My last entry described the facts leading to Amgen’s suit against FDA over denial of pediatric exclusivity for Sensipar. Below I describe what’s at issue in the case. At bottom, this litigation relates to a federal agency developing a new standard (a new interpretation of its statute) that it will apply when ruling on applications for a benefit, after its prior interpretation suffered a defeat in federal court. Rather than announcing the standard publicly, the plaintiff in this case argues, the agency applied the standard in non-public rulings for more than a decade. Not only does the standard conflict with the statute, plaintiff adds, but the agency has not been consistent in its application of the standard. Thus the dispute is more about how a federal agency is operating than it is about the law of pediatric exclusivity.
FDA’s interpretation of “fairly respond” was inconsistent with the statute. Amgen made several arguments that the Chevron rubric should not apply to FDA’s interpretation of “fairly respond,” the most interesting of which was the argument under Mead that the interpretation — being “ad hoc” and “freewheeling” — lacked “traditional hallmarks” that would entitle it to deference. Judge Moss was not persuaded; the decision was a formal letter ruling, albeit unpublished (which is normal for these rulings). As for the Chevron analysis itself, he agreed with FDA that “fairly respond” is “inherently ambiguous.” (For instance, does “fairly” relate to the extent of the response, or does it mean “reasonably well,” which is more qualitative?) Attaching meaning to the phrase, he wrote, necessarily requires a policy-laden judgment. In the end, construing “fairly respond” is “precisely the type of statutory gap-filling that ‘involves difficult policy choices that agencies are better equipped to make than courts,” for which he indirectly cited Brand X. FDA’s selection of a “clinically meaningful information” standard was reasonable, Judge Moss found, because it links the standard to the underlying goal of the pediatric exclusivity provision.
Amgen makes the intriguing argument that the clinically meaningful standard actually conflicts with the statutory provision (and is therefore unreasonable), because it injects uncertainty into the exclusivity process. As the company explained in its summary judgment memorandum, the central feature of the pediatric exclusivity scheme is that the company knows the terms of the bargain before it decides whether to invest in the research. As I noted in my first post on this topic, it has been clear for years that the exclusivity reward is for research, not results. Amgen reasons that the agency’s new standard amounts to requiring a particular result. That is, the information now has to be “clinically meaningful” — i.e., it has to result in data with clinical significance.
FDA’s denial was arbitrary and capricious. Judge Moss was not persuaded by Amgen’s argument that it was arbitrary and capricious to ignore the company’s efforts to enroll the 15 patients specified for Study 3. But the company’s argument that FDA had treated it differently from other companies was concerning. In particular, Amgen pointed out that FDA granted exclusivity to Johnson & Johnson for pediatric studies with Ortho Tri-Cyclen (norgestimate/ethinyl estradiol) even though J&J had not enrolled the required 120 adolescent women aged 12 through 17 with anorexia as defined in the DSM-IV. Indeed, Amgen had made this point to FDA, so the agency’s response was part of the record before Judge Moss — and in his view the record cast “doubt” on whether J&J had met the terms of the written request. FDA’s decision on Amgen’s request was “arbitrary and capricious” simply because it had not adequately addressed this precedent. Herein a lesson for students of administrative law: Amgen was vindicated, but what did the company get? Maybe just a better explanation! After all, there “may be an answer” to Amgen’s argument, Judge Moss wrote. Judge Moss speculated that perhaps the J&J study did comply, or perhaps FDA had a basis for “distinguishing, disregarding, or abandoning” the precedent.
What happened on remand? On February 5, FDA provided a “Remand Decision” to the court, which (1) conceded that J&J’s studies did not satisfy the relevant written request, but (2) argued that at the time the agency thought the studies did. On February 6, however, FDA found a new document showing the agency had raised questions with J&J at the time. Hearing this, the court remanded the matter to FDA again. On February 8, the agency revised the Remand Decision, reporting that J&J had responded to the agency’s questions with an argument why the written request had been satisfied. At this point, apparently, the FDA medical officer recommended the award of exclusivity on the ground that the study met its requirements.
