Amarin’s appeal of the district court’s decision invalidating the Marine patents is now fully briefed. Amarin’s reply brief is strong, but does it tip the odds in favor of Amarin prevailing on the appeal?
We blogged about Amarin’s opening brief here, and the generics’ brief here. As discussed in our earlier post, Amarin’s opening appellate brief set forth three principle arguments: (i) the Cyclobenzaprine argument; (ii) the “weighing” of the secondary considerations argument, and (iii) the “two patient populations” argument. Our short-takes on Amarin’s reply brief are that Amarin appears stronger on the first argument, about the same on the second argument, and a bit weaker on the third argument.
Cyclobenzaprine and “weighing” the secondary considerations
In our earlier posts, we discussed certain drawbacks to the Cyclobenzaprine argument. Yet, rather than backing away from this argument, Amarin’s reply brief doubles-down on Cyclobenzaprine and continues to make it the centerpiece of its appeal.
The caselaw suggests that there may not be a consensus at the Federal Circuit in favor of adopting the framework articulated in Cyclobenzaprine. That would suggest, as we previously wrote in earlier posts, that leading with this argument has risks. Nevertheless, Amarin’s reply brief has an interesting effect that highlights a tactic behind leading with this argument. Namely, Amarin is using the Cyclobenzaprine case as a compelling vehicle to persuade the panel—and whichever Judges are eventually designated to make up that panel—to reconsider all of evidence afresh.
For those Judges at the Federal Circuit who are already sympathetic to Cyclobenzaprine’s framework, Amarin will likely already have their ear. Yet, for other Judges, they need not adopt the Cyclobenzaprine framework itself to nevertheless be persuaded that this case was decided the wrong way.
In Amarin’s telling, this case is all about the objective indicia of non-obviousness—otherwise known as “secondary considerations.” Judge Du placed too much emphasis on the prior art. By doing so, the lower court improperly “pre-judged” the case. By the time the court got around to assessing the secondary considerations, it had already implicitly shifted the burden to Amarin to show that its secondary considerations should overcome that “pre-judgment.”
Amarin uses Cyclobenzaprine as a launching point to retell the invention story and take back the narrative. The generics’ narrative frontloads the prior art, starting with the premise that Lovasa was known to raise LDL-C, and Mori suggested that DHA, rather than EPA, was the culprit. According to Amarin, this is backwards. The secondary considerations must be considered alongside and part-and-parcel of any assessment of the prior art. If Judge Du had done that, the questions asked by the lower court, as well as the corresponding answers, would have been different. For instance, if using pure EPA in severe patient populations was so obvious, why did it take so long? If Mori’s teachings were so compelling, why was Lovasa® launched years after it? If it was obvious from Epadel that EPA was LDL-C neutral for severe patients, why didn’t Mochida bring it to the US market earlier?
Overall, Amarin’s reframing of the evidence in this manner is compelling. If the secondary considerations are considered alongside the prior art, rather than only after a determination of a prima facie case, the use of pure EPA in severe patients was not obvious. Viewed in that way, Amarin’s point is that the eventual composition of the panel should not matter.
Yes, it would be preferable to score a panel with Judges already sympathetic to the Cyclobenzaprine cause. But even if that doesn’t happen, Amarin’s Cyclobenzaprine discussion is an invitation to reconsider all of the evidence—and importantly, consider the objective indicia of nonobviousness alongside the prior art. That reconsideration would be de novo. Even for those Judges unsympathetic to recasting the obviousness analysis in the mold promoted by Cyclobenzaprine, Amarin is betting that that de novo review by any Judge eventually designated to this appeal should be sufficient for reversal.
