Amarin: What does the generics' appellate brief say?
Zachary Silbersher
Hikma and Dr. Reddy’s have filed their appellate brief defending Judge Du’s decision invalidating the Marine patents. What does it say?
Big Picture: “two patient populations” vs. 500
Looking at the big picture, the appeal is shaping up to be a contest between two perspectives on this case: the “two patient populations,” as argued by Amarin, and the number 500, as argued by the generics.
At the heart of this case is the following question: at the time the patents were filed, did persons of skill in the art expect that administering pure EPA to patients with triglyceride levels above 500 mg/dL would increase LDL-C? Judge Du found that experts would have not have expected an LDL-C increase. Amarin’s principle argument at trial, which it has returned to on appeal, is the “two patient populations” argument. Amarin argues that Mori studied patients with TG levels below 500 mg/dL, and Mori is therefore inapplicable to the patient population recited in the patents (namely, those above than 500). Moreover, the evidence presented at trial showed that, because of the generally understood mechanism-of-action for reducing triglycerides when the patents were filed, persons of skill understood that patients with severely high triglyceride levels would experience a spike in LDL-C. That, according to Amarin, further underscores a distinction over Mori.
The generics do not even discuss the mechanism-of-action in their brief. They do not challenge it or offer an alternative mechanism of action. By not even addressing the mechanism of action, the generics are tacitly arguing that Amarin’s “two populations” argument is a ruse. And that’s because, according to the generics, what really matters is the number 500.
The patents cover patients with TG levels at 500 mg/dL or above. Thus, the patents technically include a patient with exactly 500 mg/dL. To invalidate the patents, the generics were only required to show that it was expected that pure EPA would not increase LDL-C in patients with TG levels at exactly 500 mg/dL. Judge Du relied upon prior art showing that LDL-C purportedly did not increase in patients with TG levels as high as 400 mg/dL. (Bench Order at 60). The court also cited to testimony from Amarin’s expert (Dr. Toth) suggesting that data on patients at 400 could reasonably be used to make predications for patients above 500. (Id.). Moreover, the court also relied upon testimony from the generics’ expert (Dr. Heinecke) suggesting that the 500 threshold relates to pancreatitis, not LDL-C, and there was “no ‘magical mechanistic difference’ between having triglycerides of 400, 500, or 600 mg/dL.” (Id. at 61).
By highlighting this evidence, focusing on the number 500, and avoiding all discussion of the clearance from VLDL to IDL to LDL, the generics try to liquidate Amarin’s “two patients” population argument as a sideshow narrative divorced from what the patents actually cover. They point out that this case was a classic battle-of-the-experts, and regardless of any disagreement between them, Judge Du’s decision credited the testimony of the generics’ expert.
What is at issue here is a factual finding, and that finding is subject to clear error. Thus, the generics still have an advantage under the legal standards governing what an appellant (Amarin) must prove to prevail in an appeal. Amarin’s “two patient populations” argument, and the attendant discussion of the mechanism-of-action, may all be taken to be true. Nevertheless, the generics strategy on appeal is to contain the discussion to what the patents actually recite and require. And for better or worse, the patents were drafted to cover a patient with exactly 500 mg/dL.
The Cyclobenzaprine argument
As we previously discussed in earlier posts, Amarin’s Cyclobenzaprine argument suffers from the drawback that subsequent decisions by the Federal Circuit indicate that not all Judges at the CAFC consider that the methodology of obviousness to be a problem that needs fixing. As excepted, the generics cited to one such decision, namely, Novo Nordisk A/S v. Caraco Pharm. Labs., Ltd., 719 F.3d 1346 (Fed. Cir. 2013). (Whereas Amarin’s appellate counsel was lead counsel in the Cyclobenzaprine case, the firm representing Hikma in this case also represented the generics in the Novo Nordisk case.) Thus, Amarin’s chances of prevailing on the Cyclobenzaprine argument depends, at least in part, on the eventual composition of the panel. (The composition of the panel is, at this time, most likely not yet set, and will not become public until the day of the oral arguments.)
