Getting Up to Speed on the Direct De Novo Process

February 19, 2016

FDAnews held a webinar earlier this month in which three attorneys from Hogan Lovells, including partner Yarmela Pavlovic, explained the FDA’s direct de novo process. The following is adapted from that event.

QUESTION: What are the procedures for a direct de novo?

PAVLOVIC: The principal way to talk to FDA about whether or not a company or a product is appropriate for a direct de novo is via the presubmission process. It allows for companies to submit a regular presubmission to talk to FDA not only about the data that may be relevant that may support clearance of the product but also about the regulatory pathway.

But there are other options. A company could opt to submit a 513(g) request instead of a full presubmission if, for example, a company were not yet ready to talk to FDA about the data that will support the product or about other aspects of the product development. There are, of course, strategic decisions that go into whether a 513(g) request or a presubmission is more appropriate.

For example, it’s sometimes easier to argue to FDA that a product is appropriate for classification in Class II when the data that will support that clearance is available to FDA and is shared with FDA, or at least the plans to develop that data. When FDA can see how robust the data is going to be they may feel more comfortable with the product, more comfortable with it staying in Class II or even in Class I. But a 513(g) request is a possibility.

It is also possible for a company to submit a direct de novo request to FDA without previously talking to FDA, but it is strongly discouraged by the agency. We would typically advise clients that it is strongly discouraged, as well, because there is no opportunity then to get feedback from FDA prior to submission about what FDA wants to see in the submission package.

The presubmission does provide the opportunity to talk to FDA not only about the regulatory pathway but about the data that will support that pathway.

QUESTION: What types of clinical data are typically required for de novo applications?

PAVLOVIC: The nature of the required clinical data depends on the device type, the risks associated with that product and the concerns the FDA has about the product’s performance. Data requirements can vary from small confirmatory studies to full randomized, controlled clinical trials.

However, it is not safe to assume that a safety-only study would be appropriate for establishing the performance of a product that’s going to go through the de novo pathway. There might be circumstances where that’s appropriate, but it’s certainly not the default assumption.

The default assumption is that the study will show that the product is safe and effective for its intended use, so establishing both efficacy and safety. Whether that is done in a statistically significant manner depends on the particular product, but it is generally assessing both performance and safety.

QUESTION: How are special controls normally established for a de novo application? Is it based on the recommendation of the sponsor?

PAVLOVIC: In general, the special controls are derived from discussions between the sponsor and the FDA. During a presubmission meeting, sponsors should propose all of the appropriate validation testing, as well as mitigations for potential risks.

That could be things like specific labeling or adherence to certain standards. It could be specific testing that isn’t in line with any particular standard but was developed by the sponsor. The sponsor then should get FDA feedback. There may be additional considerations and risks that the agency would like to see addressed.