[ad_1]
Introduction
I hope that everyone reading this has a baseline knowledge of Amarin (AMRN), its sole product Vascepa, and the REDUCE-IT cardiovascular outcomes trial, or CVOT. Further, presumably everyone is aware of AMRN’s recent revelation of positive topline results of REDUCE-IT, with major adverse cardiovascular events, or MACE, reduced by around 25%.
I bought AMRN in the $10-11 range the morning it released those topline results and got an article out on Seeking Alpha as soon as I could. Given that the stock had been languishing in the low single digits for a long time, I wasn’t sure if setting a $15 price target into next year was aggressive or cautious. Well, the traders proved me cautious, at least for now. Sticking to my discipline, which was to wait for Saturday’s presentation of more complete data on REDUCE-IT at the American Heart Ass’n’s annual meeting, I scaled out between $15 and $20. I have had no position in AMRN since it first reached $20 and have no plans to go long or short.
AMRN is faced with two generic companies that have not settled with it as Teva (TEVA) has done. These are Dr. Reddy’s Labs (RDY) and Hikma Pharmaceuticals (OTCPK:HKMPY). The latter also does, or did, business in the US as West-World. BOTH RDY and HKMPY have filed ANDAs with the FDA and are in the early stages of litigation with AMRN. So it is possible that Vascepa could have generic competition by January 2020, when its most basic drug substance/drug product patent, #8188146, expires.
Now I want to get to some of my first thoughts on the REDUCE-IT CVOT, as discussed in the NEJM, a supplement to that NEJM article, and in a brief “rationale and design” 2017 article.
I do want to emphasize from the start that I am not thrilled to be presenting these initial observations. I would much rather this had been completely clean, at least at first read.
A placebo that may not be a placebo at all
The multi-year, 8000+ patient REDUCE-IT trial of high-dose eicosapentaenoic acid, or EPA, was designed to take Vascepa from a niche product for very high triglyceride levels to common usage for patients at high risk of a CV event. AMRN plans to request a label enhancement for Vascepa so that it is indicated for certain common populations to reduce the risk of MACE, and also to reduce the risk of CV death.
Given the known benefit of statins for most high-risk CV patients, the base case for REDUCE-IT was to study Vascepa as add-on to statins. Vascepa was given as two large capsules twice a day. The placebo group was given a preparation containing mineral oil.
The mineral oil is a problem for me, for reasons I will show next.
What is mineral oil?
Berkeley Wellness, in collaboration with the University of California’s (Berkeley) School of Public Health, provides the following general description of this agent:
Like its relative petroleum jelly (petrolatum), mineral oil is an inexpensive byproduct of refining crude oil to make gasoline and other petroleum products.
Mineral oil is sold OTC and typically used for the skin, as the blog goes on to say:
Like petroleum jelly, mineral oil is an effective emollient, which forms an oily layer on the skin that traps water. Thus it is a common ingredient in moisturizers and other skin care products; some people use mineral oil straight from the bottle to treat dry skin.
What about chronically ingesting mineral oil? The blog piece addresses this:
Is it okay to consume mineral oil as a laxative?
We don’t recommend it. Mineral oil, taken orally, acts as a laxative because of its lubricating effect; it also helps stool retain water and thus stay soft. It’s sometimes advised temporarily for people who have pain caused by hemorrhoids or an anal fissure or after rectal surgery. However, it can have adverse effects, notably oily leakage from the anus and malabsorption of fat-soluble vitamins and carotenoids from foods.
“Malabsorption” refers to incomplete absorption. The fat-soluble vitamins are A, D, E and K. Mineral oil may also inhibit absorption of certain drugs, such as statins.
So even though the dose of mineral oil was low, giving it for years strikes me as not a placebo.
Thus I have serious questions about the basic design of this study. I must therefore put in a caveat, which is that I have no knowledge of what FDA and the clinical trialists have said about this point. I am going to refer to the placebo also as a “placebo” to remind readers of my point of view, which is mine only at this point. Also, as I repeat later, this is an investment article. All worries matter. The truth will out after the FDA does its thing next year. As a trader, my take is to consider the mineral oil as an active agent and not a true placebo. If I’m wrong, then all that happens is that I may miss another move up in the stock, including a potential takeover.
But if I’m right, I may miss a big sell-off.
Here’s why I’m concerned that the mineral oil may have harmed the health of the placebo/”placebo” group.
Did mineral oil prove a confounder?
Some detective work led me to raise this question. I went to the tables in the supplement that is linked to above. On page 40, Table 6 shows 23 side effects that Vascepa might have been associated with. Only a few were statistically different between Vascepa and placebo. These were
- diarrhea, 2.1% more common with “placebo”
- constipation, 1.8% less common with “placebo”
- anemia, 1.1% more common with “placebo”
- peripheral edema (leg swelling), 1.5% more common with Vascepa.