Because the applicability of the Cyclobenzaprine legal framework is an open question, the generics also argue that even if the Court were inclined to adopt the framework articulated in Cyclobenzaprine, that framework was nevertheless satisfied by Judge Du in this case. The generics stress that Judge Du did not “pass judgment on the ultimate issue of obviousness” before considering the secondary considerations. Rather, Judge Du spent numerous pages assessing the secondary considerations before making a final determination under all four Graham factors.
The short answer is this: the Cyclobenzaprine argument may be a crapshoot for Amarin. If Judges Reyna or Newman are on the panel, Amarin is likely to have a sympathetic ear for this argument. But even then, reasonable Judges at the Federal Circuit could likely find that Judge Du complied with whatever is required by the Cyclobenzaprine framework even if it applies. Judge Du’s Bench Order spends considerable time assessing the secondary considerations, and for the most important one—unexpected benefits regarding LDL-C neutrality—Judge Du considered this secondary consideration within the context of assessing the prima facie case.
Amarin’s opening brief argued that, had the court considered the secondary considerations—such as unexpected benefits, praise and skepticism—in the context of the prima facie case, then the lower court would have arrived at a different outcome. While the brief is admittedly artful and compelling, Amarin is essentially asking the Court to weigh the evidence differently than the lower court. By doing so, Amarin is also challenging Judge Du’s factual findings with respect to certain secondary considerations found not to support nonobviousness. But in that case, it is not unlikely the Court falls back on the clear error standard for reversing findings of fact, and therefore refuses to bite—regardless of their views on the Cyclobenzaprine debate. It will be interesting to hear if the Judges even raise the Cyclobenzaprine issue during the oral argument.
The “weighing” of the secondary considerations argument
When we first read Judge Du’s Bench Order, we were surprised that Judge Du “weighed” the secondary considerations against one another given that doing so is inconsistent with the role that secondary considerations typically play within the obviousness analysis. We also noted that Judge Du failed to cite to any caselaw supporting this step. The question then was: on appeal, will the generics discover any legal authority to justify “weighing” secondary considerations against one another?
The answer is, apparently not. The generics’ brief does not cite to any caselaw or other authority to firmly justify Judge Du’s “weighing” of the secondary considerations against each other.
Accordingly, this is likely one of the weakest parts of Judge Du’s Bench Order, and at this point, most likely the strongest chance Amarin has for reversal. It is also likely an issue that the Judges will pick at during oral argument given what appears to be a rather straightforward error of law.
This argument also presents a clear path to reversal that does not require the Court to delve into the scientifically thorny question of “two patient populations” versus the number 500. The pathway is this: by weighing the secondary considerations against one another, Judge Du erroneously discounted the evidence of secondary considerations found in favor of non-obviousness (namely, commercial success and long-felt need.) If those secondary considerations had not been discounted, then they would have weighed more heavily in favor of non-obviousness. And that alone may justify the finding that the generics failed to satisfy their clear-and-convincing evidence burden. It will be very interesting to see if the Judges pursue this line of questioning during the oral argument.
While that seems fairly straightforward, there are nevertheless still a few wrinkles (because there are always more wrinkles). The generics make some hand-waving arguments that Judge Du did not really intend to “weigh” the secondary considerations against one another (even though, that is exactly what the Bench Order says.) As a result, the generics are forced to argue that Judge Du’s error was harmless. That means, they claim that even if Judge Du had not weighed the secondary considerations against one another, the patents would still have been invalidated.
This is where the generics focus most of their attention. Their primary argument is that the prior art was really strong and the secondary considerations were really weak. So, according to the generics, even if Judge Du had not weighed the secondary considerations against one another, the showing of commercial success and long-felt need were too weak to overcome the strong prima facie case of obviousness (namely, the prior art, Mori, Hayashi, etc.)
To get there, the generics also try to kneecap Judge Du’s finding regarding commercial success. According to the generics, the evidence at trial showed that only approximately 25% of prescriptions for Vascepa are for the indication required in the patents—namely, reducing triglycerides in patients with TG levels above 500 mg/dL. Because of that, according to the generics, any purported commercial success of Vascepa lacks a nexus to the patented inventions.
