Amarin Corp Plc (AMRN)

[Pyrrhonian]

ps.: P- I will explain / detail later (tom.) the 75% approach --HD

Okay good. While you are at it:

- you could not separate FDA approach from FDA sponsored / supported) publication regarding ARR --HD

I'm going to go off on a tangent for a second about DHA, shown to lower blood pressure and heart rate, while EPA has not been shown to do so.

(allow me please to link to another potentially 'fraudulent' study):

http://hyper.ahajournals.org/content/34/2/253

And FDA has stated:

Actuarial data and epidemiological studies such as the Framingham Heart Study have shown that elevations in blood pressure (systolic or diastolic) are associated with an increased risk of cardiovascular events. These data show that this relationship is monotonic — absolute risk increases progressively with increasing blood pressure — and approximately exponential — the absolute risk increase per mmHg increases with increasing blood pressure. Systolic pressure may be more important than diastolic pressure, especially in the elderly.

The effect of blood pressure on relative risk appears to be similar in people at high or low absolute risk. Therefore, the absolute risk increase per mmHg of blood pressure elevation is much greater in patients whose risk for cardiovascular events is high for reasons other than blood pressure, such as patients with diabetes mellitus, chronic kidney disease, a history of stroke, or cardiovascular disease.

Among adults, placebo-controlled outcome trials have been conducted with combination regimens of drugs in numerous pharmacologic classes (e.g., diuretics, reserpine, beta-adrenergic receptor blockers, direct vasodilators, and calcium channel blockers), and large trials have consistently found reductions in the risk of cardiovascular events. The largest effect has been reduction in the risk of stroke, but reductions in the risk of myocardial infarction and cardiovascular mortality also have been seen.

http://www.fda.gov/downloads/Drugs/.../Guidances/ucm075072.pdf

What about heart rate?

http://www.ncbi.nlm.nih.gov/pubmed/15644652

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4437199/

http://www.clarify-registry.com/download/CLARIFY_Heart_Rate_in_CAD.pdf

Lovaza > Vascepa imo for the above, and other, reasons. That's why it remains tier 1 on all plans. Including the largest plan in the country. While that same plan has Vas as tier 3.

Lovaza is closer to replicating high concentrated doses of the active ingredients in fatty fish. This extends beyond just EPA of course (and beyond DHA too, incidentally).

But n-3s in diet is only one part of the equation that significantly lowers risk of CVD and MCE.

Back to the point of FDA and ARR, I've already cited current thinking from their own site, despite a legal disclaimer saying they cannot be held to everything written in it (obviously), it still represents their thinking. And one reviewer in particular actually derided the idea that a 2% to 1% ARR should be called a "50% reduction in risk."

FDA has openly stated ARR and RRR must be presented together on labeling. My point in bringing up ARR had to do with the JELIS study. A 0.7% ARR is imo too small to be reliable. When we are talking about 3.5% of one group of 9,000+ patients having a MCE vs 2.8% of another group of 9,000+ having a MCE, which was only significantly reduced in "unstable angina," it raises doubts about the veracity of the claimed "19% reduction in MACE risk."

If you disagree, you disagree, that's fine. The above is just my opinion, and I believe is pretty close to FDA's as well.

- it looks like you do not have a clue about R-IT (mean TG below 200 and high LDL.C acc. to you)

When management says they "estimate" the mean TG will be "above 200," but not "above 220," or "250," etc., I assume they mean pretty close to 200. That means many patients must have enrolled with TG counts from 150 - 199. And then they also say just one other risk factor is needed. That can be elevated LDL-C. This is me just quoting them.

Whal gave greater clarity to the above via his own experience (which I personally believe). But those changes to hunting for only those >200 TG and stenting or other higher risk factor probably occurred after most patients were already enrolled. And I believe over 5,000. That makes a) trying to model placebo group performance futile, and b) a very mixed group of patients with lower and higher relative risk.

That probably doesn't matter much though as in the JELIS study these were split as 1.1%/2424 vs 1.0%/2468 in control vs EPA group with 2 risk factors that had a MCE; 2.0%/2324 vs 1.6%/2239 with 3 risk factors, and 2.5%/1205 vs 2.35%/1228 with 4 risk factors. Trying to pick out meaningful differences between groups here is not possible, imo. Sure, 2.0% vs 1.6% is a 20% RRR, but only a 0.4% ARR, and if you think that RRR is going to reproduce when the total % evented is 20%, well… good luck. Because it really will be lucky, imo.

- all patients in R-IT wre advised perfectly ... yes, if the health systems are the same in the US and (e.g.) in Ukraine ...

