Comparing the FDA 510(k) and De Novo Device Pathways
Check out the similarities and differences of two different paths (510(k) and De Novo) for getting pre-marketing authorization from the FDA for a software-as-a-medical-device.
Those determined to bring life-changing medical devices into clinical practice must navigate the complex pathways to secure the regulatory authorization that is an essential step in the journey. I have spent the past several years learning the ins and outs of the process and have already shared some of my lessons learned about the importance of building a relationship with the FDA and effectively engaging with them on this blog. Yet, I am still getting regular requests to unpack some of the distinctive differences between the primary medical device clearance path, the 510(k), and the less-common path required for truly novel devices. So let’s step through what I have gleaned from walking both routes for Fith Eye’s Analytic for Hemodynamic Instability (AHI) software-as-a-medical-device (SaMD), a medical device consisting exclusively of software.
What is a medical device anyway?
Medical devices are “things” (as opposed to chemicals that might be metabolized, aka drugs) that are intended to diagnose or treat disease or affect the structure or function of a person’s body. Medical devices can range from tongue depressors to implantable pacemakers, along a spectrum of human risk ranging from high to low. For moderate or high-risk devices, FDA review and authorization is required before marketing the device for use. Our software device, AHI, is a moderate-risk device intended to identify signs of hemodynamic instability (an important indicator of patient deterioration) or stability using a single lead of ECG (which measures the heart’s electrical signals) as its only input. Given AHI’s use as a monitor continually providing adjunctive clinical information to doctors and nurses to help them determine the best course of care for hospitalized adult patients, AHI meets the FDA’s definition of a medical device. My lesson learned here is that figuring out if your product meets the definition of a medical device, especially for a SaMD, often pivots on subtle distinctions and precedent, so it is wise to get an assessment from a knowledgeable regulatory consultant or even the FDA themselves if you are unsure. Since getting a regulatory clearance is time-consuming and expensive, you only want to do it if you are required to, however, if you are required to, you definitely do not want to do something illegal like marketing a medical device without authorization.
Is there a predicate?
Once you determine that your product meets the legal definition of a medical device, you need to determine if the FDA has already established a product classification for your device. While you can look at the regulations themselves, it is often easier to look around the marketplace and see if something similar could be considered “the same” from a regulatory perspective. Another way to think about it is to ask whether another device presents the same safety and effectiveness considerations and has a comparable technological basis as your device. If you find one or more, you have identified a potential “predicate” device that can be used as the basis for a 510(k) clearance.
The traditional 510(k) process provides clearance for marketing based on whether your device is considered “substantially equivalent” to another device that is already on the market. For a new device, it typically takes three to six months to complete the regulatory review process on top of whatever time is required to develop the product and gather the evidence required by the FDA. While many 510(k)s do not require the submission of clinical performance data since they are often new modifications of already cleared products, new SaMDs embodying a new type of analytic like AHI do.
The tricky bit here is that there is virtually never an identical device on the market for a newly developed device. Honestly, if there were, what would be the point of creating your device to simply deliver an identical value proposition? So, there is an inherent fuzziness in the “substantial equivalence” process. Since a 510(k), if possible, is the fastest, most expeditious regulatory route, it is the path that most will strive to use. For Fifth Eye, we identified several potential predicates, but we knew they were a bit of a stretch (somewhat different indications for use and definitely different technologies); however, along with our investors, the FDA was definitely encouraging us to find and make a case for a predicate. We actually tried three different possibilities, via two pre-submission meetings, with the FDA. A couple of those possibilities were even suggested by FDA itself. However, when we made our substantial equivalence argument, FDA ultimately indicated that we were not close enough. This “find a predicate” journey ultimately failed – and added about a year to our process since the communication cycles for two pre-submission meetings take many months.
Nevertheless, I highly recommend pre-submission meetings when you are charting new paths. It would have been far worse to just commit and submit given what a stretch we were trying to accomplish because it would have cost much more in fees and time, and likely would not have been successful. Once it was clear to us and to the FDA that we really could not make a case for a substantially equivalent predicate, we turned our focus to the FDA’s De Novo process.
