The regulator is looking to remedy a lack of available scientific evidence to support pediatric device indications in PMA applications and HDEs by encouraging the extrapolation from existing clinical data, hoping the approach will streamline the requirements for establishing a pediatric intended use claim.

The FDA issued guidance aiming to increase the availability of pediatric devices by explaining when existing clinical data can be leveraged, describing how it will determine whether extrapolation is appropriate — and to what extent it can be used — and going over statistical methodology. Maintaining its risk-based approach, the FDA notes the criteria on which the regulator is basing its decision of whether to extrapolate are “considered separately for effectiveness and for safety.”

The FDCA defines pediatric device patients as “persons aged 21 or younger at the time of their diagnosis or treatment.” In 2004, the CDRH issued final guidance, Premarket Assessment of Pediatric Medical Devices, which states existing data can be used to support effectiveness and, on a limited basis, safety for PMAs when it is consistent with scientific principles. The FDAAA specifically authorized the use of adult data to show pediatric effectiveness.

In releasing Leveraging Existing Clinical Data for Extrapolation to Pediatric Uses of Medical Devices, the FDA says there are also cases where extrapolating for safety may be appropriate. Extrapolation can make use of existing clinical data that may be helpful in understanding device performance in pediatric patients, but the FDA notes that sponsors are limited to extrapolating from adult data only in situations where “the course of the disease or condition and the effects of the device are sufficiently similar in adults and pediatric patients.”

The FDA will consider various types of existing data sources for extrapolation, including data from a variety of clinical investigations, historical clinical data, reference samples and published literature.

In its guidance, the regulator distinguishes full from partial extrapolation, and goes over when extrapolation is suitable to support effectiveness, safety or both — again noting that decisions relating to safety and effectiveness are made independently. The document also goes over the extrapolation decision process, using a decision tree to illustrate how to determine whether extrapolation is appropriate. Statistical methodology is also covered in the guidance, which notes that Bayesian methods are “quite applicable for partial extrapolation from prior adult studies.”