Observations on the Pathogenesis and Management of Pulmonary Hypertension William H. Muller, Jr, MD, Charlottesville,
Virginia
This paper was p .esented at a special program entitled “The Blalock Heritage in American Surgery.” The material was selected because of the encouragement that Doctor Blalock provided for its undertaking. It is now published in this Festschrift to honor a friend of long standing who has contributed so much to the profession and to society.
In 1939 Doctors A.S. Levy and Alfred Blalock [I] published a paper entitled “Experimental Observations on the Effects of Connecting by Suture the Left Main Pulmonary Artery to the Systemic Circulation,” detailing efforts to produce in experimental animals pulmonary hypertension and vascular lesions similar to those seen in certain patients with congenital heart disease. They pointed out that in certain pathologic conditions in man, blood enters the pulmonary circulation directly from systemic vessels. Included among these conditions were patent ductus arteriosus, acquired fistulas from the rupture of an aortic aneurysm into the pulmonary artery, and aberrant pulmonary arteries arising from the aorta or other large systemic arterial trunks. Noting that in all of these conditions pulmonary artery blood pressure is considerably higher than normal, they attempted to simulate the condition observed in human patients by anastomosing the end of the left subclavian artery to the end of the pulmonary artery in twelve dogs, five of which survived (Figure 1.) The animals were sacrificed at intervals, the longest of these being approximately five months, and no pathologic changes were found in the left lungs. In addition, physiologic studies revealed no alterations of gaseous diffusion in these lungs. Blood pressure in the pulmonary artery immediately distal to the From the Department of Surgery, University of Virginia School of Medicine, Charlottesville, Virginia. Reprint requests should be addressed to William H. Mutlcr, Jr, MD, Department Of Surgery, University of Virginia School of Medicine, Charlottesville, Virginia 22901. Presented at the Blalock Heritage in American Surgery Program, Baylor University Medical Center, Dallas, Texas, March 12, 1976.
302
anastomosis was approximately half that of the systemic pressure, and it is probably for this reason that pathologic changes did not develop in the time the dogs were observed. Two animals were alive at the time of the report, but no subsequent mention is made as to whether or not pulmonary vascular changes occurred in them. Had all of these animals been sacrificed at a later date, it is possible that some degree of pulmonary vascular change in the small muscular pulmonary arteries might have been observed. Although this experiment was not fruitful, a modification of the procedure was later used for the treatment of the tetralogy of Fallot. Experimental Model and Results
Because the degree and duration of pulmonary hypertension seemed a significant determinant of operability in certain types of cardiac diseases and malformations, this virtually unexplored area seemed an important one for investigation. We therefore
Figure 1. Anastomosis between the end of the left subclavian and the end of the left pulmonary arteries.
The American Journal of Surgery
PulmonaryHypertension
Figure 2. Anastomosis between the distal end of the left puhnonary artery and the side of the aorta,
attempted to produce an experimental model in which progressively increasing pulmonary resistance might be developed secondary to high pressure, or flow, or both, in the pulmonary vascular bed. It seemed likely that the resistance imposed by the length of the subclavian artery and constriction at the end-to-end anastomosis limited the level of pressure and flow in the experiments of Levy and Blalock [I]. Therefore, in an attempt to obviate this possible limiting factor in our initial experiment, the end of the left main pulmonary artery was anastomosed to the side of the aorta in twenty-one dogs, creating as large an opening as possible. (Figure 2.) The pressures measured in the aorta and the pulmonary artery before and after the anastomosis indicated that the postoperative pulmonary artery pressure was markedly elevated and that, in fact, the diastolic pressure was equal to that in the aorta. The low resistance in the pulmonary vascular bed did not allow systolic pressures to become equal. The mortality rate was extremely high; only five of the animals survived and then only because of a vigorous regimen of digitalis and diuretics. The others died of high output cardiac failure, similar to that seen in infants with large left-to-right shunts at the ventricular level. In the surviving animals, left lung biopsies after one year showed significant wall thickening and luminal narrowing in the smaller pulmonary arteries. Through a more closely controlled experiment we wished to observe the sequential nature of the development of these pulmonary vascular changes, make them more severe, and produce an experimental model with a lower mortality rate. Therefore,
Volume 135. March 1979
Fipre 3. Small muscular puknonaty artery of left upper lobe one week after anastomosis between left wbclavian and left upper lobe arteries. ihere is perflntknal henwrrhage and hemorrhage within the wall of this vessel. There are also circumferential layers of leukocytes, round cells, and red blood cells.
the pulmonary artery of the left upper lobe only was anastomosed to the end of the left subclavian artery, thus involving a relatively small vascular bed and producing a high flow and pressure. These animals did not develop high output failure, as had the first group. The progression of vascular changes followed a distinctive pattern. Biopsy from the left upper lobe, immediately after the anastomosis was opened, showed an engorged dilated vascular bed and areas of intraalveolar hemorrhage. Tissue examined microscopically after one week showed significant dilatation and engorgement and tortuosity of the vascular tree. There was widespread periintimal hemorrhage, and areas of hemorrhage were noted about and within the wall of the small pulmonary arteries, as demonstrated in Figure 3. Circumferential layers of leukocytes, round cells, and red blood cells surrounded other vessels, often invading the adventitia. Sections obtained after three weeks showed a decrease in intraalveolar hemorrhage and edema. Red blood cells previously seen in the adventitia and perivascular spaces of small muscular arteries were largely replaced by round cells and a red-staining collagen-like material. At six weeks, there was gross evidence of medial hypertrophy accompanied by a further increase in perivascular collagen and marked fragmentation of the elastica interna. Within eight weeks, nearly all of the artery was involved, as demonstrated in Figure 4. Severe pulmonary vascular disease was evident at the end of eleven weeks. Histologic examination showed greatly thickened vessels
303
Muller
Figure 4. Small pulmonary artery of left upper lobe after eight weeks, showing marked obstructive intimai fibrosis.