Of these, the P value, a measure of whether a comparative difference could have occurred by change, was very low (i.e. persuasive that the effect was not due to chance) for diarrhea, constipation and edema.
Since anemia was of marginal statistical significance and edema was more common with Vascepa, I’m going to focus on diarrhea and constipation.
Guess what? These are just what one would expect if mineral oil was having a clinical effect on some people. More diarrhea, less constipation. What we do not know from this is if mild effects on GI function occurred in more than just a couple of percent of mineral oil patients. In other words, could the “placebo” have harmed the health of patients rather than acting as pure placebo?
So far, that’s just a question. Now to p. 38 and Table 4. This gives lipid and other parameters. With this, I see a more direct problem.
The “placebo” group may have been getting an anti-statin
Using the Hopkins method of measuring LDL cholesterol (in mg/dl), these are some of the Vascepa numbers:
- baseline: 85.8
- month 4: 83.6
- year 2: 85.5.
So, a mild initial drop, then no change at steady state after 24 months on treatment.
Now, look at the same numbers for the “placebo” group:
- baseline: 86.7
- 4 months: 93.7
- year 2: 96.1.
Within just 4 months of going on “placebo,” for unknown reasons the LDL-cholesterol level in these patients jumped 7 points, or 8%. Because of the mild baseline difference in LDL-C, the difference at 2 years between Vascepa and “placebo” was larger than 7 points. It was 10.1 points. Looking at it from the “placebo” standpoint, it is as if the “placebo” group went on a mild lipid-lowering agent and lowered LDL-C by 10.1/93.7, or 10.8%.
My guess is that there is no unknown reason. I suspect that the mineral oil impeded the absorption of the statin, and that the trial design was defective in not correcting for that.
In addition, and consistent with there being an anti-statin effect, statins lower a measure of inflammation known as C-reactive protein, or CRP. This was also affected by “placebo.” To wit, Vascepa lowered CRP (listed in Table 4 as hsCRP) from 2.2 mg/L) from 2.2 at baseline to 1.8 at both follow-ups. But the “placebo” group again did not act as if it were receiving an inert agent, showing these numbers:
- baseline: 2.1
- year 2: 2.8
- last visit: 2.8.
Finally, while perhaps of no clinical importance, the levels of the drug, EPA, were measured in the EPA group and the “placebo” group. Naturally, from baseline to study conclusion, EPA levels in the Vascepa group soared, from 26 to 144 (micrograms/ml). But EPA levels dropped in the “placebo” group, from 26.1 to 23.3. This is a 10.7% decline. Could this, I wonder, have played perhaps a slight role in negatively affecting the health of the “placebo” group?
The authors of the NEJM article addressed this.
Comments from the article on this issue
From the final section, the Discussion, the authors minimize this issue, saying, in full:
Second, if mineral oil in the placebo affected statin absorption in some patients, this might have contributed to differences in outcomes between the groups. However, the relatively small differences in LDL cholesterol levels between the groups would not be likely to explain the 25% lower risk observed with icosapent ethyl, and a post hoc analysis suggested a similar lower risk regardless of whether there was an increase in LDL cholesterol level among the patients in the placebo group. Although JELIS was designed as an open-label study that did not use a mineral oil placebo, it showed a 19% lower risk of ischemic events with statin therapy plus EPA than with statin therapy alone.
The JELIS study that the authors referred to was a Japanese study that led AMRN to think that REDUCE-IT had good reasons to succeed.
I have to be cautious in making comments on a study I have just read about. The authors know the details many times better than I. That said, they may be invested intellectually (at the least) in the study. We would all wish it to succeed and thus open up an important new therapeutic modality.
That said, here’s my thoughts on the matter as a retired cardiologist with some experience in this field.
Not loving the NEJM comment on mineral oil as placebo
I am not persuaded that the authors have provided full context.
First, let’s say that they are quite sensible that the roughly 10% changes in LCL-C, CRP and EPA levels are not enough to cause a 25% reduction in MACE. But what if they are enough to explain perhaps a 10% reduction in MACE? In that case, what would clinicians make of a 15% or so reduction in MACE? What would the FDA say? That’s a big difference given the competition from ezetimibe and PCSK9 inhibitors, and very possibly from late-stage pipeline drugs. Also, such important outcomes as the 20% decline in CV mortality might fail to reach statistical significance.
