We were reading recent cases (yeah, we do other things, too) and part of the discussion of removal in In re Aredia & Zometa Products Liability Litigation, 2007 WL 649266 (M.D. Tenn. Feb. 27, 2007), caught our eye. We’re not talking about the failure of so-called “complete preemption.” Id. at *3-4. That lost and we haven’t had very much success with it ourselves in the last decade.
No, what was interesting to us was the argument that the case presented “substantial federal issues” within the meaning of Grable & Sons Metal Products, Inc. v. Darue Engineering & Manufacturing, 545 U.S. 308 (2005) – and particularly what the court said with respect to Grable and fraud on the FDA.
Grable, of course, to all those federal civil procedure fanatics out there (all half-dozen of you), was a recent Supreme Court case that relaxed the seemingly absolute standard the Court had previously set Merrell Dow Pharmaceuticals, Inc. v. Thompson, 478 U.S. 804 (1986). Thompson had held, that – absent an independent federal cause of action – the presence of a federally-based claim (such as the FDCA-based negligence per se alleged in Thompson) or defense (most prominently preemption) isn’t enough to create federal question jurisdiction. Grable punched a little hole in that reasoning.
Grable involved a state-law quiet title action that turned on whether a federal agency (the IRS) had failed to carry out its responsibilities under federal law (to provide adequate notice before seizing the property of a tax delinquent). A federal claim wasn’t required to create federal question jurisdiction where the “state-law claim necessarily raise[s] a stated federal issue, actually disputed and substantial, which a federal forum may entertain without disturbing any congressionally approved balance of federal and state judicial responsibilities.” 514 U.S. at 314.
Thus even though there was no pretense of a federal cause of action, the Court recognized federal question jurisdiction where: (1) the federal question “important,” (2) it was the “only” seriously contested issue in the case, (3) a federal forum was needed to “vindicate [federal] administrative action,” and (4) the likely recurrence of the question was “rare.” 545 U.S. at 315.
The last observation – rarity – was probably the most important, since the court went on to distinguish Thompson as a case involving a “missing welcome mat” because a contrary finding in Thompson “would have attracted a horde of original filings and removal cases raising other state claims with embedded federal issues.” Id. at 318. See also Empire Health Choice Assurance, Inc. v. McVeigh, 126 S. Ct. 2121, 2137 (2006) (describing Grable as one of a “special and small category” of cases); Nicodemus v. Union Pacific Corp., 440 F.3d 1227, 1237 (10th Cir. 2006) (affirming Grable jurisdiction over railroad land grant litigation where effect upon federal caseload would be “microscopic”).
With the Court’s explicit reference to “rarity” efforts to construe Grable as in effect overruling Thompson as to negligence per se-type allegations have not been successful. E.g., Zahora v. Precision Airmotive Corp., 2007 U.S. Dist. Lexis 17155 (E.D. Pa. Mar. 8, 2007) (allegations of compliance with comprehensive federal regulations (FAA) and a preemption defense do not fit Grable); Merchant ex rel. Berry v. Hueser, 2006 WL 2246436, at *5 (W.D. Mo. Aug. 4, 2006) (allegations that drugs were “adulterated,” “misbranded” and “mislabeled” did not confer Grable jurisdiction); Weiss v. Fujisawa Pharmaceutical Co., 415 F. Supp.2d 720, 721 n.3 (E.D. Ky. 2005) (FDCA-based negligence per se allegations alone did not confer Grable jurisdiction); Caggiano v. Pfizer, Inc., 384 F. Supp.2d 689, 691 (S.D.N.Y. 2005) (same).
On the other hand state-law based suits filed by states that seek reimbursement for Medicaid payments have been successfully removed to federal court under Grable. State of West Virginia ex rel. McGraw v. Eli Lilly & Co., 2007 WL 685587, at *4 (E.D.N.Y. Mar. 6, 2007); Franklin v. Merck & Co., 2007 WL 188264, at *3 (D. Colo. Jan. 24, 2007) (transferring rather than remanding); In re Zyprexa Products Liability Litigation, 375 F. Supp.2d 170, 172 (E.D.N.Y. 2005); contra Alaska v. Eli Lilly & Co., 2006 WL 2168831, at *3 (D. Alaska July 28, 2006). These cases involve compliance with a host of federal Medicare reimbursement statutes as well as FDCA violation claims.

