Originally Published MDDI March 2004
In recent years, courts have tended to view FDA approval as a shield against litigation. But a recent ruling bucks that trend.
James G. Dickinson
A Minnesota federal judge made waves last December when he ruled that FDA approval of a medical device doesn't mean its maker is not responsible for injured patients. The ruling went against a tide of recent federal court decisions on the Constitution's preemption clause. Those decisions held that federal regulation trumps state courts in some medical device litigation. The ruling, by judge John Tunheim, came in consolidated cases brought by injured patients against St. Jude Medical Inc. over its Silzone heart valve.
St. Jude had asked the court for a summary judgment dismissing the patients' suits on the ground that federal law preempted their claims. The patients had argued they should go forward because of key unresolved questions of material fact.
The Silzone valve has a coating of silver on the sewing cuff, the portion of the valve that is attached to a patient's heart. The company has said it introduced the coating because silver's antimicrobial properties combat endocarditis, a complication that occurs in some valve patients.
FDA approved the Silzone valve in 1998. However, it did not allow the company to claim that the coating would reduce endocarditis, because there were no clinical tests validating that claim. The company then sponsored the Artificial Valve Endocarditis Reduction Trial (AVERT) to answer the question.
But in January 2000, the AVERT monitoring board reported that recipients of the Silzone valve were more prone to a complication known as a paravalvular leak than those who had received a conventional valve. Those suffering the leak had to have the valve removed and replaced. On January 21, 2000, the monitoring board suspended enrollment in the AVERT trial because of an increase in the number of such leaks. At that time, St. Jude recalled all unimplanted Silzone products. FDA then said it was treating the event as a recall.
In asking for summary judgment, St. Jude argued that approval of the valve through a PMA supplement was enough to show that the device is reasonably safe and effective for its intended use; thus, it said, any patient claims based on safety or efficacy were preempted. The patients said the FDA approval had been improperly secured. The company, they alleged, had concealed pertinent information from FDA that would have led to a different decision.
Tunheim's ruling was based on a detailed and complex analysis of precedents in earlier medical device FDA preemption cases, most of which ruled for federal preemption. He found that the Silzone case more closely resembled the exception, Medtronic v. Lohr, in which the Supreme Court disallowed preemption. The key, he said, is not that the existence of a federal requirement must preempt any state requirement. Rather, it is whether the state requirement authorizing tort litigation actually conflicts with a federal requirement—in which case the federal requirement preempts.
“The submission of a product to the FDA for premarket supplement approval does not, in itself, amount to a specific federal requirement meriting preemptive effect,” he wrote. Instead, he would examine “whether the FDA, in this instance, has promulgated any ascertainable precondition to regulatory approval.”
Tunheim found that FDA was not aware of the risks presented by the Silzone coating and that the agency had furthermore ordered St. Jude not to claim the coating was effective in preventing endocarditis. Therefore, he ruled there is no conflict between state and federal requirements in this case.
In January, two former employees of the now-defunct medical device sterilizer manufacturer, Abtox Inc. (Mundelein, IL) entered guilty pleas to one misdemeanor count of distribution of illegal devices. The charge resulted from a criminal indictment filed in February 2003 in the Northern District of Illinois federal court. Clinical services director Marilyn M. Lynch and marketing director Mark E. Schmitt each pled guilty. The court deferred sentencing until the trial of two additional codefendants is complete. At the time of Lynch's sentencing, said her attorney, Ted Poulos, the other counts against her will be dismissed.
The other codefendants in this case are Ross A. Caputo, Abtox president and CEO, and Robert Riley, former vice president for regulatory affairs. Their trial date, previously set for February, has been vacated; at press time no new date had been set.
The original indictment contained 19 counts, including mail and wire fraud and making false statements to an FDA investigator.
The number of electronically filed new device submissions rose in 2003 by a modest 20 or so, said CDRH director of device evaluation Daniel G. Schultz. He hopes for a bigger increase this year. In 2002, the Office of Device Evaluation (ODE) received 73 complete 510(k), PMA, and IDE submissions from 14 different companies; for 2003, Schultz was expecting final numbers to be near 100, including annual reports and “a couple of 510(k)s.”
