Irvine Firm’s Product Is Center of Safety Question : Heart Valve Failures Prompt Concerns
At least 147 times in the last eight years, a small metal strut has cracked or broken in a mechanical heart valve made by Shiley Inc. of Irvine and implanted in 84,600 patients worldwide. In 65% of the cases when this happened, the patients have died.
When the strut breaks, the valve stops working, impairing the flow of blood out of the heart. Unless immediate surgery is performed to replace the valve, the patient is unlikely to survive more than a few hours.
No other heart valve now marketed in the United States has a comparable problem with mechanical failures. No one knows how many more valves may fail in the years ahead. No one knows how many patients and their families are aware of the potential problem and when and how to quickly seek medical help.
Despite years of discussion between the manufacturer and the U.S. Food and Drug Administration, it was not until last month that the large sizes of the valve, in which struts have broken 112 times, were permanently taken off the market--voluntarily by Shiley. Some of the valves had been recalled in 1980, 1982 and 1983, although sales continued.
Smaller sizes of the valve, which are sold twice as often, have broken 35 times. They remain on the market.
A close study of the Shiley case calls into question the adequacy of warnings about defective products to physicians and patients. It also raises questions about the effectiveness of the FDA’s regulation of the medical device industry.
“Our institutions to protect patients are not functioning well,” said Henry Piehler, a metallurgical engineer at Carnegie-Mellon University in Pittsburgh and an expert in medical devices who has studied the valves. “There is a frustrating inability to respond to a problem once it appears,” said Piehler, who has been retained as an expert witness by plaintiffs in suits involving the valve malfunctions.
James S. Benson, deputy director of the FDA’s Center for Devices and Radiological Health, said his agency was “teetering on the edge” of seeking removal of the valve from the market several times, but he repeatedly made the “close call” against such action.
“No other product has been followed so closely by us,” Benson said. “Shiley has been very reassuring over the years that they had solved the problem.”
But Dr. Sidney Wolfe of the Public Citizen Health Research Group in Washington charged that Shiley “conned” the FDA into inaction. All of the valves of that type, not just the large sizes, should have been taken off the market in 1984, he said. “The FDA must share responsibility for strut fractures in valves implanted in the last 15 or 16 months.”
Symptoms of a strut fracture include sudden shortness of breath, loss of the “click” sound of the mechanical valve, dizziness and chest pain. The diagnosis can usually be confirmed within minutes by X ray. Patients can survive several hours after the valve breaks, but immediate surgery to replace the valve is essential.
Risk Deemed Small
Because the risk of immediate death from elective surgery to replace a functioning artificial valve is about 5%, and the chance of a strut fracture in the highest-risk valves is between 0.3% and 0.4% a year, according to Richard P. Chiacchierini of the FDA, no one has recommended that they be removed from those patients who already have them.
But that is small comfort to those whose survival depends on the valve.
“My wife and I have been in a state of shock and mental anguish,” said one such patient, a 64-year-old man from Wilmington, Del., who asked that his name not be used. “I can hear that thing clicking, and that little strut is always on my mind. I try not to get excited because if I do, I’m afraid it will break.”
The patient had a large Shiley valve implanted in Philadelphia in November, 1983. He learned about the problems with the valve in July, 1984, when his daughter saw a television report. The patient’s surgeon did not call to inform him of the problem until March, 1985.
The man said he never travels without a list of hospitals that perform cardiac surgery and never closes the door to his condominium when his wife goes out, so the neighbors will hear him if any problems develop.
‘I Would Be Up a Creek’
“We should have been advised about this at the time,” he said. “If I didn’t know about this and the valve broke, I would be up a creek without a paddle.”
Shiley, the world’s largest manufacturer of heart valves, is now a defendant in more than 25 strut-fracture lawsuits in the United States and additional cases abroad, Allen Greenberg of the Public Citizen Health Research Group reported. Settlements have ranged from $350,000 to nearly $1 million, he said. No case has gone to trial. The company, a subsidiary of Pfizer Inc. of New York, would not comment on the litigation.
