Generic and branded pharmaceutical manufacturers squared off again this week in Caraco Pharmaceutical Laboratories v. Novo Nordisk. Anybody who had dismissed the case as an obscure patent case was disabused of that notion by the incisive tone of the Justices’ questions.
The question here involves a generic manufacturer’s efforts to sell a drug for a particular use that is approved by the FDA but not covered by the branded manufacturer’s patent. The specific problem involves the interface between the FDA and the Patent and Trademark Office (PTO). Despite the importance of patents to the pharmaceutical industry, the FDA traditionally has taken a hands-off approach to matters involving the interpretation or assessment of patents. Thus, generally speaking, the role of patents in the FDA process for approving generics is that the branded manufacturer files a list of its patents and a summary of their scope with the FDA. If the branded manufacturer has filed patents claimed to cover the drug, a generic manufacturer that nevertheless wants to sell the drug must certify either that the patent does not extend to the drug or that the patent does not extend to “an approved method of using the drug.”
This case involves a drug (repaglinide) approved by the FDA for three separate uses related to diabetes. Novo has a patent that covers one of the three uses but not the other two uses. Still, for reasons that the parties dispute, the summary of patent scope that Novo submitted to the FDA covers all three uses. Because the FDA abjures any power to revise the description, however patently erroneous it might be, no generics can be sold for the non-infringing uses. Accordingly, Caraco seeks an order under 21 U.S.C. § 355 “correct[ing]” Novo’s description of its patent. The Federal Circuit ruled that this power is not available, which means that Novo’s overbroad description allows it to unilaterally extend the scope of its patents, with no possibility of administrative or judicial correction.
The central textual question turns on the distinction between the words “an” and “any.” Under Section 355, the generic can prevail only if it can prove that the branded manufacturer’s patent does not extend to “an” approved method of use. In the Federal Circuit’s view, this requires proof that there is no approved method to which the patent extends. By contrast, Caraco argues that it need only prove that there is a single (“an”) approved method of use to which the patent does not extend.
Arguing for Caraco, James Hurst was interrupted almost as soon as he began his presentation by Justice Alito, who blandly asked him whether he would agree that his “brief does not cite a Supreme Court decision.” Apparently stumbling for a moment at the odd question (his brief cites more than two dozen Supreme Court decisions), Hurst then agreed that he believed it would be a correct statement. Justice Alito quickly joined with him in agreeing that this was a reasonable interpretation of his question, in context, emphasizing how sensible it would be for a judge to dismiss a prosecution because the prosecutor “failed to prove an element of the crime.”
Justice Scalia, however, was not nearly so agreeable. Admitting that the language could bear the reference to “any” use code that Hurst asserted, he insisted that he found this an unaccountably odd and unexpected way to write something quite simple. Hurst attempted to deflect the problem by pointing out that respondent Novo’s reading of the statute is much more problematic. He argued emphatically that Novo would read the “correction” remedy out of the statute entirely: if generics can’t get orders to “correct” a use code, the only thing the statute authorizes is correction of a patent number – an absurdly unlikely event, in Hurst’s view. But Scalia drew laughs when he rejoined: “Well, it’s not much, but it’s something.” Scalia continued to push Hurst doggedly for the remainder of his time, clearly unpersuaded that Novo’s reading is untenable. Indeed, even when Justice Sotomayor attempted to ask a question, Justice Scalia pressed Hurst unceasingly until his time expired with Justice Sotomayor’s question unheard.
The most noteworthy point in the argument came almost immediately after Benjamin Horwich rose to argue for the United States. He attempted to start by picking up on a question from Justice Kennedy about whether the use codes are written by the FDA or the branded pharmaceutical companies. This question is important because the greater the FDA’s role in drafting the use codes, the harder it is to blame the branded companies for inaccuracies in the codes; Novo argues with considerable force that its use code, however inaccurate it might be from a patent perspective, is perfectly in line with FDA requirements. But before he could finish a few sentences on that subject, Justice Ginsburg interrupted to ask Horwich what the FDA’s position was here.
This is a startling question at first, because Horwich is appearing on behalf of the United States, and the FDA obviously is the agency most affected by the case. Which is what Horwich says, predictably enough. But Justice Kagan (no stranger to the dynamics of the Solicitor General’s office) asks why the FDA didn’t sign the brief. Horwich did not answer directly, but responded, somewhat awkwardly, that the names on the brief should not be a guide to any deference question and that he wasn’t arguing in any event that the FDA should get deference. All in the audience surely were wondering exactly why the FDA refused to sign, but when the Justices didn’t press Horwich farther, that interesting question was left unanswered.
