Etiology and pathogenesis of primary pulmonary hypertension: a perspective – Brenot Memorial Symposium on the Pathogenesis of Primary Pulmonary Hypertension
Alfred P. Fishman
In recent years, considerable advances have been made in treating primary pulmonary hypertension (PPH). These have provided a series of therapeutic options, ranging from the oral administration of calcium channel blockers to the continuous infusion of prostacyclin and/or lung transplantation. These therapeutic advances have highlighted the need for the better understanding of etiology and pathogenesis. Among the key uncertainties, the following are defined as leading uncertainties: (1) the nature of the initiating lesion; (2) the shared pathogenetic mechanisms that culminate in the pathologic lesions of PPH; (3) the molecular genetic bases for familial PPH and for susceptibility to PPH; (4) understanding of the obliterative-proliferative occlusive process in the small muscular pulmonary arteries; and (5) redefinition of “primary” and “secondary,” ie, a revised nomenclature of pulmonary hypertension. A revised classification based on etiology is presented. (CHEST 1998; 114:242S-247S)
This symposium was held in honor of Francois Brenot, who contributed so importantly to the present understanding of primary pulmonary hypertension. A major goal of the symposium was to define uncertainties in the present understanding of the disease and to pinpoint areas for exploration and clarification.
Of the many topics dealt with in the preceding articles, five can be singled out as directly related to this goal: (1) identifying the early lesion; (g) discovering common denominators among the diverse etiologies of primary pulmonary hypertension; (3) examining familial forms of the disease with respect to defining genetic predisposition; (4) exploring mechanisms involved in the pathogenesis of the obliterative-proliferative stage of the disease; and (5) clarifying the meaning of the designation “primary.”
THE EARLY LESION
Probably the greatest challenge facing research on primary pulmonary hypertension is uncertainty concerning the nature of the initiating, or early, lesion. By the time of autopsy or biopsy of the lung, both of which occur late in the course of the disease, all that can be discerned is that the process has affected the small pulmonary arteries and arterioles and that one or more of four separate mechanisms has been involved: (1) vasoconstriction; (2) proliferation of components of the vascular wall, ie, intima and media; (3) inflammation; and (4) thrombosis. Two of the four seem far removed from the initiating or early lesion: thrombosis is generally held to be an epiphenomenon, secondary to an interplay between the abnormal vascular lining and blood constituents; inflammation of the vessel walls is neither a consistent or prominent feature of the histologic findings at biopsy or autopsy. When evidence of inflammation is present, it is usually in the form of a necrotizing arteritis that is more apt to represent a complicating burst of malignant pulmonary hypertension than an initiating mechanism for the pulmonary hypertension.
For decades, the idea of pulmonary vasoconstriction as the initiating mechanism for primary pulmonary hypertension has dominated thinking about pathogenesis and clinical approaches to diagnosis and treatment. This view is intuitively attractive as an extrapolation from systemic hypertension in which vasoconstriction plays a key role in the pathogenesis of systemic hypertension. More concretely, the idea has been reinforced by the interpretation of thickened media in the resistance vessels of the lungs in patients, especially in children, who died of primary pulmonary hypertension as evidence of vasoconstriction.[1] However, the current consensus is that although heightened pulmonary vascular tone is present in an appreciable number of patients with primary pulmonary hypertension, it is more apt to be a contributing factor than the initiating mechanism.
The predominant current view is that injury to the vascular endothelium of the small pulmonary arteries and arterioles initiates the pulmonary hypertensive process (see article by Rich, p 237S). Several lines of evidence have converged to support this view. For example, an abrupt increase in pulmonary arterial pressure, produced experimentally by anastomosing a systemic artery to a segment of the pulmonary vascular tree elicits, over time, many of the lesions of primary pulmonary hypertension, including plexiform lesions.[2] Dietary pulmonary hypertension, produced by feeding crotalaria to rats, evokes endothelial proliferation followed by proliferation of smooth muscle in the media (see article by Rabinovitch, p 213S). However, that endothelial injury may not be the full story is suggested by occasional instances of primary pulmonary hypertension in which the small muscular arteries and arterioles only manifest hypertrophy of the media with no visible alterations in the intima.
