Neither Du nor Heinecke are the bad guys
I’m afraid Du will not reverse her decision based on mistake or fraud, and all the statistical hand-wringing will not change her mind. I am not a patent law expert, and have forgotten way more than I retained from the one International Intellectual Property Law class I took. But I have scanned through much of the transcript from the case. Much emphasis is appropriately placed on the POSA (person of ordinary skill in the art) and what they would have thought about the ‘obviousness’ of the prior art. Valid patents, of course, must be non-obvious when considering the prior art. While the parties differed slightly, both Amarin and Hickma basically said that a POSA should be someone with a medical degree and several years of experience in the treatment of lipid blood disorders. Neither party suggested that the POSA should have a level of statistical expertise that was worth mentioning.
So, the question is not whether there were statistical mischaracterizations made, it is whether a POSA with the designated skill set would have looked at results from Mori and Kurabayushi and said “c’mon man, it is obvious that EPA is not going to increase LDL-C like DHA does. Mori showed, in individuals with mild hyperlipidemia, the DHA group significantly increasing in LDL-C, while the increase in LDL-C with EPA was not significant. This data point alone would not come anywhere near the clear and convincing standard required of the overall decision, but might a lipidologist with rudimentary statistical training say “it’s obvious that EPA might lower triglycerides without raising LDL-C”?
Further, it was Mori and not Heinecke that said, without proper foundation, that DHA and EPA had differential effects on lipids and glucose metabolism. Heinecke might have capitalized on this, but a POSA might have reached the same conclusion.
A POSA might have had an even easier time saying that the findings from Kurabayushi would have made development of EPA for hyperlipidemia obvious. Nonsignificant reductions in Apo-A and Apo-B in hyperlipidemic, menopausal women with estriol alone (known to elevate triglyceride levels) but highly significant reductions at each time point in women with EPA added to the estriol (27% reduction in triglyceride levels). Might a POSA have said, “c’mon man, it’s obvious someone should develop high purity EPA to combat hypertriglyceremia” knowing the results WITHIN these two groups.
Testimony at the trial also indicated that Amarin told investors in 2010 that results from Mori and Kurabayushi bode well for the development of AMR-101 (Vascepa), which certainly did not help them with their non-obvious arguments.
Heinecke did not tell blatant lies (see testimony below). He did what Amarin’s expert witness did as well – couched his answers in the most favorable terms to who he represented.
Really, the only thing that Amarin might have done and been successful would have been to put a statistical expert on the stand to get into an unfettered discussion about what Mori and Kurabayushi specifically did NOT say. I don’t think anything less than that would have worked on Du. But I do think that if Professor H. Kush has been put on the stand for a mer 60 minutes to explain the ins and outs of these findings, Du would have listened and maybe even ruled accordingly. It would have been a small investment for Amarin, the need should have been anticipated, and interviewing your own expert beats the hell out of mudwrestling the other side’s medical expert to a draw.
While Du would have listened to statistical expertise before the decision, doing so post-verdict is another kettle of fish. The Curfman paper is very strong. It is not so much opinion as statement of fact. Du will come to understand the statistical principles that Marjac puts forward, but she can still tell herself that the verdict should stand because a POSA still would have found the invention obvious. I think ‘mistake’ might have been the way to go with Kurabayushi too. Telling Du she was defrauded by an omitted legend containing redundant information sends the message that she was incapable of interpreting the ‘NS’s’ in the table without a legend holding her hand. She probably is not willing to accept that version of herself.
Let’s not forget that the Examiner at the Patent Office granting Amarin patents indicated that they were prima facie obvious based on the prior art. It was only based on secondary considerations that the patents were granted. Relatedly, Du did say that the ‘long-felt but unresolved need’ indicia for non-obviousness worked in Amarin’s favor, but she was viewing this factor as secondary to her analysis of the invention and the prior art, per the Federal Circuit’s obviousness test. Whether those secondary factors are to be integrated into the primary analysis is the issue before the Supreme Court.
If (huge if) the Supreme Court took the case, indicated that the non-prior art considerations deserved primary consideration, and remanded to the District Court, then Marjac’s brief well might provide her with the rationale to tip the scales to rule in Amarin’s favor. But that would be a long time off and she is not going to stay her original opinion in the interim. If there is a denial of cert on April 19th or the immediate weeks following, Du would likely deny Marjac’s Rule 60 motion shortly thereafter. I don’t think Marjac lives in an alternate reality. And that is why he deserves such high kudos for expending so much in the way of time and energy in the face of overwhelming odds against him. Thanks again Marjac/HK.
------ Some Heinecke testimony ------------(DDX's refer to exhibits)
Q. Did Kurabayashi report any changes in apo B levels in patients taking purified EPA?
A They did. They state that the apolipoprotein B level in the EPA group was significantly lower at week 48. They have those results shown in the table. What we're looking at here is the percent change in apo B at week 48, and what is found is that there is a nonsignificant change in the control group of minus 1.5 percent, and a highly significant change, a decrease, in the EPA group of 6.9 percent. The P value there is less than .001. That's very significant. And this was -- the analysis in this study was carried out by the one way repeated analysis of variance. So I might mention one of the advantages of this study is that they had multiple determinations of the apo B levels over the course of the study, and so they were able to use all of those in concert to determine the change in the apo B levels. And this is a more powerful approach than simply measuring the values at the beginning and the end of the study.
