The FDA’s three September draft guidances for the 510(k) program provide much-needed clarity regarding premarket notifications under the 510(k) program and are part of a set of reforms the Agency proposed in 2018. One of these—the draft guidance for 510(k) clinical data—brings several critical questions into focus, such as the circumstances in which a change in the indication for use may require clinical testing.
The draft guidance for clinical data is supplementary to the 510(k) program guidance released in 2014 and provides a more complete set of insights into the use of clinical data to establish substantial equivalence (SE) to the predicate device.
Randomized, Controlled Trial Not Always Necessary
The 510(k) program guidance states that clinical data may be necessary when the FDA determines that non-clinical data are insufficient to establish that the device is substantially equivalent to the predicate. However, the term “clinical data” is not limited to a randomized, controlled trial (RCT), but includes real-world data (RWD) sources, such as:
- Data obtained from registries;
- Data from medical records, such as electronic health records;
- Claims data from payer databases; and
- Adverse event data.
It is important to note that the FDA has aligned itself with the definition of clinical data as described in the International Medical Device Regulators Forum (IMDRF) guidance for key definitions and concepts for clinical evidence. The IMDRF states that clinical evidence for a device should be reviewed and updated throughout the total product life cycle (TPLC), both for that manufacturer’s device and for comparable devices. One difference between the IMDRF guidance and the FDA clinical data draft is that the FDA states that data obtained from human factors testing is not considered clinical data while the IMDRF does not take a position.
Different Indication for Use Does Not Always Require Clinical Data
The FDA’s 510(k) clinical data draft states that the agency will not automatically deem a device not substantially equivalent (NSE) when the indication for use is not the same as that of the predicate. However, the proposed indication for use can create problems if it conflicts with the device type’s intended use statement. There are several factors the Agency will take into consideration in this scenario, such as:
- Differences in the patient population or the disease;
- Differences in the anatomical site or structure;
- An expansion of the indication for use relative to the predicate; and
- A change to the benefit-risk profile associated with the proposed indication for use.
Another factor to bear in mind with regard to the relationship between the intended use and the indication for use is a guidance issued in 1998, the general/specific intended use guidance. The clinical data draft guidance refers to this general/specific intended use guidance as a guideline for determining whether a new indication for use remains within the governing intended use statement. The general-specific guidance lists several considerations in this context, such as whether there is a body of knowledge or evidence available to the FDA to determine whether the new indication for use still falls within the intended use statement. That body of knowledge could be the medical literature or clinical practice guidelines published by professional medical societies.
Sizes Outside of Predicate’s Design Envelope Potentially Troublesome
The second scenario in which clinical data might be required is when the device is designed with technological characteristics that differ from the predicate—such as in materials—in the design of the device, or the energy source used in the device. However, it should be noted that the overall design of the device can be identical to that of the predicate and still require clinical data if the new design exceeds the size envelope of the version cleared by the predicate.
A new size that falls within the range of sizes described in the predicate will not need clinical data so long as no new questions of safety and effectiveness are raised, and non-clinical test methods may be adequate to address any concerns. The Agency said, however, that a new device that is either smaller or larger than any existing cleared versions may in some circumstances prompt the need for clinical data because of the new boundaries of device size/volume.
The third scenario described in the clinical data draft guidance is when substantial equivalence cannot be determined by methods such as analytical, bench, or animal testing models. Should the FDA determine that non-clinical testing methods fail to fully assess device performance or are not predictive of clinical outcomes, clinical data may be needed even though a non-clinical testing model seems fit for purpose for that device. There may also be questions that are specific to the anatomy that cannot be answered without clinical evidence, such as would be seen in an application for an imaging system to be applied to a part of the anatomy for which there are no validated phantoms.
An errant indication statement could set back a project by months and add enormously to the cost of bringing a device to market.
In the fourth scenario, the FDA may request that applicants provide new clinical data regarding device safety and performance along with the 510(k) submission to address any new risks that have been identified for the predicate device. This would also apply if the rate of occurrence of a known risk has risen, and the FDA said it would discourage the use of a predicate if the predicate exhibits any new risks or an increased rate of an existing risk. Should an appropriate predicate not be available, clinical data may be required to provide a reasonable assurance of safety and effectiveness.
There are several reasons the FDA arrives at a determination of NSE for 510(k) submissions, and some of those are not entirely within the control of the developer, such as when troubling postmarket signals suddenly emerge for the predicate device. However, this new draft guidance reinforces the need to carefully consider the proposed indication for use, because an errant indication statement could set back a project by months and add enormously to the cost of bringing the device to market.
