Commentary Authors Urge Improvements to Device Approval Process

Medical device regulation is in need of widespread reform—and a prime example of how the US Food and Drug Administration (FDA) approval process could be optimized is encapsulated in the experience of the Watchman device, asserts a commentary piece published online July 11, 2011, ahead of print in Archives of Internal Medicine. Despite a positive clinical trial and backing from an FDA advisory panel, the device for stroke prevention in atrial fibrillation has yet to gain approval.

Based on the Watchman’s regulatory setbacks, Sanjay Kaul, MD, MPH, of Cedars-Sinai Medical Center (Los Angeles, CA), and colleagues propose “greater rigor, clarity, efficiency, and transparency in device evaluation, approval, and surveillance.”

Mitchell W. Krucoff, MD, of Duke University Medical Center (Durham, NC), told TCTMD in a telephone interview that he appreciated the commentary’s thoughtfulness. “I think it’s positively oriented,” he said. “It [asks], how can we do this better? Where are there some areas that we could concentrate?”

The Watchman Waits

The Watchman device (Atritech, Plymouth, MN), intended to prevent stroke in patients with nonvalvular atrial fibrillation, first showed promise when data from the randomized 700-patient PROTECT-AF trial were released in March 2009 at the American College of Cardiology Scientific Session/i2 Summit in Orlando, FL. Percutaneous left atrial appendage closure with Watchman was found to be noninferior to ongoing warfarin therapy in terms of preventing stroke, systemic embolism, and cardiovascular death. However, its use was accompanied by a significantly higher risk of complications, mainly pericardial effusion and procedural stroke.

Despite an FDA advisory panel’s 7-to-5 vote favoring approval in mid-2009 and subsequent publication of PROTECT-AF in the August 2009 issue of the Lancet, “the FDA nevertheless decided against approval and requested a confirmatory study to further substantiate the safety and effectiveness of the device,” Dr. Kaul and coauthors report.

Approval was hampered by a number of issues arising from PROTECT-AF, they say, including:

• Enrollment of a highly selective patient population that failed to represent who would be given the device in clinical practice
• Low number of events
• Short follow-up duration
• Unconventional noninferiority design for a device vs. medical therapy trial
• Definition of efficacy that is driven by reduced bleeding rather than thrombotic risk
• Failure to reassure that warfarin can be safely stopped after successful device implantation
• Increased risk of procedural complications

“Thus, it is questionable whether benefits of the device clearly outweighed its risks,” the coauthors note, adding, “Consequently, the nonapproval decision by the FDA and its request for additional studies clarifying procedural safety and durability of efficacy for its intended use appear well justified.”

Asked whether the Watchman experience highlighted problems with the current system or whether it in fact demonstrated the strength of the FDA approval process, Dr. Kaul replied: “That the FDA did not approve the device clearly points to the fact that the process is working.

“However, one can argue that a more timely feedback by the FDA to the sponsor and more transparent communication regarding insufficiency of trial design and analysis and data interpretation could have avoided premature submission of the Watchman PMA,” he told TCTMD in an e-mail communication.

David R. Holmes Jr, MD, of the Mayo Clinic (Rochester, MN), who served as principal investigator for PROTECT-AF and is currently president of the American College of Cardiology, offered a succinct reply to the critique of his study.

“The authors have their own specific slant on the process. I find it rather objectionable and based on presumptions that I do not agree with,” he commented in an e-mail communication with TCTMD. “This article is hardly a triumph of science that the authors should have been trying to achieve.”

For his part, Dr. Krucoff said that when reading the commentary, “I started with the impression that they were going to use the Watchman as an example of how the process did not work, but actually by the end, you get the feeling that [it illustrates] how the process did work,” he noted. “I think . . . the authors were trying to give a balanced overview of what is a very complex area.”

Device vs. Drug Approval

By way of background, Dr. Kaul and colleagues describe the 2-pronged path toward device approval in the United States: low- or intermediate-risk Class II devices require demonstration of substantial equivalence to a predicate device via premarket notification or the 510(k) exemption process, while potentially high-risk Class III devices—such as the Watchman—require reasonable assurance of safety and efficacy via the premarket approval (PMA) pathway.

However, the commentary authors also cite recent research indicating that a substantial number of Class III devices enter the market without being tested in randomized controlled trials with hard endpoints and refer to controversies over fast-track 510(k) approvals and post-market failures. “Thus, the standards for approving devices appear to be less rigorous than those for approving drugs, which require substantial evidence of effectiveness based on an adequate number (≥ 2) of well-controlled clinical studies,” the coauthors note.

This leniency can have real-world effects, they argue, in that “[d]ata that would never be sufficient to support the approval of a drug is strong enough to support approval of a device used to treat the same condition—potentially diverting patients from effective drugs such as warfarin to less-effective devices such as Watchman.” On the other hand, Dr. Kaul and colleagues note that serious safety problems are rare, with fewer than 5% of devices receiving FDA approval between 2004 and 2009 ultimately being recalled.

