Eylea: How do courts solve a problem like a patent thicket?

Zachary Silbersher

Along with other members of Congress, Senator Elizabeth Warren recently wrote to Director Kathi Vidal of the USPTO about a patent-thicket Merck is growing around its blockbuster drug Keytruda®.  Senator Warren’s letter emphasized that, by one analysis, the biggest grossing drugs were protected by about 125 patents.  That’s a problem, argued Senator Warren, because “no matter how weak some patent applications may be, challenging them is nevertheless estimated to cost millions per patent in a thicket.”  Senator Warren’s letter concluded by asking what the Patent Office can do to curb Keytruda®’s patent thicket? 

While Senator Warren asked what the Patent Office can do, it is also worth asking what the courts can do.  The primary business of the Patent Office is to issue patents.  Yet, the actual place where patents are used to delay generic competition is in the courts.  The courts are in the unique position of having to contend with scores of patents asserted within a single suit.  To put some perspective on this, patent suits involving two or three patents can easily take upwards of two to four years to litigate.  And because patent cases often involve complicated technologies that courts and juries are unfamiliar with, they create unique scheduling challenges for Judges’ dockets.  When a brand pharmaceutical company shows up with 20, 60 or 100 patents in a single suit, those challenges can be exacerbated by orders of magnitude.

 How have courts managed dockets asserting patent thickets?

So, how have courts face this challenge?  Regeneron’s pending patent suit against Mylan over its prospective biosimilar for Eylea® is an interesting example.

Regeneron’s complaint against Mylan was filed in August 2022 and asserted 24 different patents.  At the outset, Mylan argued that the sheer number of patents and various technologies covered by those patents dictated that a 24 month trial schedule was necessary to work through the voluminous discovery required to get to trial. 

In itself, Mylan’s concession reflects the practical obstacle of bringing a lower-cost medication to market when faced with a patent thicket.  Because Regeneron showed up with 24 patents, Mylan (the prospective lower-cost biosimilar) started off by telling the court that it will take at least two years to get to trial.  That trial will likely be followed by several months for the court to issue an opinion, which will then be followed by at least a year for the losing party to appeal to the Federal Circuit.  All of that is required before Mylan can enter without risk.  Furthermore, Regeneron’s complaint was preceded by more than eight-months for the “patent dance” required by the BPCIA to play out.  (Mylan first notified Regeneron that its BLA for Eylea® was accepted in January 2022.) 

In this case, however, Regeneron had other reasons to avoid unnecessary delay.  The FDA’s exclusivity period for Eylea® is scheduled to expire in May 2024.  Because of that, Regeneron is motivated to get to trial and appeal any subsequent decision long before that.  Regeneron’s goal is to win an injunction against Mylan’s entry before the expiration of Eylea®’s regulatory exclusivity.  Yet, to do that, Regeneron only needs to prove infringement on a single patent.  It doesn’t need to show infringement of all 24 asserted patents.

Given all this, Regeneron proposed a case schedule wherein trial would occur by June 2023, but as a compromise, Regeneron would only pursue twelve patents in the first phase of the litigation.  The remaining twelve asserted patents would be asserted in a second phase.  Mylan balked at that proposal.  Just like Regeneron, Mylan is looking for certainty—i.e., freedom to operate without patent liability.  Mylan did not want to go to trial on twelve patents, win on those patents, but still have Regeneron hanging another twelve patents over its head indefinitely.

Regeneron, however, was determined to get to trial on some patents by June 2023.  So, it proposed that, for the remaining twelve patents, it would not seek injunctive relief against Mylan, and therefore, those “second wave” patents could theoretically not act as an obstacle to Mylan’s entry.  Mylan could, however, still have to pay sizeable royalties if and when its biosimilar enters the market.

The court and the parties eventually struck a compromise.  At the outset of the case, Regeneron narrowed the case to six patents from only three families, i.e., three general areas of technology.  Then, later in the case, but before trial, Regeneron will once again narrow the case to three patents, but asserting only 25 claims.  (Each patent has multiple “claims,” which are essentially different embodiments of the same invention.)  Further, Regeneron promised that, for the 18 patents left out of the case at the outset, it will not seek injunctive relief against Mylan on the basis of those patents in the future.  See Regeneron Pharmaceuticals, Inc. v. Mylan Pharmaceuticals Inc., Case No. 1:22-00061 (N.D.W.V.) (Dkt. 87). 

