Can Amarin’s patents protect Vascepa from generics?
Zachary Silbersher
Amarin Pharmaceuticals ($AMRN) will be going to trial soon to protect Vascepa® against generic competition. Amarin has asserted 15 claims from six patents against ANDAs filed by Hikma, Dr. Reddy’s and Teva. What are the issues to be addressed at trial?
This post provides an overview of the issues to be litigated at the upcoming bench trial scheduled for January 2020. There are numerous other questions currently surrounding this case. Will the parties settle before the trial? What dynamics are motivating the likelihood of settlement? Why did Teva settle, and does anything from that settlement read through to the trial? How does the REDUCE-IT study and Amarin’s bid to add a cardiovascular indication to Vascepa® impact this case? Can Amarin use the REDUCE-IT study in the upcoming litigation trial to prevail? If the generics launch at-risk, could they face exposure for off-label prescriptions of the cardiovascular indication? We will potentially address these and other questions regarding Vascepa® in upcoming blog posts.
Background
Vascepa® is indicated to treat patients with hypertriglyceridemia, which are severely high levels of triglycerides (“TGs”) above 500 mg/dl. The active ingredient in Vascepa® is highly purified omega-3fatty acid called ethyl-eicosapentaenoic acid (“ethyl-EPA” or “EPA”).
Amarin was not the first drug to treat hypertriglyceridemia with highly purified omega-3fatty acid. An earlier drug, Lovasa®, also used EPA in combination with another omega-3 fatty acid called DHA, to treat hypertriglyceridemia by lowering TGs. Another drug, Epadel, distributed in Japan, used only EPA to reduce TGs and was supposedly in clinical trial use before the priority date of Amarin’s patents.
Nevertheless, Amarin claims that it was the first to patent a formulation of EPA that lowered TGs without also increasing “bad” cholesterol (LDL-C). Though Amarin’s patent applications faced numerous rejections from the Patent Office as obvious, the patents were eventually allowed after Amarin showed that its dosage regime for pure EPA yielded unexpected results, namely, it allegedly does not increase of LDL-C levels and it reduces apolipoprotein B (“Apo-B”) levels.
Amarin’s case against the generics has been pending since 2016, and the case is scheduled to go to trial in January 2020. In late October, the district court in Nevada, where the case is pending, issued a ruling on cross-motions for summary judgment. That ruling essentially narrowed the issues to be litigated at the upcoming trial down to two principle disputes. First, the generics argue they do not induce infringement of the patents, and second, the generics argue that even if they do infringe, the patents are invalid as obvious.
Infringement
The patents are directed to methods of treatment. Because those methods are technically performed by the administering physician, and not by a generic pharmaceutical drug company, Amarin has been compelled to argue that the generics’ infringement is indirect rather than direct. Specifically, Amarin has argued that the generics will induce doctors to infringe the patents. The generics’ proposed labels are substantially similar to Amarin’s FDA-approved label. Thus, to show induced infringement, Amarin will effectively argue that the generics’ proposed label instructs doctors to perform each claimed element of the asserted patents.
The problem for Amarin, however, is that the patents require administering the drug for at least 12 weeks, but the label does not specify any duration for drug administration. Accordingly, the generics have argued that they cannot induce doctors to infringe the patents—including the 12-week limitation—because their proposed labels, which basically copy Amarin’s label, says nothing about how long to administer the drug.
Amarin’s response is that the label says more than the generics suggest it does. Specifically, the label refers to the Phase 3 MARINE clinical trial, where patients were administered the drug for 12 weeks. Based on that, Amarin argues that prescribing physicians would be encouraged to follow the time-period used in the clinical trial that backed up approval of the drug. Amarin also argues that because hypertriglyceridemia is a chronic condition, thus requiring indefinite treatment, doctors would understand that the drug should be prescribed indefinitely. That would therefore be at least 12 weeks.
The court has ruled that this issue is essentially a fact question on which it must hear expert testimony—namely, would doctors really read the label and determine the drug should therefore be prescribed for at least 12 weeks? More importantly, the court indicated it was persuaded, at least initially, by Amarin’s argument that because hypertriglyceridemia is a chronic condition, doctors would be encouraged to prescribe the drug indefinitely, which could necessarily encompass twelve weeks.
The court cited earlier precedent from a Hatch-Waxman case involving the drug Multaq® that also held that a label lacking a duration for treatment nevertheless induced infringement of a patent requiring 12-month administration for a condition that was also chronic. At the upcoming trial, the court will no doubt pay attention to the expert testimony on this issue, but barring any surprises, the court has for now indicated that it is inclined to side with Amarin on the issue of infringement.
Some of the patents require additional effects, such as specific lipid effects through administration of the drug (e.g., reducing Apo-B). For similar reasons, the court has allowed the issue of inducement on these claims to also go to trial. Either way, it bears emphasis that Amarin is not required to succeed on all 15 patent claims at trial to be entitled to an injunction against the generics launch. All of Amarin’s patents expire in 2030, and therefore, success on only a single patent claim theoretically could suffice.
Invalidity
The patent claims generally require, among others, the following elements: pure EPA, administered to patients with TG levels in excess of 500 mg/dl, 4 grams/day, for 12 weeks, without co-administration of DHA, where LCL-C levels are not increased. Among other things, some of the patent claims also require reduction of Apo-B. To show that the dosing regime and method of treatment was obvious, the generics rely principally on four prior art references: Lovasa®, Mori 2000, Hayashi and Kurabayashi.