Having separate processes for drug and device approval is not necessarily bad, Dr. Krucoff said. “I think we have to be careful that we don’t say that because the way we evaluate devices is different than the way we evaluate drugs means that somehow it’s intrinsically inferior.”

For example, Dr. Krucoff noted, because drugs are metabolized, genetic factors can play a major role in how well they work. “In devices, it’s interesting because if you put a piece of steel in a coronary, it probably matters more what size the coronary is and whether you are male or female than whether you are Japanese or American or African or whatever,” he said. “So the details between drugs and devices are importantly different, and the processes have to reflect the attention to these details.”

“What this commentary points to is that there are some fundamentals to just good clinical trial data, which are shared across drug and device [trials],” Dr. Krucoff commented. “In their [list of recommendations], I don’t think a lot of those things are even changes. They’re fundamentals that need to be adhered to.”

Raising the Bar

Dr. Kaul and colleagues recommend a number of reforms for Class III devices going through the PMA pathway:

• Approval process: more transparency with consistent prespecified criteria, larger and longer randomized controlled trials, inclusion of broad patient population
• Surveillance process: mandatory recording of details on implanted devices, better premarket surveillance through registries with extended follow-up and better access to data, greater authority to conduct postapproval trials and to remove unsafe and ineffective devices from the market, independent surveillance office within the FDA to adjudicate postapproval safety
• Operational standards: robust trial design and statistical methodology, collaboration with reimbursement agencies and professional organizations

“The motivating force behind our proposals is to improve the evidentiary standard for device evaluation and approval, making it more solidly grounded to public health practice, and to encourage innovation without compromising public safety,” they note.

Dr. Kaul acknowledged that implementing some of the operational standards might prove challenging.

“For example, aligning reimbursement with clear-cut evidence of benefit by incentivizing ‘on-label’ use and disincentivizing ‘off-label’ use could become a thorny issue if there are no well-established standards,” he said, noting that a shift in that direction has already started within the overall US health care system that might extend to the device arena.

However, Dr. Kaul saw few hurdles in the areas of approval and surveillance, suggesting that simple steps such as tracking device use could make a difference. “It is interesting to note that there is a digitally retrievable code describing in detail every prescription drug dispensed but no such requirement is applied to devices,” he pointed out. “This makes it very difficult to conduct a systematic assessment, notification, or recall in case of device malfunction. . . . Why shouldn't this be a priority for both the device manufacturers and the regulators?”

Innovation as an ‘Ecosystem’

“Right now,” said Dr. Krucoff, “there is clearly a broad concern about device innovation in the United States—that it is too expensive and too slow, that we have delayed access to new devices and that an economic environment exists which reimburses on such a low margin [compared] with the high cost of research and development, that a lot of industry will say the financial equation means that they should just marginalize the United States altogether, and clearly no one wants that to happen.”

But Dr. Krucoff cautioned against pointing fingers. “The common scapegoat for our crisis in device innovation is the FDA,” he said. “The FDA is clearly a critical part of the device innovation process, but it’s also important for us to step back and recognize that device innovation is really an ecosystem.” For example, Dr. Krucoff noted, the US Congress risks hobbling the FDA by cutting funding, while clinicians need to recognize that trial enrollment in the United States can be cumbersome, expensive and slow compared to what occurs internationally.

“[N]ow we have to recognize that you can get very high quality data from outside the United States—faster and less expensively. And that’s got nothing to do with the FDA. That is a research infrastructure issue at a clinical site level,” he commented.

While the specifics, such as appropriate length of follow-up and superiority vs. noninferiority trials, can be debated, Dr. Krucoff stressed, the conundrum being wrestled with is “how to facilitate innovation and still provide reasonable assurance of safety.”

In this regard, the commentary’s balanced tone is apt, Dr. Krucoff concluded. “The difficulty we get into is, when it just comes down to the medical device industry and the regulatory authorities trying to figure out all the answers, the atmosphere tends to be very contentious,” he said. “The more we get academics and professional societies into that dialogue, the more the balance between patient advocacy and clinical expertise can provide a focus on solutions and thoughtfulness, not just controversy and headlines.”

Dr. Kaul similarly commented, “It would help if all the stakeholders were equally committed to abiding by the general principles and standards collectively referred to as ‘Good Clinical Research Practice’ to generate valid and credible evidence to inform clinical practice and guide health care.”

 

---

Sources:
1. Cingolani E, Stevens S, Shehata M, et al. Medical device regulatory reform: Insights from the Watchman left atrial appendage closure technology for stroke prophylaxis in atrial fibrillation [commentary]. Arch Intern Med. 2011;Epub ahead of print.

2. Holmes DR, Reddy, VY, Turi ZG, et al. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: A randomised non-inferiority trial. Lancet. 2009;374:534-542.

 

 

Related Stories:

• Ongoing Experience Improves Safety for LAA Closure with Watchman Device
• Data Supporting FDA Approval for Cardiovascular Devices Called into Question
• FDA Panel Narrowly Recommends Approval of Watchman