Korlym: What do the post-trial briefs say about what happened at trial?
Zachary Silbersher
In late September, Corcept Therapeutics finally went to trial against Teva in the parties’ long-running patent dispute over Teva’s prospective generic for Korlym. The Hatch-Waxman litigation between the two companies has been waging since 2018. Numerous patents have come in and out of the case over that time. Yet, by the time of the trial a few weeks ago, Corcept had narrowed its case down to alleging Teva’s infringement of only two patents: U.S. Patent No. 10,195,214 and 10,842,800. The parties’ post-trial brief have been filed. What do they say?
The post-trial briefs indicate that the trial focused on two primary issues. Both issues related to infringement of ‘214 and ‘800 patents. Corcept has the burden to prove infringement, and as a result, the trial focused primarily on Teva’s non-infringement defenses to the patents. Importantly, the trial does not appear to have focused on invalidity.
The two issues litigated trial included: (1) will doctors directly infringe the patents? (2) will Teva encourage or instruct doctors to infringe the patents?
With one minor distinction, both questions apply equally to both patents. Thus, if the court finds in favor of Teva on either of these questions, then Teva will likely prevail in the case.
(I’ve previously blogged several times about this case over the past five years. For more background and prior posts, see here.)
Will doctors directly infringe the patents?
Teva’s first defense is that doctors will not directly infringe the asserted patents. Teva is the only named defendant in the litigation, and there are no doctors who are otherwise named as defendants. Yet, to understand why Teva argued that doctors will not directly infringe the patents requires a brief background into patent law’s doctrine of indirect infringement.
The ‘214 and ‘800 patents are each method patents. Generally, they are directed to methods of treating Cushing’s syndrome. There are minor differences between the patents, but they are both generally directed to methods of reducing the dosage of mifepristone for patients also taking a strong CYP3A-inhibitor, such as ketoconazole. (The motivation behind the patents rests on Corcept’s claim that it discovered how to safely to administer both mifepristone and ketoconazole.)
Because the asserted patents are directed to methods of treatment, Corcept was compelled to rely upon a theory of indirect infringement against Teva. This is because Teva is a pharmaceutical company; it does not actually treat patients for Cushing’s syndrome. Rather, doctors treat patients. As a result, because the patents are directed to methods of treating patients, Teva cannot directly infringe them. Yet under an indirect theory of infringement, otherwise known as “inducement,” Teva can still be liable if it encourages or instructs doctors to infringe the patents.
Accordingly, Corcept’s burden at trial was to prove that Teva will provide instructions to doctors that, if followed, will result in the doctors directly infringing the patents. Those instructions are embodied within the label for Teva’s prospective generic formulation of Korlym. Because Teva is compelled by the FDA to copy Korlym’s label, the trial essentially focused on whether the label for Korlym encourages or instructs doctors to infringe the ‘214 or ‘800 patents.
However, for Corcept to hold Teva liable for indirectly infringing the patents, Corcept must also prove that some party is directly infringing the patents. Here, if Corcept’s theory is that Teva is liable for indirectly infringing the patents by instructing doctors to do so, then Corcept must also prove that doctors actually practice the methods claimed in the patents.
In many Hatch-Waxman cases involving method-of-use patents, there is often no dispute that doctors will directly infringe the patents. That is because the patents often read directly upon the label. For instance, if a patent covers a method of treating condition X by administering Y grams of medication Z, and the label is indicated for the same condition at the same dosage of the same drug, then there is typically no dispute that the label doctors following the label will directly infringe the patents.
Here, however, the patents do not read directly upon the label. The label for Korlym is directed only to administering mifepristone, whereas the patents cover the scenario where a doctor is co-administering mifepristone along with a strong CYP3A-inhibitor. In fact, the label cautions against co-administering the two drugs due to potential safety concerns. This is because strong CYP3A-inhibitors increase the blood levels of mifepristone, which means that a safe dosage of mifepristone when taken as a monotherapy could be toxic when co-administered with ketoconazole.
On the other hand, the label is not silent on co-administration of the two drugs. Rather, the label contemplates that there may be times when it is medically necessary for a patient to take both mifepristone and ketoconazole. In such cases, the label generally instructs doctors to lower the dosage of mifepristone.
