On December 18, 2018, FDA published in the Federal Register a new docket entitled, “Notice of Intent to Consider the Appropriate Classification of Hyaluronic Acid Intra-articular Products Intended for the Treatment of Pain in Osteoarthritis of the Knee Based on Scientific Evidence.”
This document is quite unusual, to say the least. Without fanfare, FDA has suddenly announced that products regulated as Class III devices for 23 years “may” actually be drugs. FDA invites holders of device approvals to seek a classification decision the next time they want to make a change to the product.
Let’s back up and start at the beginning: Under section 201(h) of the Federal Food, Drug, and Cosmetic Act (FD&C Act) (21 U.S.C. 321(h)) a device “does not achieve its primary intended purposes through chemical action within or on the body,” among other things. HA is introduced to the synovial fluid of the affected knee joint via injection, with the intention of treating pain.
Since 1995, FDA has classified these injectable products as devices. The basis for classifying HA as a device has been FDA’s conclusion that this product achieves pain reduction in an osteoarthritic knee via mechanical or physical actions at the joint (e.g., shock absorption). The first premarket application (PMA) was filed in 1995 and approved in 1997 (P950027). Since that time, there have been 198 separate approvals of either PMAs or PMA supplements (mostly the latter). The most recent one was 12 days ago, on December 6, 2018 (P080020 S031).
Now, FDA has suddenly issued a notice suggesting a new scientific theory that would necessarily require classifying these products as drugs. FDA’s summary of the notice (p. 1) states:
The Food and Drug Administration (FDA) is announcing our intent to consider the appropriate classification of hyaluronic acid (HA) intra-articular products intended for the treatment of pain in osteoarthritis (OA) of the knee. Although HA products intended for this use have been regulated as devices (Procode MOZ; acid, hyaluronic, intra-articular), the current published scientific literature supports that HA achieves its primary intended purpose of treatment of pain in OA of the knee through chemical action within the body [which would require classification as a drug]. Because HA for this use may not meet the definition of a device, sponsors of HA products who intend to submit a premarket approval application (PMA) or a supplement to a PMA for a change in indications for use, formulation, or route of administration are encouraged to obtain an informal or formal classification and jurisdiction determination through a Pre‑Request for Designation (Pre-RFD) or Request for Designation (RFD), respectively, from FDA prior to submission. If a sponsor believes their product meets the device definition, they may provide relevant evidence in the Pre-RFD or RFD.
As a procedural matter, this approach does not seem sensible. If FDA believes that new science requires the agency to re‑classify HA injectables for the knee as drugs, then it would make more sense to announce the concern, conduct a public proceeding to gather all the appropriate information and views, and make a class‑wide decision, with clear guidance on the transition and implementation of the decision. The sudden change without public input will deprive FDA of receiving broad feedback.
Instead, FDA has expressed doubt about the classification of all these products but only “encourages” PMA holders to seek classification decisions via the Pre‑RFD or RFD processes, and only if they are making a change to their product. This approach seems likely to result in inconsistency and expense, with little clarity. Will these RFD reviews really have pre‑determined outcomes, based on the notice? Nothing in the notice suggests that the science FDA is citing would apply to some HA products but not others. If so, why bother with case‑by‑case decision making at all? Why not take a class‑wide approach, and solicit comments?
Practical questions come to mind. If a PMA‑holder complies with the Quality System Regulation (QSR) for device manufacturing, would they be required to revamp their operations to comply with the Good Manufacturing Practice (GMP) regulation for drug manufacturing? That could be expensive and time‑consuming. How long would they have? Would companies submit Medical Device Reports (MDRs) for one version and Adverse Event Reports (AERs) for a product with new labeling deemed a drug? It is more than a little strange that the notice does not try to anticipate and answer the obvious practical questions that PMA‑holders will immediately ask.
Under the FD&C Act, a Pre‑RFD/RFD request is voluntary. It will be interesting to see whether the Pre‑RFD/RFD process will remain truly voluntary or if the Center for Devices and Radiological Health (CDRH) will refuse to process new PMAs and PMA supplements for HA products unless the applicant first obtains a Pre‑RFD/RFD decision. If CDRH does continue to process PMAs and PMA supplements for HA products, it would probably be advisable for PMA holders to simply skip submitting a Pre‑RFD/RFD request, since there is only a downside, which is reclassification as a drug.
Finally, if new science may necessitate upsetting long‑settled industry regulatory expectations, can we now expect that devices now inappropriately regulated as drugs will be moved to their rightful domain as well? If this type of notice can be issued announcing an intent to consider reclassifying devices as drugs, then perhaps we will see additional notices in the coming years, announcing an intent to consider reclassifying drugs as devices. Over the years, we have expressed our concerns about the product jurisdictional process (for example, here). This notice only reinforces our concerns.