Unlike the situation addressed by the Merrell Dow court, where finding federal jurisdiction over a state cause of action that implicated a federal standard might have led to a significant rise in the caseload of the federal courts, this case does not pose such a threat. At issue here is not simply a federal standard, but also the added factor of an intricate federal regulatory scheme including detailed federal funding provisions, requiring some degree of national uniformity in interpretation.

West Virginia v. Eli Lilly, 2007 WL 685587, at *4 (Grable citation omitted). These Medicare reimbursement actions are a relatively discrete area of litigation and thus don’t threaten the congressional “balance” the way mere compliance or negligence per se as a basis for federal question jurisdiction would. Cf. In re Pharmaceutical Industry Average Wholesale Price Litigation, 457 F.Supp.2d 77, 80 (D. Mass. 2006) (coming to similar conclusion with litigation over federal Medicare statutory definitions); but see Hawaii v. Abbott Laboratories, Inc., 2006 WL 3457617, *12 (D. Haw. Nov. 30, 2006) (wholesale price definition insufficient to confer Grable federal jurisdiction); State of Wisconsin v. Abbott Laboratories, 390 F.Supp.2d 815, 823-24 (W.D. Wis. 2005) (same).
The fraud on the FDA claim in Aredia falls somewhere in between, and the court gave the claim a long, hard look before ultimately deciding to remand. There was no set prohibition, rather “the Supreme Court has made it clear that each case involving an alleged violation of the FDCA must be reviewed to determine whether it should be distinguished on its facts from those in Grable and Merrell Dow.” 2007 WL 649266, at *6. The court then conducted an in depth analysis of preemption of punitive damages claims alleging fraud on the FDA and held that such claims were almost certainly preempted. Id. at *8-9. Another Grable element was satisfied because “both parties concede that the federal defense is the only question truly at issue.” Id. at *9.
But all that still wasn’t enough to sustain federal question jurisdiction because there was no necessity that the state court decide the ultimate fraud on the FDA question:

Where Plaintiffs seek punitive damages under the New Jersey statute in either state or federal courts and the issue on fraud-on-the-FDA is raised, the potential would exist for the FDA’s personnel to be drawn into those controversy on a case-by-case basis over and over again. This would generate a wholly impractical situation. . . . whether in federal or state court, the FDA will need to make a finding on the matter of fraud before the Plaintiffs will be entitled to recover punitive damages under New Jersey law. With all of this said, it is the Magistrate Judge’s belief that Plaintiffs’ claims for punitive damages will not require the New Jersey state court to find fraud-on-the-FDA and as such, no substantial federal issue is present which warrants federal jurisdiction. Rather, the Defendant will have to present a defense to the New Jersey state court.

Id. at *9-10 (emphasis added). So the case went back to state court.
Suppose, however, that the critical assumption was wrong. What then? One reason defendants like being in federal court is that federal judges are more likely to respect federal defenses like preemption. If that’s right (and we know from experience it is), consider what happens when, after remand, a state court rejects preemption of the fraud on the FDA-based punitive damages claim and sets the cause for trial. At that point the punitive damages claim will require the state court to adjudicate fraud on the FDA, with all the attendant issues that such claims raise as to FDA primacy, diversion of resources, discovery, etc.
One thing to consider – remove it then. This is federal question jurisdiction we’re talking about here, not diversity. The one-year limit in 21 U.S.C. §1446(b) doesn’t apply to federal questions. Once the adjudication of fraud on the FDA becomes “necessary” instead of just possible, the case is a lot closer to Grable. It’s also a lot rarer too, which is perhaps even more important. Many state courts will respect preemption of fraud on the FDA claims, and in many more cases, there won’t be any evidence to support the allegations and the claim will fail at the summary judgment stage (if not before) for simple failure of proof. The denial of the preemption summary judgment motion becomes the “paper” establishing the criteria for removal. We don’t particularly like “snap” removals on the eve of trial (we’re old enough to remember that drill from before the amendment to §1446(b)), but we like going to trial in front of a state-court judge with a mindset that allows fraud on the FDA go to a jury even less.
Once the need to actually litigate fraud on the FDA on the merits becomes unavoidable, then there’s a much better argument that Grable’s strict requirements are met than at the outset of litigation when all we have to go on is a few allegations in a pleading. So don’t forget about removal just in such cases just because it doesn’t succeed (or isn’t possible) at the outset of litigation.