This is not a terrible argument, but it is also not a slam dunk. It is also an argument that the generics pressed at trial, and which Judge Du does not appear to have found persuasive. Either way, the generics are attempting to challenge a finding of fact—Vascepa®’s commercial success—and to successfully do that, they must show clear error. Given that Judge Du’s analysis of the commercial success factor in the Bench Order was quite extensive, the generics face an uphill battle overturning this finding.
Overall, even if the Judges agree that “weighing” the secondary considerations against one another was legally erroneous, the Court still must still review the evidence de novo to determine if reversal of Judge Du’s decision is warranted. In that case, the Court is likely to reconsider whether the secondary considerations of commercial success and long-felt need do, in fact, outweigh the prima facie case. The rub will be whether or not that invites reconsideration of the large factual issue at the heart of this case—namely, whether persons of skill expected that pure EPA would be LDL-C neutral. And that is why, even though “weighing” the secondary considerations was most likely a clear error of law, reversal may still hinge on the Court’s consideration of the big factual debate at the heart of this case—“two patient populations” versus the number 500.
Other issues
The generics stuff their brief with additional points they made at trial, even though these points were not necessarily addressed in Judge Du’s Bench Order. First, the generics repeatedly emphasize that before filing for the patents, Amarin told the FDA and investors that the prior art suggested that pure EPA would be LDL-C neutral. According to the generics, that is essentially the opposite of what Amarin told the Patent Office and what it is now telling the courts. To put this in perspective, it is not common or typical for these types of patent-holder admissions about the prior art to be available to generics attempting to invalidate a patent at trial. Even when cases are decided upon supposed “technicalities,” Judges often prefer to decide matters based upon what is viewed to be fair. The generics are therefore attempting to paint Amarin’s inconsistent statements about the prior art as fundamentally unfair. These admissions continue to haunt this appeal.
Second, the generics also emphasize that the Marine patents have no data. Although the patents disclose a method of administering pure EPA to patients with severe hypertriglyceridemia without reducing LDL-C, the patents lack any data showing that. The patents were filed long before the results of the Marine study came out. This is a nuanced point because, while it is true that a patent application is not per se required to include data substantiating a discovered therapeutic benefit, Judges typically search for a meaty invention story illustrating that the patents were not merely an exercise in life-cycle management.
Amarin has already attempted to preempt some of this by leading its opening brief with the narrative of its invention story. That move was instrumental. Nevertheless, the Federal Circuit sees more patent cases than any other court, and they are very well-versed in different types of pharmaceutical patents. While the foregoing issues will not necessarily be dispositive on the issue of the validity of the patents, they cannot be discounted completely because they may potentially inform the Judges’ overall impression of this case. (The generics are likely gambling that these issues may have informally informed the lower court’s overall impression of this case.)
Alternative grounds for affirmance
The generics brief also includes numerous alternative grounds for affirmance. The generics are essentially asking the Court—if the Court is inclined to reverse Judge Du’s decision on invalidity, the Court should still affirm the judgment because there exist alternate grounds for affirmance.
The first question is a procedural: if the generics did not file a cross-appeal challenging Judge Du’s infringement ruling, then why can they make collateral attacks on it in their responsive brief? The rules are slightly technical. A court’s ruling that a patent is invalid is considered to be broader than a ruling that a patent is infringed. A non-infringement holding is limited to the specific products accused in that particular case, whereas an invalidity holding reads through to all subsequent allegations that the patent is infringed. Likewise, a party defending a lower-court’s judgment on appeal (in this case, the generics) is required to file a cross-appeal if doing so would expand the lower-court’s judgment. Here, because the lower court invalidated the patents, rather than finding them not infringed, then the generics’ alternative grounds for affirmance do not attempt to expand the scope of the judgment, and a cross-appeal was evidently not required.
The next question is whether any of the generics’ “alternative grounds for affirmance” are likely to have any traction.
First, the generics argue that seven of the 10 patent claims permit co-administration of a lipid-altering therapy, such as a statin. According to the generics, these claims are clearly obvious. At the time the patents were filed, surely experts expected that administering pure EPA along with a statin would not increase LDL-C. Judge Du never reached this issue because all of the patents were invalidated for the same unrelated reasons. This is likely the strongest among the generics’ “alternative grounds for affirmance,” but it only addresses a subset of the claims. Thus, technically, Amarin could lose on this issue—if the Court reaches it—and still technically prevail in this case. In its response brief, we anticipate that Amarin is likely to make a legal argument regarding what is required to make an “obviousness to try” argument, but exactly how Amarin handles this will be worth paying attention to.
