| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1997;64:1549-1554
© 1997 The Society of Thoracic Surgeons
Department of Surgery, New York University School of Medicine, New York, New York
Abstract
The development of valvular heart surgery over the past 50 years has required the efforts and creative genius of many surgical pioneers. It has been filled with exhilarating short-term successes and some devastating failures. This article traces the 50 years of persistence and determination that have brought us to a time when the majority of patients with heart valve disease can be returned to a happy and fulfilling life by valvuloplasty or by valve replacement.
I deeply appreciate the opportunity to participate in this symposium with the Muller Surgical Society to celebrate the 40th anniversary of the beginning of open heart surgery at the University of Virginia. The dedication of this symposium to Dr W. H. Muller, Jr, properly honors his creative imagination and leadership in developing the modern sophisticated cardiac center that we see today. I was requested to highlight significant developments in cardiac valvular surgery over the past four decades.
1947–1948
In June 1947, I graduated from Vanderbilt University School of Medicine and came to Johns Hopkins to begin a surgical internship. During that internship year (1947–1948), Charles Bailey in Philadelphia came to a Hopkins weekly surgical conference and presented his initial exciting experiences with digital commissurotomy. Dwight Harkin in Boston had independently done similar work. Few in the audience realized that this was the birth of modern adult cardiac surgery.
Unknown to anyone, many of the house staff and faculty at Johns Hopkins at that time would independently make major contributions to cardiac surgery in the decades ahead. The exciting first successful "blue-baby" operation, the Blalock-Taussig subclavian pulmonary anastomosis, had been performed at Hopkins by Dr Alfred Blalock, the Chairman, on November 29, 1944, only 3 years before [1]. The house staff included myself, Dave Sabiston, and Jim Maloney of Los Angeles in the intern group; Henry Bahnson and Andrew Morrow were junior house staff, and C. R. Hanlon was chief resident part of that year. During the year W. H. Muller, our honoree today, and Denton Cooley each returned to the senior house staff from military service in the Army. The faculty, under the Chairmanship of Alfred Blalock, included William Longmire, William Scott, and Mark Ravitch. The collegial friendships formed at this time, evolving and maturing over the next five decades, produced a close communication of new ideas and experiences. This close communication individually helped everyone in their individual investigations of new areas in cardiac surgery.
1948–1955
Thousands of digital commissurotomies were performed for mitral stenosis. Although recurrent stenosis developed in at least 50% of patients within 5 years, initial results were dramatic. At that time rheumatic mitral stenosis was a common cause of death, especially in the northeastern United States, frequently leading to death in young women during their first pregnancy. Digital commissurotomy was soon found ineffective in some densely scarred valves, so different mechanical dilators were developed. The most successful was developed by Tubbs [2] in England and still remains in limited use today. Some degree of mitral insufficiency developed in at least 15% to 20% of patients operated on.
An aortic valve dilator for aortic stenosis was developed by Bailey, used briefly, and quickly abandoned because of prohibitive mortality caused by either insufficiency or recurrent stenosis.
In retrospect, knowledge was very meager, for the beating heart had not been electively operated on since the brief unsuccessful efforts of Cutler and Levine with mechanical mitral valvotomy in Boston on May 20, 1923 [3]. In these early years, arrhythmias were a much feared event. The first successful defibrillation of the heart was done by Claude Beck in 1947, but effective defibrillators for use in the operating room were not manufactured for several years.
In 1955 the epochal work of Lillehei at the University of Minnesota and John Kirklin at the Mayo Clinic successfully launched the clinical use of the pump oxygenator [4, 5]. The major contribution, of course, was made by John Gibbon, who patiently and persistently worked over 22 years since 1931 to develop a heart-lung machine that finally resulted in the first successful operation in a man in 1953 [6]. The screen oxygenator used by Gibbon had been developed through the generosity of IBM. Gibbon, in turn, generously gave the blueprints for his machine to Kirklin, who then constructed a similar machine at the Mayo Clinic, firmly launching the screen oxygenator for clinical use on a sound physiologic basis. Without Gibbon's incredible investigation over 22 years, this would never have been possible.
Independently, C. W. Lillehei in Minneapolis initially performed open heart operations in small children in 1954 with cross-circulation from an adult [4]. Using this experience, he then first performed an open heart operation with the bubble oxygenator developed by DeWall [7]. For the first time the dream of operating within the human heart became a reality. The magnitude of these achievements is best understood by realizing that a few years earlier, research funding from both the National Institutes of Health and the American Heart Association had been sharply curtailed with the concept that the possibilities of developing a heart-lung machine were too visionary to warrant use of limited research funds!