In the normal pulmonary circulation, a balance favoring vasodilation and inhibiting proliferation is maintained between the vasodilators prostacyclin and nitric oxide, on the one hand, and endothelin I, a potent pulmonary vasoconstrictor, on the other (see article by Christman, p 205S). In primary pulmonary hypertension, the balance is tilted toward the production of endothelin I, increased quantities of which can be detected in the circulating blood; both cyclooxygenase and nitric oxide synthase activity are concomitantly decreased (see article by Giaid, p 208S). Clearly, redressing the balance is an attractive therapeutic prospect; for example, antibodies to endothelin I might favor pulmonary vasodilation.
If endothelium is the predominant site of injury, can the resultant derangement in endothelial function be detected in vivo? In primary pulmonary hypertension, the degree of downregulation of nitric oxide synthase varies with the severity of the disease; also, small muscular pulmonary vessels and plexiform lesions contain diminished concentrations of nitric oxide synthase (see article by Voelkel, p 225S). Accordingly, determinations of levels of this enzyme may be useful as a marker of abnormal endothelial function in primary pulmonary hypertension. Moreover, nitric oxide synthase may be a suitable therapeutic target, ie, to administer the ecNOS gene via gene transfer. One difficulty inherent in interpreting levels of enzymes with respect to etiology and pathogenesis is whether an abnormal level reflects the cause of deranged endothelial function or an effect.
Another approach to detecting abnormal pulmonary endothelial function is basically an extension of the indicator-dilution principle using tracer sub stances that are normally removed by the endothelium in a single circulation through the lungs. One such tracer substance is serotonin.
In the face of all the attention focused on endothelium, mounting evidence is calling attention to the possibility that a direct effect of an injury on vascular smooth muscle may set into play the cascade that ends in primary pulmonary hypertension. The strongest evidence along this line is provided by recent observations on the molecular bases for the hypoxic pressor response. Although experiments involving acute hypoxia may, at first blush, appear to be somewhat tangential to the etiology and pathogenesis of primary pulmonary hypertension, recent observations on isolated small pulmonary arteries indicate that the same [K.sup.+] channels are involved in the vasoconstrictor responses to acute hypoxia and to appetite-suppressing agents (eg, aminorex and fenfluramine): reversible inhibition of KDR channels causes membrane depolarization, activation of voltage-operated calcium channels, and vasoconstriction (see article by Weir, p 200S). A similar picture can be drawn for oxygen-induced constriction of the ductus arteriosus.[3]
However, certain key missing links remain in the case for direct injury to vascular smooth muscle as the initiating mechanism for the pressor response. (1) Where is the sensor for the hypoxic stimulus? (2) Accepting that there is a sensor that triggers pulmonary vasoconstriction, how does the resultant increase in tone of the small muscular arteries bring about the proliferative and obliterative responses that narrow and/or occlude these vessels? (3) How can the evidence for a pivotal role for endothelium in the pressor response to hypoxia and injury be reconciled with the experimental observations indicating that smooth muscle can be directly stimulated by the anorexigens? (4) Although the chronic hypoxic pressor response is associated with pulmonary vascular remodeling, the structural changes elicited by chronic hypoxia bear no resemblance to those evoked by the appetite-suppressant drugs. (5) How do endothelium, media, and adventitia interplay in both the acute and chronic responses to injury?
Such uncertainties underscore a dominant theme of the conference, ie, previous preoccupation with vasoconstriction as the initiating mechanism of primary pulmonary hypertension must be succeeded by intensified research into the structural changes of primary pulmonary hypertension, ie, the proliferative and obliterative aspects. Exploration of these new avenues requires that the basic concepts and techniques in cell growth, development, and division be applied to studies of the responses of the pulmonary circulation to diverse stimuli.
