Novartis is trying to block generics for Entresto with a patent Novartis nearly admits it did not invent.
Zachary Silbersher
Novartis is currently embroiled in a sprawling multi-district patent-litigation campaign to block generic entrants for Entresto®, which is Novartis’ blockbuster heart medication. Numerous generics have filed ANDAs, and Novartis has asserted a variety of patents against them. In the fall of 2022, Novartis went to trial on the validity of one of the patents, U.S. Patent No. 8,101,659 (“the ‘659 patent”). The trial briefs suggest Novartis is asking the court to block lower-cost generic entrants based on a patent that it didn’t really invent. How can this be?
The ‘659 patent will not expire until July 2025. Entresto® is one of Novartis’ highest-earning drugs, with over $4.5 billion in sales in 2022. A sizeable portion of that is likely generated from U.S. patients. Thus, if Novartis can successfully use the ‘659 patent to keep out generics for another two years, the upside will be enormous.
The ’659 patent is directed to a combination of two drugs to treat hypertension or heart failure. In particular, the patent claims a combination of valsartan, which is an angiotensin receptor blocker, and sacubitril, which is a neutral endopeptidase inhibitor. For Novartis to leverage the ‘659 patent against generics, it must prove the generics will infringe the patent. Yet, to do that required that Novartis convince the court to interpret the ‘659 patent in a very specific way.
The patent claims a combination of two drugs. The generics argued to the court that the patent should be interpreted as a combination of the two drugs as individual components. By contrast, Novartis argued that the claimed combination should not include a simple combination of the two drugs as individual components. Rather, by implication, Novartis argued patent should be interpreted to also encompass non-covalently bound complexes. (Entresto® is itself a combination of valsartan and sacubitril as a non-covalently bound complex, and by listing the ‘659 patent in the Orange Book, Novartis indicated that the ‘659 patent covers Enesto®.)
The court agreed with Novartis. This was a boon for Novartis. According to the generics, if the court had found otherwise, Novartis would not have been able to show that the generics will infringe the ‘659 patent.
Fast-forward to the trial. To succeed on using the ‘659 patent against the generics, Novartis next had to defeat the generics’ challenge to the patent’s validity. At trial, the generics argued that the patent, as previously construed by the court, is not enabled.
There are many different criteria that must be established for a patent to be valid and enforceable. For instance, the primary criteria is that the claimed invention is novel and not obvious. Yet, another, lesser-known criteria, is that the patent must be enabled. Essentially, that means that the patent must sufficiently teach a manner of making and using the invention without too much work (usually described as “undue experimentation”). A patent that claims a wheel that turns infinitely without any power would undoubtedly be novel and not obvious. Yet, if the patent does not teach a manner of making such an inventive wheel, then the patent would not be enabled, and it would therefore be unenforceable.
At trial, the generics demonstrated that the ‘659 patent does not, by itself, include any teachings on how to make a valsartan/sacubitril complex. In fact, when Novartis filed for the ‘659 patent, nobody knew how to make a valsartan/sacubitril complex. Rather, expert testimony solicited at trial showed that, as of the date that Novartis filed its application for the ‘659 patent, no one would have even contemplated, foreseen or envisioned such complexes. On its face, that suggests the invention claimed in the ‘659 patent is not enabled.
Novartis’ response to the generics’ enablement argument is curious. First, Novartis appears to concede that the ‘659 patent does not teach how to make valsartan/sacubitril complexes. Indeed, Novartis admits that such complexes were not discovered until much later, i.e., long after Novartis had filed for its patent. Novartis told the court that it discovered “valsartan and sacubitril in the form of a complex” in a purported “later, nonobvious discovery” after the ‘659 patent. (See Case No. 1:19-cv-0201-RGA (D. Del.) (Dkt. 231, Dec. 16, 2022) p. 28). Novartis also told the court, “[t]he only example of a complex of valsartan and sacubitril cited by [one of the experts] was undisputedly invented years after the ’659 patent priority date.” (See Case No. 1:19-cv-0201-RGA (D. Del.) (Dkt. 231, Dec. 16, 2022) p. 37). In other words, Novartis appears to concede that a valsartan/sacubitril complex was invented after the ‘659 patent. Indeed, Novartis has two other patents filed after the ‘659 patent, namely, U.S. Patent Nos. 8,877,938 and 9,388,134, that purport to cover valsartan/sacubitril complexes.