Corcept argued at trial that because the patents claim reducing the dosage of mifepristone when co-administered with ketoconazole, and that’s what the patents cover, then doctors following the label will directly infringe the patents. Corcept also argued that, during FDA approval process, the FDA acknowledged that patients suffering from Cushing’s syndrome may require co-administration of mifepristone and ketoconazole. Corcept presented testimony from certain doctors in the field who claimed to have frequently co-administered the two drugs to their patients suffering from Cushing’s syndrome. Corcept presented additional testimony suggesting that, because of certain opportunistic infections resulting from Cushing’s syndrome, patients taking mifepristone may also need to take ketoconazole. Based on this collective evidence, Corcept argued to the court that doctors have directly infringed the patents in the past, and they are therefore likely to continue to do so if Teva’s generic enters the market.
Teva hit back with its own evidence. Teva attempted to show that, despite Corcept’s evidence, it still failed to show that doctors would directly infringe the patents. That is because, according to Teva, evidence that doctors have co-administered both mifepristone and ketoconazole to certain patients is insufficient to show direct infringement. The patents do not simply cover co-administration of the two drugs. Rather, they are directed to patients who are already taking high dosages of mifepristone, either more than 900 or 1200 mgs, depending on the patent. They also require the doctors to also administer a strong CYP3A-inihbtor as well as reduce the dosage of mifepristone.
Teva’s principle defense is that Corcept provided no proof that this has ever happened or will ever happen. The Korlym label instructs doctors to start at a low dose, 300 mg, and titrate upwards. The label also warns about the risks of co-administering mifepristone with ketoconazole. Doctors know that strong CYP3A-inhibitors exacerbate the blood levels of mifepristone. As a result, Teva argued it is unlikely any doctor would titrate a patient all the way up to 900 or 1200 mgs of mifepristone while also co-administering ketoconazole.
Teva claimed that not only was this unlikely, but there is no evidence it has ever happened. According to Teva, Corcept did not present testimony from a single doctor who could claim that they had previously co-administered the two drugs at the specific dosages required by the claims.
Will Teva encourage or instruct doctors to infringe the patents?
The second issue litigated at trial also comprises a non-infringement defense by Teva. As discussed above, for Corcept to prove that Teva induces infringement of the patents, Corcept must prove that Korlym’s label instructs doctors to practice the patented methods. The parties generally agree on what the label says, but they disagree over what it means. Unlike the first issue discussed above, the second issue hinges less on doctors who testified at trial because it is really more of an academic legal debate.
Corcept’s argument is straightforward. It claims that the label expressly addresses what to do in the event patient is taking both mifepristone and ketoconazole at the same time. It specifically instructs doctors to lower the dose of mifepristone. That’s what the patents cover, and therefore, according to Corcept, the Korlym label clearly instructs doctors to practice the patented methods.
Teva, on the other hand, acknowledges that the label provides instructions to doctors in the event a patient is taking both mifepristone and ketoconazole at the same time. But, according to Teva, that does not mean that the label encouragesdoctors to administer both drugs at the same time. On the contrary, the label includes bold warnings about the risks of such co-administration. If anything, the label only tells doctors what to do if those doctors first decide, on their own, to forgo those risks and administer both drugs at once. But, according to Teva, instructing doctors what to do if they decide to co-administer is categorically different from encouraging doctors to co-administer in the first instance.
The post-trial briefs filed by both Corcept and Teva reveal that this debate is more of a rhetorical one rather than a scientific one. (That is primarily a function of how the patent claims were drafted by Corcept.) Yet, Teva claims the law is on its side. It points to a precedential decision from the Federal Circuit, HZNP Medicines v. Actavis Labs. Teva claims this case stands for the proposition that where a label that instructs doctors to do X only if they first choose to do Y, that does not constitute an instruction to do Y, and it is insufficient to find infringement of a method patent that requires performing Y. By contrast, Corcept claims the HZNP case is distinguishable and inapplicable to its patent allegations against Teva.
***
Because this is a Hatch-Waxman litigation, and there are no damages at issue, the case was tried before a judge rather than a jury. Accordingly, the court will decided the tried issued by itself. Yet, regardless of the outcome, the case is almost certain to be appealed by the losing party.
The first issue is more fact-intensive, and thus, will be subject to more deference on appeal. The second issue, however, is more legal, and will therefore be subject to considerably less deference on appeal. Thus, regardless of what the court decides, the final outcome of this case—barring a settlement—is likely to be decided by the Federal Circuit.