Following Kirklin's and Lillehei's groundbreaking achievements, Alfred Blalock at Johns Hopkins generously gave the responsibility for developing open heart surgery to Henry Bahnson and two younger members of the faculty, myself and Dave Sabiston.
UCLA: 1949–1951
As this symposium is dedicated to Dr Muller, my association with Dr Muller, a mentor and friend for decades, is briefly described in the following paragraphs.
Cardiac surgery began at UCLA not long after W. P. Longmire came in July 1949 to become Chairman of the new Department of Surgery. W. H. Muller came with him as a member of the junior faculty. I also came to continue residency training. Closed heart surgery was soon performed in different community hospitals by Longmire and Muller. Soon, after a generous donation to the new Department of Surgery, a fledgling cardiovascular laboratory was begun in temporarily constructed frame buildings at UCLA with Muller as Director and myself developing the laboratory. Despite the meager surroundings, the pulmonary artery banding operation was successfully developed by Muller and Dammann [8] within the next few months, and remains a useful surgical procedure today. The unexpected outbreak of the Korean War in the late spring of 1951 abruptly ended my UCLA career as I entered military service with the United States Navy and Marine Corps for the next 2 years, after which I returned to Johns Hopkins. During that time, in 1954, Harry Muller accepted the Chairmanship at the University of Virginia.
With the increasing safety of the new heart-lung machines, "open" mitral operations were performed with increasing frequency, especially when mitral stenosis existed with concurrent insufficiency. The lethal nature of air embolism from air trapping within different cardiac chambers was not well recognized for several years; it undoubtedly was a major cause of the high operative mortality. A number of ingenious procedures were attempted for mitral insufficiency, but all failed with the exception of the posteromedial annuloplasty developed by several independent investigators, especially Merendino and Bruce [9]. The pursestring annuloplasty of Glover enjoyed a brief popularity, but the basic concept proved erroneous. The pathologic changes in many diseased mitral valves were clearly irreparable, so prosthetic replacement was the only solution.
Intense investigation of prosthetic mitral valves began. Within 4 years a large symposium on the development of prosthetic valves was held at the Edgewater Beach Hotel in Chicago in 1958. The proceedings, edited by Garrot Allen, were published in a large book of more than 600 pages [10]. The reports were very discouraging, for all prostheses had failed.
During this time replacement of individual aortic cusps with cusps constructed of Dacron was developed and used briefly by Bahnson, myself, and our associates [11]. Others developed similar prostheses. A high operative mortality due to both inadequate myocardial preservation and the difficulties in obtaining a competent valve by constructing individual aortic cusps precluded their widespread use. The few successful operations failed within a few years because of progressive fibrosis and stiffening of the cloth cusps.
1961: The Starr-Edward Ball-Valve Prosthesis
The landmark report by Starr and Edwards [12] of the first four successful operations in humans firmly launched the era of prosthetic cardiac valves. This development evolved from the fortunate liaison between Starr, a young cardiac surgeon at the University of Oregon, and Edwards, a semiretired mechanical engineer. The ball-valve prosthesis was constructed from concepts used by Hufnagel [13] years earlier with a ball-valve prosthesis placed in the descending thoracic aorta. Significant support to this epochal achievement was given by J. E. Dunphy, the Chairman of the Department of Surgery. This strong support by the Departmental Chairman was especially crucial, because valvular replacements in dogs had a high mortality due to thromboembolism. The cause of the high canine mortality only became apparent years later when the chronic bacteremia present in many dogs was recognized.
1962–1997: Ball-Valve Prostheses
Problems with thromboembolism from thrombus formation on the cloth and metallic surfaces of the ball-valve prostheses unless anticoagulation was used were quickly recognized. With effective permanent anticoagulation, thromboembolism could be controlled to a major degree, but at present, 36 years later, thromboembolism remains the major hazard with metallic prostheses, with a reported frequency of at least 1% to 2% per 100 patient-years. This basic problem has remained for more than three decades because a durable lining of prosthetic valves has not been found that does not initiate the coagulation cascade with prompt thrombus formation.