The Shiley valve--its full name is the Bjork-Shiley 60-degree convexo-concave valve--is used to replace a human heart valve that has become diseased. Costing about $2,000, it most often replaces one of the two valves on the left side of the heart, the side that pumps blood to the body.
The valve has a tilting disk that opens and closes to control the flow of blood. Two metal struts hold the disk in place. If one of the struts cracks or breaks, the disk usually floats free and the valve stops working.
Invariably, the fractures have involved the “outlet” strut, which is welded to the valve’s metal ring, not the “inlet” strut, which is made from the same piece of metal as the ring.
“This problem is welding the outlet strut,” said Piehler, the Carnegie-Mellon professor. “Very simply, it shouldn’t be done. It is possible to make the entire valve from one piece of metal. This problem was easily avoidable.”
Robert Curtis, Shiley’s president and chief executive officer, defended the company’s handling of the problem. He said that the “overall benefit” of the valve has justified its continued marketing and that “tighter” quality control procedures have led to the manufacture of the “best ever” valves since June, 1982. Curtis said that strut fractures are “a multifaceted problem” and that the welding faults were only one of several causes the company had identified.
Piehler, however, contends that Shiley “tried to put a Band-Aid on a gaping sore” by initiating “questionable” quality control procedures that, in retrospect, did not work. He said he has examined more than 10 of the failed valves under an electron microscope and detected flaws associated with the welding process in all of them.
Since 1981, there have been at least 12 reports in medical journals of individual fractures of Shiley valves, according to a review of the literature by The Times. But it was outsiders--the Public Citizen group’s Wolfe and Greenberg--who alerted the general medical community to the problem in January, 1985, through a letter published in the New England Journal of Medicine, and who persuaded the FDA to mention the strut fractures in the December, 1984, FDA Drug Bulletin.
“The importance of being aware of this potentially fatal complication cannot be overemphasized,” Wolfe and Greenberg wrote. “Patient survival clearly corresponds to prompt recognition of the problem.”
Shiley has confined its direct communications about the valve recalls to cardiovascular surgeons and chiefs of cardiology, even though a patient with a strut fracture is often first seen by a paramedic or emergency room physician. Curtis said an international advisory panel of cardiac surgeons is helping draft a statement that will be sent to major medical journals next year.
Curtis said the company “discussed at length” the possibility of contacting patients directly but concluded with advice from the advisory panel that direct notification might be interpreted as “some sort of practice of medicine” and was a responsibility of physicians. The FDA’s Benson said warning patients directly would “do more damage than good” and cause unnecessary alarm.
Instead, the company has given the serial numbers of recalled valves, based on its shipping records, to hospitals and surgeons. And last month Shiley, for the first time, gave the names of some patients who had been implanted with the recalled valves to their surgeons. The company’s information on patients is based on valve registration forms, which are supposed to be completed by the surgeon. One problem, Curtis noted, is that Shiley receives completed forms only 25% to 30% of the time.
Hundreds of Calls
Some surgeons, such as Dr. Edward Rittenhouse of Providence Hospital in Seattle, have called in all patients with recalled valves to discuss the situation directly. But many patients have first learned about the valve failures from media reports of the recalls and have taken it upon themselves to contact Shiley, the Public Citizen group or their physician. Shiley has received between 500 and 1,000 patient phone calls in the last month and Public Citizen more than 50. Many of those who called had heart valves not involved in the recalls.
Rittenhouse said that because the incidence of strut fractures is “very low,” he has counseled patients with the convexo-concave valves not to dwell on the subject excessively or to alter their life styles. “These people tend to be in situations with people around them” he said. “They are also more cognizant that they can have problems of all different kinds.”
Chiacchierini, director of the FDA’s division of biometric sciences, emphasizes that the risk of fractures is small contrasted with the combined risks of other complications, such as infections, blood clots and leaks, that valve patients may have. Their chance of dying from all causes, irrespective of which heart valve a surgeon chooses, is 4.4% a year, he said.
With the benefit of hindsight, one might ask why it has taken so long to sort out this problem in the United States. The first strut fracture was reported to the FDA in 1978, while the valve was undergoing pre-marketing testing.