CDRH is lagging seriously behind other centers in processing electronic submissions, Schultz said. Why? Because it would need a major IT upgrade to do much better. A major problem has been the amount of scanning needed to convert paper submissions into image files.
Schultz said ODE is working on a plan to attract more all-electronic submissions in 2004. This could relieve the scanning workload and enable ODE to train more of its reviewers to do electronic reviews.
In January, FDA issued a guidance for industry on “Clinical Study Designs for Percutaneous Catheter Ablation for Treatment of Atrial Fibrillation.” It addresses study design issues associated with catheter ablation devices intended for treatment of atrial fibrillation (AF). The agency says the devices covered by the guidance are significant-risk devices.
In general, according to the guidance, randomized, controlled trials reflect the least burdensome means of collecting clinical data in support of safety and effectiveness for these devices. FDA believes that alternative study designs using patients as their own control or using a historical control can complicate demonstration of effectiveness. But sponsors wishing to use an alternative study design may do so if it is scientifically sound and addresses the relevant safety and effectiveness questions.
“For example,” FDA says, “a self-controlled study may be appropriate to evaluate treatment effectiveness among patients with truly refractory permanent AF, and a study with a sham control arm may be considered for a study involving subjective criteria such as perceived symptoms.”
Appropriate control therapy for a randomized, controlled study depends on indication for use, patient population, and device design considerations. Potential control therapies can include best medical therapy with antiarrhythmic drugs, therapy with approved medical devices, patients as their own controls, and sham therapy.
The guidance says that control to a previously approved atrial fibrillation device therapy may be a particularly desirable option. In such a case, study design, patient enrollment, and data analysis may all be straightforward.
The guidance also covers study endpoints, including primary effectiveness and safety endpoints; data collection forms; study groups; sample size; patient follow-up; blanking period; anticoagulation; and study monitoring. The guidance may be accessed at http://www.fda.gov/cdrh/ode/guidance/1229.html
The science behind silicone gel breast implants continues to evolve, and thus FDA's best advice to device manufacturers also must evolve. With this statement, the agency released on January 8 a revised draft guidance for breast implant sponsors. The guidance supersedes one it issued 11 months earlier, in February 2003.
CDRH director David Feigal explained the rationale at a news media teleconference announcing the revised guidance. “We're issuing the revised guidance because we want to make clear to the sponsors of breast implant products what must be in marketing applications in light of new data,” he said.“FDA has been discussing these products for more than a decade and we've learned quite a bit. Patients and surgeons need this information to be able to make appropriate decisions and react if problems arise.”
Substantive new recommendations in the draft guidance involve mechanical testing, modes and causes of implant rupture, clinical study information, postapproval requirements, and labeling. Feigal said the most important issues deal with ruptures and getting a clear understanding of the extent of the problem, when and how it occurs, what are the predictors of a rupture, what are the impacts of a ruptured gel implant, and what steps need to be taken if rupture occurs.
The specific new recommendations cited in an FDA release on the new guidance include the following:
•The general information section under mechanical testing has been modified. It now recommends that mechanical testing be designed so it can predict clinical outcomes, such as how long breast implants will last before rupturing in the body.
•In the section on fatigue rupture testing of the total device, FDA now recommends that a sponsor develop a new test methodology that can accurately predict rates of rupture over time. According to FDA, methods currently used do not simulate observed rates of rupture.
•The section on bleed testing (relating to gel bleeding out of an implant) has also been modified. It now recommends that a sponsor develop a new gel-bleed test that more closely mimics conditions in the body. The goal is to quantify the chemicals that may leach out of an implant shell over time.
•A new section on modes and causes of rupture has been added to the guidance. It clarifies FDA's recommendation that a sponsor characterize the modes and causes of rupture. The agency recommends a retrieval study involving examination and testing of breast implants that have been removed from patients; an assessment of a sponsor's manufacturing processes for the shell to determine whether any allowances for imperfections may be related to device rupture; an assessment of surgical techniques that increase the risk of rupture to better guide doctors on the best way to implant the devices; and a comprehensive literature review of durability based on studies of explanted devices.