Figure 5. Small pulmonary artery of left upper lobe after eleven weeks, showing marked thickening of the arterial wail, almost obstructing the lumen.
surrounded by vascular, thin walled alveoli, and there was a decrease in the total number of capillaries, as indicated in Figure 5. Excessive medial hypertrophy and intimal proliferation in many of the vessels, including these, were sufficient to cause complete obstruction of the lumens. In other vessels, the micro-
scopic anatomy was partially or totally destroyed. Changes in peripheral areas of the lobe were less rapid than near the hilum and, at three to five weeks, appeared comparable to those seen proximally at one week. An injection cast of the upper lobe and remaining
Figure 6. injection cast of left upper lobe and left lower lobe after subciavian artery-left upper lobe artery anastomosis. The let? upper lobe is devoid of small vessels except for small segment not subjected to systemic arterial pressures. The lower lobe is normal.
The American Journal of Surgery
Pulmonary
ELEVATED
PULMONARY
ARTERY
Hypertension
PRESSURE
‘6 r&2 % .& ‘.-3 5 Vessel
Wall
Wall
Acute
4 Injury
Artktir
Elevated
Vessel
Stretch
Intimal
Vorc.
Activated
Increased
Prolif.
Resistance w-----
Pulmonary
; Prewar
Artery
Stretch
Receptors
; Varomotor
Medial
I Active
Wall
Tone
4 Hypertrophy
/ Y Hypertension
Figure 7. Diagram showing possible explanation of the sequence of events which occurred in this model as compared to the naturally occurring sequence of events in patients who have congenital heart disease and pulmonary hypertension.
normal lobe at approximately eleven weeks shows advanced changes in the entire pulmonary vascular tree. At eleven weeks the upper lobe was almost devoid of small vessels. (Figure 6.) Comments A possible explanation of the sequence of events which occurred in this model, as compared to the naturally occurring sequence of events in patients, is demonstrated in Figure 7. Although it is virtually impossible to obtain the patient data to support this, experimental and clinical observations indicate that the similarity of the two courses is plausible. One of the most important questions related to the clinical implication of these pathologic changes is whether or not they are reversible. If regression oc-
Figure 8. Small muscular pulmonary artery of normal infants. Left, from newborn, showing fetaiized artery with thick wail and narrow lumen. Right, at six months, showing relatively thin wail and large lumen.
curs after the systemic stimulus is removed, closure of the ventricular defect with a high pulmonary resistance might be indicated. Studies in which the systemic stimulus was withdrawn by reanastomosing the distal pulmonary artery to the proximal pulmonary artery showed that minimal to moderate changes would cause a reversion to relatively normal-appearing arteries. When advanced lesions had developed, however, reconstitution of the pulmonary circulation resulted in no change. Although the vascular changes produced in this manner appeared similar to those in patients with certain types of congenital heart disease, the pathogenesis is not entirely compatible. The end result, however, is essentially the same.
texture of the vascular tree. There are numerous small vessels and capiiiaries, or so-called arborization.
Volume 135, March 1979
305
Muller
Ftgure 10. Hemodynamic relationships in a small infant wit h congenital heart disease and a large left-to-right shunt. There is a low pulmonary resistance and a high pulmonary blood f/o w.
Figure 17. Hemodynamic relationships in an infant with congenital heart disease and a large ventricular septal defect. The pulmonary resistance has increased and there is a balanced shunt.
were fortunate enough to survive for a period of several months, the high output failure gradually subsided and lung biopsies showed that pulmonary vascular changes had developed which narrowed the lumens and reduced blood flow. These changes ultimately progressed to an irreversible status. In an effort to better understand the development of pulmonary hemodynamic and pathologic alterations in certain human cardiac malformations, one should consider first the normal course of events in the fetal and postnatal lung. Fetal and Postnatal Lung Development
i Figure 12. Hemodynamic relationships in an older child with congenital heart disease and a large ventricular septai defect. Shunting has reversed to a predominant/y rightto-left direction. The pulmonary blood flow is lesser than the systemic blood flow and the pulmonary resistance is high.