The convexo-concave valve was thought to be a significant advance on an earlier Bjork-Shiley design called the spherical valve, and is still considered so by many physicians. The convexo-concave valve was engineered to offer less resistance to blood flow, and Shiley has been able to convince the FDA and many surgeons that it is less likely than the spherical valve to cause life-threatening blood clots.
The spherical valve, which is still marketed, has been, according to Rittenhouse, “very reliable over the years” and was the “gold standard” among all mechanical valves available in the 1970s. It has been implanted more than 175,000 times worldwide since 1975, with no strut failures, according to Curtis.
The FDA approved marketing of the convexo-concave valve in April, 1979. In February, 1980, Shiley voluntarily recalled all sizes of the valve after a second strut fracture was reported. As the fractures continued to be reported, a second recall followed in May, 1982, and a third in July, 1983. These involved large-size heart valves made between May, 1980, and February, 1982. The recalls resulted in some valves’ being taken off the market, but manufacturing and sales of others continued.
FDA Backs Down
In June, 1982, the FDA asked Shiley to voluntarily cease distribution of the valve until new manufacturing and testing procedures were evaluated, but backed down after Shiley refused.
In the spring of 1984, a study by the FDA refuted the convexo-concave valve’s superiority over the spherical valve in avoiding blood-clotting complications. In June, the FDA told the company to either provide new data on the clotting or to cease distribution of all sizes of the convexo-concave valve and withdraw all unimplanted valves on the market.
In July, 1984, after reviewing new data, the FDA again changed its mind. “Even though there is an apparently continuing problem of strut fracture . . . the incidence of these fractures appears to be low when compared with the advantage of fewer thromboembolic (clotting) complications,” the FDA’s Benson wrote to the company.
Shiley immediately sent a letter to cardiac surgeons and cardiologists criticizing “the recent dissemination of inaccurate and misleading information through the news media” and claiming that the new studies “amply demonstrated the safety and efficacy of the valve and answered all questions.”
Increase in Failure Rate
Shiley’s claim soon proved wrong. Additional fractures began to be reported in large valves welded between March and June of 1982. Shiley notified the FDA of this trend in late 1984 and cardiac surgeons and cardiologists in February, 1985, according to the FDA Medical Devices Bulletin.
The latest available figures show that 28 valves made after March, 1982, have broken and that 19 patients have died, according to the company. In mid-October, after discussing these valve failures with the FDA, Shiley agreed to take all large valves made after March, 1982, off the market.
According to the FDA’s Chiacchierini, the risk of fracture in large valves welded between March and June, 1982, is now no different than in those welded earlier--three to four per thousand implants per year.
For valves welded after June, 1982, less than one fracture has been reported per thousand implants per year, but the figure is not precisely comparable, because the time since surgery for these patients is shorter.
FDA officials, in multiple interviews, said they faced many difficult decisions in the Shiley case.
The case against the marketing of the valve was too weak to justify a withdrawal by the FDA of the permission to sell it granted Shiley in 1979 or a finding that the valve presented “an unreasonable risk of substantial harm to the public health,” according to William H. Damaska, director of the division of compliance operations.
The FDA, in fact, has never tried to withdraw such a permission of sale in more than 300 cases it has processed since 1979. In the case of the Shiley valve, Benson said, it would probably have required “cumbersome” public hearings and the concurrence of a review panel.
Instead, the FDA chose to use what Benson called an “arm-twisting” approach in the hope of achieving “expedient voluntary action” by Shiley to get the valve off the market.
“The level of evidence required for the FDA to take action as compared to the company (taking action) are worlds apart,” Chiacchierini said.
The government also investigated allegations of manufacturing deficiencies at Shiley’s plants and identified some problems, Damaska said, but it concluded that manufacturing deficiencies were “not causing” the strut fractures.
A possible flaw in the FDA’s review of the convexo-concave valve was that it judged the valve mainly by comparison with Shiley’s own spherical valve, rather than with the valves of other manufacturers. When choosing among valves, surgeons are not limited to the products of one manufacturer.
FDA officials said they lacked the data to make meaningful comparisons among manufacturers. They also said the comparison to Shiley’s own product was the basis for approving the convexo-concave valve in the first place.
