Amarin Corp Plc (AMRN)

[HDGabor]

K-

Granowitz in the CC has already acknowledged that the placebo rate will be approximately 5% ( Co had previously stated that it could be as low as 5.2% )
Gabors response ( from memory ) was. ..." approx 5% could be anywhere between 4% and 6% ."...but I think Granowitz is skilled enough in modeling to be pretty accurate given the data they have access to .

Full with mistakes, way off ... e.g. CG did NOT say anything about what WILL be the rate .. he said "was assumed in the design of the study to incur major adverse cardiovascular events at a rate approximately 5% per year"

Please understand some basic things:
a.) study design isn't equal with actual study / data
b.) study design is fixed by SPA in 2011 ... the p rate had not been changed ...
c.) The study design is about number of enrollees ... assuming RRR (R), Power (P), p rate (pr), etc (E) ... you will get the required number of patients (NoP) ... it was 6,990. "To protect against the possibility that the actual placebo event rate is lower than estimated" they added additional 1,000 to the design ("without having to modify the other sample size assumptions") ... resulted in "event rate in the placebo group could be 5.2% per year". eg.

R x P x x pr x E = Nop

If you increase the NoP, do not change R, P and E ... the pr will be lower, but both cases, pre and post change, is a fix / calculated number.

It is not debatable that the designed p rate is 5.2% (somewhere between 5.150% or 5.249% ... due to rounding) ... the assumption, the design was fixed 6.5 years ago.

Hopefully it helps you to understand what is the design, why the p rate is 5.2% and why CG's words refer to this ...

Re. Jardiance and other "improvements"
- you constantly forget that 40% only are in the US ...
- patients were randomized 1:1 between arms, so any effect affects both arms ... could change the ARR, but "not" the RRR (e.g. 6% vs 3% and 4% vs 2% have the same RRR ... 50%)
- despite your reference to "worldwide" Kaiser follow-up (SMS, etc.) the statistics do not reflect the "expected" improvements

Best,
G