A serious complication developed with the first silicone ball prostheses from absorption of lipids from body fluids. This absorption produced changes in contour and stability that led to fracture of the silicone poppet within a few years in a significant percentage of patients. Fortunately, changes in the chemical composition of the silicone ball led to the development of a durable silicone poppet by 1966. This remains widely used today, with 30-year follow-up data now published [14].
A number of different poppets were tried to minimize the problem of thromboembolism. A steel ball prosthesis was used for a period of years. The struts of the prosthesis had to be covered with cloth, however, to prevent prohibitive noise. Eventual erosion of the cloth-covered struts from the repeated impact of the steel ball gradually led to removal of the prosthesis from the market. In some patients this did not occur for 10 to 15 years [15].
Another important experience occurred with the Braunwald-Cutter cloth-covered silicone ball prosthesis [16]. This ingenious valve successfully solved the problem of cloth erosion on the struts, but unfortunately the silicone ball gradually eroded from serial impact of the ball on the cloth struts. This progressive erosion of the silicone gradually decreased the diameter of the poppet, resulting within 3 to 4 years in eventual extrusion of the poppet from the cage of the prosthesis, usually with immediate death. This tragic unforeseen development well emphasizes the basic principle that all prosthetic valves are experimental until implanted in the human body for about 5 years. Changes that occur in the human body can only be partly duplicated in a pulse duplicator.
Disc Prostheses (1969–1997)
Minor physiologic disadvantages of ball-valve prostheses were soon recognized. There was an inevitable low-grade but tolerable hemolysis from the serial impact of the ball on the cage. Also, the effective cross-sectional area of the valve orifice was not ideal because a significant percentage of the orifice area was occupied by the poppet. These considerations led to the development of disc prostheses. Among many tried, the most successful was the Björk-Shiley tilting-disc prosthesis, first reported in 1969 [17]. This excellent initial prosthesis was developed by Viking Björk at the Karolinska Institute in Stockholm, Sweden, working in conjunction with the Shiley Corporation in California. Several other disc prostheses were developed and used briefly, but the Björk valve quickly became the disc valve of choice and was widely used for many years. Unfortunately, the problem of thromboembolism requiring permanent anticoagulation remained.
Different modifications of the Björk valve were done to minimize the problem of thromboembolism. Unfortunately, in the last model used, the concavo-convex prosthesis, a fracture of the strut of the prosthesis developed after several years in a small number of patients [18]. This tragic complication was often fatal because of extrusion of the disc and resulting total aortic insufficiency. The resulting medicolegal complications eventually led to withdrawal of all Björk prostheses from the market. This unfortunate experience reemphasizes the experimental nature of all prosthetic valves for at least 5 years. The concavo-convex valve simply had a slightly wider opening angle than its predecessor, designed to minimize thrombus formation. Apparently the increased mechanical stress from the slight increase in the opening angle led to the tragic complication of strut rupture.
The St. Jude pyrolytic carbon disc prosthesis was first clinically used in 1977 and remains 20 years later one of the most popular and durable prostheses [19]. To my knowledge, clotting of the prosthesis has occurred in a few instances, but mechanical failure has not occurred. Similar effective disc prostheses in current use include the Medtronic-Hall, Omniscience, Carbomedics, and others. There seems to be little physiologic difference among the different prostheses. Although a more effective orifice area is present with disc prostheses, major physiologic advantages of disc prostheses over ball-valve prostheses have not been demonstrated.
Tissue Valves: Autologous, Extracardiac
The insoluble problem of thromboembolism with metallic prostheses quickly led to the investigation of tissue prostheses. Initially different tissues in the human body were used, especially pericardium and fascia lata. Senning in Zurich, Switzerland, probably had the most extensive experience with aortic valves constructed with fascia lata [20]. It was quickly found that thromboembolism was much less common than with valvular prostheses. This important fact remains the major reason for the use of tissue valves of any kind. Initial results were excellent, but within 5 to 10 years progressive fibrosis of the fascia lata led to increasing stenosis and insufficiency. At present, with the exception of autologous cardiac valves, progressive fibrosis and contraction has been found in all autologous tissues used for valvular cusp construction.