No predicate?
If there is no predicate for your device, then you may be eligible for FDA’s De Novo pathway. This is a possible route to create a new classification for novel devices of low to moderate risk. The De Novo pathway is newer, much less traveled, and requires more information to establish that your device’s benefits outweigh its risks, rather than just that it is substantially equivalent to another device (that presumably has already established that such a device has benefits that outweigh the risks). You must carefully evaluate the risks of your device and make the case to the FDA that it is both safe and effective. Clinical data is much more commonly expected for this argument. Because of the additional complexity and importance of the FDA’s evaluation of a novel device, the evaluation process takes much longer (a year plus is typical, but the variability for a De Novo is much greater than for a 510(k) and can stretch years).
Ultimately, it took Fifth Eye about six months to prepare our submission with the required clinical study (which was quite streamlined) and human factors study. Then another 11 months of FDA review (with at least five major FDA interactions) to secure the granting of a new De Novo classification. This process created an entirely new regulatory category, with our device as the first of a kind. By the time it was over, we felt like the FDA had asked us to slice and dice our data in every conceivable way – and honestly, it ended up building our confidence in our analytic ever higher as every time we did yet another post-hoc analysis, the data confirmed that our device was performing as expected and in a way that demonstrated that our device was measuring a valid and significant signal from the patients we had tested it on. It also made the rigor of the FDA’s review very clear.
More than one predicate?
For complex devices, especially in SaMDs, sometimes you combine functions that relate to more than one predicate in what is known as a combination device. At Fifth Eye, now we are marketing our first product as we continue work on our next-generation product. We are nearly ready to submit our first 510(k) for an expanded AHI System, and this time not only will we be submitting a modification to our existing AHI analytic (new features and functionality) with our own De Novo-granted device as a predicate, we will also add a new analytic, with another on-the-market device as a secondary predicate. This adds complexity to the submission, but can result in broader and better FDA-cleared Indications for Use. We will know how the FDA views it in six months or so.
Advantages and Disadvantages of the 510(k) Path
The clear and obvious advantages of the 510(k) path are that it is more known, predictable, and shorter. These reduce risk. However, for a newly developed device, the whole structure of the “substantially equivalent” case that must be made means that the predicate’s intended use, indications for use, and special controls define the basis of comparison. Those characteristics may be a poor fit for the new device and may even mean that the unique advantages of the new device must be minimized. At the end of the day, the goal of a substantial equivalence case is to show that your device is “the same” as the chosen predicate.
Advantages and Disadvantages of the De Novo Path
Many seek to avoid the De Novo path because it is less well-understood, takes longer, and they fear the FDA’s rejection. Investors often amplify these fears. However, my experience indicates that, if there is no good fit from a predicate perspective to support a 510(k), you are likely to get declined on your 510(k), and the De Novo path may well be a better, less risky fit.
In addition, the De Novo path allows you to propose your own, custom-tailored intended use, indications for use, and special controls (within FDA’s norms). That means that you get to put your device’s best foot forward, showcase its best capabilities, and establish a standard that others will need to match to leverage for follow-on 510(k). It also shows that your device is distinct and novel while creating a competitive barrier. In many ways, while more material had to be prepared for our De Novo submission, it was easier to prepare because we were not trying to fit our device into someone else’s mold, and we could tailor our clinical study to fit our device’s inherent characteristics.
While this post distills some of what we learned over many interactions with the FDA and thousands of pages of submission material, we know we continue to have much to learn. I will continue to invest in learning and growing our regulatory knowledge and capability – and commend this approach to others. Keep in mind that there is a reason that investors fret about the risks of the FDA. Missteps can create significant delays, and it is certainly possible for the FDA to decide you have not met their standards and send you back to the drawing board. On the other hand, because the process of securing the FDA’s authorization to market your product is challenging, it also creates a competitive barrier and value when you can execute it well.
