How long can Roche keep back biosimilars for Avastin®, Herceptin® or Rituxan®?
The biosimilars are biting. And they are biting at three big biologics distributed by Roche through its biotech subsidiary, Genentech. Roche currently faces pending biosimilar competition against Avastin® (bevacizumab), Roche’s biggest selling cancer drug, Herceptin® (trastuzmab), Roche’s breast-cancer drug, and Rituxan® (rituximab), Roche’s immunotherapy drug.
Roche is currently embroiled in biosimilar patent litigation for all three drugs. For Avastin®, Genentech commenced a suit against Amgen ($AMGN) in the District of Delaware in October 2017 (Case No. 17-1407). For Rituxan®, Genentech commenced a case against Celltrion and Teva ($TEVA) in the District of New Jersey in January 2018 (Case No. 18-574). Genentech commenced another Rituxan® case against Sandoz ($NVS) in December 2017 in the District of New Jersey (Case No. 17-13507). For Herceptin®, Genentech commenced a case against Celltrion and Teva in January 2018 in the District of Delaware (Case No. 18-95), a case against Pfizer ($PFE) in November 2017 in the District of Delaware (Case No. 17-1672), and a case against Amgen in June 2018 in the District of Delaware (Case No. 18-924).
The biosimilars, for the most part, did not follow the patent dance.
These cases have, for the most part, followed a similar pattern. In 2017/2018, the biosimilars respectively filed their aBLA’s with the FDA for their proposed biosimilars. Once those applications were accepted for review by the FDA, that triggered the “patent dance” under the BPCIA (Biologics Price Competition and Innovation Act.) The BPCIA is the statute that regulates resolving infringement of patents owned by the reference-product sponsor (innovator) prior to the biosimilar’s launch. (It is essentially the analogous Hatch-Waxman Act for biosimilars.) Under the “patent dance,” the biosimilar applicant is supposed to disclose its aBLA application as well as its manufacturing process. After that, the parties are supposed to negotiate the patents that should be litigated in a first phase of litigation. The biosimilar, rather than the innovator, can actually control the number of patents asserted in this first phase. The biosimilar is required to give the innovator six months notice before launching. That notice triggers the second phase of litigation, wherein the innovator can sue on any other patents not included in the first phase.
The rub with the “patent dance” is that it’s optional. The Supreme Court held that biosimilars are not required to participate in the patent dance. That is what has essentially happened in Roche’s biosimilar cases. While most biosimilars disclosed their aBLA to Genentech, none of them appear to have disclosed their full manufacturing process. As a result, Genentech claims to have been stifled in figuring out which of its patents are actually infringed. That is because, unlike small-molecule drugs, the process for manufacturing antibodies and biologics can itself be a distinguishing characteristic, and brand biologics likewise have built out extensive portfolios covering the manufacturing process itself. AbbVie’s Humira® is a case in point, where AbbVie claims to have at least 24 patents covering manufacturing of the drug.
The result is that Genentech was forced to effectively sue blind. Because of this, each of Genentech’s case have asserted a very high number of patents. Genentech’s case against Amgen asserts approximately 25 patents. Its cases for Rituxan® and Herceptin® assert between 40 and 50 patents each.
These high number of asserted patents are not necessarily an indication that all of these patents are infringed, or that Genentech intends to go to trial on all of them. Rather, Genentech has indicated that it is open to narrowing down the number of patents in these cases to a more manageable size. That likely cannot happen, however, until after Genentech collects discovery of the biosimilar’s respective manufacturing processes. Indeed, in the Avastin® case against Amgen, the Court has ordered Genentech to reduce the case to eight patents by the end of this summer. Something similar is likely to happen in the other cases.
The cases are young, and likely to last 3-4 years, barring settlement.
The cases are all in their early stages. They were all filed in late 2017 and early 2018. These are all comparatively large cases, as far as patent cases go. They will likely take at least two or three years to litigate, barring exceptional circumstances. For instance, in the Avastin® case, the Court has currently scheduled a 15-day trial in July 2020. The other cases, which were filed after the Avastin® case, are likely to receive trials scheduled also in 2020 or within that timeframe.
Fifteen days is a long trial for a patent case, but more importantly, the trial itself is not scheduled for another two years. That date is also not set in stone, and depending on how discovery proceeds, that date can change. Importantly, while it would not be uncommon for the trial date to be pushed back, it is highly unlikely it will be moved forward to an earlier date.
