On August 31, 2017, the FDA issued a final version of a guidance entitled, “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices[i]” and, while the industry is still digesting it, I wanted to examine it from a different perspective, where I won’t repeat the many well-written reviews that distill its regulatory contents. Rather, I wanted to focus on the questions I used to receive from business leaders at a global medical device manufacturer when they heard about an FDA guidance that was coming or had been issued: “Why did they write it, what does it mean in English, and does it help our product teams in any way?”
At the end of 2016, major changes to the FDA rules were passed when the 21st Century Cures Act[ii] (Act) became law. At nearly 1000 pages,[iii] the Act contains a host of amendments that are beyond what we are discussing here, but this next part is where things get a little odd. Section 3022 of the Act, entitled “Real World Evidence,” instructs the FDA to “establish a program to evaluate the potential use of real world evidence” and, within 5 years of the Act, issue a draft guidance to industry on the subject. So, you might be thinking, as I was, that this new final guidance fulfills that obligation, stunningly ahead of schedule.
Well, if you thought that, then we were both mistaken.
When Congress spoke about real world evidence, they were talking about drugs, not devices. For starters, the Act defines “Real World Evidence” as “data regarding the usage, or the potential benefits or risks, of a drug derived from sources other than randomized clinical trials.” As for the required guidance to industry, the Act tells the FDA that they are supposed to explain “the circumstances under which sponsors of drugs . . . may rely on real world evidence” for two specific purposes:
1) to help to support the approval of a new indication for a drug approved under section 505(c); and
2) to help to support or satisfy post-approval study requirements.
If there were any remaining doubt that even the FDA does not view their new medical device real world evidence guidance as fulfilling any obligations from the Act, you only have to look at their public “deliverables” chart, in which the FDA transparently keeps track of their progress toward the 80-or-so specific deliverables required of them under the Act.[iv] If you scroll down to Section 3022, you’ll find about 4 items that pertain to the framework and industry guidance about real world evidence, except that you’ll see the “owner” of those items is CDER (FDA’s drug division), not CDRH (FDA’s device division). And FDA states that they intend to have a “plan” around the issue by the end of 2018. The draft and final guidance deliverables boxes are blank.
With that backdrop, we return to those questions I used to receive from my business colleagues.
Why did they write it?
Truthfully, I’m not sure, but I can speculate. I think it shows that the FDA is progressive-thinking and responding to the industry’s complaints that data generated outside of clinical trials can still be valid data for FDA purposes.
Take the most famous example from the guidance regarding how Edwards Life Sciences received approval for its Sapien 3, an artificial heart valve[v], in which the company pointed out that “FDA approval of the indication expansion was supported by real-world data collected from the Society of Thoracic Surgeons and American College of Cardiology (STS/ACC) Transcatheter Valve Therapy (TVT) Registry.”[vi] As is industry practice, Edwards issued a press release to announce FDA approval on June 5, 2017, the day the FDA notified them of it.
For those of you who have worked on similar medical device press releases, can you name a time in which the FDA issued its own statement, touting your product’s approval, nine days after you issued your press release? Now that’s progressive.
As you probably know, the FDA has maintained a blog for many years called FDA Voice.[vii] In an official blog post entitled, “How Creative FDA Regulation Led to First-in-the-World Approval of a Cutting-Edge Heart Valve,”[viii] Jeff Shuren, MD, JD, Director of CDRH, explains how the FDA was the first regulatory agency in the world to approve this new device, a vast improvement over an earlier version, where the FDA was the 42nd country to approve the device. With regard to the FDA’s approach, Shuren explained, “we worked closely with the industry on creative clinical trial designs and the use of other sources of clinical evidence that could demonstrate that the device is safe and effective when used in the intended patient population.” They even included a photograph of the lead reviewer and medical officer that handled the submission.
For this, we must give the FDA some credit, as Shuren is publicly embracing new ways of doing things that the industry likes and which can be a “win-win-win” for patients, the FDA, and the manufacturer. In the post, Shuren acknowledges that many companies launch first in Europe and then in the US, and this story justifies at least some hope that the day may come when American device-makers choose to launch first in the US when American patients represent the largest potential market segment.
What does it mean—in English?
At a high level, the guidance says that the FDA may accept non-clinical trial evidence for regulatory purposes as long as the underlying data are of appropriate quality. What might meet that standard? The only example given is that of a very robust medical device registry for a Class III implant. To be fair, there are a half-dozen examples provided, but only one of them definitively states that the FDA issued an approval decision that one can infer was viewed favorably by the manufacturer.