Tissue Valves: Homografts
Aortic homograft valves were initially used by Donald Ross in London, England, in 1962 [21], and also by Barratt-Boyes in Auckland, New Zealand, in 1962 [22]. The importance of different methods of preservation of the homograft valves was unknown at that time. Initial methods included radiation, an antibiotic solution, or freeze-drying. All are now known to be inferior for long periods of preservation. Years later, in 1987, Mark O'Brien and associates [23] reported from Brisbane, Australia, long-term experiences with cryopreservation. Their remarkable data suggest that a few cells remain viable even after 10 years of preservation. Cryopreservation is now generally accepted as the best method for long periods of preservation. This technique also greatly augments homograft availability, for the limited availability of homograft valves in many areas of the world has also long restricted their widespread use.
Homograft valves fortunately remain strikingly free from thromboembolism, but fibrosis produces a gradually increasing frequency of aortic insufficiency after 10 to 15 years.
In Brazil, Zerbini's group [24] used aortic valves constructed from homografts of dura mater for several years, but these were eventually abandoned because of progressive insufficiency.
Tissue Valves: Heterografts
Once the low frequency of thromboembolism with tissue valves was recognized, heterograft valves were investigated because of the limited availability of homograft valves. Experiences in the 1960s with formaldehyde-preserved valves were initially encouraging, but the valves failed in 2 to 3 years. Fortunately, a few years later it was discovered that glutaraldehyde, a chemical with a structure similar to formaldehyde but with a longer carbon side chain, was an excellent tissue preservative. This led Carpentier in Paris to explore the use of porcine aortic valves preserved with glutaraldehyde. In contrast to all other experiences with different forms of heterograft preservation, the durability of glutaraldehyde-preserved prostheses was dramatically better [25]. More than 90% functioned satisfactorily 5 years after implantation, and 75% to 85% 10 years later.
The significant influence of age on durability was soon recognized; in children, a high percentage of heterografts fibrose and calcify within 2 to 3 years. Presumably this early calcification is related to body growth and increased calcium metabolism. Long-term studies have also found that failure rates are significantly greater in adults in their second to third decade as compared with those in the sixth to seventh decade. In general, the older the patient, the better the durability.
Bovine (calf) pericardial valves were also initially used by Ionescu and others [26], but for uncertain reasons had a significantly higher failure rate than porcine valves. In recent years, with a different method of preservation, Carpentier [27] has reintroduced the bovine pericardial valves because the hemodynamic characteristics are somewhat better than those of porcine valves. Enough data are not yet available to confirm, however, that long-term durability is equal to or better than that of porcine valves.
Autologous Aortic Valves
Donald Ross [28] in London, England, in 1967, reported the ingenious concept of replacing the aortic valve with the patient's pulmonary valve, followed by inserting a homograft pulmonary valve. With the low pressures in the pulmonary artery, durability of the homograft valve is better than in the aortic position. This complicated operation was not widely accepted because of its complexity and the limited efficacy of myocardial preservation at that time. Despite the operative mortality, however, Ross applied the procedure in a significant number of patients. Long-term studies 19 years after operation were reported in 1988 [29]. These valves have a striking freedom from thromboembolism but a vulnerability to endocarditis, similar to all other valve prostheses. Aortic insufficiency has gradually developed in a few patients 10 years after implantation, but the overall frequency is significantly less than with aortic homograft valves.
With better techniques of preservation, Elkins and associates [30] in Oklahoma City reintroduced the "Ross Switch operation" in children with impressive results. With the great advantage of freedom from anticoagulation and good durability for at least 10 to 15 years, the operation has become increasingly popular throughout the United States in recent years, and is now often considered the prosthesis of choice when valve replacement is required in children. The availability of cryopreserved grafts for insertion in the pulmonary position has helped significantly. The "Ross switch" is now being used in young adults with increasing frequency.
Mitral Valve Reconstruction
The four hazards of any type of prosthetic valve have always been thromboembolism, the need for anticoagulation, the lifelong risk of endocarditis, and durability. Metallic prostheses now have excellent durability but require lifelong anticoagulation. Tissue valves often do not require anticoagulation but have limited durability, especially after the first 10 years after implantation. Both valves are equally vulnerable to endocarditis, averaging about 1% to 2% per 100 patient-years.
For these reasons, in the late 1960s and throughout the 1970s, Carpentier's group [31] in France and Duran's group [32] in Spain explored different forms of mitral valve reconstruction. Different forms of annuloplasty were tried in the late 1950s before prosthetic mitral valves were developed. The posteromedial commissural annuloplasty developed by Merendino and others [9] had the best results but could only be applied in a limited percentage of patients. A tailored mitral annuloplasty, narrowing both the annulus of the aortic and mitral leaflets, was developed by Reed and associates at New York University [33], creating a mild degree of mitral stenosis, an orifice near 1.0 cm2. This was useful in children but not widely applicable in adults. Long-term results were reported in 196 patients by Reed and associates [34].