Yet, as compelling as it is, there nevertheless remain risks to this tactic. Judges are people, and to justify reversal, they typically search for a simple, clean error of law or fact that reads through to the outcome. Amarin’s Cyclobenzaprine argument requires the Judges take up the invitation to reconsider the entire balance of the evidence. That is a tall request that some Judges may decline if there are easier ways either in or out of this appeal. While Judge Du’s decision has spawned considerable discussions among investors, it has not attracted much significant attention within the patent community. That is reflected in the low number of amicus briefs filed in support of either side. Thus, the risk for Amarin remains that its Cyclobenzaprine argument, as artful as it is, may fail to earn the level of attention required to get to a reversal.
This is where the “weighing” of the secondary considerations argument remains important, and why it may be Amarin’s best chance for reversal. For those Judges either unsympathetic to the Cyclobenzaprine framework, and disinclined to conduct a broad sweeping review of all of the evidence, the “weighing” of the secondary considerations argument offers a much easier and cleaner way out of this appeal. While Amarin does not add much to this argument in its reply brief, it remains a sound and fortunate complement to the Cyclobenzaprine centerpiece. Oral argument could likely tell which of the two approaches the Court may be more interested in.
The “two patient populations” argument
Judge Du found that, before Amarin’s patents, it was expected that EPA would be LDL-neutral if administered to severe patients. The lingering obstacle to Amarin’s chances of reversing Judge Du on this point is Dr. Heinecke. He was the generics’ expert, and he testified that the Mori study showed that EPA might reasonably be expected to be LDL-C neutral in patients with severe hypertriglyceridemia. We can all go back and forth about what Mori did and did not teach, but ultimately, Judge Du’s decision hinges on Dr. Heinecke’s opinion, and overturning that decision requires undermining his opinion.
The Federal Circuit does not perform fact-finding. From its perspective, the question is not necessarily what does Mori teach or not teach. Instead, its task is to determine whether Judge Du’s findings about Mori are supported by evidence. If a Judge credited the trial testimony of one expert over another, that is typically afforded considerable deference on appeal. That is what makes overturning a find finding based upon expert testimony so difficult.
Amarin knows this. In its reply brief, it attempts to cut out the problem at the root—namely, by arguing that Dr. Heinecke’s trial testimony was off the reservation. Amarin now claims that Dr. Heineck’s testimony was conclusory and not supported by any extrinsic evidence. Amarin cites to caselaw holding conclusory testimony cannot rise to the level of clear-and-convincing evidence. (Reply at 18). Amarin does not appear to have mad this argument in its earlier brief, and so showing up for the first time on reply is not likely to make an enormous impact.
Either way, the Federal Circuit may question whether Dr. Heinecke’s testimony was, in fact, conclusory and unmoored from an extrinsic evidence. In the generics’ telling, Dr. Heinecke testified that the 500 mg/dL threshold was set for a risk related to pancreatitis, not anything related to LDL-C. (Bench Order at 60). (We can debate whether the pancreatitis threshold is, in fact, related to LDL-C, but Amarin does not appear to have forged that point with great force.) Studies presented at trial showed that patients with TG levels at 400 did not experience increases in LDL-C. Dr. Heincke also testified that there was no “magical” mechanistic difference between patients at 400 and those at 500. (Id. at 61). Given that, the generics argue that extrapolating from patients at 400 to patients at exactly 500 is not enormous leap. Buttressing this leap, Judge Du cited Dr. Toth’s testimony that for other effects, and it was not unreasonable to make predictions for patients at 500 based upon observations of patients at 400. Whether this leap had a basis in the evidence versus being clearly erroneous is what this entire case was all about.
It is not until page 24 that Amarin finally addresses the generics’ principle contention—namely, that all that matters here is whether the evidence supported an expectation that pure EPA would not increase LDL-C in a patient with TG levels at exactly 500 mg/dL. Amarin’s principle retort is that the Lovasa® label warned that patients above 500, but also including patients at exactly 500, could experience significant LDL-C increases. Amarin’s point is that evidence in the art suggested a material patient-population departure at 500 mg/dL. At that level, according to Amarin, persons of skill believed that LDL-C increases could be markedly higher. And that is precisely why, according to Amarin, prior art references such Mori and Hayahshi are irrelevant.
