If a surgeon combines a PMA device with a 510k device, what does that do to federal preemption? We ask this question because one of the more fascinating aspects of the medical device world is modularity. We have heard of modular furniture, modular homes, modular computers, and modular hamster cages—collections of components that can be matched together in different ways to form the desired end result. The highest and best use of various components could be different for each user, depending on his or her particular needs and vision. After all, why should two different families own the same three-seat sofa when one would prefer four seats, with a recliner, and perhaps a built-in cooler, and maybe add the matching foot stool?

When you look at it that way, many medical devices (and particularly orthopedic devices) would have to be modular to some extent to offer the greatest benefit to patients. A defining principle of healthcare—and the laws that govern it—is that different patients have different needs that require a physician’s best judgment and exercise of discretion. That was the situation in Nagel v. Smith & Nephew, Inc., No. 3:15-cv-00927, 2016 U.S. Dist. LEXIS 98408 (D. Conn. July 28, 2016). In Nagel, a surgeon treated his patient with a hip replacement system. But instead of using the plastic cup liner that came with the system, the surgeon elected (with his patient’s consent) to use a metal liner from a similar system made by the same manufacturer, thus forming a “metal on metal” system.

The fascinating part arose because the FDA cleared the hip system under the 510k process as substantially equivalent to a predicate device. But the metal liner was approved under the FDA’s rigorous premarket approval process. Faithful readers know where this is headed—express preemption, which is strongest when applied to PMA devices like the metal liner.

Turning back then to our introductory question, what do we do with an implanted device with “mixed levels of approval”? The short answer is that the plaintiffs’ claims were all preempted, even though significant components used to treat the patient were 510k products. Here is how the court came to that result.

The court first walked through the now-familiar dance between express and implied preemption. On the one hand, the FDCA expressly preempts any state law requirement that is different from or in addition to federal requirements. See Section 360k(a). Thus, when suing over a PMA device, a plaintiff must allege the breach of a state-law duty that is the same as federal requirements—the so-called parallel claim exception. On the other hand, the plaintiff’s claim must not rely only on a requirement already imposed under the FDCA, or else the claim is impliedly preempted under Buckman. That’s the “narrow gap” through which a plaintiff must navigate. Id. at **10-11.

The key of course is that the “narrow gap” closes down on the plaintiff only when dealing with a PMA device (with exceptions that we won’t cover here). When dealing with “mixed levels of approval,” different courts have gone different ways. Id. at **11-12. But the court in Nagel decided that express preemption applied:

I am persuaded by the initial line of cases that applies the MDA’s express preemption clause to claims involving device components subject to mixed levels of approval, particularly in the context of the allegations that are present in this case, which focus on harm allegedly caused either independently by the PMA-approved R3 metal liner, or by the R3 metal liner’s interaction with the R3 system (and as distinct from any harm stemming from the § 510(k)-approved R3 system not related to the PMA-approved R3 metal liner). Because the FDCA does not prohibit off-label use of a component within a PMA-approved device in another medical device, I conclude that a manufacturer should not lose the protections of express preemption under § 360k(a) because of a surgeon’s permissible choice to use the PMA-approved component in a § 510(k)-approved device, even if the interaction of the two components causes problems.

Id. at **12-13. Let’s unpack that a little. It’s basically a two-step analysis. The court first reasoned that the harm allegedly flowed from the PMA-approved metal liner, either independently or from its interaction with other components. On this score, the court took care to distinguish (without deciding) a situation where the alleged harm was not related to the PMA device. Id.

Second, the court explained the policy behind its decision: The law does not prohibit off-label use, which could include the use of a PMA device with another system for which it was not approved. According to the court, a surgeon’s “permissible choice” to use a PMA-approved component within a 510k cleared system should not undermine the express preemption protection to which the product’s manufacturer would normally be entitled. Id. at *13. The FDCA defines “device” to include “any component, part, or accessory.” Id. at *14 (citing 21 U.S.C. § 321(h).

So express preemption applied, which led to a claim-by-claim analysis of the plaintiffs’ complaint. On strict product liability and failure to warn, the plaintiffs did not allege facts constituting the violation of any FDA requirement, leaving them no basis for a “parallel claim.” Id. at **16-20. They alleged only that metal-on-metal hip replacement systems had higher-than-desired revision rates, not that the design or manufacture of the devices violated federal requirements. Id. at **17-18. On failure to warn, the plaintiffs alleged that the manufacturer had failed to comply with the FDA’s requirements for adverse event reporting (an apparent attempt to echo the Ninth Circuit’s wrongly decided Stengel v. Medtronic opinion), but they alleged no facts to back up that allegation, nor facts that any alleged failure to report caused the plaintiffs’ injury. Id. at **19-20.

There was no parallel claim for misrepresentation based on off-label promotion because “[t]here is no federal claim for off-label marketing, so no state law claim can survive.” Id. at *21. Plus, even if there were a federal claim for false and misleading off-labeling marketing, the plaintiffs had not sufficiently alleged facts that the marketing actually was false or misleading. Id. at **21-22. The negligence and implied warranty claims fell for similar reasons—they sought to impose requirements in addition to federal law—testing requirements, reporting requirements, manufacturing requirements, and the like. Id. at **22-24. None of those could support “parallel claims.”

The final blow was an order denying leave to amend as futile. The plaintiffs had already taken one opportunity to amend, and they had “not indicated what changes [they] would make to survive a future challenge . . ., but rather steadfastly assert[ ] that [their] first amended complaint is sufficient.” In an environment in which we find courts granting leave to amend all too liberally, the court’s reasons here is sound. Plaintiffs already had two chances, and having stood on their amended complaint, their journey was over.