During this period, the pump oxygenator had not yet been developed for clinical use, and we were concerned about devising some means to control the high output cardiac failure in small infants with large left-to-right intracardiac shunts, because a high percentage of the uncontrollable group died. If they
306
In the partially collapsed unventilated lungs of the fetus, the muscular pulmonary arteries are hypertrophied, constricted and coiled, and supported by alveolar amniotic fluid, so that they offer a marked resistance to blood flow. After birth, extravascular pressures actually are reduced when the amniotic fluid is replaced with air. Expansion of the lungs partially straightens the tortuous vessels and the effective transmural pressure sharply increases, leading to the rapid dilatation of patent vessels and the opening of those heretofore closed. Both pulmonary resistance and right ventricular pressure, therefore, rapidly diminish as the decrease in pressure exerted on the walls leads to a decrease in vasomotor tone and eventually to dilated thin walled blood vessels. Thus, at birth the pulmonary vascular bed changes from a high pressure, high resistance system with a relatively small cross-sectional area to
The American Journal of Surgery
Pulmonary Hypertension
Figure 13. Roentgenogram of radiopaque injecth and an injection cast of the King of a four year oM chikl with a ventricular septai defect. There is loss of pulmonary vascuiar markings on the x-ray film and a diminished number of capiiiaries and small arteries in the iung cast.
Figure 14. Roentgenogram of radiopaque injection and injectbn cast of the hrng of an eight year oki chiki with a ventricular septai defect, showing marked loss of vascuiarity on the x-ray fiim and few small arteries and virtually no capillaries in the iung cast.
Figure 15. Roentgenogram of radiopaque injection and injection cast from the iung of a twelve year old child with ventricular septai defect. There are virtuaiiy no small vessels demonstrated.
Volunm 135, March 1978
low pressure, low resistance system with a large cross-sectional area. The small muscular pulmonary arteries in the newborn have relatively thick walls and small lumens, but at approximately six months, the walls are thin, the lumens large, and the resistance to blood flow low. (Figure 8.) The increase in vessel size after birth alters the ability to withstand the effect of continuous high pressure. Because the tension required in the vessel wall to prevent further dilatation is equal to the intravascular pressure times the radius, it is apparent that the larger the diameter, the stronger the walls must be, because now the vessel is more susceptible to increased pressure. Thus, the pulmonary arteries are less able to withstand stress after birth, not only because the muscular wall is thin and the extravascular pressure reduced, but because the diameter is larger. Radiopaque injection and a polyvinyl cast of the lung of the normal infant show the delicate texture of the lung with numerous small vessels and capillaries, or so-called arborization, as shown in Figure 9. a
Causes of Complications
The sequence of events just described does not occur if a significant left-to-right intracardiac shunt is present at birth. The normal medial thickening persists and progresses, and the high blood flow and pressure stretch the vascular bed maximally. This high flow and pressure appear to be the most prominent factors in producing pulmonary hypertension and the progressive, obstructed, associated vascular lesions. Congestive heart failure may appear subsequently because of the excessively high cardiac out-
Figure 16. Cross section of a small pulmonary artery from the lung of a twelve year old child with a large ventricular septal defect, showing marked obstructive medial hypertrophy and intimal fibrosis.
308
put and great workload placed on the heart. The resistance to blood flow is low and therefore a large amount of left to right shunting occurs through the septal defect, greatly increasing the blood flow through the pulmonary circulation, as indicated in Figure 10. The compensatory increase in vasomotor tone and progressive medial hypertrophy reduce the shunt volume, and the patient improves clinically. A physiologic balance has been achieved and may be maintained for an indeterminate period, as indicated in Figure 11. Eventually, additional vascular changes increase the pulmonary resistance enough to reverse the shunting, and the patient becomes cyanotic and dyspneic, develops frequent respiratory infections, and finally expires. The shunt flow has reversed, so there is now a lesser total blood flow through the pulmonary than through the systemic vascular bed, and there is a greater volume of unsaturated blood. (Figure 12.) Roentgenograms of opaque media injections and pulmonary casts show what happens in patients with large ventricular septal defects. Figure 13 shows the lung of a four year old child; there is a loss of pulmonary vascular markings on the x-ray film and a markedly diminished arborization in the injection cast. Figure 14 shows the lung of an older child with a very large ventricular septal defect showing more advanced changes. Finally, Figure 15 shows the lung of a twelve year old child with a similar lesion. Note that small pulmonary arteries are virtually nonexistent. A cross section of one of his pulmonary arteries (Figure 16) shows both marked obstructive medial hypertrophy and intimal fibrosis. It is well known that such changes develop far later in patients who have supraventricular shunts because of the relatively low pressure and the buffering effect of the atrioventricular valves. Figure 17 shows the lung from an older adult patient with an intraatrial septal defect. The pulmonary artery itself is markedly dilated; however, the lung is not as devoid of pulmonary vascular markings in the periphery as it is in the twelve year old patient with ventricular septal defect, and the same can be said of the cast. The degree to which these changes occur is quite variable and is influenced by a number of factors, especially the size and position of the defect, which determine initially the volume of the shunt. We studied a series of patients with patent ductus arteriosus, single ventricle, and a large ventricular septal defect, and categorized them into three phases according to the severity of their clinical symptoms. The ratio of the wall thickness to the size of the lumen was plotted according to age. Most of the
The American Journal of Surgery
Pulmonary Hypertension
younger patients fell into phase I, and their lumenwall ratios were within normal limits. The older ones fell into the more severe clinical phases and were below the limits of normal. Management
