Can Novartis defend the validity of the ‘405 patent covering Gilenya?
Zachary Silbersher
Novartis’ ($NVS) blockbuster drug, Gilenya®, is facing an onslaught of prospective generic competition. In response, Novartis has wielded one of its patents – the ‘405 patent. The patent survived an IPR challenge, and that decision has been appealed to the Federal Circuit. The appeal is fully briefed and heard oral argument earlier this year in January. If Novartis prevails in the appeal, that means it could potentially exclude generics until the ‘405 patent expires in 2027. What is the likelihood Novartis will prevail?
Background
Gilenya® is a drug for treating Relapse-Remitting multiple sclerosis (RRMS). The active ingredient in Gilenya® is fingolimod. The ‘405 patent (U.S. Patent No. 9,187,405) does not cover fingolimod itself. Rather, the patent covers a dosing regimen for administering a daily dose of 0.5 mg of fingolimod to treat RRMS. The patent claims themselves refer to FTY720, which is fingolimod. (Novartis was recently granted an additional patent, U.S. Patent No. 10,543,179, which we will address in a separate blog post.)
The ‘405 patent does not expire until 2027. Given that, if Novartis can successfully show that the patent is infringed, as well as defend against an invalidity challenge, then Gilenya® could theoretically be safe from generic competition for another seven years. Infringement does not appear to be too much of a concern. Indeed, Novartis has filed numerous district court litigations against prospective generics in which it has asserted the ‘405 patent. Within those cases, Novartis moved for a preliminary injunction to prevent any generics from launching at risk. In the course of arguing against a preliminary injunction, the generics do not appear to have contested infringement.
Instead, the generics have staked much of their case on proving that the ‘405 patent is invalid. To that end, numerous generics filed petitions for inter partes review (IPR) with the PTAB that each challenged the validity of the ’405 patent. Those IPRs were consolidated into one, and in July 2018, the PTAB issued a Final Written Decision upholding the validity of the ‘405 patent.
Presumably, that would put Novartis in a fairly strong position. Indeed, Novartis has successfully defended an IPR challenge to the ‘405 patent, and infringement of that patent does not appear to be strongly contested. Shouldn’t that mean that Novartis can be expected to sail through the Federal Circuit with an easy win upholding the validity of its patent?
Not necessarily. Although the PTAB held the ‘405 patent was not invalid, the reasoning in that decision arguably rests on shaky ground. The recent oral argument bears this out. Before the Federal Circuit, the generics advance two separate reasons why the PTAB’s decision should be reversed.
The Invalidity Arguments
The ‘405 patent is directed to administering 0.5 mg of fingolimod (FTY720) daily to treat RRMS. The problem for Novartis is that the prior art appears to have taught this much. During prosecution of the patent, the Examiner asserted that the Virley prior art reference taught administering fingolimod for treatment of RRMS, and Kovarik taught administering S1P receptor antagonists at 0.5 mg for treatment of autoimmune diseases, which includes MS. (FTY720 is an S1P receptor antagonist.)
To get around this prior art, Novartis added another limitation to the claims. Specifically, Novartis amended the claims to require that the 0.5 mg daily dose of fingolimod is given in the absence of an immediately preceding loading dose. Kovarik taught administration of fingolimod only after a loading dose. Thus, apparently because of this distinction, the ‘405 patent was originally allowed by the Examiner during prosecution.
During the subsequent IPR proceedings, the generics raised a prior art reference, Kappos 2010. Kappos 2010 disclosed a clinical study where patients with RRMS were administered daily 0.5 mg of fingolimod without a loading dose—in other words, Kappos 2010 appears to directly anticipate what is required by the ‘405 patent. Indeed, during the IPR, Novartis did not dispute that Kappos 2010 anticipated the ‘405 patent. (See IPR2017-00854, Paper 109 at 40). Instead, Novartis argued that Kappos 2010 was not a prior art reference because the priority date for the ‘405 patent dates back to a British application filed in 2006, whereas Kappos 2010 was not published until 2010.
The problem for Novartis, however, is that the negative limitation in the ‘405 patent claims – namely, fingolimod is administered without a loading dose – did not appear in the ‘405 patent application until an amendment filed in 2014, several years after Kappos 2010 was published. To overcome this problem, Novartis has therefore argued that the original priority patent application for the ‘405 patent, which was filed in 2006, nevertheless disclosed the negative limitation that fingolimod is administered in the absence of a loading dose.
Yet, this raises another problem for Novartis, which really forms the crux of this dispute. The ‘405 patent itself says nothing about administering the drug without a loading dose. In fact, it does not say anything about giving the drug with or without loading doses whatsoever. Nor does it provide any reasons for eliminating loading doses. Typically, for a “negative” limitation to be a patentable distinction over the prior art, the patent’s specification must explain the reason for eliminating that element. The generics have cited precedent to support that. (See e.g., Santarus, Inc. v. Par Pharm., Inc., 694 F.3d 1344, 1351 (Fed. Cir. 2012)).