Q Now, based on these data, would a person of skill in the art have attributed the reduction in apo B to the estriol or to the EPA in Kurabayashi?
A The EPA therapy.
Q And why is that?
A Because a control group had a nonsignificant decrease in apo B, and the addition of EPA led to a significant decrease in apo B levels. So in this study design that would indicate that the effect was due to the EPA treatment.
Q. Could you identify this document for the record.
A Yes. This is the 2000 publication of Mori, entitled "Purified EPA and DHA Have Differential Effects on Serum Lipids and Lipoproteins."
Q Now, turning to DDX 6.44, there is another snapshot of the Mori study, which is DX 1538, from pages 1 to 2. What was the objective of the Mori 2000 study?
A The objective was to determine whether EPA -- and they say "and," but I would say "or DHA" -- have differential effects on serum lipids and lipoproteins. It was a double-blind, placebo-controlled trial. That means that the investigators were not aware of the intervention, that they used two interventions, and they compared the effects of those interventions with a group that received no therapy, a placebo group. The treatment was 4 grams of purified EPA a day, or purified DHA. And I'll note that none of the subjects were regularly taking lipid-lowering drugs, or other drugs known to a lipid -- known to affect lipid metabolism.
Q So were any of the patients in Mori 2000 taking concurrent lipid-altering therapy?
A No, they were not, and that would include the estriol or other estrogenic drugs.
Q So turning to DDX 6.45, with another snapshot of DX 1538, this time from pages 2 to 3, what was the purity of EPA used in the Mori 2000 study?
A They note that it was approximately 96 percent pure, and, again, it was given at a dose of 4 grams a day.
Q And according to Mori, what were the effects of 4 grams a day of purified EPA and purified DHA on lipid levels?
A Again, they observed a significant reduction in -- here they call them triacylglycerols, but that's triglycerides -- of 18.4 percent with EPA and 20 percent with DHA. In my opinion, the difference between 18 percent and 20 percent reduction is indistinguishable and of no clinical significance. They also note that LDL cholesterol increased significantly with DHA, by 8 percent, but not with EPA, strongly suggesting that these two Omega-3 fatty acids could have distinct effects on LDL cholesterol levels.
Q Now, does the 3.5 percent higher value for EPA indicate that there was an increase in LDL-C levels for the EPA group?
A No, it does not, and I think this comes back to trying to understand what the significance of changes is in terms of statistical testing. If one repeatedly measures a variable over and over again, there will always been variation, and so by chance alone you can have a small increase or decrease. The way one assesses a significance of a change is by making the assumption that it is different from chance. And so this is a key concept. I think it's very important to realize that small variations around the mean do not mean that those are necessarily significantly different
Q Now, turning to DDX 6.73, did the key prior art that you rely on disclose the elements that appear in all of the asserted claims?
A In my opinion, yes. The first key element is for the patient to have a triglyceride value of greater than or equal to 500 milligrams per deciliter, and I believe the Lovaza and Hayashi publications support this. The duration of therapy had to be at least 12 weeks, and, again, the Lovaza and the Kurabayashi publications would support this. The dosage needs to be about 4 grams a day, and, again, Lovaza and Mori. And, finally, the purity of the preparation of the EPA should be about 96 percent, and Mori and Kurabayashi both use EPA preparations of this purity.
Q And turning to DDX 6.74, did the key prior art that you rely on disclose the remaining elements that appear in only some of the asserted claims?
A Yes. So some of the claims require no increase in LDL cholesterol, and this would be addressed by Mori, Hayashi, and Kurabayashi. There was another requirement for a triglyceride reduction of at least 20 percent, and Hayashi addresses this issue. There was a requirement for an apo B reduction, and Kurabayashi indicates this. Moreover, Mori, et al., demonstrated that unlike DHA, purified EPA, 4 grams a day, reduced triglycerides without increasing LDL cholesterol levels. And it's -- it's obvious to me a clinician interested in this situation would have considered using 4 grams of purified EPA to treat patients with triglycerides greater than 500 with the desirable goal of not increasing LDL cholesterol level.
Excerpts from Heinecke cross examination ---
THE COURT: Counsel, I'm going to ask you to pause for a moment. I keep hearing this odd scratchy noise. Miss Clerk is going to find out what's going on.
THE WITNESS: Oh, I'm sorry, that's me. I apologize, Your Honor. I'm getting a little agitated over here.
THE COURT: All right.
MR. SIPES: I'm doing my best, Your Honor. I don't mean to agitate any one.
THE COURT: Perhaps you should take a deep breath between each sentence. That would also help the court reporter. All right. Let's resume.
...
Q Dr. Heinecke, in 2000, you –
A This might be the FDA's conclusion, but it's not my conclusion.
THE COURT: So, Dr. Heinecke –
THE WITNESS: Yes.
THE COURT: You should listen to the question asked. There was no question asked whether you agree with the FDA or your opinion of the FDA. You should take your advice about breathing before you answer.
THE WITNESS: Thank you.
THE COURT: So let's wait for the question.
THE WITNESS: Thank you.
THE COURT: All right?
THE WITNESS: Thank you. I will try and do that. Thank you very much, Your Honor.