In 1949 Civin and Edwards [2] reported on a patient, who died in his late 40s who had had a large ventricular septal defect accompanied by pulmonary stenosis to such a degree that the volume of the shunt was limited and the two circulations were balanced. He had no symptoms referable to his heart and died of other causes. The authors suggested that in certain instances of ventricular septal defect the creation of pulmonary stenosis might serve to control high output failure and, in addition, serve as a protective mechanism for the vascular bed against the ravages of high flow and pressure. Partial obstruction of the outflow from the right ventricle did appear to be a logical means of managing the hemodynamic alterations and the pulmonary vascular bed in patients with subvalvular left-to-right shunts producing pulmonary hypertension. It seemed to us that narrowing the pulmonary artery might be an easier, more controllable, and more effective means of producing such an obstruction than attempting to narrow the valves or the infundibulum of the right ventricle, and we devised a method of producing pulmonary artery stenosis in the laboratory. Holman and Beck [3] had indicated that bands, particularly rigid bands, placed about the pulmonary artery would erode through it in a certain percentage of instances, and for this reason we initially removed
a segment of the pulmonary artery before placing the band. We then applied the procedure [4] to an extremely small, emaciated three month old infant in persistent high output failure, who was thought to have a large ventricular septal defect but who was later found to have an atrioventricular canal. At operation (Figure 18), the pulmonary artery was several times the size of the aorta; a segment of the artery was removed (Figure 19), and a band was placed about it (Figure 20). The patient improved considerably, but not to the extent that we had hoped, and subsequent catheterization studies indicated a continued high shunt and pulmonary blood flow. We therefore reoperated approximately two and a half years later, increasing the stenosis; the patient showed considerable improvement in that his growth rate increased rapidly and his weight became normal. This operation was not used widely until the advent of the pump oxygenator and the realization that infants did not tolerate perfusion well It was then used with increasing frequency in infants as a palliative procedure until they reached an age at which total bypass presented a lower risk. Today, however, the use of refined perfusion alone in infants and particularly the widespread application of profound hypothermia and cardiac arrest has caused the operation to be used only infrequently. It is indicated currently as a palliative operation for certain otherwise irreparable intra- and extracardiac defects, when right-sided cardiac output is excessively high. That protection of the pulmonary vascular bed and regression of the pulmonary vascular changes do indeed occur after the banding procedures is sup-
F&we 17. Injection cast and roentgenogram of radiopaque injection of tfte lung from an older adult patient with an intraatrial septal defect. The pulmonary arteryis markedlydilated, but a moderate number of small arteries and capillaries is demonstrated.
Volume 135, March 1978
309
Muller
Figure 18. Operative findings in a three month old infant 4th a large atrkventricular canal and kft-to-rkht shunt. The puimonary artery is markedly larger than the aorta.
Figure 19. A clamp is applied to the large pulmonary artery and a segment ts excked. The arterial defect k closed with a conttnuous suture.
Figure 20. A band has been placed about tfte aorta at the point where the arterial segment was excised. The dktat pulmonary artery pressure is approximately half that of the proximal pressure.
310
The American Journal of Surgery
Pulmonary
Hypertension
Figure 21. Comparison of arteries before and six years after banding in the first patienf upon whom the bandlng procedure was performed, demonstrating the protective effect of the banding operation. The artery on the right has a much thinner wall and larger lumen than that on the left.
ported by a comparison of lung biopsies taken at the time of definitive surgery with those taken at the time of the banding procedure. Figure 21 shows a comparison of an artery at the second operation in our first patient with one obtained during operation for definitive repair six years later; in this and another comparison of arteries taken in a similar situation from a child with a large ventricular septal defect, the protective effect of the previously placed band is demonstrated by the thin walled pulmonary artery at the time of definitive repair. Numerous other similar comparisons of pre- and postbanding lung biopsies confirm this protective effect. Summary
Experimental subclavian-pulmonary and aortapulmonary anastomoses were performed in attempts to produce progressively increasing pulmonary vascular changes and to observe their sequential development. At six weeks after anastomosis there was medial hypertrophy, at eight weeks total involvement of the artery, and at eleven weeks severe pulmonary vascular disease, total decrease in the number of capillaries, and complete obstruction of the lumen.
Volume 135. March 1976
These pathologic changes are compared with the normal course of events in the fetal and postnatal lung. A series of patients with patent ductus arteriosus, single ventricle, and a large ventricular septal defect were studied and categorized into three phases according to the severity of their clinical symptoms. A method for producing pulmonary artery stenosis in the laboratory as applied to a three month old infant is discussed.
References 1. Levy SE, Blalock A: Experimental observations on the effects of connecting by suture the left main pulmonary artery to the systemic circulation. J Thorac Surg 8: 525, 1939. 2. Civin WH, Edwards JE: Pathology of the pulmonary vascular tree: a comparison of the intrapulmonary arteries in Eisenmenger’s complex and in stenosis of ostium infundibuliassociated with biventricular origin of the aorta. Circulation 2: 545, 1950. 3. Holman E, Beck CS: The physiologic response of the circulatory system to experimental alterations: effect of aortic and pulmanic stenosis. J C/in Invest 3: 283. 1926. 4. Muller WH Jr, Dammann JR Jr: The treatment of certain congenital malformations of the heart by the creation of pulmonic stenosis to reduce pulmonary hypertension and excessive pulmonary blood flow: a preliminary report. Surg Gynecol Obstet 95: 213, 1952.
