(Cross posted from Objective Intent) Recently I spoke at the annual meeting of the Food and Drug Law Institute (FDLI) on Eagle v. Azar, which is currently on appeal to the D.C. Circuit.  At first blush the case seems of limited importance, because Eagle Pharmaceuticals is simply challenging FDA’s interpretation of statutory language that has since been amended.  But reading through the litigation papers reveals a more interesting disagreement between the parties, about what a court should consider, when assessing whether a statute clearly answers a particular legal question.  I will unpack this after the jump.  Warning: this is more of an essay than a blog post. I start with a TL;DR. TL;DR:  This is a fight over whether Eagle gets 7 years of market exclusivity for its new drug for two rare cancers.  FDA says no, because Teva already had 7 years of exclusivity for exactly the same drug.  The case isn’t all that important for FDA law, because Congress has since changed the statute.  But administrative law scholars may like the case.  The question is whether the statute permits serial awards of exclusivity.  The court of appeals will first ask whether the statute clearly answers that question.  If the court looks just at the narrow provisions of law at issue, it will agree with Eagle.  If the court dives more deeply into the overall statute, how the various provisions fit together, and the purpose of the scheme, it may well agree with FDA.  So the case tees up a basic issue in administrative law: what does it mean to say that a statute clearly answers a question? Unfortunately the court of appeals probably won’t answer that question, but it will at least illustrate its approach. Background The dispute relates to whether Eagle is entitled to “orphan exclusivity,” a seven-year period of protection from competition under the Federal Food, Drug, and Cosmetic Act (FDCA). “Orphan” drugs are intended to treat rare diseases and conditions.  During research and development, a company applies to have its drug “designated” as an orphan drug.  Ordinarily when FDA approves an orphan-designated drug, it awards orphan drug exclusivity.  This is conventionally described as meaning FDA cannot — for seven years — approve another application for the same drug for the same disease, meaning no generics and also no innovative products. Congress added orphan drug exclusivity to the FDCA in 1983.  The undisputed purpose of this provision is to provide a financial incentive for companies to invest in the development of drugs that would otherwise not be developed, because the market for their use is too small to be profitable. The Statute and Regulations The statutory exclusivity provision, § 527 of the FDCA, codified at 21 U.S.C. § 360cc,  was short:

If the Secretary approves an application filed pursuant to section 355 of this title . . .  for a drug designated under section 360bb of this title for a rare disease or condition, the Secretary may not approve another application  . . . for such drug for such disease or condition for a person who is not the holder of such approved application . . . or of such license until the expiration of seven years from the date of the approval of the approved application . . .

There are two exceptions.  FDA can approve a second product that is “such drug” for “such disease or condition” if (1) the first company consents in writing, or (2) the first company cannot produce enough to meet the needs of the orphan population.  The agency’s implementing regulations (which date to 1992) focus on whether the active moiety was previously approved and add that a second drug is not the “same” (and thus “such drug”) if it is clinically superior. In short, if FDA approves a new drug application (NDA) . . . for a drug designated under § 526 of the FDCA [which is 21 U.S.C. § 360bb] . . . for a rare disease or condition, . . .  it may not approve another application . . .  for the same drug for the same disease or condition . . . until the expiration of seven years from the date of the approval of the approved NDA.  But a clinically superior product with the same active moiety is not considered the same drug, so it is not blocked by the orphan exclusivity. How does a company earn orphan designation? As indicated, that’s governed by a different statutory provision, § 526, and different regulations.  FDA will designate a drug as an orphan drug if the drug is intended to treat a rare disease or condition and there is a medically plausible basis to expect the drug will be effective in doing so.  Also, if the drug is otherwise the same as an already approved drug for the same orphan disease, FDA requires a medically plausible hypothesis that it’s clinically superior — i.e., not actually the “same” drug.  With orphan designation, a company can get subsidies for its clinical trials as well as a tax credit, and at the end of the process FDA will waive the user fee for its marketing application. The Disputed Agency Decision There are two products at issue: Treanda and Bendeka.  Both contain bendamustine, and both are labeled for the same rare lymphocytic cancers.