The trial itself will not be the end of the case. It appears that none of the biosimilars have yet to launch, which means there currently are no damages at stake in these cases. Because these cases primarily seek injunctive relief (Roche is requesting the biosimilars are kept off the market until its patents expire), then the cases will most likely be tried before the Judge, rather than a jury. Juries are favorable from a timing perspective because they render a decision immediately. By contrast, Judges presiding over a bench trial often take three to nine months to issue a decision. (Technically, jury trials are not always faster because the Judge usually still has to decide post-trial motions, which can take months.) Even after the trial, whoever loses is most certain to appeal the decision to the Federal Circuit. Those appeals typically last about a year.
In sum, if these cases do not settle, and they play out according to what’s normal, they are not likely to be resolved with certainty—through appeal—for at least another three or four years.
Mylan has also filed a biosimilar application for Herceptin®. There is no pending litigation with Mylan, however, because Roche and Mylan appear to have entered into a licensing deal and agreed on an entry date. The terms of that deal, and the entry date are not public. Some commentators have suggested an entry of Mylan’s biosimilar in 2019. That date, however, is questionable. And it depends upon how concerned Roche is that the biosimilars will launch at risk. Because, assuming they won’t launch at-risk, then it is unlikely Roche would agree to an entry date (2019) that is almost certain to be long before the hypothetical litigation against Mylan would wrap up.
Will the biosimilars launch at risk?
An at-risk launch by any of the biosimilars would technically be before conclusion of the any appeals to the Federal Circuit. That, however, could be four years away. Further, at-risk launches are unlikely right now for many of Roche’s encroaching biosimilars. That is because several have yet to receive FDA approval. Amgen’s biosimilar for Avastin® was approved in September 2017, but the no biosimilars for Rituxan® or Herceptin® have yet to be approved (with the exception of Mylan for Herceptin®). Indeed, many applications have been rejected. In April 2018, the FDA rejected Celltrion’s biosimilars for Rituxan® and Herceptin®. In May 2018, the FDA rejected Sandoz’s biosimilar for Rituxan®. In April 2018, Pfizer received a complete response letter for its Herceptin® biosimilar. In June 2018, Amgen’s biosimilar for Herceptin® was rejected.
At the end of 2017, we performed an analysis of when biosimilars launch at-risk based upon existing biosimilar launches. Statistically, a biosimilar at-risk launch is still less-likely than more likely, but the chances of at-risk launches are higher for biosimilars compared to small-molecule generics.
Two cases for Roche currently stand out on this point. First, in the Avastin® case, where Amgen has received FDA approval, Amgen has indicated that it would not launch until December 2018, but then changed its commitment to April 2018. That said, April came and went without an Amgen launch. But Amgen has clearly made waves that it is contemplating an at-risk launch.
Second, in the Rituxan® case against Sandoz, Genentech has filed a preliminary injunction. The briefing on that motion remains under seal, but it appears that it may be heard before the court, in a mini trial, in July 2018. This will be an interesting event for potential read-through to other Rituxan® cases. While it is not clear from the docket, the reason that Genentech may have moved for a preliminary injunction against Sandoz is because it has indicated it will launch at-risk. That motion, however, appears to have been filed before the FDA recently rejected Sandoz’s application. That means, even if Sandoz had intentions to launch at-risk, those plans are indefinitely delayed. It will be necessary to monitor the docket for further developments on this point.
Will there be settlements?
Settlements are always very likely in biosimilar cases. Roche appears to have prepped itself well against encroaching biosimilars by stockpiling large portfolios of patents covering each drug. This is evident in the number of patents asserted in each case – between 25 to 50 patents. This follows AbbVie’s model for Humira®, where AbbVie surrounded Humira® with more than 70 patents, making Amgen’s chances of circumventing all of them extremely difficult. This is in contrast to Amgen’s Enbrel®, which is really only protected by two patents—albeit two very interesting patents.
Coupled with the litigations, some of the biosimilars, including Pfizer and Celltrion, have started pre-emptively attacking some of Roche’s patents in IPRs. Those IPRs, if successful, would read-through to all biosimilars for the same drug. And presumably, the biosimilars are attacking those patents against which they believe they are most vulnerable, i.e., they are less likely to design around. Therefore, the success or failure of these IPRs is likely a moderate leading indicator of potential settlements, and deserve to be monitored.
Given the large breadth of patents that Roche has built around each drug, and assuming it fares relatively well against the IPRs, then Roche is unlikely to settle for entry dates before the natural expiration of the litigations—i.e., at least two or three years from now. That means, we are not likely to see a major push into the market by biosimilars against Roche’s biologics—at least within the U.S. That said, these cases are in their early stages, and they will require continued monitoring to determine when settlement leverage shifts.