311
Virginia
This paper was p .esented at a special program entitled “The Blalock Heritage in American Surgery.” The material was selected because of the encouragement that Doctor Blalock provided for its undertaking. It is now published in this Festschrift to honor a friend of long standing who has contributed so much to the profession and to society.
In 1939 Doctors A.S. Levy and Alfred Blalock [I] published a paper entitled “Experimental Observations on the Effects of Connecting by Suture the Left Main Pulmonary Artery to the Systemic Circulation,” detailing efforts to produce in experimental animals pulmonary hypertension and vascular lesions similar to those seen in certain patients with congenital heart disease. They pointed out that in certain pathologic conditions in man, blood enters the pulmonary circulation directly from systemic vessels. Included among these conditions were patent ductus arteriosus, acquired fistulas from the rupture of an aortic aneurysm into the pulmonary artery, and aberrant pulmonary arteries arising from the aorta or other large systemic arterial trunks. Noting that in all of these conditions pulmonary artery blood pressure is considerably higher than normal, they attempted to simulate the condition observed in human patients by anastomosing the end of the left subclavian artery to the end of the pulmonary artery in twelve dogs, five of which survived (Figure 1.) The animals were sacrificed at intervals, the longest of these being approximately five months, and no pathologic changes were found in the left lungs. In addition, physiologic studies revealed no alterations of gaseous diffusion in these lungs. Blood pressure in the pulmonary artery immediately distal to the From the Department of Surgery, University of Virginia School of Medicine, Charlottesville, Virginia. Reprint requests should be addressed to William H. Mutlcr, Jr, MD, Department Of Surgery, University of Virginia School of Medicine, Charlottesville, Virginia 22901. Presented at the Blalock Heritage in American Surgery Program, Baylor University Medical Center, Dallas, Texas, March 12, 1976.
302
anastomosis was approximately half that of the systemic pressure, and it is probably for this reason that pathologic changes did not develop in the time the dogs were observed. Two animals were alive at the time of the report, but no subsequent mention is made as to whether or not pulmonary vascular changes occurred in them. Had all of these animals been sacrificed at a later date, it is possible that some degree of pulmonary vascular change in the small muscular pulmonary arteries might have been observed. Although this experiment was not fruitful, a modification of the procedure was later used for the treatment of the tetralogy of Fallot. Experimental Model and Results
Because the degree and duration of pulmonary hypertension seemed a significant determinant of operability in certain types of cardiac diseases and malformations, this virtually unexplored area seemed an important one for investigation. We therefore
Figure 1. Anastomosis between the end of the left subclavian and the end of the left pulmonary arteries.
The American Journal of Surgery
PulmonaryHypertension
Figure 2. Anastomosis between the distal end of the left puhnonary artery and the side of the aorta,
attempted to produce an experimental model in which progressively increasing pulmonary resistance might be developed secondary to high pressure, or flow, or both, in the pulmonary vascular bed. It seemed likely that the resistance imposed by the length of the subclavian artery and constriction at the end-to-end anastomosis limited the level of pressure and flow in the experiments of Levy and Blalock [I]. Therefore, in an attempt to obviate this possible limiting factor in our initial experiment, the end of the left main pulmonary artery was anastomosed to the side of the aorta in twenty-one dogs, creating as large an opening as possible. (Figure 2.) The pressures measured in the aorta and the pulmonary artery before and after the anastomosis indicated that the postoperative pulmonary artery pressure was markedly elevated and that, in fact, the diastolic pressure was equal to that in the aorta. The low resistance in the pulmonary vascular bed did not allow systolic pressures to become equal. The mortality rate was extremely high; only five of the animals survived and then only because of a vigorous regimen of digitalis and diuretics. The others died of high output cardiac failure, similar to that seen in infants with large left-to-right shunts at the ventricular level. In the surviving animals, left lung biopsies after one year showed significant wall thickening and luminal narrowing in the smaller pulmonary arteries. Through a more closely controlled experiment we wished to observe the sequential nature of the development of these pulmonary vascular changes, make them more severe, and produce an experimental model with a lower mortality rate. Therefore,
Volume 135. March 1979
Fipre 3. Small muscular puknonaty artery of left upper lobe one week after anastomosis between left wbclavian and left upper lobe arteries. ihere is perflntknal henwrrhage and hemorrhage within the wall of this vessel. There are also circumferential layers of leukocytes, round cells, and red blood cells.
the pulmonary artery of the left upper lobe only was anastomosed to the end of the left subclavian artery, thus involving a relatively small vascular bed and producing a high flow and pressure. These animals did not develop high output failure, as had the first group. The progression of vascular changes followed a distinctive pattern. Biopsy from the left upper lobe, immediately after the anastomosis was opened, showed an engorged dilated vascular bed and areas of intraalveolar hemorrhage. Tissue examined microscopically after one week showed significant dilatation and engorgement and tortuosity of the vascular tree. There was widespread periintimal hemorrhage, and areas of hemorrhage were noted about and within the wall of the small pulmonary arteries, as demonstrated in Figure 3. Circumferential layers of leukocytes, round cells, and red blood cells surrounded other vessels, often invading the adventitia. Sections obtained after three weeks showed a decrease in intraalveolar hemorrhage and edema. Red blood cells previously seen in the adventitia and perivascular spaces of small muscular arteries were largely replaced by round cells and a red-staining collagen-like material. At six weeks, there was gross evidence of medial hypertrophy accompanied by a further increase in perivascular collagen and marked fragmentation of the elastica interna. Within eight weeks, nearly all of the artery was involved, as demonstrated in Figure 4. Severe pulmonary vascular disease was evident at the end of eleven weeks. Histologic examination showed greatly thickened vessels
303
Muller
Figure 4. Small pulmonary artery of left upper lobe after eight weeks, showing marked obstructive intimai fibrosis.
