Amarin Corp Plc (AMRN)

[HDGabor]

K-

1.) First of all the numbers re. Secondary Prevention Analysis

- Patients with established CAD were observed with a mean follow-up of 4.6 years
- Laboratory data (p / EPA)
> Total cholesterol, mean (SD), mg/dl: 270 / 270
> LDL-cholesterol, mean (SD), mg/dl: 178 / 178
> HDL-cholesterol, mean (SD), mg/dl: 55 / 55
> Triglyceride, median (IQR), mg/dl: 163 / 160
- Data from all 3,664 patients in the secondary prevention strata were used for the analysis.

EPA vs control (RRR)
- Total: 8.7% vs. 10.7% (adjusted hazard ratio = 0.77, 95% confidence interval (CI) 0.63–0.96, P=0.017). RRR: 23%
- 1,050 patients with prior MI: 15.0% vs. 20.1% (adjusted hazard ratio = 0.73, 95%CI 0.54–0.98, P=0.033). RRR: 27%
- 895 patients with prior coronary intervention (PTCA or CABG): 14.7% vs. 22.0% (adjusted HR 0.65, P=0.007), RRR: 35%
- 537 patients with prior MI and coronary intervention: 15.2% vs. 24.7% (adjusted HR 0.59, P=0.008). RRR = 41%
- 2,903 patients with stable AP: 7.8% vs. 9.4% (adjusted HR 0.76, P=0.036). RRR = 24%
- 484 patients with stable AP and a coronary intervention: 16.8% vs. 24.6% (adjusted HR 0.66, P=0.044). RRR = 34%

Please note:
- all p-value was less than 0.05
- sum of the 5 subgroup is more than a total, since some of them have overlap (eg.: "prior MI" and "prior MI and coronary intervention")

EPA vs control (yearly event rate with a mean follow-up of 4.6 years)
- Total: 1.89% vs. 2.33%
- 1,050 patients with prior MI: 3.26% vs. 4.37%
- 895 patients with prior coronary intervention (PTCA or CABG): 3.20% vs. 4.78%
- 537 patients with prior MI and coronary intervention: 3.30% vs. 5.37%
- 2,903 patients with stable AP: 1.7% vs. 2.04%
- 484 patients with stable AP and a coronary intervention: 3.65% vs. 5.35%

2.) R-IT placebo rate

The current design allows the 5.2% as minimum, but it does not mean it is the expected. Don't forget: originally they designed as 6,990 patients, which allows the 5.9% as minimum -> they expected rate should be 5.9%+. We will see the reality, but I will use 5.2% and 5.9%.

-----
Hereafter, I will deal with "patients with prior MI" and "patients with prior coronary intervention" only. (Please note: adjusted hazard ratio is higher than the RRR below, due to KM method.)

a) Without any modification: R-IT placebo vs JELIS placebo rate
MI:
> 5.2% vs. 4.37% = RRR 15.97% (37.29% vs JELIS active arm)
> 5.9% vs. 4.37% = RRR 25.94% (44.73% vs JELIS active arm)
PCI:
> 5.2% vs. 4.78% = RRR 8.03% (38.55% vs JELIS active arm)
> 5.9% vs. 4.78% = RRR 18.94% (45.84% vs JELIS active arm)

b) With modification ["Dropping LDL levels from 178 to under 90 will definitely lower the event rates" & "high intensity Statins (Crestor and Lipitor ) reduce risk more then moderate intensity Statins"], which effects the JELIS placebo rate only, since R-IT placebo assumption included these. Modification: Lower (JELIS) rate by 20% - R-IT placebo vs JELIS placebo rate
MI:
> 5.2% vs. 3.50% = RRR 32.78% (49.83% vs JELIS active arm)
> 5.9% vs. 3.50% = RRR 40.75% (55.78% vs JELIS active arm)
PCI:
> 5.2% vs. 3.83% = RRR 26.42% (50.84% vs JELIS active arm)
> 5.9% vs. 3.83% = RRR 35.15% (56.67% vs JELIS active arm)

Of course it was a simple math, but suggest a stop with high probability.

The difference between you and me:
- you are "dealing" with RRR (calculate R-IT RRR from JELIS RRR)
- I am using event rates comparison (R-IT placebo vs JELIS placebo / active)

Please note: high intensity statins / lower LDL affect both JELIS arm equally. The ARR will be lower, but RRR will remain the same.

Best,
G