Instead, the Justices allowed Horwich, largely unimpeded, to emphasize his principal point, which is that Congress wrote the statute in question simultaneously with the agency regulations, which require submission of the information that petitioner wants to “correct.” His central theme, reiterated throughout the argument, is that the FDA does not want to interpret patents, but rather wants courts to determine what patents say for purposes of ensuring that an appropriate line is drawn between patented pharmaceuticals sold by branded manufacturers and nonpatented pharmaceuticals sold by generics.
When Mark Perry rose on behalf of Novo, he went straight to Justice Kennedy’s point, arguing that the use codes are in truth written by the FDA, and thus that the only sensible remedy is litigation against the FDA. Chief Justice Roberts immediately scoffed at the practicality of that suggestion, pointing out that the FDA would “have to hire an awful lot of patent lawyers” to review use codes for compliance with patents. Justice Sotomayor was similarly skeptical, asking Perry “[w]hat odds” he would “put as a betting lawyer on them winning a challenge to the FDA policy decisions of what it’s capable of doing and not doing”?
When Perry responded that generics had in the past won Administrative Procedure Act challenges against the FDA, Justice Breyer stepped in, paralleling Justice Alito’s inquiry from the opening moments of the argument. He read the relevant statute at length and then challenged Perry to explain “looking at the literal words, why doesn’t it fit” to say that Caraco is proving that Novo’s patent does not cover “an” approved method of use when it proves that there is an approved method to which Novo’s patent does not extend.
To an external observer, this seemed the crucial point of the argument. Justice Breyer directly served up to Perry the central question in the case and sat back to hear if Perry could rebut Caraco’s central textual point. Strangely, instead of directly answering the question with his own textual argument (the centerpiece of the Novo brief), Perry instead chose to avoid the question, arguing that it reflects a basic misunderstanding of the statute. Unfortunately, for Perry, the tack he chose to avoid the question was to argue that the confusion rested on an interpretation of the statutory reference to “patent information” that turned on a failure to understand the statutory basis for the relevant rulemaking. When the Justices asked for a citation to the rulemaking on which he was relying, he was unable to locate it quickly, so the discussion foundered.
Justices Kagan and Breyer kindly intervened, however, to take him back to his secondary textual argument: that the only patent information that is relevant is the patent number and expiration date (not the use code). The Chief Justice in particular found this argument hard to swallow, emphasizing how unlikely it would be that a brand manufacturer that mistakenly misidentified its patent to the FDA would litigate rather than immediately identify the correct patent.
Justice Sotomayor then took him back to the central policy argument, asking him why it made sense for his company to retain the exclusive right to sell the drug for a particular use after the patent expired. His answer, as in his brief, was simply that Congress had failed to provide a remedy for that situation. Apparently unpersuaded by that response, Justice Sotomayor then took a different tack, asking what “parade of horribles” would ensue if she accepted the government’s view and voted for the generic Caraco. When he responded that the obvious problem would be increased litigation and consequent expense, Justice Kagan quickly stepped in to summarize the parade of horribles on the other side: “You basically read ‘correct’ out of the statute.” And she went on to emphasize that “[t]he statute read your way essentially allows you to unilaterally expand your patent. . . . So why should we read the statute so that it effects a purpose that is entirely antagonistic to the purpose that Congress had . . . , given that the statute is at best from your perspective ambiguous”?
The Justices’ dissatisfaction with Perry’s position became, if anything, even clearer as the argument proceeded. Apparently offering Perry an opportunity to explain his position more fairly, Justice Ginsburg asked whether he really argued that a generic could not sell for an off-patent use. Perry did not back down. Rather, he stuck to his guns, insisting that Congress had simply failed to think of the possibility and that the Court (and FDA) were powerless to permit the generic to go forward.
Another interesting aspect of Perry’s argument is what he pointedly did not mention: the jurisdictional argument with which Novo started its main brief, raised in the Court for the first time but not pressed for a moment at the argument.
By the end of the argument, a large group of the Justices had taken the view that Novo’s view imputed to Congress an impossibly bizarre perspective. Only Justice Scalia had affirmatively supported the possibility of accepting Novo’s reading. Whether the Justices will find the statute clearer when they return for deliberations will be a question for the months to come, but unfortunately for Novo, the Court as a whole certainly will be reading the statute to see if it forces them to vote for Novo, not to see if it will permit them to vote for Novo.
CLICK HERE FOR FULL VERSION OF THIS STORY