CLUES TO ETIOLOGY
Even Sherlock Holmes would have been perplexed by the surfeit of clues to the etiology of unexplained pulmonary hypertension and stimulated by the fact that, to date, each clue has led to a blind alley in the search for a common denominator. How to reconcile the diverse etiologies of unexplained pulmonary hypertension with the spectrum of histologic changes in the small muscular arteries and arterioles found at autopsy and by biopsy? Some of the leads would have been very tantalizing. What is the hidden meaning of the association of portal and pulmonary hypertension? What is the relation of autoimmune disease, such as scleroderma or lupus erythematosus, to so-called “primary” pulmonary hypertension? Do the high levels of circulating antinuclear antibodies in these autoimmune diseases and in unexplained pulmonary hypertension indicate that some instances of primary pulmonary hypertension are, in reality, consequences of autoimmune diseases?
Promising leads have come from documented instances of the accidental ingestion of toxic materials (toxic oil), from side effects of therapeutic agents (appetite suppressants), and from serologic evidence (lupus erythematosus). Another lead stems from the history of research on the hypoxic pulmonary pressure response.[4] In the late 1890s, Francois Frank and Bradford Dean demonstrated that asphyxia elicited pulmonary vasoconstriction and that the sympathetic nervous system was involved in this pressor response. In the 1940s, Euler and Liljestrand diverted attention from this integrative response by demonstrating that the pulmonary pressor response could be elicited in the open chest animal and that the pulmonary vasoconstriction could be accounted for by local regulatory mechanisms. Subsequently, others, including Duke and Nissel, showed that the hypoxic pressor response could even be elicited in the isolated (denervated) lung.
Although the bulk of experimental attention had shifted to the local response, I. deB. Daly and coworkers, using a variety of elaborate experimental preparations, kept alive the idea that the sympathetic innervation could play a role in the control of the pulmonary vessels. Spotty evidence from other laboratories and clinics, based on more intact preparations, was in keeping with the elaborate experiments of I deB. Daly and coworkers on highly artificial preparations: (1) the observations that systemic and pulmonary hypertension often coexist in humans; (2) the experimental finding in the dog that stimulation of the carotid body elicits pulmonary vasoconstriction; and (3) the demonstration that intense pulmonary and systemic hypertension follows direct stimulation of the “defense area” in the region of the hypothalamus of the brain of the dog.[5] These observations suggest that even though the sympathetic nervous system may not, per se, initiate pulmonary hypertension, it may well play an enhancing role (eg, during periods of anxiety and stress).
FAMILIAL PULMONARY HYPERTENSION AND INHERITED SUSCEPTIBILITY
Of all the leads to etiology and pathogenesis, probably none is more promising than the familial occurrence of unexplained pulmonary hypertension (see article by Barst and Loyd, p 231S). To date, approximately 60 families with familial pulmonary hypertension have been identified in the United States (see Barst and Loyd, p 231S). This discovery already has paid large dividends by showing that in patients with familial pulmonary hypertension of the pulmonary vascular lesions, at autopsy, these are apt to be quite heterogeneous. Indeed, in some instances, plexiform lesions are missing. This variety in the types of occlussive proliferative vascular lesions has challenged the notion of a pathologic hallmark of primary pulmonary hypertension.
Although the familial nature of some instances of primary pulmonary hypertension was recognized years ago, only with recent access to a considerable number of families could the genetic nature of the disorder be explored scientifically. In recent years, familial primary pulmonary hypertension has been characterized as an autosomal dominant disease, with incomplete penetrance and genetic anticipation; the genetic defect has further been localized to a particular chromosome and the genetic explanation for familial pulmonary hypertension has invoked a mutation involving greatly expanded trinucleotide repetition. In this regard, a similar mechanism has been shown to operate in a miscellany of other diseases, including the fragile X syndrome, myotonic dystrophy, and Huntington’s chorea.
To date, the genetic studies have yielded promising leads in two different, but related, areas: (1) on a fundamental level, the genetic bases for familial primary pulmonary hypertension; as indicated above, the affected chromosome has been identified and the nature of the mutation is being explored; and (2) on the clinical level, inherited susceptibility to pulmonary hypertension is being investigated in families of parents with familial pulmonary hypertension.