• FDA approved Treanda (bendamustine) in 2008 and awarded seven years of orphan drug exclusivity.  This means FDA could not approve Bendeka during the Treanda exclusivity term unless (1) Teva couldn’t supply the market, (2) Teva agreed, or (3) Eagle proved that Bendeka was clinically superior.  What did Eagle do?  It secured a waiver from Teva.
• But Eagle didn’t just want approval during Treanda’s exclusivity term.  It also wanted exclusivity itself.  At Eagle’s request, in July 2014, during Treanda’s exclusivity term, FDA designated Bendeka as an orphan drug for the same cancers.  Eagle had presented a plausible hypothesis that Bendeka was clinically superior, due to formulation differences.

The disputed decision: when FDA approved Bendeka in 2015, it declined to award the drug orphan exclusivity.  The statutory exclusivity provision describes how exclusivity operates (i.e., what it blocks).  FDA’s implementing regulations added a requirement relating to receiving exclusivity in the first instance: that a company is not entitled to orphan drug exclusivity if it is the same as an earlier approved drug.  Eagle had not proven that Bendeka was clinically superior to Treanda.  Its drug was therefore the same as Treanda and thus not entitled to exclusivity.  This reflects FDA’s interpretation of § 527. The Framework For Review The courts will (usually) apply a two-step analysis when reviewing an agency’s interpretation of its own statute, deriving from the Supreme Court’s 1984 Chevron decision and decisions that followed it.  Step 1 asks whether Congress “directly spoke” to the question, whether the intent of Congress is clear, and whether Congress unambiguously expressed its intent.  If the reviewing court decides the answer is no, it moves to Step 2, asking whether the agency’s answer is based on a permissible or reasonable interpretation of the statute. The World According to Eagle Eagle’s analysis is straightforward and ends at Step 1.  In my Administrative Law class, we call this a “hits you over the head with a hammer” answer.

• If FDA approves an NDA for a drug designated under § 526
  • Bendeka was designated under § 526
  • FDA approved the NDA for Bendeka in 2015
• then it may not approve another application for such drug for such disease until the expiration of seven years from the date of the approval of the approved NDA
  • then FDA can’t approve another application for bendamustine for these cancers until 2022 (i.e., Bendeka gets seven years of exclusivity)

It’s not complicated.  The plain language of the statute, read literally (“hyper-literally,” the agency complained in the lower court), resolves the controversy in Eagle’s favor. The World According to FDA FDA’s briefing in the district court was muddled, especially with respect to the governing administrative law principles.  The agency did not seem to have a single coherent theory.  Sometimes it argued the statute was ambiguous, for instance, and sometimes it argued the statute was silent.  Sometimes it argued from legislative history and statutory purpose to support the reasonableness of its interpretation (Step 2), sometimes it argued that these need to be considered at Step 1.  The government’s brief in the court of appeals is much cleaner. The key to FDA’s position is the earlier Treanda approval and exclusivity. First, FDA argues that § 527(a) is ambiguous on its face:

• If FDA approves an NDA for a drug designated under § 526
  • Treanda was designated under § 526
  • FDA approved the NDA for Treanda in 2008
• then it may not approve another application for such drug for such disease until the expiration of seven years from the date of the approval of the approved NDA
  • FDA may not approve another application for bendamustine for these cancers until 2015 (but “after 2015” is up for grabs)?
  • The prohibition on approving another application for bendamustine for these cancers ends in 2015?

Let’s put this another way.  “FDA may not do X until DATE” could mean “the prohibition on doing X ends on DATE.”   That is, the prohibition on approving the same drug for the same disease ended in 2015.  If this is correct, then Eagle cannot get orphan drug exclusivity for the same drug.  Why not?  Because that would block FDA from approving applications for bendamustine for these cancers until 2022.  And the prohibition on approving bendamustine for these cancers was supposed to end in 2015. This argument didn’t get much traction in the district court (ruling here).  But second, FDA argues that § 527(a) cannot be read in isolation.  If the court takes into account the rest of the orphan drug provisions and indeed the rest of the FDCA, FDA argues, the court will reach a different conclusion.  And thus the provision is ambiguous.