Figure 5. Small pulmonary artery of left upper lobe after eleven weeks, showing marked thickening of the arterial wail, almost obstructing the lumen.
surrounded by vascular, thin walled alveoli, and there was a decrease in the total number of capillaries, as indicated in Figure 5. Excessive medial hypertrophy and intimal proliferation in many of the vessels, including these, were sufficient to cause complete obstruction of the lumens. In other vessels, the micro-
scopic anatomy was partially or totally destroyed. Changes in peripheral areas of the lobe were less rapid than near the hilum and, at three to five weeks, appeared comparable to those seen proximally at one week. An injection cast of the upper lobe and remaining
Figure 6. injection cast of left upper lobe and left lower lobe after subciavian artery-left upper lobe artery anastomosis. The let? upper lobe is devoid of small vessels except for small segment not subjected to systemic arterial pressures. The lower lobe is normal.
The American Journal of Surgery
Pulmonary
ELEVATED
PULMONARY
ARTERY
Hypertension
PRESSURE
‘6 r&2 % .& ‘.-3 5 Vessel
Wall
Wall
Acute
4 Injury
Artktir
Elevated
Vessel
Stretch
Intimal
Vorc.
Activated
Increased
Prolif.
Resistance w-----
Pulmonary
; Prewar
Artery
Stretch
Receptors
; Varomotor
Medial
I Active
Wall
Tone
4 Hypertrophy
/ Y Hypertension
Figure 7. Diagram showing possible explanation of the sequence of events which occurred in this model as compared to the naturally occurring sequence of events in patients who have congenital heart disease and pulmonary hypertension.
normal lobe at approximately eleven weeks shows advanced changes in the entire pulmonary vascular tree. At eleven weeks the upper lobe was almost devoid of small vessels. (Figure 6.) Comments A possible explanation of the sequence of events which occurred in this model, as compared to the naturally occurring sequence of events in patients, is demonstrated in Figure 7. Although it is virtually impossible to obtain the patient data to support this, experimental and clinical observations indicate that the similarity of the two courses is plausible. One of the most important questions related to the clinical implication of these pathologic changes is whether or not they are reversible. If regression oc-
Figure 8. Small muscular pulmonary artery of normal infants. Left, from newborn, showing fetaiized artery with thick wail and narrow lumen. Right, at six months, showing relatively thin wail and large lumen.
curs after the systemic stimulus is removed, closure of the ventricular defect with a high pulmonary resistance might be indicated. Studies in which the systemic stimulus was withdrawn by reanastomosing the distal pulmonary artery to the proximal pulmonary artery showed that minimal to moderate changes would cause a reversion to relatively normal-appearing arteries. When advanced lesions had developed, however, reconstitution of the pulmonary circulation resulted in no change. Although the vascular changes produced in this manner appeared similar to those in patients with certain types of congenital heart disease, the pathogenesis is not entirely compatible. The end result, however, is essentially the same.
texture of the vascular tree. There are numerous small vessels and capiiiaries, or so-called arborization.
Volume 135, March 1979
305
Muller
Ftgure 10. Hemodynamic relationships in a small infant wit h congenital heart disease and a large left-to-right shunt. There is a low pulmonary resistance and a high pulmonary blood f/o w.
Figure 17. Hemodynamic relationships in an infant with congenital heart disease and a large ventricular septal defect. The pulmonary resistance has increased and there is a balanced shunt.
were fortunate enough to survive for a period of several months, the high output failure gradually subsided and lung biopsies showed that pulmonary vascular changes had developed which narrowed the lumens and reduced blood flow. These changes ultimately progressed to an irreversible status. In an effort to better understand the development of pulmonary hemodynamic and pathologic alterations in certain human cardiac malformations, one should consider first the normal course of events in the fetal and postnatal lung. Fetal and Postnatal Lung Development
i Figure 12. Hemodynamic relationships in an older child with congenital heart disease and a large ventricular septai defect. Shunting has reversed to a predominant/y rightto-left direction. The pulmonary blood flow is lesser than the systemic blood flow and the pulmonary resistance is high.