The question of susceptibility keeps on resurfacing. The aminorex epidemic showed that of the thousands of people who took the drug, relatively few developed the disease. The recent epidemic due to the appetite-suppressant fenfluramines and fenfluramine-phentermine (“fen-phen”) has raised the same question. Studies of members of affected families who are phenotypically normal hold promise of clarifying the bases of susceptibility and of developing methods for early detection. The need for early detection is critical for the understanding of the pathogenesis of primary pulmonary hypertension since clinical manifestations generally appear when the disease is far advanced and the remote initiating lesions have become blurred by time and relentless anatomic changes in the pulmonary vasculature.
The heightened interest in the inheritable form of the disease has already been edifying by indicating that some instances of primary pulmonary hypertension that are apparently sporadic are, in reality, familial in nature. One promise, as yet unfulfilled, is that of developing animal models of the disease that. can be studied from their inception. As yet, no such model exists, although a variety of approximations, such as the fawn-hooded rat and monocrotaline-induced pulmonary hypertension, are providing relevant information.
THE OBLITERATIVE-PROLIFERATIVE RESPONSE
Understandably, clinical attention until now has been focused at the vasomotor and thrombotic response because “that is where the light is shining.” However, the administration of vasodilator and anticoagulant substances has proved effective in only about one third of patients with primary pulmonary hypertension. In contrast to the success in treating the vasoconstrictive component by administering vasodilator agents, neither a prophylactic antiproliferative therapy nor a therapy that will arrest and/or reverse the fibrotic process has as yet materialized.
However, momentum along this line is currently gaining. In particular, attention is being directed at components of the pathogenetic cascades that begin with injury to endothelium and culminate in proliferative and obstructive pulmonary hypertension. The experimental approaches have taken four tacks: (1) the search for the “early lesion” based on extrapolation from experimental models to the human disease; (2) stepwise elucidation of the pathogenetic sequence that begins with endothelial injury and then enlists smooth muscle migration and proliferation into the process; (3) the search for biological markers of endothelial dysfunction; and (4) identification of biological sites in the cascade that may be amenable to therapeutic interventions.
One fruitful approach to dissecting the proliferative, obliterative response has been a rat model of pulmonary hypertension induced by monocortaline (see article by Rabinovitch, p 213S). In this model, hypertrophy of the media of the small pulmonary arteries is accompanied by increased synthesis of insoluble elastin along with an increase in the number of elastin fragments, suggesting an elastolytic process (see article by Rabinovitch, p 213S). The research has identified an endogenous vascular elastase in serum, a novel enzyme related to the serine proteinase adipsin; this enzyme dissociates elastin from vascular smooth muscle and promotes proliferation and migration of vascular smooth muscle.
The end result of the elastase-induced cascade is remodeling of the vessel wall entailing proliferation, migration, and apoptosis. This process involves tenascin-C, an extracellular matrix glycoprotein, which is induced during the remodeling process, and cellular fibronectin. Experimental studies in the rat have shown that inhibitors of elastase can prevent or retard the proliferative and migratory responses. Clearly, this approach suggests a variety of therapeutic applications. But, it still remains to be learned how much of the experimental success is relevant to the human disease and the extent to which these results apply to other pulmonary diseases that involve pulmonary vascular remodeling.
A different experimental approach to the obliterative response begins with three assumptions: (1) that the vessel wall contains intrinsically all of the biologically important ingredients and information that lead to the obliterative vascular disease; (2) that severe pulmonary hypertension is the result of misguided angiogenesis; and (3) that monoclonal proliferation of endothelial cells is the hallmark of primary pulmonary hypertension that distinguishes it from polyclonal secondary pulmonary hypertension (see article by Voelkel, p 225S). This model pictures primary pulmonary hypertension as a disorder predominantly of endothelial cell proliferation that can begin either as a growth of an endothelial tumorlet or as a circumferential monoclonal proliferation of endothelial cells. This concept and model are currently being tested in samples of tissue from normal and pulmonary hypertensive lungs using such techniques as immunohistochemistry (eg, for factor VIII, vascular endothelial growth factor, prostacyclin synthase).