Nothing is "perfectly." My point is they are all advised, as higher risk patients, to consume more fatty fish. And that has lots of EPA and DHA in it. And we tend to absorb 2x the EPA and 9x the DHA in fish vs fish oil (even pharma grade). I think that may confound a difference between groups, as some other study authors have suggested is possible, also noting the possible absence of a dose-dependent response to n-3 supp (beyond serum). Or, more simply, they fall by the law of diminished returns, just like most nutrient supplements. Go ahead and take 10,000 mcg of B-12. Your serum may be elevated from active dosing but your cells won't actually utilize hardly any of it. If you check your serum MMA levels after high B-12 dosing you'll find that to be true. Elevated serum levels of EPA (esp things like >300 ug/mL as in MARINE study) probably do not correlate with significantly higher utilization than, say, 70ug/mL. You just don't need that much of the stuff. You will excrete it. Unused. Even if that means it passes through your liver first (and look at the elevated liver enzymes in the JELIS study--clue).

- refer to MARINE / ANCHOR / JELIS data full with mistakes ... e.g. meanwhile M EPA baseline level was higher (app. 2 times than A) it was still app. 50% of JELIS

The actual numbers are:

As reported by Itakura, mean plasma EPA in the Epadel group (n=8321) increased from 97µg/mL (baseline) to 170 µg/mL. (Itakura 2010). Similar effects were observed in patients treated with 4 g/day Vascepa in the MARINE study, in which mean plasma EPA increased from 61 µg/mL to 327 µg/mL (n=69), and in the ANCHOR study, in which mean plasma EPA increased from 28 µg/mL to 183 µg/mL (n=71).

Interesting they say "similar." Relative to dose, perhaps. My point was the baseline in the MARINE study was a lot higher than in the ANCHOR (which was still a lot higher than avg US levels), and over 60% of the baseline levels of the JELIS study. And it's most likely because higher risk patients are advised to consume a diet rish in n-3 from fatty fish. The baseline EPA serum levels of MARINE study patients is more than double that of ANCHOR. And both are far above the claimed:

"Plasma EPA was 2.9 mol% at baseline in our study population, which is similar to reports by Iso and co-workers that serum EPA composition was 4.1 mol% in rural Japanese and 2.4 mol% in urban Japanese; these values are much higher than those recorded in the USA, which are about 0.3 mol%."

If you compare JELIS with MARINE baseline (which may be closer to R-IT baseline) it's 2.9 mol% vs 1.8 mol%. MARINE baseline was 6x higher than the "rest of the US," apparently.

Personally I don't think much benefit (MACE risk reduction) exists beyond consuming 3 average-sized potions of wild salmon or equivalent fatty fish per week, as data back up. Especially not now with all of the statin+aspirin use out there. And I think all of the fish consumption in the R-IT study will confound whatever effect there may be from high dose EPA supp on top of it.

- meanwhile you could find one "relevant" study only, you do not think that try to figure out (not modelling!) R-IT event rate is waste of time ...

I consider all the meta-analyses that show more than 3 fish meals a week procure no additional benefit, and the meta analysis that led to a practice changer wrt n-3 supplementation and MCE risk also "relevant." But yeah, the 300n placebo controlled, randomized, double blind trial testing 4g n-3/d vs corn oil, with corn oil numerically beating n-3 group, is definitely part of the AMRN bear thesis. Do you think lots of linoleic acid every day should beat Lovaza? Must be all that horrible DHA, lowering blood pressure and heart rate, and increasing LDL-P size from taking Lovaza… Or maybe it's just because more isn't better. A few fish meals a week will give you all the benefit you will get. And very high doses might even cause some inflammatory side-effects.

And I do think modeling is in this case futile. But that's jmo, and I've been wrong before.

- Because you think the final p-value will be worse than p=0.17, you don't think the RRR will be better than 5% ... what? ... you have a strong opinion about p, but not about event rates ... Furthermore p determines RRR acc. to you ...

Sure, it's possible to predict no significant separation between groups because the "active drug" is little better than placebo, without hammering down a precise placebo group event rate.

- You could not find anything about CHERRY trial other than AMRN's PR, so it was sponsored by Amarin ...

I said it wasn't sponsored by AMRN, according to AMRN. I also believe that. And I would think the CHERRY study was more interesting if we had any color on baseline characteristics per group, especially % with ACS, and of those the % that also had diabetes. Because they will respond to pitavastatin much better wrt plaque regression. And especially if there was any difference in MACE between groups. But there wasn't.

All the above is Jmo.

GL