• For instance, § 527(b) allows FDA to approve another drug during exclusivity, if the “holder” (singular) of exclusivity cannot assure the availability of sufficient quantities.  This makes no sense if exclusivity can be stacked, the agency argues, because in that case (as here, with Treanda and Bendeka) there are multiple holders of exclusivity.  FDA makes several similar points about § 527(b) that, in its view, are inconsistent with Eagle’s reading of § 527(a).
• FDA also points to the fact that allowing multiple exclusivity periods would allow companies to collaborate and stack the exclusivity, as Teva and Eagle effectively did.  The canon favoring narrow interpretation of statutory monopolies, it argues, counsels against Eagle’s reading.

What is this “stacking”? 

• Bendeka was approved during Treanda’s exclusivity, thanks to an agreement between the companies, and if Eagle won this dispute, there should be no generic copies of either drug until at least 2022.  The agency recently confirmed that Eagle’s exclusivity will block generic copies of Treanda as well.  (FDA says Eagle’s interpretation “would turn the orphan-drug exclusivity provision into a gift that keeps on giving successive seven-year monopolies to formulation changes that provide no material benefit to patients over an existing drug.”)
• There’s no question that Eagle’s interpretation permits this stacking.  But this happened only because Eagle (a) had a plausible hypothesis of superiority at the time it sought designation, and (b) failed to prove superiority. Without the plausible hypothesis it couldn’t have gotten the designation, and without the designation it couldn’t have earned exclusivity.  Unless FDA changes its rules governing designation, stacking will require a plausible hypothesis of superiority followed by a failure to show superiority.  (If Eagle had showed the superiority of Bendeka, then copies of Treanda would not be blocked by Bendeka’s exclusivity.  There would be no stacking.)

Further, FDA argues, § 527 is best read as authorizing only a single orphan drug exclusivity term.

• Still in Step 1, the agency effectively claims that it has settled on the best reading of the provision, relying largely on other provisions of the Orphan Drug Act and the broader FDCA.
• Eagle’s approach to § 527 doesn’t work with § 526 (the designation provision), it argues.  Designation is supposed to happen early, so that companies can get tax credits and clinical trial subsidies.  If Eagle (and the district court) are right, then when FDA receives a request for orphan designation for a drug that has the same active moiety as a previously approved orphan drug, the agency has two equally problematic choices.
  • First,  it could refuse orphan designation under § 526 without firm proof of clinical superiority. This isn’t workable, because companies don’t have this proof until they are done with clinical trials.
  • Second, it could grant designation under § 526 and then award exclusivity regardless of whether the drug turns out to be different.  This makes exclusivity under § 526 different from all other exclusivities under the FDCA.  In every other instance, FDA argues, the statute ties exclusivity to contributions to the public interest (“exclusivity-for-development,” the agency says).
• Indeed, the agency contrasts the Hatch-Waxman 180-day exclusivity provisions (which can create a six month duopoly when a generic company challenges an innovator’s patent) with the Orphan Drug Act exclusivity provisions (which would, under Eagle’s reading, create a seven year duopoly even when a second innovator’s drug is not different).
• And FDA points to legislative materials after enactment of the orphan drug provision — such as quotes in the Congressional record describing the orphan drug exclusivity framework as limited to the “first” drug designated and approved for a rare disease.

And then the agency makes its Step 2 argument, that its interpretation of the statute is reasonable.

• FDA’s Step 2 argument has nothing to do with how its approach can or cannot be squared with the wording of the statute.  Instead, in four brief paragraphs the agency defends its approach as a reasonable policy call.  Eagle’s approach, it explains, would allow “infinite, successive 7-year periods of exclusivity,” which is “a practice known as ‘evergreening.”  This would also allow exclusivity periods without any benefit to patients over existing therapies.