Second, the generics argue that the remaining three patent claims require administering EPA without concurrent lipid-altering therapy, such as a statin. According to the generics, these claims are not infringed because the Vascepa® label is essentially mum on this issue—it does not explicitly encourage doctors to administer EPA with or without a concurrent lipid altering therapy. Judge Du nevertheless found that the label was still instructive because the Marine Study, which is disclosed in the label, indicated that only 25% of the participants were administered a statin. Yet, the generics claim there are other lipid-altering therapies other than statins. Either way, Judge Du’s Bench Order identified textual support in the label backed up by expert testimony. Exactly how Amarin handles this in response will be instructive, but as of now, the generics chances of prevailing on this argument—if it is reached by the Court—is most likely not in their favor.
Third, the generics once again mount their primary non-infringement argument, namely, that the 12-week limitation is not satisfied. Judge Du’s reasoning was that hypertriglyceridemia is, for the most part, a chronic condition, and thus, doctors would understand treatment to extend indefinitely. Nevertheless, while Judge Du found that hypertriglyceridemia is for the most part chronic, the court also found that there are some cases where it can be acute. And in those cases, treatment might be less than 12 weeks. The generics seize upon that finding to argue Judge Du improperly inferred infringement from the label. The problem, however, with this argument is that evidence of non-infringing uses (administering Vascepa® to patients with an acute condition for less than 12 weeks) is not a defense to an allegation of induced infringement. And no matter how they cut, that appears to be essentially what the generics are arguing. (Indeed, this was a major legal consideration for why Judge Du found Amarin’s patents to be infringed. If evidence of non-infringing uses could be a defense to induced infringement, Amarin would have likely lost this case on infringement.)
Finally, the generics snuck in an argument that was not even at issue during the trial—namely, the patents are invalid for a lack of written description. Judge Du previously struck this defense from the case because the generics had failed to raise it within their expert report. That means, to convince the Federal Circuit to bite into their written description argument, the generics face two hurdles—first, they would have to show that Judge Du improperly dismissed the generics’ written description defense from the case (that would be an evidentiary argument), and second, they would also need to persuade the Court that the defense has merit.
On the first issue, the generics argue that their written description is not an “expert” issue because it was only intended to be a Plan B argument in the event Judge Du agreed the prior art did not teach an expectation that EPA was LDL-C neutral. And thus, the generics complain that there was no basis to include the defense within their expert report. Yet, the reason the defense was stricken appears to be because the generics raised it too late in the game. Judge Du wrote a relatively lengthy analysis for why striking the defense was warranted given the prejudice it would work on Amarin. For the generics to overcome this evidentiary holding with such terse one-page argument on appeal is, at this point, unlikely.
But even so, what about the merits of the argument? It is an interesting argument that has some traction in the law. Amarin’s patents essentially claim using pure EPA for patients with severe hypertriglyceridemia without raising LDL-C, but the patents themselves contain no data proving that Amarin had actually discovered that as of the 2008 filing date for the patents. Rather, the Marine Study was not complete until several years later.
According to the generics, if the Court is inclined to agree with Amarin’s “two patient populations” argument, and thus find the prior art did not teach pure EPA to be LDL-C neutral, then the patents must still be invalid because the inventors themselves lacked the data to prove that that EPA was, in fact, LDL-C neutral. This is another spin on the generics’ argument that Amarin did not do anything more than patent the prior art.
Either way, given that the generics’ written description challenge likely demands further fact-finding, it is unlikely that the Federal Circuit will invalidate the patents on this basis without at the very least remanding the issue back to the district court. (That is if the Court even reaches this issue.)
Yet, the argument nevertheless serves the corollary purpose discussed above. Namely, the generics are attempting to telegraph to the Federal Circuit their narrative for this case, which is that these patents were nothing more than a series of life-cycle management hunting licenses designed to eventually jam up any generic release of Vascepa®. Everybody has their own narrative.
Amarin’s reply brief is due later this week, and it will be important to review Amarin’s responses to the points brought up in the generics’ brief.