Most credit for the development of an effective method of mitral valve reconstruction belongs to Carpentier. He first demonstrated the crucial fact that a quadrilateral resection of diseased portions of the mural leaflet of the mitral valve could be safely done, excising as much as 2 to 4 cm of the mural leaflet. This was a remarkable achievement, first demonstrating that the thin mitral leaflet tissue could be excised and successfully sutured without dehiscence of the suture line in the reconstructed leaflets when cardiac contractions began and generated systolic pressures of 100 to 120 mm Hg. This was primarily accomplished with a careful annuloplasty, supported by a prosthetic ring, which removed tension from the suture line in the leaflets. Further techniques for specific problems were developed, principally shortening of elongated chordae to the aortic leaflet and the technique of chordal transposition. In 1983, Carpentier reported experiences with more than 1,500 patients [35]. Most of his experiences were with patients with valves injured by rheumatic fever. The follow-up was limited.
The application of the Carpentier concepts of mitral valve reconstruction for all types of mitral insufficiency was explored with increasing frequency in the United States, especially after 1980. It was unknown whether the fragile leaflet tissue found with mitral valve prolapse would be as amenable to reconstruction as the fibrosed valves of mitral insufficiency.
Starting in 1980, the techniques were vigorously explored at New York University, principally by S. Colvin [36]. Similar studies were soon undertaken by the surgical group at Cleveland, led by Cosgrove [37]. The impressive durable results over the subsequent years have gradually led to widespread adoption of mitral valve reconstruction.
At New York University more than 700 valve reconstructions have been performed between 1980 and 1996 (Grossi EA, Galloway AC, Miller JS, et al; unpublished results). A continuing follow-up has been done because the main uncertainty with reconstruction has not been short-term results but long-term durability. A recent follow-up, 98% complete, found 88% of reconstructed nonrheumatic valves functioning well 8 years later. A somewhat higher frequency of recurrent insufficiency (72% functioning well at 8 years) was found with rheumatic valves, presumably due to diffuse fibrosis after the rheumatic inflammation.
The operation is especially attractive in young women in their child-bearing years. With the excellent long-term durability, the operation is now applied at an early stage with significant mitral insufficiency, preferably before significant enlargement of the left atrium has developed and while the patient is still in sinus rhythm. Patients who remain in sinus rhythm after the operation remain strikingly free from thromboembolism and also have an extremely low frequency of endocarditis.
Summary
In summary, the remarkable development of valvular prostheses over the past five decades has been reviewed. Initially the concept of replacing a human heart valve with a prosthetic one seemed visionary and impractical. Multiple investigations by many investigators have made the present age of valve prostheses possible. The most significant advances were the development of safer pump oxygenators, permitting long periods of perfusion, and the development of effective forms of myocardial preservation. The lethal hazards of air embolism were simply unknown at the beginning of open heart surgery, but now can be effectively prevented in the vast majority of patients, especially with the development of operative monitoring with transesophageal echocardiography. Years of clinical investigation were required to find a valvular prosthesis that remained functional for long periods of time in humans with the mechanical stresses resulting from at least 30 million cardiac contractions per year. At this time both the available disc valves and the ball valves have admirable records of durability for more than 20 years. The ideal valve for the future should possess the durability of the present prostheses but be constructed of materials that do not initiate the coagulation cascade and thromboembolism unless anticoagulants are used. Thus far this has remained an elusive unsolved problem for more than three decades.
Finally, the presentations by different participants in this W. H. Muller symposium well demonstrate the basic importance of his many contributions since beginning the cardiac surgical program at the University of Virginia four decades ago. The presentations also clearly demonstrate the basic importance of collegial sharing of information by colleagues throughout the world. The cross-communication through publications, scientific meetings, and conversations leads to remarkable advances that otherwise might not occur.
Footnotes
Presented at Cardiovascular Surgery—Then and Now, University of Virginia Medical Center, Charlottesville, VA, April 26, 1997.
Address reprint requests to Dr Spencer, Department of Surgery, New York University School of Medicine, 550 First Ave, New York, NY 10016.
* Some information is anecdotal, from conversations over past decades, and may or may not be found in print. 
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