“It is almost impossible at this point in time to compare one manufacturer’s valve to another manufacturer’s valve with regard to complications,” Chiacchierini said. Indeed, Benson said, the FDA began just this spring to assemble the data necessary for such comparisons. Based on the preliminary data, Benson said, he is “not convinced the convexo-concave valve was superior to other valves on the market.”
The data the company furnished on the comparative safety of the convexo-concave and spherical valves may also have been flawed because of the way it was obtained.
Shiley based its conclusions on the experiences of prominent surgeons with the two valves, a method acceptable under FDA regulations.
There is, however, a better way to obtain comparative data, according to Dr. Shahbudin Rahimtoola, chief of cardiology at USC Medical Center and an international authority on the subject. That method is to design a study that eliminates all variables other than choice of valves, he said.
Rahimtoola said in an interview that the most important factor predicting the rate of valve complications is the health of the patients who receive the surgery and that that variable was not eliminated by Shiley’s method.
Chiacchierini called Shiley’s retrospective data on the experiences of prominent surgeons with the two valves “the next best thing” to the test favored by Rahimtoola.
The original Bjork-Shiley heart valve was designed by American engineer Donald P. Shiley and Swedish cardiac surgeon Dr. Viking O. Bjork and was first implanted by Bjork at Stockholm’s Karolinska Hospital in 1969.
Donald Shiley, who founded the company in a Tustin garage in 1966, has not not been involved in the day-to-day operations since Pfizer bought the firm in 1978.
Shiley Inc. controlled 29% of the world’s market for heart valves in 1984 and sold a total of 27,000 valves, according to the newsletter Biomedical Business International. Shiley’s total sales in 1984 were $120 million.
Bjork said he was initially a paid consultant but has been “completely independent” financially of the Shiley firm since 1978. He has, however, continued to work closely with company engineers on all significant modifications of the original design. Bjork is now director of research for the Heart Institute of the Desert at Eisenhower Medical Center in Rancho Mirage and an emeritus professor at the Swedish hospital.
Since May, 1982, Bjork and his colleagues in Sweden have used a special monostrut valve made from a single piece of nickel-cobalt alloy, eliminating the weld.
Monostrut ‘Best Version’
The monostrut is “the best version developed, and includes all improvements,” Bjork said in an interview last month. The valve has been approved for use in more than 20 countries and about 12,000 have been implanted, mostly in Western Europe, with no strut fractures reported, according to the company.
Bjork criticized FDA pre-marketing requirements, the anticipation of which has kept Shiley from even filing an application to sell the monostrut in the United States, although the company has announced its intention to seek such approval.
“In Sweden, I am a government adviser on heart surgery,” Bjork said. “I had to ask myself what had to be done. If you have a better car, you buy the better car. If you have a better valve, why not use it?”
FRACTURES IN BJORK-SHILEY HEART VALVE
Sales of large sizes of the Bjork-Shiley heart valves were halted last month as a result of fractures found in the outflow strut, below. When the strut breaks, the tilting disk, which opens and closes to control the flow of blood, can float away, forcing emergency surgery. At right, how the valve functions in the heart.
CONVEXO-CONCAVE 60 VALVE
SEWING RING
METAL RING
DISK
BLOOD FLOW
OUTFLOW STRUT
Structural problems have occurred at the welds joining the outflow strut to the valve’s metal ring.
REPLACEMENT LOCATION FOR MITRAL VALVE
Mitral valve, between atrium and ventricle, is pushed open by rising atrial pressure. Ventricle fills with blood.
Ventricles contract. Increased pressure closes the mitral valve. Aortic valve opens and oxygenated blood flows to the body.
BJORK-SHILEY HEART VALVE FAILURES
Between 1976 and 1985, there were 147 fractures among all sizes of the Bjork-Shiley 60-degree convexo-concave heart valves implanted worldwide. The valve was recalled in 1980, 1982,1983 and 1985.
YEAR FRACTURES
1978 1
1979 2
1980 3
1981 7
1982 29
1983 40
1984 42
1985 (eight months) 23
Source: Shiley, Inc.