FDA’s final position, in the revised Remand Decision, seems to be that: (1) it applied the same standard to J&J in fact, or perhaps (2) the requirement at issue wasn’t really a requirement, or (3) maybe the agency had some other reason for thinking J&J qualified, and anyway (4) even if it made a mistake, that doesn’t justify making the same mistake with Amgen’s studies. (I’m — truly — not making this up; see pages 6-7 of Judge Moss’s opinion.) On February 17, Judge Moss issued a second opinion. He seems to have concluded — based on his review of the new documents including one document reviewed in camera — that FDA staff at the time viewed the particular requirement at issue as flexible, which led the agency to find J&J had satisfied its written request. Although FDA “acknowledges some uncertainty” about what occurred, he added, “this account makes sense of the documentary record.” In the end, “FDA has offered a reasoned explanation for why it reached different outcomes in response to the Ortho Tri-Cyclen and Sensipar applications for pediatric exclusivity.”
FDA’s application of its new interpretation of “fairly respond” to Amgen “without any notice” contravened both procedural due process and APA principles. Amgen argued that it had no way to predict — when it agreed to the written request — that FDA would adopt the interpretation that it later adopted. The company invoked a doctrine that precludes an agency from applying a legal interpretation that cannot be discerned in advance with “ascertainable certainty.” Judge Moss wrote, however, that it is “common” for agencies to set forth their statutory interpretations in the adjudicative process. The rule simply cannot be what Amgen urged; it would “rework the regulatory process” to apply the “ascertainable certainty” standard to every adjudication in which regulated parties have a substantial stake or reliance interest. To bolster the ruling, Judge Moss added that Amgen’s argument also fails on the facts — in his view, the company’s arguments before FDA show the company was well aware its exclusivity was in doubt. Nor, he added, can the company point to a “manifest injustice” in its retroactive application of a “new” interpretation; after all, with the 2001 court ruling in place (that FDA could not hold companies to perfect compliance with every word of a written request), Amgen could hardly claim it was acting in reliance on some other interpretation of the law.
The Policy Question
While the case was pending, several professors wrote a “Viewpoint” piece for the Journal of the American Medical Association, arguing that the courts should defer to the agency’s new interpretation of “fairly respond” — because there is a “strong public interest in ensuring that, in exchange for the commercial advantage conferred by pediatric exclusivity, companies conduct trials that are likely to provide robust clinical evidence.” After the ruling, they followed up with a blog post praising the agency’s “clinically meaningful” standard, arguing that this allows FDA to ensure that companies conduct studies “that are likely to guide clinical decision making for pediatric patients.”
But there are two responses. First, FDA doesn’t need the “clinically meaningful” standard to ensure that companies conduct studies that are well designed to generate clinically meaningful information. I’ve noted before that FDA controls the pen on written requests. It will ask for trials only if it believes they have scientific merit – meaning the potential to generate useful information. Second, this is not what FDA is doing with the new “clinically meaningful” standard. It is not using the standard to steer companies into the kind of studies that will guide clinical decisions. It is using the standard to measure whether the results earn the company exclusivity.
It seems to me that — at bottom — we are talking about a meaningful change in agency policy affecting innovation incentives and that more transparency about the change — indeed a public dialogue about the policy change — would have been proper. It’s fine to discuss the merits of tying pediatric exclusivity awards to the results of the research, but this represent a significant change in policy. And there are a lot of considerations to weigh, including whether uncertainty about the reward will undermine efforts to encourage this badly needed research. On March 22, FDA updated its list of pediatric studies conducted in response to either the exclusivity incentive or the (separate) pediatric assessment requirement. I’d urge everyone to look closely at the 40 products studied only because of the reward of exclusivity — look at the drugs, the studies, and the results — before concluding that a change was (or is) a good idea.
This Litigation Is Not About That Question
Back to Amgen’s suit. It does not present this policy question. Here is one way to think about it. A federal court told FDA that it could not interpret “fairly respond” to mean that companies must comply exactly with every word of the written request. For the next 16 years, FDA ruled on pediatric exclusivity requests applying a new standard that presumably complied with the court order — although, who knows? — without sharing the standard publicly. Amgen accepted a written request and invested about $20 million in studies , told FDA it was having trouble with enrollment, and received pressure to persevere. There doesn’t seem to be a dispute that the company did the best it could. Once FDA received the data (and required labeling changes based on what it received, see here), it denied the reward for the research and only then told the company the standard that it was applying. And the new standard requires the pediatric research to generate information that is clinically meaningful — which is inconsistent with the way pediatric exclusivity had previously been understood. And it still is not entirely clear (to me, at least) that FDA really applied this standard to J&J.
This is plenty for the D.C. Circuit to chew on. Oral argument before the court of appeals (docket no. 18-5046) is scheduled for May 17.