[Posted on behalf of Jeffrey Skopek, Academic Fellow, Petrie-Flom Center (with the same disclaimer about the off-the-cuff nature of live blogging)]
Jeffrey Shapiro (Hyman, Phelps, McNamara PC), Why the 510(k) Pathway is the Right Approach for Most Medical Devices
Jeff Shapiro began with an introduction to FDA approval processes for medical devices. He explained that all devices reach the marked based on a finding of reasonable assurance of safety and efficacy, whether it is through the PMA or 510(k) process.
His focus was on the 510(k) clearance process for Class I/II devices, which is based upon FDA’s finding of substantial equivalence to a “predicate device.” This predicate device is another Class I/II device that (1) has the same intended use and (2) has the same technological characteristic, or has difference characteristics but does not raise new questions of safety. He explained that this review process creates a chain of linked comparisons through which the FDA allows advances only far enough that likely clinical impact can be predicted. Novel devices are shifted to PMA process.
He argued that the 510(k) approach has many benefits. Like common law, it is precedent driven. It provides an open regulatory architecture. It allows leapfrogging within industry sectors, avoiding re-inventing the wheel. It is efficient, focusing on modifications, rather than complete evaluation. Each clearance adds richness to the body of potential baseline technology, allowing fine comparisons in a wide variety of device types. The PMA approval, by contrast, has closed regulatory architecture. Each iteration must re-invent the wheel, and thus is only practical for a small number of high risk devices.
He closed by suggesting that the 510(k) process can be improved in ways discussed in his paper, but that it works well for most medical devices and should thus be around for a long time.
Kayte Spector-Bagdady/Elizabeth Pike (Presidential Commission for the Study of Bioethical Issues), Device-ive Manuevers: FDA Regulation of the Bifurcation of Direct-to-Consumer Genomic Data and Information
Kayte Spector-Bagdady began with a historical overview of direct-to-consumer (DTC) genetic testing, which has grown in size, scope, and power in recent years. For example, in 2007, DTC genetic testing services provided access to only 0.03 percent of their genome. Today, they provide access to the entire set of whole genome sequence (WGS) data. Originally, the FDA was not involved, but in 2009, it declared that the tests are medical devices and subject to FDA requirements.
She explained that in response to the threat of FDA regulation, some companies are choosing to bifurcate: some are providing genetic data only (without any analysis of this information); others providing only analysis, without conducting any laboratory procedures.
She then explored two new developments that are posing challenges to FDA. The first is that the analysis of data is being provided not only by companies, but also by community based websites, such as openSNP, which provide analysis for free. This poses a significant regulatory challenge for the FDA and raises tough questions about the FDA’s ability to regulate publicly available medical information. The second is the move from single discrete gene tests to WGS, which poses challenge to FDA’s interest in differentiating between high and low risk tests.
She concluded with a discussion of ways in which FDA might try to address these challenges, including mandating clinician involvement. Noting difficulties, she concluded that FDA will have to become as dynamic as the devices it aspires to regulate.
Thomas McLean started with a discussion of historical context, noting that a number of recent medical device failures (e.g., St. Jude Medical’s Riata leads for its internal cardiac defibrillators) have prompted calls for the FDA to update its medical device regulations.. In specific, there have been calls for post market surveillance (PMS).
He then provided an overview of the current FDA premarket approval (PMA) process under §510(k) and §360 of the Medical Device Amendments to the FDA Act. He noted that 90% of class III devices enter the market via the 510K process and suggested that this might not be best way, as it provides no substantive review of risks and benefits (e.g., involves on average only 20 hours of review time). Turning to the §360 route, he explained that it involves a formal engineering evaluation, in which risks and benefits are considered, in which thousands of hours are spent on the review. Although expensive, the manufacturer gets the benefit of federal preemption of state tort law claims. He then turned the question of how a manufacturer decides which route to take, suggesting that they use a multifactorial process based on benefit to the company—and that this might not be in the public interest.
He concluded by suggesting that we should close the 510(k) window and force all manufacturers to obtain §360 PMA, and then perform PMS through MDEpiNet. In this way: (1) physicians and consumers are likely to have greater confidence in the safety of medical devices; (2) fewer medical devices are likely to gain market access thereby reducing the rate of medical inflation; and (3) §360 preemption should reduce medical device litigation.
Nicholson Price asked whether there was role for FDA in controlling monetization of informational analysis in DTC tests, and asked Jeff and Thomas to respond to each other?
Kayte Spector-Bagdady: The FDA can help ensure the clinical validity of devices, which will be central to their effective monetization.
Jeff Shapiro: Part of the problem identified by Thomas is that Class III devices are going through 510(k) even thought they shouldn’t be. Manufacturers don’t get to choose which pathway since 1990. The 20 hour figure for 510(k) is from 1984 and is out of date. This picture is 30 years out of date. The watershed year is 1990.
Thomas McLean: Jeff’s account is representing the manufacturer’s perspective, as opposed the physician’s perspective.
David Korn asked Jeff to talk about metal-on-metal hip transplants.
Jeff Shapiro: The FDA has made a call for PMAs. They are a Class III transition problem.
Glenn Cohen asked if the FDA is the right party to handle this problem. Is the 23andWe part troubling?
Kayte Spector-Bagdady: There are two ways to look at it. One way is to look at WGS. Need to make sure it is clinically valid, although CLIA might be better. Another is that FCC also should have role in marketing. Regarding 23andWe and genomics research generally, if it falls under the Common Rule and HIPAA, it would not just be a matter for FDA.
Unnamed question: Would PMS have helped prevent the metal-on-metal problem?
Jessica Flanigan: Post-Myriad, what direction will the DTC industry take?
Jeff Shapiro: Regarding PMS for metal-on-metal hips, the devil is in the details. Doing PMS surveillance on every device will be too extensive. How to select devices is also difficult. Finally, certain device failures occur after more than 5 years, and then failure goes up. Thus, although it sounds good, it is hard to implement.
Kayte Spector-Bagdady: Post-Myriad, 23andMe includes a limited interpretation of the mutations on BRCA1 and 2 genes.
Lewis Grossman: What about 510K creep? Isn’t is a problem that we are getting further and further away from the original device?
Jeff Shapiro: No. We want creep. It is progress. We don’t want to compare new device to original 1976 device. This doesn’t make sense. We want to compare it to the newer device, for which we have clinical experience. We only need pre-market clearance when we don’t know how it will behave. The baseline is moving because our understanding is evolving.