•The section on clinical studies general information has been modified to clarify data requirements. Although a sponsor may submit a PMA with a minimum of two years of clinical data, the data may not demonstrate a reasonable assurance of safety and effectiveness. Thus FDA may recommend submission of additional premarket data to describe the rates of complications, reasonably predict the safety profile of the device over its lifetime, and address safety concerns.
•A new clinical studies section on rupture of silicone gel–filled breast implants updates the previous section on silent rupture. The update focuses on rupture as a whole and not just silent rupture. FDA recommends that as part of its core study, a sponsor provide data on the rate of rupture and rate of change of rupture over the expected lifetime of the device, the frequency of rupture observed, and characterization of any local health consequences of ruptured implants.
•The clinical trials connective tissue diseases section now clarifies the agency's recommendations regarding data collected on connective tissue diseases and signs and symptoms of such diseases. It recommends that sponsors collect information on diagnoses of connective tissue diseases as part of the device.
FDA officials declined any comment on the agency's simultaneous action determining that an Inamed PMA for silicone gel breast implants was not approvable. (In a letter to the company, FDA described the information needed to reconsider the PMA. Inamed officials say the FDA letter indicates that submission of the requested information and data will place the PMA in approvable form and allow the agency to further review it.)
In response to the nonapprovable letter, Public Citizen Health Research Group director Sidney Wolfe said FDA “acted wisely (in) asking for more data upon which to judge the safety of the implants. The agency is thereby rejecting its own flawed ‘guidance,' which encouraged companies such as Inamed to think that dangerously inadequate amounts of data could be used as a basis for approval.”
Wolfe called FDA's decision “a rejection of what would have been the most dangerous, defective medical device ever approved by FDA despite advanced knowledge of its dangers. Although saline implants are also defective in their frequent ruptures, there is an extraordinary difference between rupture and leakage of salt water—which is immediately and safely absorbed by the body—and the spread of highly reactive silicone gel, causing local and regional scar tissue formation as well as gel migration to other parts of the body.”
To access FDA's news release on the new guidance, visit http://www.fda.gov/bbs/topics/NEWS/2004/NEW01003.html.
Staar Surgical Cited for Reporting Failures
Staar Surgical (Monrovia, CA) failed to report to FDA malfunctions related to intraocular and implantable contact lenses, according to a warning letter issued in December. “In one case, a physician had to perform a vitrectomy on a patient after the lens delivery system failed,” the letter charged. “In another incident, a physician had to suture a patient after the cartridge failed. Your company had knowledge of these serious injuries but failed to forward the information to the FDA.”
In addition, FDA's letter took issue with Staar's complaint handling. FDA said the firm failed to perform “root cause analysis” on complaints involving diopter shift, blurred vision, cloudy vision, and posterior capsule tears. FDA also cited the firm for not assuring that its outside test laboratory uses validated methods to test raw materials and finished devices.
FDA noted that it had received Staar's response to the FDA-483 (investigator observations) and that the corrections promised by the company will be verified during an upcoming reinspection.
In a January statement, Staar said it is “aggressively pursuing corrective actions to remedy issues cited,” and said the costs involved “will not be material.” Staar said the complaints cited in the warning letter's first point (diopter shift, cloudy vision, blurred vision and capsular tears) were reported for intraocular lenses used in cataract surgery and not in the company's pending intraocular contact lens (ICL) for myopia. “The company is extremely confident that all of the ICL-related complaints in the U.S. FDA Study have been accurately and completely reported to the FDA, and were disclosed prior to the October 3, 2003, panel meeting,” Staar's statement said.
CDRH has issued a new guidance entitled “Replacement Reagent and Instrument Family Policy.” It provides information about validating and documenting “changes to cleared test systems that do not significantly affect the safety and effectiveness of the device.” The guidance applies
to previously cleared reagents and instruments.
It supersedes a June 1996 guidance, “Data for Commercialization of Original Equipment Manufacturers (OEM), Secondary and Generic Reagents for Automated Analyzers.”
The new guidance describes a process for adding either a cleared reagent to a previously cleared instrument, or a new instrument family member to a previously cleared instrument family.
To view the new guidance, visit www.fda.gov/cdrh/oivd/guidance/950.html.
Copyright ©2004 Medical Device & Diagnostic Industry