During prosecution, Amarin argued that it was “unexpected” that administering pure EPA, without concomitant administration of DHA, to patients with high TG levels, did not increase LDL-C levels. The Examiner initially refused to allow the patents on this basis because the prior art suggested that a former drug, Epadel® (highly purified ethyl-EPA) could be administered to patients with TG levels under 500 mg/dl without a major impact on LDL-C levels. Thus, according to the Examiner, it was not “unexpected” that the same result could occur for patients with higher LDL-C levels.
Instead, the Examiner allowed the claims after concluding that it was “unexpected” that administering 4 g/day of EPA to patients with high TGs would decrease Apo-B levels. In particular, the Examiner concluded that the prior art showed no change in Apo-B levels from administering EPA to patients with TG levels below 500 mg/dl. In addition, Amarin’s MARINE study showed that this result was shown for 4 g/day, but not for 2 g/day, which evidenced certain criticality of the 4 g dose. The Examiner also concluded that there was a long-felt need for a drug for patients with high TG levels that did not also increase LDL-C levels, which is associated with cardiovascular disease, especially given that the two prior drugs approved for treatment of high TGs (Lovasa® and Triplix®) appeared to raise LDL-C levels.
EPA was previously sold as Lovasa® to treat high TGs. Lovasa® was not pure EPA, but rather a combination of EPA and DHA. Lovasa® was apparently administered to patients with high TG levels, with an attendant decrease in those levels. One of Lovasa®’s side-effects was increased LDL-C (bad cholesterol) levels, especially at high TG levels. Amarin argues that increased LDL levels were correlated to the severity of TG levels—thus, higher TG levels led to increased bad cholesterol levels when taking Lovasa®.
The generics rely upon Mori 2000 to show that, allegedly, DHA was responsible for increased LDL-C levels, rather than EPA. Mori 2000 discloses a study in which three groups of approximately 20 men were administered EPA, DHA or placebo, respectively. Thus, Mori 2000 is one of the few studies that examined the differences between individual administration of EPA versus DHA. On the one hand, the study showed lower increase of bad cholesterol with DHA versus EPA, namely, 8% for DHA and 3.5% for EPA. (The decrease of 3.5% for EPA was not considered statistically significant.) Given this, the generics argue that it was obvious that DHA, rather than EPA, was responsible for increasing LDL-C levels for Lovasa®, and thus, it would have been obvious to a person of skill that administering pure EPA would not increase LDL-C levels.
On the other hand, Amarin criticizes the study for its relatively small sample size, where none of the patients allegedly had TG levels in excess of 500 mg/dl, and both drugs yielded increased LDL levels. Specifically, Amarin claims that the baseline TG levels for the EPA group was 11% lower. This was significant, according to Amarin, since they also claim that increased LDL-C levels from taking EPA or DHA (principally from Lovasa®) was a function of baseline TG levels. Thus, the fact that none of the patients in the Mori 2000 study had baseline TG levels as high as required by the claims, according to Amarin, mitigates the showing of obviousness.
Hayashi was a study that disclosed 1.8 gm per day of EPA-E over eight weeks to patients with TG levels of 300 ± 233 mg/dl. (Based on this description, the parties dispute whether the Hayashi study would have statistically included patients with TG levels in excess of 50 mg/dl.) The study showed that TG levels decreased (41%) as well as LDL-C (7%) and apolipoprotein B (7%). Amarin critiques Hayashi because it was a relatively small size, the purity of EPA was not disclosed, and there is no proof that any patients had TG levels in excess of 500 mg/dl, or any data on patients that might fit into this category. This is also likely to be an issue of dispute at the upcoming trial.
Kurabayashi was a study that disclosed pure EPA (96.5%) along with estirol at 1.8 g/day for 48 weeks to women. The results showed decrease in Apo-B levels, and apparently also showed significant reduction in TG levels. Because the patients did not apparently include high TG levels (in excess of 500 mg/dl), Amarin argues the study did not undermine the expectation that purified EPA in patients with high TG levels would not experience a corresponding increase in LDL-C levels. Also, because the study administered EPA along with estirol, which is purportedly itself a lipid-altering agent, that did not change any expectations about the impact of pure EPA on LDL-C or Apo-B. The parties also dispute whether the study actually showed that EPA alone, rather than in combination with estirol, actually showed reduction of TG levels.
The obviousness question in this case clearly involves highly-factual scientific questions and the trial is likely to boil down to a battle of the experts. Whether or not it was “unexpected,” in light of the available prior art, that administering pure EPA to patients with high TG levels would not increase bad cholesterol, while simultaneously reducing Apo-B levels, is going to be a relatively close call. Nonetheless, there are a few obstacles that the generics will face even if they can mount a strong case of obviousness.
First, all of the primary prior art references relied upon by the generics were previously before the Examiner or discussed by the Examiner during prosecution of the patents. That, in itself, is not determinative that the district court will side with Amarin. There is precedent of courts invalidating patents based upon prior art previously considered during prosecution. Nevertheless, it does not help the generics, and adds an extra weight to their load.
Second, the burden of proving obviousness is on the generics, and it is a relatively high one. The generics must show that the patents are invalid based upon a “clear-and-convincing-evidence” standard. In short, if the case really comes down to a close call, and the court correctly applies the standard, that suggests that the court should in theory rule against invalidating the patents.
Overall, there are numerous issues in this case, and we will continue to address them in upcoming blog posts.