Scientific discussion – JL, et al.
While we have this lull in news from Amarin, I would like to engage the minds on this board to discuss some scientific issues that I have been mulling over / researching.
First, let me say that I am not trying to revive the ghost of Pyrrhonian , but I do think that one should always have an open mind and constantly evaluate a position in a stock.
In my humble opinion, Amarin trades in the 16s now, as opposed to mid 20s, for one reason: mineral oil. Whether we like it or not, it has created enough doubt to really hamper the company’s perceived market value. Markets hate uncertainty is an axiom. In Amarin’s case, MO has been the MO to create the uncertainty. (Sorry but I love puns.)
So, here are a few things I’d like to discuss to hear opinions on the issue.
The first is anemia. There were 45 more cases of anemia in the placebo arm than Vascepa arm (NEJM appendix p 40.) This had a P value of .03. At first blush, this concerned me as I thought that the 4 grams of mineral oil per day may have hindered vitamin uptake in the small intestine. Worse, is that anemia is a large factor in CV events (lots of literature on this). After more thought, I wondered if Vascepa might actually have another beneficial effect – aiding the body in production of red blood cells? 4 grams of mineral oil is a very small amount to have much of an effect, especially since it doesn’t get absorbed. Could such a small amount (2 grams with two meals a day) coat 20 plus feet of small intestine sufficiently to inhibit vitamin absorption to the point that the placebo arm was having greater rates of anemia? To get to the point, my conjecture is that EPA must somehow be assisting the body in red blood cell production and it wasn’t that the placebo arm had higher than normal anemia, but rather the Vascepa arm was having lower than normal anemia. Love to hear some educated thoughts on the hypothesis.
Second, can we learn anything from the Number Needed to Treat of 21? This number is less than half on any NNT that I know of from a statin trial. (I don’t like NNT as measure as it does not include time. I personally think all NNTs should be standardized to a “per annum basis.” While that doesn’t mean that a medicine may be effective in the first year of use, it would at least take the randomness out of NNT. NNTs are usually calculated over the average patient time in a clinical trial, so the time factor can vary greatly. ) Next, of course is that patient populations vary greatly from trial to trial – the entry criteria can be quite different. Having said all that, is there anyway to compare length and patient populations in statin trials to REDUCE-IT? I ask that question because even a ballpark comparison could eliminate the mineral oil “uncertainty” factor. Some have suggested (foolishly IMO) that mineral oil accounted for all the RRR in Reduce-it - i.e. give the patients a different placebo and Vascepa would have done exactly the same as those on statins only. What does an NNT more than twice as good as any statin trial say about the "mineral oil accounted for all differences" argument? Again would love to hear the opinions of those who like to read trial reports and do lots of math. (I have been doing a lot of both in the last few weeks, but I don’t want to “lead the witnesses” with my opinions / conclusions.)
Again, I am not doing this to throw doubt on Vascepa. I own almost 1 million dollars worth of shares. I just want to continue to think about and evaluate my investment and I think these are two good topics to develop.
Best wishes,
AI
