SAN DIEGO – The chief of the FDA device center painted a picture of a world freed from hoary regulatory tools and positioned to bring 21st Century devices into a world that is already a seventh of the way through the century in a town hall talk at AdvaMed.

Jeff Shuren, director of the FDA Center for Devices and Radiological Health (CDRH), reiterated the oft-stated CDRH vision of access to high-quality, important devices in a first-in-the-world manner. "It's not about beating other countries," Shuren cautioned, but simply an effort to reduce regulatory hurdles that add no value.

Shuren recapped the center's strategic priorities for 2014-2015, and said the CDRH does pay attention to issues such as IRB problems despite that those are not within the agency's formal purview. The agenda items for 2016 and 2017 seem to double down on ta mission to align its policies with tools that enable a different approach to the clinical trial enterprise, among other priorities.

"We're going to look at growing our early feasibility study program," Shuren said, pointing to electronic sources of routine clinical care data that can be used to reduce the premarket liability "in lieu of more traditional approaches." He indicated an interest in trimming down the volume of data as well, stating, "most of the things measured are never used, so can we get to measuring fewer things, which means [clinical] studies that cost less."

"We effectively down-classified class II devices to class I," on a number of occasions as well, Shuren bragged, adding that the expedited approval program (EAP) has generated 19 requests. The agency has approved 11 of these and demurred on five others, while the remaining three are at present under review. However, Shuren posed the question of whether the EAP program should apply to more than breakthrough devices.

"Reimbursement models ... have become a barrier" to knowledge generation, Shuren stated as he commenced with a description of a novel paradigm for medical device life cycle management.

"We need more of a learning medical device system based on a neural network model," he said, which "with the support of reimbursement models is a game-changer." Moving into this next phase of regulation and device development requires work on regulatory paths, the further establishment of knowledge commons, and a national evaluation system based on the ongoing national surveillance system effort.

In contrasting this new regulatory paradigm with the existing order, Shuren said the next phase would revolve more around the concept of a determination of the extent of patient access rather than the current binary pre-/post-market model. Were such a model "adopted on the reimbursement side, you [would] change how you think about clinical studies," as well, Shuren mused.

Patients not P values

In keeping with the patient-centric focus of the modern FDA, Shuren said, "let's not be looking at p values. Let's think about patient values." This requires that the agency remind itself, "if it doesn't add value" to regulate the product, "don't regulate it." Taking this idea into clinical trials would lead one to "tailor it with a risk-benefit paradigm," which Shuren noted is what fostered the early feasibility study program. This mindset requires that a regulatory agency routinely ask itself, "when can you accept more uncertainty about benefits and risks?" he acknowledged.

Among the possible next steps in this rework of the regulatory system is development of a progressive approval approach, or perhaps some sort of adaptive licensing concept. Shuren said the potential value of a device should play a larger role in how the FDA regulates the device's entry into the market than is currently the case, but he said the national evaluation system – essentially a more ambitious version of the nascent national surveillance system – is needed to facilitate this kind of regulatory overhaul.

The next-gen sequencing dilemma is well described, Shuren said, acknowledging that interpreting the data from massively parallel and other sequencing technologies will be at least half the task. "The answer for us is a knowledge commons" with quality checks, a notion that is already in place in at least a preliminary sense thanks to existing repositories of analyzed genetic data. The idea going forward is to continue "letting the clinical community crowdsource" the science, thus allowing a claim for a diagnostic to be backed with these accumulations of knowledge, he said.

Shuren said the national evaluation system will need unique device identifier data as well as claims data to provide the information needed to make this grand scheme work. "This is not an FDA system," he cautioned, reminding the audience that clinicians and patients are at the table as well, but the information thus obtained is critical if the FDA will ever find the breathing room it needs to defer a greater chunk of the data collection task into the post-market period.

Shuren brought up again the general reluctance of patients to enroll in studies of devices that are already widely available, but he pointed out that a national evaluation system can reduce or eliminate this headache. He acknowledged that the cost of the TVT registry for transcatheter aortic valve devices is no trifling matter, but he reminded attendees, "we have stopped asking for post-approval studies" for such devices because of the TVT registry.

As is the case with electronic health records, registries have an interoperability overhang, a problem Shuren acknowledged would not lend itself to quick resolution. Nonetheless, Shuren made an impassioned case for this med tech learning system. He presented a chart that depicted the volume of data collection as shifting from a roughly half-and-half ratio at the present to a system in which a sizeable majority of the data collection would take place after a device makes it to the clinic with a more or less broad set of indications for use.

"Our ecosystem, we think, can have an even brighter future if we're willing to take some risks," Shuren said. He made reference to the ongoing series of federal budget tribulations, however, and said the 21st Century Cures program could exacerbate the resource problem at CDRH. Should the legislation be signed into law without an appropriate fiscal bolus, the agency would have to "divert resources, because if Congress says do it, everything [else] gets trumped," he said.

"If we're going to get things done," industry has to recognize that "we don't do anything in a vacuum. We need partners," Shuren concluded.

Published: October 9, 2015