

Leaders of outstanding research training programs (Julius Comroe is the obvious example) have taught that interested young people, even those with no significant scientific training, could, given the opportunity to learn how to do science, become successful contributors to the fund of scientific knowledge. At the CVRI, Comroe transformed his NIH pulmonary training grant and greatly expanded it into a multidisciplinary training program (H5251) that, if the numbers are meaningful, was the first pulmonary training grant awarded by the NIH (8). That program included not only bench research, but also provided reviews of mathematics, physiology, biochemistry, pharmacology, statistics, experimental design, electronics, scientific writing, and teaching. In-house programs were supplemented with visiting lecturers, scientific specialists and professors. In that one program, over the years, thousands of young men and women learned how to do science and many became excellent investigators, both basic and clinical. The costs of the training were largely borne by the NIH over a period of more than 30 years.

Sidney Burwell is supposed to have said that the hardest part about teaching medicine was that he knew half of what he was teaching would be proven wrong in the future, but he didn't know which half. We know that 20% of the scientists we are training will make 80% of the future discoveriesbut we can't tell which 20%. Here is an example from each of our laboratories.

Dr. Richard Bland came to Staub's laboratory in 1973, straight from his military service. In a personal communication to Staub (letters of July and August, 1997), he writes:

I became interested in pulmonary edema . . . as a U.S. Army pediatrician . . . at Tripler Army Medical Center in Hawaii. . . . I did clinical studies showing that newborn infants with respiratory distress were . . . hypoproteinemic, but derived little . . . benefit from . . . albumin. . . . The lungs of infants who died had protein rich edema. . . . These observations led me to pursue a research career focusing on the pulmonary circulation and fluid balance in the developing lung. . . . I joined Norman Staub's laboratory . . . I [studied] the effects of alveolar hypoxia on lung fluid balance in adult sheep as my first research project. . . . My next step was to learn the basics of fetal and neonatal cardiovascular surgery. . . . I spent 6 months in the laboratory of Dr. Abraham Rudolph. . . . My first independent research projects were funded as part of a Pediatric Pulmonary SCOR (HL19185) . . . directed by Dr. John Clements. . . . My first independent. . . grant [was supported] by California Lung Association at $10,000 per year. . . . I subsequently joined Norman Staub and Abraham Rudolph . . . on their Program Projects. . . . All of the research . . . related to the developmental biology of the lung and to . . . issues . . . that seemed relevant to my clinical endeavors in neonatal intensive care.

Bland had no laboratory research experience, but he had determination and drive. He soon forged new vistas in fetal and neonatal lung research that could not have been predicted. He continues an extremely successful research career supported by the DLD.

Dr. James Loyd came into the research laboratories in the Pulmonary Medicine Division at Vanderbilt University in 1979, having had no laboratory experience and having graduated from a medical school that was not among the top echelon of research universities. About that experience he writes:

I would like to claim that I had some great vision of a research career in academic medicine. In truth, I entered blindly into the world of academic medical research, having been given an opportunity for research in a fellowship program. Therein I was captivated by the commitment of individual colleagues and mentors, the excitement, and the enjoyment of discovery, and a vision of contributions which might have lasting impact on the outcome of patients with unique and tragic diseases.

In fact, because I enjoy working with patients so much, I still occasionally wonder why I have not yet gone into full-time clinical practice. The answer is that I am unable to imagine that any other endeavor is as fun or rewarding as what I do now. The opportunity and support to contribute in an environment where bright and dedicated minds are encouraged to examine questions that each judges as most important is an absolutely unparalleled privilege. I cannot imagine a more rewarding feeling than that of contributing to the discovery of a significant advancement of understanding about a disease, which then has impact on every current and future patient with that disease, and others like it.

In summary, having been given the opportunity for a year of research in a pulmonary fellowship program, I entered blindly into a world of academic research, where I was captivated by the commitment of individuals, the excitement and joy of discovery, and an opportunity to change the outcome of patients with unique and catastrophic diseases. This opportunity is the legacy of former and current division directors and department chairmen who emphasize knowledge and truth, and colleagues at all levels who revel in their pursuit. (J. Loyd, personal communication [Note, October 1997])

Three months into his first research year, Loyd was depressed about his potential as a scientist and wanted to quit the program; obviously, he did not quit. The subsequent story is a long and fascinating one that will not be recounted here, but his research has been continuously supported by the NHLBI since he finished his fellowship in 1981. One of his latest papers published in Nature Genetics (14) localizes the genetic abnormality in patients with familial primary pulmonary hypertension to a 25-megabase region of chromosome 2q31. He has organized an international consortium of investigators focused on cloning the primary pulmonary hypertension gene (15) and that outcome is imminent. An Institutional Training Grant from NHLBI made all of that possible. Talk about return on an investment!