To overcome this problem, Novartis put forward expert testimony that a person of skill reading the ‘405 patent would immediately understand that it taught administration of fingolimod to treat RRMS without a loading dose. Among the reasons cited by Novartis’ experts is that MS is a chronic condition, and as such, it was known not to require a loading dose. Indeed, this was allegedly shown in an earlier study from 2005. Another reason was that a person of skill would avoid a loading dose due to the adverse risk of first-dose bradycardia.
The Oral Argument at the Federal Circuit
Oral argument was held at the Federal Circuit on January 9, 2020. During that argument, the panel of Federal Circuit Judges did not appear very eager to agree with Novartis’ arguments. On the one hand, two of the Judges appeared to acknowledge that a patent claiming a negative patent limitation need not necessarily articulate a reason to exclude that particular limitation within the specification.
On the other hand, however, the Judges appeared to acknowledge that the specification must nevertheless articulate some sort of “hook” suggesting that the patent teaches the negative limitation. Novartis claimed there is such a hook. Specifically, Novartis argued that the patent describes that the 0.5 mg dose of fingolimod is “daily” dose. That, according to Novartis, would immediately communicate to any person of skill that if only 0.5 mg is to be administered per day, that necessarily leaves no room for more medication in the form of a loading dose. The Judges repeatedly questioned Novartis on this point, but did not appear satisfied with the answers.
For good reason. Novartis’ explanations raise more questions than they answer. The key point of novelty in the ‘405 patent—namely, the reason the patent was granted in the first instance—is that the 0.5 mg dose is administered in the absence of a loading dose. Because the patent does not specifically disclose that—namely, it does not expressly teach the key point of novelty—Novartis is compelled to argue that a person of skill reading the patent application filed in 2006 would immediately understand that the drug is administered without a loading dose. Why? Well, because, according to Novartis, MS is a chronic condition, and the adverse risks of first-dose bradycardia would counsel against a loading dose.
But by making that argument, Novartis comes very close to suggesting that the very point of novelty in the patent—the absence of a loading dose—was previously understood by persons of skill in the art. But if the key point of novelty was previously understood by persons of skill in the art, then why isn’t the patent obvious? Why is Novartis entitled to another seven years of monopoly pricing for something that it is essentially admitting was previously known in the art? Why, again, are consumers deprived of lower-cost generics for a patent where Novartis could not be bothered to actually articulate the key distinction over the prior art until eight years after the patent was filed?
The generics also argued that two additional prior art references should invalidate the ‘405 patent, namely, Kovarik and Thompson. The PTAB overlooked this prior art because it failed to disclose any clinical data showing that the 0.5 mg dose of fingolimod was somehow critical. Rather, the references allegedly taught a range of doses without showing criticality for the dose required by the patent (0.5 mg). On this point, the problem for Novartis is that the ‘405 patent itself suffers from the same deficiency—namely, it has no clinical data showing that criticality of the 0.5 mg dose. On the contrary, the ‘405 patent merely discloses a proposed clinical trial with a range of potential doses, including 0.5, 1.25 or 2.5 mg. This point was not addressed during the oral argument, but it could serve as a separate basis to reverse the PTAB’s decision.
There was one final issue addressed during oral argument. In a final attempt to save the ‘405 patent, Novartis attempted during the IPR proceedings to amend the claims to eliminate the negative limitation. In its place, Novartis proposed revising the claim to require that the daily dose “consist[s] of” only 0.5 mg. In patent drafting, using the phrase “consisting of” means that the limitation is exclusive. By contrast, using the phrase “comprising of” means the limitation is non-exclusive, and other non-claimed elements can be included.
At oral argument, at least one of the Judges struggled with whether this proposed amendment would warrant remanding the proceeding back to the PTAB, rather than simply reversing the decision and deeming the patent invalid. Novartis gained some traction by arguing that the amended claim would be more narrow than the issued claim, because it would exclude more than just a loading dose, and therefore, the PTAB would be compelled to address the validity of that new claim scope.
On the other hand, the amended would appear to suffer from the same problems that the issued claim has. Novartis would still have to argue that the “consisting of” limitation renders the claim distinguishable over Kovarik, and to do so, the primary basis for that distinction would be the absence of a loading dose. And that would reintroduce all the same problems with the issued patent claims that compelled the Federal Circuit’s remand in the first instance.
This is why drug prices are so high.
Why are drug prices so high? This case illustrates a tangible reason why. Novartis is quibbling with words and explanations to try to squeeze out seven more years of monopoly pricing to the detriment of consumers. Novartis claims it is entitled to that windfall because of an invention in a patent. But it is questionable whether Novartis actually invented anything in this instance.
The point of alleged “novelty” in the ‘405 patent did not even show up in an amendment until eight years after the patent was filed. To get around that, Novartis is forced to argue that the point of novelty was inherently disclosed in the original patent application. But that argument practically forces Novartis to admit that the alleged invention in the patent was immediately known to persons of skill, without being expressly articulated. But that is sort of another way of saying — it was obvious.
This is life-cycle management. If the public wants to know why lower-cost generics take so long to come to market, this is the answer.