When Should a 510(k) Include Clinical Data?

Although it seems not widely known outside of the medical device industry, FDA can require sponsors to include clinical data as part of a 510(k) submission.  Such data may be needed to demonstrate substantial equivalence to a previously-marketed predicate device or, less frequently, to show that new or modified indications for use fall within the same intended use as the predicate device.

Despite previous efforts by FDA to shed light on this issue (see The 510(k) Program: Evaluating Substantial Equivalence in Premarket Notifications [510(k)] (510(k) Program Guidance), sponsors still often find there is disagreement with FDA when it comes to determining if clinical data are needed, especially when the sponsor believes there is a reasonable justification that non-clinical data are sufficient to reach a substantial equivalence determination.

As part of a multipronged effort to strengthen and modernize the 510(k) program (see our recent blog post here), FDA recently issued a draft guidance, Recommendations for the Use of Clinical Data in Premarket Notification [510(k)] Submissions (Draft Guidance), describing situations in which clinical data may be necessary to demonstrate substantial equivalence in a 510(k) submission.  When necessary, data from clinical studies is submitted in addition to standard non-clinical data such as bench performance testing; sterilization and shelf-life testing; biocompatibility testing; animal studies; electrical, mechanical, and thermal safety testing; electromagnetic compatibility testing; software testing; and cybersecurity testing.

The 510(k) Program Guidance provides several scenarios that illustrate the most common situations in which clinical data may be needed to demonstrate substantial equivalence. The Draft Guidance provides additional context and clarity to these three scenarios and adds a new scenario in which clinical data may be needed.  The scenarios are intended “to provide broad considerations to be used by industry and FDA to help determine whether clinical data may be necessary to demonstrate that a new device is substantially equivalent to a predicate device.”

The four scenarios, described in the new Draft Guidance, in which clinical data may be needed in a 510(k) include:

1. There are differences between the indications for use of the new device and the predicate device;
2. There are differences between technological characteristics of the new device and the predicate device;
3. SE between the new device and the predicate device cannot be determined by non-clinical testing (analytical, bench, and/or animal); and
4. A newly identified or increased risk for the predicate device suggests clinical data may be needed for the new device.

The Draft Guidance provides specific examples illustrating the application of each scenario.  The examples include both diagnostic (including in vitro diagnostic) devices and therapeutic devices, since there are significant differences in the types of clinical data that may be needed for these two categories of devices.

While the first three scenarios, and the accompanying examples, seem fairly straightforward and reasonable, the newly-added fourth scenario is perplexing, as it describes a situation where clinical data may be needed for a new device even though no clinical data were needed for the predicate device. This situation is where most surprises arise as a sponsor reviews the 510(k) summary for the predicate device, sees that clinical data were not needed for clearance, and assumes that clinical data will not be needed for their similar device.

The Draft Guidance suggests not using predicate devices with newly identified risks but acknowledges that there may be some cases where there is not a more recently cleared predicate device that does not share the newly identified risk.  FDA’s examples for this scenario show a good faith attempt for a level playing field between new devices and existing devices by describing situations where new device submissions will require clinical data while the currently marketed devices with newly identified risks may be subject to postmarket surveillance studies or recalls with new submissions for modifications to the recalled device that include clinical data.  However, we can envision situations where clinical studies of the previously cleared devices with the new risk have not yet occurred and a sponsor with a new, similar device may be surprised at the need for clinical data.

Although there is some discussion in the examples that relates to whether the predicate device required clinical data for 510(k) clearance, we found that the Draft Guidance does not explicitly address the question of whether or not inclusion of clinical data in the predicate device 510(k) provides any bearing on whether clinical data will be needed for a new device.

The Draft Guidance also describes the types of data that may constitute clinical data supporting a 510(k), drawing on previous recommendations of the International Medical Device Regulators Forum  (see “Clinical Evidence – Key Definitions and Concepts”).  These include:  results of pre- and post-market clinical investigation(s) of the device (i.e., traditional clinical trials); results of pre- and post-market clinical investigation(s) or other studies reported in the scientific literature of a comparable device; published and/or unpublished reports on clinical experience of either the device in question or a comparable device; and other sources of clinical experience such as registries, adverse event databases, and medical records (e.g., electronic health records, claims).

When considering whether data from a comparable device can be used, the Draft Guidance indicates that “adequate justification regarding the applicability of such data should be provided demonstrating why such data would be representative of the new device.”  Given FDA’s current emphasis on data being collected on the final, finished device, it may be beneficial to use the pre-submission process to discuss the applicability of data collected with a comparable device with FDA to allow time for a new clinical study, if they do not agree that the justification is adequate.

Finally, it is not surprising that the Draft Guidance notes that there may be other scenarios not described where clinical data may be necessary and that the need for clinical data may also change as information on the device type is accrued.  We think that the Draft Guidance may prove helpful, especially in situations where the sponsor concludes that clinical data are necessary, at which time they can engage with FDA via the pre-submission process to discuss clinical study design.  However, sponsors should be cautious when review of the Draft Guidance suggests that clinical data are not necessary.  If there are differences in indications for use, novel technology, or the potential for differences in risk profile, a pre-submission may be warranted to confirm FDA agrees with the sponsor’s application of the scenario to the specific circumstances and the conclusion that no clinical data are needed to support substantial equivalence.