Subsequent Legislative Change In 2017, Congress amended § 527 of the FDCA, but the change is broader than this dispute and thus not directly on point.  The statute now provides that a company cannot earn orphan drug exclusivity if FDA has previously approved the same drug for the same disease, unless its drug is clinically superior (not the “same drug” in FDA’s parlance).  This is broader, because it precludes an award of exclusivity even if the first drug did not have exclusivity. FDA had denied exclusivity in just such a situation in 2007.  The district court in the Depomed case ruled against the agency, requiring an award of exclusivity.  Although FDA gave the drug in question exclusivity as ordered, the agency took the view that the district court’s reading of the statute was wrong, and it expressly chose to otherwise ignore the ruling. In the district court the parties of course drew contrasting inferences from this subsequent amendment of § 527, but the court mostly declined to speculate, as will I. Once More, From an Administrative Law Angle With the change in § 527, the case is of limited interest to FDA law junkies — except insofar as it provides another example of the agency’s non-acquiescence to judicial rulings.  It is more interesting from an administrative law angle. As noted Chevron Step 1 asks if Congress has directly spoken to the question, if the intent of Congress is clear, if the intent of Congress is unambiguously expressed. But, what does it mean to say that the statute is clear?  Or ambiguous? In this case, a strict parsing of the narrow statutory provision at issue — § 527(a) — leads to one conclusion at Step 1: that Eagle wins.  A more thorough review of the statute, taking into account other provisions of the statute, and its purpose,  may lead to a different conclusion.  Does it lead to the opposite reading?  Or to a finding of ambiguity?  FDA generally calls the result ambiguity and then argues for a win at Step 2.  But it could just as easily have argued that the thorough review led to a different conclusion at Step 1, especially because its Step 2 briefing isn’t really about whether its interpretation is reasonable (i.e., is one of the several permissible readings of the statutory language). We have seen this movie before. This should remind astute readers of Brown & Williamson v. FDA, decided by the Supreme Court in 2000.  FDA had decided to regulate tobacco products under its drug and device authorities, reasoning that nicotine was a drug and that cigarettes (and other tobacco products) were drug delivery devices.  A strict parsing of the narrow statutory provisions at issue, defining drug and device, led to one conclusion at step 1: that FDA was right.  For instance, nicotine satisfied the definition of drug, because it was intended to affect the structure or function of the body.  But a more thorough review that took into account other provisions of the FDCA and how this ruling would interact with those provisions, plus the purpose of the statute’s drug and device provisions, and other subsequent legislative enactments, led to precisely the opposite conclusion.  At Step 1. So what happens next in the Eagle litigation?  Gary Lawson has argued (in his excellent casebook for Administrative Law, which I use) that the courts do not seem to be aware that there are different approaches to Step 1 from which they are apparently choosing.  I can’t see the lack of awareness working in FDA’s favor here.  And the simple syllogism offered by Eagle for Step 1 is quite seductive. No matter how FDA presents its case, it has to persuade the court that the real legal issue is the meaning of § 527 as applied to Treanda — the legal effect of the exclusivity awarded to Treanda.  The agency has not done a great job of emphasizing this to date.  But it’s necessary to get past Eagle’s seductive Step 1 argument.  If the court just focuses on Bendeka, the agency will lose at Step 1.  Eagle’s case is a slam dunk based on the plain meaning of the words at issue.  If Congress had meant to limit orphan drug exclusivity to the first drug approved, it could easily have said so.  (And, in truth, Congress did exactly the same thing with biological product data exclusivity as Eagle claims it did with orphan drug exclusivity: each innovative BLA gets 12 years of exclusivity, if it is that company’s first licensure.  It doesn’t matter that another innovator got the same biologic approved previously.  So this is not entirely out of sync with the broader statutory framework in which FDA operates.) Regardless of the outcome, for the administrative law enthusiasts, this may be another good example of a situation in which the two approaches lead to differing conclusions — where the court’s approach to Step 1 dictates the outcome of the case.  And the stakes here are the reverse of the stakes in Brown & Williamson: here the strict parsing of the narrow provision at issue leads to the agency losing, rather than the agency winning.  This may strike some people differently.