Despite all of this, Novartis points to a very old case (In re Hogan, 559 F.2d 595 (C.C.P.A. 1977)) that purports to hold that after-arising technology cannot be used to invalidate a patent for lack of enablement. In other words, when a court examines whether a patent is or is not enabled (i.e., whether the patent teaches how to actually make the claimed invention), the court should not look at evidence that came after the patent application. Rather, the question of enablement must be assessed against the state-of-the-art as of the date of the patent application. Accordingly, Novartis argues that evidence showing that it figured out how to make valsartan/sacubitril complexes years after it applied for the ‘659 patent should be ignored when assessing whether the patent is enabled.
It is hard to make sense of Novarits’ argument. In fact, it sounds a lot like “heads-I-win-tails-you-lose.”
To block generics for Entresto®, Novartis asserted the ’659 patent. To prove the generics would infringe that patent, Novartis explicitly asked the court to interpret the patent to encompass non-covalent complexes. That’s another way of asking the court to find that the “invention” in the ‘659 patent includes a combination of valsartan and sacubitril as a complex. Novartis won that argument. In response, the generics then argued, well, if that’s the case, then the patent is invalid for lack of enablement. The patent simply doesn’t teach how to make a valsartan/sacubitril complex. In fact, Novartis admitted at trial that no scientist reading the ‘659 patent would have even contemplated that it included valsartan/sacubitril complex or that they even existed. (See Case No. 1:19-cv-0201-RGA (D. Del.) (Dkt. 231, Dec. 16, 2022) p. 38).
But then that begs the question, what did Novartis actually invent here? In the first instance, Novartis expressly advocated that the court interpret the patent to include valsartan/sacubitril complexes. Yet, in the next breath, Novartis admits that the patent doesn’t actually teach you how to make those complexes. And when it filed for the patent, it didn’t even contemplate such valsartan/sacubitril complexes. So, Novartis is claiming the generics are infringing an invention that Novartis admits it didn’t even contemplate when it filed for the patent. How can that be?
Novartis’s reliance on the Hogan case seems like smoke and mirrors. Even if the court ignored the after-arising discovery of valsartan/sacubitril complexes, shouldn’t the question of whether the ‘659 patent is enabled end with the admission that the patent in no way teaches or contemplates valsartan/sacubitril complexes? The after-arising discovery of valsartan/sacubitril complexes is not required to prove the patent does not teach how to make those complexes. Instead, it just confirms that, if no one even discovered how to make those complexes until long after the patent, then that patent most certainly could not have taught that.
Novartis adds one further justification against invalidating the ‘659 patent for lack of enablement. It argues, “[s]uch a holding would frustrate the well-settled practice of obtaining appropriate patents on base inventions and subsequent appropriate patents on improvement and selection inventions, which build on that innovative base.” (See Case No. 1:19-cv-0201-RGA (D. Del.) (Dkt. 231, Dec. 16, 2022) p. 28). In other words, Novartis appears to be suggesting that it should not be penalized for drafting its patent claims in a manner that sweeps in a future, uncontemplated discovery. On its face, this justification makes some sense. Inventors cannot be prescient of future, uncontemplated inventions, and penalizing their patent drafting for encompassing those inventions seems like it might be misguided.
Yet, in this instance, Novartis is not simply being penalized by a future invention that is swept into the wording of its current patent. It is being penalized by a future invention that Novartis expressly asked the court to include in its current patent. And it did so because it otherwise it would not be able to prove infringement of the ‘659 patent. Novartis is essentially arguing that the patent must include complexes for the purpose of proving infringement, but not for the purpose of defending against invalidity. In other words, heads-I-win-tails-you-lose.
It is possible that the court will not reach the enablement question at all. This is because the generics primary argument presented at trial is that the patent is invalid based upon the prior art. Yet, it makes you wonder how this could happen in the first place. A brand pharmaceutical company is earning billions a year charging monopoly prices for a drug, but lower-cost generics can’t get in based on a purported “invention” that the company practically admits that it didn’t invent. And consumers must way months or even years for this issue to be litigated. Shenanigans like this are just one reason that drug prices remain too high.