During this period, the pump oxygenator had not yet been developed for clinical use, and we were concerned about devising some means to control the high output cardiac failure in small infants with large left-to-right intracardiac shunts, because a high percentage of the uncontrollable group died. If they
306
In the partially collapsed unventilated lungs of the fetus, the muscular pulmonary arteries are hypertrophied, constricted and coiled, and supported by alveolar amniotic fluid, so that they offer a marked resistance to blood flow. After birth, extravascular pressures actually are reduced when the amniotic fluid is replaced with air. Expansion of the lungs partially straightens the tortuous vessels and the effective transmural pressure sharply increases, leading to the rapid dilatation of patent vessels and the opening of those heretofore closed. Both pulmonary resistance and right ventricular pressure, therefore, rapidly diminish as the decrease in pressure exerted on the walls leads to a decrease in vasomotor tone and eventually to dilated thin walled blood vessels. Thus, at birth the pulmonary vascular bed changes from a high pressure, high resistance system with a relatively small cross-sectional area to
The American Journal of Surgery
Pulmonary Hypertension
Figure 13. Roentgenogram of radiopaque injecth and an injection cast of the King of a four year oM chikl with a ventricular septai defect. There is loss of pulmonary vascuiar markings on the x-ray film and a diminished number of capiiiaries and small arteries in the iung cast.
Figure 14. Roentgenogram of radiopaque injection and injectbn cast of the hrng of an eight year oki chiki with a ventricular septai defect, showing marked loss of vascuiarity on the x-ray fiim and few small arteries and virtually no capillaries in the iung cast.
Figure 15. Roentgenogram of radiopaque injection and injection cast from the iung of a twelve year old child with ventricular septai defect. There are virtuaiiy no small vessels demonstrated.
Volunm 135, March 1978
low pressure, low resistance system with a large cross-sectional area. The small muscular pulmonary arteries in the newborn have relatively thick walls and small lumens, but at approximately six months, the walls are thin, the lumens large, and the resistance to blood flow low. (Figure 8.) The increase in vessel size after birth alters the ability to withstand the effect of continuous high pressure. Because the tension required in the vessel wall to prevent further dilatation is equal to the intravascular pressure times the radius, it is apparent that the larger the diameter, the stronger the walls must be, because now the vessel is more susceptible to increased pressure. Thus, the pulmonary arteries are less able to withstand stress after birth, not only because the muscular wall is thin and the extravascular pressure reduced, but because the diameter is larger. Radiopaque injection and a polyvinyl cast of the lung of the normal infant show the delicate texture of the lung with numerous small vessels and capillaries, or so-called arborization, as shown in Figure 9. a
Causes of Complications
The sequence of events just described does not occur if a significant left-to-right intracardiac shunt is present at birth. The normal medial thickening persists and progresses, and the high blood flow and pressure stretch the vascular bed maximally. This high flow and pressure appear to be the most prominent factors in producing pulmonary hypertension and the progressive, obstructed, associated vascular lesions. Congestive heart failure may appear subsequently because of the excessively high cardiac out-
Figure 16. Cross section of a small pulmonary artery from the lung of a twelve year old child with a large ventricular septal defect, showing marked obstructive medial hypertrophy and intimal fibrosis.
308
put and great workload placed on the heart. The resistance to blood flow is low and therefore a large amount of left to right shunting occurs through the septal defect, greatly increasing the blood flow through the pulmonary circulation, as indicated in Figure 10. The compensatory increase in vasomotor tone and progressive medial hypertrophy reduce the shunt volume, and the patient improves clinically. A physiologic balance has been achieved and may be maintained for an indeterminate period, as indicated in Figure 11. Eventually, additional vascular changes increase the pulmonary resistance enough to reverse the shunting, and the patient becomes cyanotic and dyspneic, develops frequent respiratory infections, and finally expires. The shunt flow has reversed, so there is now a lesser total blood flow through the pulmonary than through the systemic vascular bed, and there is a greater volume of unsaturated blood. (Figure 12.) Roentgenograms of opaque media injections and pulmonary casts show what happens in patients with large ventricular septal defects. Figure 13 shows the lung of a four year old child; there is a loss of pulmonary vascular markings on the x-ray film and a markedly diminished arborization in the injection cast. Figure 14 shows the lung of an older child with a very large ventricular septal defect showing more advanced changes. Finally, Figure 15 shows the lung of a twelve year old child with a similar lesion. Note that small pulmonary arteries are virtually nonexistent. A cross section of one of his pulmonary arteries (Figure 16) shows both marked obstructive medial hypertrophy and intimal fibrosis. It is well known that such changes develop far later in patients who have supraventricular shunts because of the relatively low pressure and the buffering effect of the atrioventricular valves. Figure 17 shows the lung from an older adult patient with an intraatrial septal defect. The pulmonary artery itself is markedly dilated; however, the lung is not as devoid of pulmonary vascular markings in the periphery as it is in the twelve year old patient with ventricular septal defect, and the same can be said of the cast. The degree to which these changes occur is quite variable and is influenced by a number of factors, especially the size and position of the defect, which determine initially the volume of the shunt. We studied a series of patients with patent ductus arteriosus, single ventricle, and a large ventricular septal defect, and categorized them into three phases according to the severity of their clinical symptoms. The ratio of the wall thickness to the size of the lumen was plotted according to age. Most of the
The American Journal of Surgery
Pulmonary Hypertension
younger patients fell into phase I, and their lumenwall ratios were within normal limits. The older ones fell into the more severe clinical phases and were below the limits of normal. Management