This response is unfortunately not as crisp as someone sympathetic to Amarin’s plight may have hoped for. Notably, Amarin does not point to any admissions by Dr. Heinecke that were solicited at trial. Again, Amarin’s principle task here is to dissemble Dr. Heineck’s opinion. His opinion was the lynchpin of Judge Du’s decision to invalidate the patents. The reason the generics have placed so much focus on a patient with exactly 500 mg/dL is because that is what exactly Judge Du did in the Bench Order. (Bench Order at 58). Judge Du cited to caselaw holding that if a patent covers a range of patients (eg., TG levels from 500 to 1500 mg/dL), all that is required to invalidate the patent is to show prior art anticipating any patient within that range, including a patient at exactly 500 mg/dL.
Amarin cannot challenge this caselaw because it is firmly established. And thus, it is forced to try to take apart Dr. Heinecke’s opinion. That is a more difficult task. Its main point is that Dr. Heinecke’s opinion is an outlier when viewed within the context of all of the available prior art—including, other TG-lowering drugs and the understood mechanism-of-action.
But, at the end of the day, Amarin cannot hang its hat on any clear admission by Dr. Heincek undercutting his own opinion. That is the best type of evidence usually required to show that a finding of fact based upon an expert’s opinion was clear error. Instead, what Amarin can and does point to is a wealth of evidence to support its “two patient populations” argument. Indeed, this is what made this case a close call in the first instance.
Yet, the problem that Amarin now faces is that it is no longer trying the case to a fact-finder. Rather, the case is on appeal. Amarin’s contention that scientists did not expect pure EPA to be LDL-C neutral in severe patients is not the issue. The issue, rather, is whether Judge Du’s finding that it was expected was clear error.
Thus, the appeal remains a dispute between the prior art adequately taught enough about the “two patient populations,” on the one hand, or whether what it taught was sufficient for a patient at exactly 500 mg/dL. While it may be intuitive from a medical perspective to discuss two patient populations, the generics’ argument, which has a basis in law, is that all that matters is what the patents require. And to anticipate the patents, all that is required is a showing relevant to a patient at exactly 500. Whether Amarin has, or even can, do enough to undermine that line of argument simply highlights the fact that, for better or worse, the Judge credited the testimony of the generics’ expert over that of Amarin’s.
What about Dr. Bhatt’s article?
There has been much discussion about whether Dr. Bhatt’s article can be a savior and blow the case wide open on this particular issue—namely, the fact issue and Judge Du’s interpretation of Mori. We previously weighed in on the utility of Dr. Bhatt’s article at this stage in the comments section of our earlier post. The short answer is that the article and the argument have likely been waived and are not admissible in this appeal.
Since the article’s publication, there have been no amicus briefs filed attempting to introduce the article, and Amarin did not raise either Dr. Bhatt’s article or the line of argument from Dr. Bhatt’s article in its reply brief. Thus, Amarin has already side-stepped its best opportunity to try to sneak it in—namely, in its reply brief.
Given all that, it would be very, very surprising at this point if Amarin’s appellate counsel chose to raise, during the oral argument, which is limited to 15 minutes per side, an article that is not included in the Appendix, or a line of argument that was not raised in any of the briefs to the Federal Circuit. The generics would raise a stink about it, the Court would likely have very little patience for it, and therefore, doing so could tax Amarin’s credibility in front of the panel. The topics actually discussed during oral argument are typically dictated by the Judges’ questions, and there can be no guarantee that Amarin’s counsel would even have the time to raise this discussion.
Getting a reversal on appeal is hard. Yet, Amarin has undoubtedly made compelling arguments. While its odds of prevailing may still be less than likely, its briefing has raised those odds high enough to have a veritable fighting chance. The next big milestone will be the oral argument. Each side is technically permitted only fifteen-minutes of argument, but the Judges can extend that time if they wish to. During the oral argument, the Judges often ask questions, and those questions can sometimes be a tell into which issues the Court views to be most important.