To understand this better, we need to add a couple of the definitions that the FDA used in their guidance. Remembering that this guidance is about medical devices and not drugs, you know it wasn’t possible to use the drug-oriented definition of “Real World Evidence” from the Act as written, so the FDA created two variations:
Real-World Data (RWD) are data relating to patient health status and/or the delivery of health care routinely collected from a variety of sources.
Real-World Evidence (RWE) is the clinical evidence regarding the usage and potential benefits or risks of a medical product derived from analysis of RWD.
While my first reaction was to wonder why it was necessary to create a data-versus-evidence distinction, I found it helpful to think of RWD as the underlying raw data while RWE is the clinical claim or conclusion that flows from the data. These new definitions also narrow the data that can be considered, from the Act’s “sources other than a randomized clinical trial” to data that are “routinely collected” and relate to “patient health status and/or delivery of healthcare.” It seems that, in anticipation of the guidance about RWE in the drug context that needs to be written before 2022, the FDA wants industry to think of data sets that are almost as good as those of a clinical trial, and it’s not clear whether EHR or billing data could pass muster.
When I read the guidance, I found about 13 places where the FDA makes essentially the same point, and, with the above definitions in mind, that point (from Section IV.A on page 9) is:
FDA will consider the use of RWE to support regulatory decision-making for medical devices when it concludes that the RWD used to generate the RWE are of sufficient quality to inform or support a particular regulatory decision. The threshold for sufficient quality will depend on the specific regulatory use of the evidence.
Now that we know what the guidance means in a general sense, we move to the last question.
Does it help our product teams in any way?
If you are an implant manufacturer that participates in a robust hospital-based American College of Cardiology registry to track the clinical usage of your device, the answer is “yes.” If you don’t participate in a Class III registry or are focused on Class II devices where there are no registries on par with clinical trial data, the answer is “probably not.” All of the FDA’s examples in the guidance involve registries and they are almost all Class III.
One way to solidify these conclusions is to examine the six examples the FDA shared:
Example #1- Registry- Potential New Indications for Use
Without speaking in terms of any specific device or manufacturer, the FDA names the Cath-PCI Registry[ix], one of the eight hospital-based registries managed by the American College of Cardiology as a type of Real-World Data that “might provide sufficient evidentiary support” for expanded labeling (indications for use), “depending on the specific devices, indications, and analyses.”
Example #2- Registry- Potential Labeling Changes
Under an approved IDE, a Class III device manufacturer is using medical society registries to supplement its data and the FDA “is hopeful that the Real-World Data may be of sufficient quality” to meet safety assurances needed to support a new use.
Example #3- Registry- To meet mandatory post-market surveillance
In cases where the FDA has safety concerns about a device already on the market, the FDA may order a Class II or III manufacturer to generate post-market surveillance reports. The FDA may order these so-called Section 522 reports for 3 years or longer, and the FDA has a guidance on this topic[x] in addition to a page showing the 26 open orders.[xi] In a Section 522 order that was issued to multiple competitors, the manufacturers of Class II devices opted to collaborate with a medical society to create a registry that the FDA says “could support future premarket submissions.”
Example #4- Registry- Earlier FDA approval
For a breakthrough Class III device that was approved based on a full clinical trial, the manufacturer, in order to support post-market commitments, supported launching a registry that would capture all patient use, allowing the FDA to provide an earlier approval “conditioned on further robust Real-World Data collection and reporting in the post-market setting.”
Example #5- Pre-existing Registry- Control Group supporting Investigational Device Exemption (IDE)
For a manufacturer following an approved IDE after the FDA determined that “clinical evidence was needed to support an approval decision,” an existing registry of similar devices was evaluated by the FDA and deemed acceptable for use as a control group within the clinical study design.
Example #6- Pre-existing Registry- Control Group
For a Class III device under review for a new indication where the manufacturer provided clinical trial data but the FDA felt there was inadequate data from the control group, the manufacturer was able to supplement with data from a pre-existing registry that was already collecting Real-World Data on the control therapies.
Overall, it appears that the FDA may have developed this guidance as preparation for a future mandated guidance on the use of Real-World Evidence to support drug approvals. Therefore, it’s no surprise that, with the regulatory and data standards of drugs in mind, a medical device analogue was focused on Class III products. If you are a Class III device maker whose products are tracked in a robust registry, this guidance is welcome news. As for the Class I and Class II products that comprise 90% of medical devices[xii], you’ll have to hold on to hope.