Second, the authors refer to a “post hoc” analysis. This is a retrospective, non-specified look back at the data. I’m skeptical of that one. That’s because if I read the sentence correctly, they ignore the fact that if the “placebo” group were in fact on placebo, the LDL-C and CRP likely would have changed little at 4 months and throughout the study. Thus, some patients would have seen declines in LDL, some would have stayed stable, and some would have increased. What I care about is one prespecified group comparison to the other group.
Third, this was a brief article. Every word counts, and some and maybe most of what the authors want to say is edited out. Given that, why did they even bother mentioning JELIS? It’s almost completely irrelevant to interpreting REDUCE-IT; different populations, different doses, etc. That it survived in the final draft suggests to me that these are authors who are partisans of Vascepa. I could be wrong, but I’m writing a stock market article, not doing scientific research, and as investors and traders, we need to think of what might be the case without knowing for certain.
What about the FDA?
The FDA will look hard at this study
If I had to guess, the FDA was aware of the study’s design. But whether it raised any caution about the use of a mineral oil placebo and its potential confounding effects in case it showed a drug-like effect is something completely unknowable to me. We may never know that sort of detail.
I expect the FDA to perform complex data analyses that is beyond any analyst’s ability to suss out at this point. As an example of an old set of shockers, in the early ’90s, two companies simultaneously came up with drugs for septic shock. The companies were Centocor, later acquired by J&J (JNJ); and Xoma (XOMA), which is still hanging around with about a $150 MM market cap. If I remember correctly, each company’s drug went to an advisory committee, which was either unanimous or nearly so in favor of the drug’s safety and efficacy. You can imagine how robustly the stocks were trading. Then, the FDA nixed both drugs. What FDA had done was a careful and complex reanalysis of the data and decided that, clearly, the drugs were not effective, even if they were safe.
When I was in the industry, one of my roles involved reading the Summary Basis of Approval, or SBOA, of numerous approved, branded drugs. Basically in those days of paper records, the FDA would send the company a box of many hundreds of pages of its deliberation, and then it would be shipped to my house. Reports from every department, from toxicology to one or two different statistical analyses, were shown. I would read or at least gaze at every page, skipping none, and use my grad school level statistics knowledge to try to keep up with how the FDA’s statisticians sliced and diced the data. These people were good, really good. In addition being good, they had all the granular detail based on individual patient records that allowed them to do analyses no outsider can do. I’m confident that the FDA will comb through the REDUCE-IT results in complex ways and that the study will certainly enter the Prescribing Information. But…
My bottom line here: I would not even guess at how the FDA will look at the REDUCE-IT study. This may be a toughie.
Thus, simply, I see risk here to AMRN’s trading patterns and its ultimate value.
Conclusions
AMRN is in my view risky from the start because of the two ANDAs that have been filed with the FDA and that are being litigated. At Friday’s closing price of $21.05 and a market cap around $6.5 B, AMRN needs years of large profits to simply get to that level of earnings.
My take for now is that REDUCE-IT needs very careful study by the FDA. The strong topline CV outcomes numbers that AMRN and the investigators emphasized would indeed be great if the placebo were indeed just a placebo. However, my reaction from the data I have seen is that the side effect profile fits with the mineral oil in the placebo being present in enough quantity to have affected the GI physiology of an unknown percentage of recipients. The trend toward more anemia in the placebo/mineral oil group is also disconcerting. A lower level of hemoglobin could well predispose a patient with atherosclerotic disease to suffer a MACE.
In addition, I read over the linked articles and supplement more than once, but I saw no mention of comparative blood pressure changes. Did blood pressure worsen even a little bit in the placebo group? Did that group end up receiving more antihypertensive therapy than the Vascepa group? And so on. Lots of questions, no answers. The FDA will know everything.
In summary, perhaps unexpectedly to the designers of the study, the placebo group may actually have been on an agent that had drug-like effects. Those effects strike me as being anti-statin effects on lipids, and to have caused adverse GI side effects. Thus it may be that part of the difference between Vascepa and placebo was because it was as if the placebo group had their statin dose lowered. If so, how big a part? Again, I’ll opine: a toughie.
I have no firm idea of how traders will look at AMRN now, but I have a firm idea that after running up 10X, AMRN is a risky long given the genericization risk. I’m glad I sold. If it trades higher, if it is acquired by a big player, congrats to the longs. Yours truly is going to stay on the sidelines and wait first for commentary from the CV intelligentsia, but all that really matters is what the FDA opines. I think that all the rest is guesswork.
Again, I do wish AMRN and Vascepa the best. The world could use a safe and effective new oral agent for a large, needy section of the population.
Thanks for reading and sharing any comments you wish to contribute.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Not investment advice. I am not an investment adviser.
Editor’s Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.
[ad_2]
Source link
