January 19, 2024
U.S. FDA 510(k) Clinical Data Requirements
I always reminded clients that clinical data may not be required at all for a 510(k) subject device even though clinical data may have been required for the predicate. A pre-sub(s) is generally a good idea and in all parties’ best interest for unusual scenarios rather than cutting corners, rushing, and/or guessing.
Ultimately, a 510(k) sponsor is expected to employ clinical data when necessary to meet the statutory “substantially equivalent” threshold compared to the predicate. But that is easier said, than explained or done. For example, FDA says that the need for 510(k) device clinical testing typically depends on many factors including device type, intended use, design, safety profile, and clinical experience. I’ll give some basic principles below. But a careful expert regulatory analysis is needed for the exact subject device in order to help flush out the various attributes of each case.
In guidance, FDA explains that, while substantial equivalence is generally a “comparative” concept between subject and predicate rather than a PMA’s independent demonstration of the subject’s safety and effectiveness, FDA reminds us that the principles of safety and effectiveness still underlie the substantial equivalence determination in every 510(k) review. FDA reminds us that its 510(k) review decision must reflect a determination of the level of control (e.g., clinical data) necessary to provide a “reasonable assurance of safety and effectiveness”. Therein lies a junction where FDA’s requirements for clinical data (and/or increased non-clinical data) can be triggered for a 510(k) device. Thus, the substantial equivalence concept can be awkward and ambiguous for all parties, as the goal of a 510(k) is not to show safety and effectiveness, yet, in a sense, actually is. In practice over the years, this underlying safety and effectiveness requirement for 510(k) devices appears to have caused 510(k)s to evolve into something more like mini-PMAs as compared to a few decades ago. But I digress.
The simplest and most reliable way I can think of to explain when clinical data might be needed for an unusual 510(k) device is for us to approach it statutorily; a 510(k) is after all a statutory instrument.
Specifically, clinical data may be needed when the subject device has a different intended use and/or different technological characteristics that raise new questions of safety and effectiveness as compared to the predicate. The expert regulatory strategy/analysis for each case needs to carefully draw out these considerations for the subject device as compared to the predicate.
FDA has a basic thought process it applies when reaching a decision that clinical data are needed for a 510(k) device. In a nutshell, it is when descriptive and non-clinical data are insufficient to meet the statutory substantial equivalence threshold. FDA in guidance tries to bring further clarity on this threshold by giving some examples, including new or modified indications, technological differences, and limited or inappropriate non-clinical data.
We also need to be keenly focused on the output from the subject device’s design validation step. The aforesaid purely theoretical/statutory principles need to be combined with the subject device’s own organic requirements as driven by its design control process.