THE MEANING OF `PRIMARY’ AND THE NONUNIQUE PLEXIFORM LESION
With respect to pulmonary hypertension, the designation “primary” has been used as a synonym for “unexplained.” By tradition, the designation simply means that no cause can be found during life or at autopsy; by usage, the term has also come to be applied when a constellation of pulmonary hypertensive vascular lesions, especially the plexiform lesion, is found at autopsy. Also, as noted above, the focus on the plexigenic lesion has led to the widespread impression that “plexogenic arteriopathy” is the histologic hallmark of unexplained pulmonary hypertension.
This usage has promoted ambiguity. For example, it has been known for years that the constellation of pulmonary hypertensive lesions and plexiform lesions can also occur in certain types of “secondary” pulmonary hypertension (eg, Eisenmenger’s disease). Similar lesions have also been found in such noncardiac diseases as AIDS and scleroderma and following the use of appetite-suppressant drugs. The final nail in the coffin was driven in by the observations on familial primary pulmonary hypertension in which the designation “primary” is misleading on two accounts: (1) the cause of the disease is known, ie, the disease is genetic in origin, and (2) plexiform lesions need not be found at autopsy. A relevant question can then be asked: “What does the plexiform lesion represent?” The current consensus seems to be that the plexiform lesion may be a consequence of mechanical damage at a site of turbulence, generally at a bifurcation of a small muscular pulmonary artery; such turbulence might be produced by a sustained burst of hypertension. For many years, the etiology of the plexiform has defied attempts at detection. However, a new direction seems promising: recent evidence indicates that the lesion is made up primarily of endothelial cells, and that perivascular inflammatory cells are involved in its pathogenesis (see article by Voelkel, p 225S).
Reliance on “plexogenic arteriopathy” as the hallmark of primary pulmonary hypertension may also entail several other disadvantages: (1) it relegates the diagnosis to the autopsy table even though the histologic criteria for the disease are now known not to be specific; (2) by singling out one morphologic feature as distinctive, instead of recognizing that the histologic picture of the end-stage disease may be exceedingly heterogeneous, it suggests a single etiology and pathogenesis that are not likely to be true; (3) by relying on one morphologic feature of endstage disease, it diverts attention from the need for functional markers of the evolving disease during life that culminate in the histologic appearance at autopsy; and (4) except in diseases in which an etiologic agent, such as a causative organism can be identified in the histologic lesion (eg, the tubercle bacillus), it is generally unwise to infer the cause of a disease from the pathologic findings at autopsy.
Since current usage of the designation “primary” cloaks an assortment of diseases for some of which the cause is known (eg, toxic ingestants, autoimmune disorders), it would seem reasonable to categorize those of known or suspected cause as “secondary” and to restrict the designation “primary” to those instances of pulmonary hypertension for which the cause cannot be found during life or at autopsy. As shown in Table 1, a separate category might be created to include distinctive associations between unexplained pulmonary hypertension and other pathologic processes; such associations will probably become etiologically and pathogenetically meaningful as their interrelationships become better understood (eg, portal and pulmonary hypertension).
Table 1–Classification of Pulmonary Hypertension
According to Etiology
Examples
Secondary (known or likely
etiology)
Cardiovascular
Left ventricular failure Myocardial disease
Valvular disease of left Mitral stenosis;
side of heart mitral regurgitation
Obstruction to pulmonary Pulmonary veno-occlusive
venous system disease; fibrosing
mediastinitis
Congenital Eisenmenger’s disease
Developmental Persistent fetal circulation
Pulmonary
Obstructive airways disease COPD
Restrictive lung disease Diffuse interstitial fibrosis
Dietary (related to GI tract) Toxic oil; eosinophilia
myalgia
Infectious Schistosomiasis; AIDS
Pulmonary vascular Thromboembolic disease;
autoimmune diseases
(scleroderma; lupus
erythematosus); drugs
(cocaine)
Genetic Familial pulmonary
hypertension
Environmental Hypoxia
Primary (unexplained)
Without comorbid disease “Primary pulmonary
hypertension”
With comorbid disease Portal-pulmonary hypertension
EPILOGUE
Although this perspective deals solely with etiology and pathogenesis, it also suggests many therapeutic possibilities. To date, of the four key morphologic elements in primary pulmonary hypertension, ie, proliferation, vasomotor activity, thrombosis, and inflammation, therapy has focused on vasoconstriction. The bulk of experimental evidence favors the idea that the obliterative process begins with endothelial injury. However, in some instances of primary pulmonary hypertension, the endothelium appears normal histologically whereas the media is greatly thickened. Whether this normal appearance of the endothelium masks endothelial dysfunction that is not evident histologically is uncertain. Moreover, although the hypoxic pulmonary pressor response does not seem directly relevant to the chronic obliterative and proliferative disease known as primary pulmonary hypertension, it does raise the possibility that the media, rather than the endothelium, may be the site of initial injury.