Training programs are among the most difficult to assess credibly for those in positions to determine their support. While not confident that we could convince skeptics with irrefutable objective data, we believe that support of wide-ranging training programs will be remembered as one of the great achievements of the NIH during the last 50 years.

	SUMMARY AND CONCLUSIONS

TOP INTRODUCTION CONCLUSION REFERENCES

The last 50 yr have seen major advances in understanding of the pathophysiology of pulmonary edema and mechanisms and consequences of acute lung injury. This remarkable progress has occurred because the search for new knowledge is irresistible to bright, young (and not so young) people, and because over much of that period there has been generous support from the NIH (through the NHLBI and the DLD in recent years) for the training of investigators and the conduct of research. The imagination of scientists fostered by support of investigator-initiated research has fueled technical innovation and conceptual advances. Continuing support of such programs is critical. The design of more structured funding mechanisms has contributed in a major way to development of clinical research and the coordination of clinical databases, resulting in well-designed investigations in critically ill patients. Multidisciplinary programs have permitted development of large collaborative groups. Such groups are fertile ground for scientific discovery and for nurturing growth of the next generation of basic and clinical investigators.

	Footnotes

Correspondence and requests for reprints should be addressed to Kenneth L. Brigham, Center for Lung Research, Vanderbilt University School of Medicine, T-1217 Medical Center N., Nashville, TN 37232-2650.

	References

TOP INTRODUCTION CONCLUSION REFERENCES

1. Laennec, R. T. H. 1821. A Treatise on Diseases of the Chest. T. and G. Underwood, London. 98-99.

2. Guyton, A. C., and A. W. Lindsey. 1959. Effect of elevated left atrial pressure and decreased plasma protein concentration on the development of pulmonary edema. Circ. Res. 7: 649-657 [Abstract/Free Full Text].

3. Staub, N. C., H. Nagano, and M. L. Pearce. 1967. Pulmonary edema in dogs, especially the sequence of fluid accumulation in the lungs. J. Appl. Physiol. 22: 227-240 [Free Full Text].

4. Kanigel, R. 1992. Apprentice to Genius: The Making of a Scientific Dynasty. Johns Hopkins University Press, Baltimore.

5. Ashbaugh, D. G., D. B. Bigelow, and T. L. Petty. 1967. Acute respiratory distress in adults. Lancet 2: 219-223 .

6. Bernard, G. L., J. M. Luce, and C. L. Sprung. 1987. High dose corticosteroids in patients with adult respiratory distress syndrome. N. Engl. J. Med. 317: 1565-1570 [Abstract].

7. Brigham, K. L., W. C. Woolverton, L. H. Blake, and N. C. Staub. 1974. Increased sheep lung vascular permeability caused by pseudomonas bacteremia. J. Clin. Invest. 54: 792-804 .

8. Gross, P. L. 1971. Analysis of Current Pulmonary Research Programs. Office of Lung Programs, National Heart and Lung Institute, Bethesda, MD.

9. Program Announcement for Pulmonary Specialized Centers of Research, National Heart and Lung Institute, Division of Lung Diseases, March 1, 1975.

10. Comroe, J. H. 1983. CVRI: the early years. In J. H. Comroe, editor. The First Twenty-five Years, 1958-1983. University of California, San Francisco. 1-30.

11. Comroe, J. H., and R. D. Dripps. 1976. Scientific basis for the support of biomedical science. Science 192: 110-111 .

12. 1976. Report on National Heart, Lung, and Blood Institute Division of Lung Diseases Workshop on Mechanisms of Acute Respiratory Failure, October 20-21, 1976.

13. National Heart Lung and Blood Institute Division of Lung Diseases. 1977. Program Announcement for Pulmonary Specialized Centers of Research in Adult Respiratory Failure, December 15, 1977.

14. Nichols, W., D. Koller, B. Slovis, T. Foroud, V. Terry, N. Arnold, D. Siemieniak, L. Wheeler, J. Phillips, J. Newman, M. Conneally, D. Ginsburg, and J. Loyd. 1997. Localization of the gene for familial primary pulmonary hypertension to chromosome 2q31-32. Nat. Gen. 15: 277-280 [Medline].

15. Loyd, J. E., J. H. Newman, K. Lane, L. Wheeler, J. F. A. Phillips, T. Faroud, D. Koller, P. Conneally, W. Nichols, D. Ginsburg, and R. Trembath. 1997. Progress in a gene search for familial primary pulmonary hypertension by an international consortium (abstract). Am. J. Respir. Crit. Care Med. (In press)