In 1949 Civin and Edwards [2] reported on a patient, who died in his late 40s who had had a large ventricular septal defect accompanied by pulmonary stenosis to such a degree that the volume of the shunt was limited and the two circulations were balanced. He had no symptoms referable to his heart and died of other causes. The authors suggested that in certain instances of ventricular septal defect the creation of pulmonary stenosis might serve to control high output failure and, in addition, serve as a protective mechanism for the vascular bed against the ravages of high flow and pressure. Partial obstruction of the outflow from the right ventricle did appear to be a logical means of managing the hemodynamic alterations and the pulmonary vascular bed in patients with subvalvular left-to-right shunts producing pulmonary hypertension. It seemed to us that narrowing the pulmonary artery might be an easier, more controllable, and more effective means of producing such an obstruction than attempting to narrow the valves or the infundibulum of the right ventricle, and we devised a method of producing pulmonary artery stenosis in the laboratory. Holman and Beck [3] had indicated that bands, particularly rigid bands, placed about the pulmonary artery would erode through it in a certain percentage of instances, and for this reason we initially removed
a segment of the pulmonary artery before placing the band. We then applied the procedure [4] to an extremely small, emaciated three month old infant in persistent high output failure, who was thought to have a large ventricular septal defect but who was later found to have an atrioventricular canal. At operation (Figure 18), the pulmonary artery was several times the size of the aorta; a segment of the artery was removed (Figure 19), and a band was placed about it (Figure 20). The patient improved considerably, but not to the extent that we had hoped, and subsequent catheterization studies indicated a continued high shunt and pulmonary blood flow. We therefore reoperated approximately two and a half years later, increasing the stenosis; the patient showed considerable improvement in that his growth rate increased rapidly and his weight became normal. This operation was not used widely until the advent of the pump oxygenator and the realization that infants did not tolerate perfusion well It was then used with increasing frequency in infants as a palliative procedure until they reached an age at which total bypass presented a lower risk. Today, however, the use of refined perfusion alone in infants and particularly the widespread application of profound hypothermia and cardiac arrest has caused the operation to be used only infrequently. It is indicated currently as a palliative operation for certain otherwise irreparable intra- and extracardiac defects, when right-sided cardiac output is excessively high. That protection of the pulmonary vascular bed and regression of the pulmonary vascular changes do indeed occur after the banding procedures is sup-
F&we 17. Injection cast and roentgenogram of radiopaque injection of tfte lung from an older adult patient with an intraatrial septal defect. The pulmonary arteryis markedlydilated, but a moderate number of small arteries and capillaries is demonstrated.
Volume 135, March 1978
309
Muller
Figure 18. Operative findings in a three month old infant 4th a large atrkventricular canal and kft-to-rkht shunt. The puimonary artery is markedly larger than the aorta.
Figure 19. A clamp is applied to the large pulmonary artery and a segment ts excked. The arterial defect k closed with a conttnuous suture.
Figure 20. A band has been placed about tfte aorta at the point where the arterial segment was excised. The dktat pulmonary artery pressure is approximately half that of the proximal pressure.
310
The American Journal of Surgery
Pulmonary
Hypertension
Figure 21. Comparison of arteries before and six years after banding in the first patienf upon whom the bandlng procedure was performed, demonstrating the protective effect of the banding operation. The artery on the right has a much thinner wall and larger lumen than that on the left.
ported by a comparison of lung biopsies taken at the time of definitive surgery with those taken at the time of the banding procedure. Figure 21 shows a comparison of an artery at the second operation in our first patient with one obtained during operation for definitive repair six years later; in this and another comparison of arteries taken in a similar situation from a child with a large ventricular septal defect, the protective effect of the previously placed band is demonstrated by the thin walled pulmonary artery at the time of definitive repair. Numerous other similar comparisons of pre- and postbanding lung biopsies confirm this protective effect. Summary
Experimental subclavian-pulmonary and aortapulmonary anastomoses were performed in attempts to produce progressively increasing pulmonary vascular changes and to observe their sequential development. At six weeks after anastomosis there was medial hypertrophy, at eight weeks total involvement of the artery, and at eleven weeks severe pulmonary vascular disease, total decrease in the number of capillaries, and complete obstruction of the lumen.
Volume 135. March 1976
These pathologic changes are compared with the normal course of events in the fetal and postnatal lung. A series of patients with patent ductus arteriosus, single ventricle, and a large ventricular septal defect were studied and categorized into three phases according to the severity of their clinical symptoms. A method for producing pulmonary artery stenosis in the laboratory as applied to a three month old infant is discussed.
References 1. Levy SE, Blalock A: Experimental observations on the effects of connecting by suture the left main pulmonary artery to the systemic circulation. J Thorac Surg 8: 525, 1939. 2. Civin WH, Edwards JE: Pathology of the pulmonary vascular tree: a comparison of the intrapulmonary arteries in Eisenmenger’s complex and in stenosis of ostium infundibuliassociated with biventricular origin of the aorta. Circulation 2: 545, 1950. 3. Holman E, Beck CS: The physiologic response of the circulatory system to experimental alterations: effect of aortic and pulmanic stenosis. J C/in Invest 3: 283. 1926. 4. Muller WH Jr, Dammann JR Jr: The treatment of certain congenital malformations of the heart by the creation of pulmonic stenosis to reduce pulmonary hypertension and excessive pulmonary blood flow: a preliminary report. Surg Gynecol Obstet 95: 213, 1952.
311