Important clues to early lesions have surfaced from several different directions: familial disease, appetite-suppressing agents, hepatic cirrhosis, and AIDS. Whether these different etiologies converge on common pathways is unclear in large measure because extrapolations from experimental models and tissues and from end-stage pathology to initiating agents can only be speculative.
Among the attractive prospects raised in the Brenot symposium is the possibility of developing biological markers and new therapeutic modalities based on increasing insights into pathogenetic mechanisms and histologic mediators (eg, antibodies to particular elements of the pathogenetic sequence leading to primary pulmonary hypertension). These afford the prospect of arresting and reversing the proliferative and obliterative pulmonary vascular disease.
However, as will be indicated in the article that Follows, much remains to be learned about the biological activities of agents, such as prostacyclin and nitric oxide, which are currently used therapeutically entirely for their vasodilator properties. It may well be that prolonged use of these vasodilator agents also elicits antiproliferative effects that lead to favorable remodeling of pulmonary vessels; the opposite seems to be true for the vasoconstrictor endothelin I that has proliferative effects. What is still lacking for therapeutic purposes are biological interventions that can arrest and reverse proliferative and obliterative pulmonary vascular disease.
Finally, the control of the normal and hypertensive pulmonary circulations is a complex process that can go awry at different loci in the integrative mechanism. The popular experimental use of organs, tissues, and animals does not provide for assessing integrative mechanisms, such as the autonomic nervous system. At present, research relating to primary pulmonary hypertension is justifiably preoccupied with identifying individual molecules, mediators, and pathways. However, it must be kept in mind that once these components are identified, the pieces of the puzzle will have to be put in place before the integrative machinery becomes evident. At present, it seems reasonable to conclude that until the early lesion that starts the pathologic processes is understood and takes its place at the head of the cascade that leads to vasoconstriction, proliferation, and thrombosis, it is unlikely that the mystery of primary pulmonary hypertension will be dispelled.
REFERENCES
[1] Wagenvoort CA, Wagenvoort N. Primary pulmonary hypertension: a pathologic study of the lung vessels in 156 clinically diagnosed cases. Circulation 1970; 42:1163-84 [2] Downing SE, Vidone BA, Brandt HM, et al. The pathogenesis of vascular lesions in experimental hyperkinetic pulmonary hypertension. Am J Pathol 1963; 43:739-65 [3] Tristani-Firouzi M, Reeve HL, Tolarova S, et al. Oxygen-induced constriction of rabbit ductus arteriosus occurs via inhibition of a 4-amidopyridine-, voltage-sensitive potassium channel. J Clin Invest 1996; 98:1959-65 [4] Cournand A. Air and blood. In: Fishman AP, Richards DW, eds. Circulation of the blood: men and ideas. Bethesda, MD: American Physiological Society, 1982; 3-70 [5] Szidon JP, Fishman AP. Autonomic control of the pulmonary circulation. In: Fishman AP, Hecht HH, eds. The pulmonary circulation and interstitial space. Chicago: University of Chicago, 1969; 239-65(*) From the Department of Rehabilitation Medicine, University of Pennsylvania, Philadelphia.
Correspondence to: Alfred P. Fishman, MD, FCCP, University of Pennsylvania School of Medicine, 1319 Blockley Hall, 418 Guardian Dr, Philadelphia, PA 19104-6021
COPYRIGHT 1998 American College of Chest Physicians
COPYRIGHT 2000 Gale Group