The scheduling compromise struck in Regeneron’s Eylea® case mirrors similar scheduling compromises struck in other suits.  For instance, Genentech has recently been litigating against prospective biosilimilars for Herceptin®.  In a series of cases, Genentech asserted anywhere between 26 and 40 patents per case.  Importantly, although many of the cases were consolidated, and essentially litigated together, they were complicated by the fact that Genentech did not assert the same patents against each defendant.

In the Herceptin® cases, the parties also struck a scheduling compromise that reduced the number of patents going to trial.  Cases asserting around 30 or 40 patents were cut down to 10 patents each.  For each patent, only two claims per patent could be asserted.  Among the 10 patents, at least six of the patents had to be the same across all of the defendants.  In total, only 30 claims could be asserted.  Genentech, Inc. v. Pfizer Inc., Case No. 1:17-01672 (D. Del.) (Dkt. 48 at 5) (along with consolidated cases.)

In another case involving Avastin®, Genentech narrowed the case from 26 patents to 8 patents with a cap of 25 patents to go to trial.  Genentech, Inc. v. Amgen Inc., Case No. 1:17-01407 (D. Del.) (Dkt. 201 at 3).

What themes emerge from scheduling compromises?

There are a few important themes that emerge from the biosimilar patent suits for Eylea®, Avastin® and Herceptin®. 

First, court schedules that have tamed patent thickets do not just focus on the number of patents, but also the number of families.  In the Eylea®, Regeneron agreed to assert only six patents at the outset, but those six patents could only come from three families.  Regeneron could not assert six patents from six different families. 

That difference matters.  Patents from the same families typically cover the same general field of technology as well as the same general overarching invention.  That yields tremendous economies of scale.  The patents asserted in the Eylea® case demonstrate this.  The patents belong to 11 different families and cover a variety of different technologies.  Those included patents directed to methods of treatment for different indications, methods of purifying proteins, oxidized species of aflibercept, protein formulations, cell culture media and cell culture methods, cell line patents, DNA analysis patents, among others.  If Regeneron could pick a single patent from each family, that could easily prolong the litigation.

Second, scheduling compromises involving multiple different defendants focus on trying, as best as possible, to increase the number of patents commonly asserted across all defendants.  If the Herceptin® cases, six of the 10 patents had to be asserted against all defendants.  That also greatly increases economies of scale if multiple defendants can collaborate on invalidity challenges and claim construction issues for the same patents. 

Third, scheduling compromises also focus on the number of claims eventually asserted.  Every patents has claims.  Those claims are technically what the “invention” is within the patent.  A single patent can have anywhere from one to over a hundred different claims.  In the context of a patent case, each “claim” is technically litigated individually.  It would not make sense to limit the scope of asserted patents if plaintiffs could still assert hundreds of claims from those patents.  This is why in the Eylea® case, the number of asserted claims eventually going to trial was capped at 25, and in the Herceptin® cases, the number of claims was capped at 30. 

The final theme relates to a knotty constitution problem.  In all of these cases, although the schedules compelled the brand pharmaceutical companies to limit the number of patents asserted in each case, the schedules did not compel the plaintiffs to just dismiss the other patents.  In the case of Eylea®, Regeneron reserved the right to assert the unasserted patents at a later time.  In the Herceptin® cases, the scheduling order was explicitly “[s]ubject to [Genentech’s] reservation of due process rights stated on the record” at the status conference.  Genentech, Inc. v. Pfizer Inc., Case No. 1:17-01672 (D. Del.) (Dkt. 48 at 5) (along with consolidated cases.) 

Over the lifetime of a patent case, it is not unusual for plaintiffs to narrow the number of patents that will eventually go to trial.  That may be done out of strategic and tactical prudence.  If you are trying to persuade someone of your case, be it a judge or jury, sometimes less is more.  But the important point is that Judges cannot necessarily force plaintiffs to drop patents.  Due process constraints deriving from the Constitution typically prohibit courts from just dismissing some of patents out-of-hand.  And because courts can’t force plaintiffs to drop patents, even when plaintiffs agree to narrow the number of patents litigated at the outset, they often reserve the right to assert the remainder at a later time.  And the generics or biosimilars remain effectively hostage to those decisions by the plaintiffs.  And that calculus affects settlements—i.e., when the generics will ultimately agree to enter.  In short, courts can tame the problem of patent thickets, to more effectively manage their dockets.  Yet, outright curbing their leverage, they often cannot do.

And so we’re left with the question, if courts have only limited tools to solve the problem of patent thickets, then who wins?  Who wins when lower-cost generics are forced to litigate scores of patents that have overlapping technologies and only minor tweaks from one patent to the next?  Not consumers.