KIWI...
Just to refresher message. Re: REDUCE-IT
The idea is to extrapolate the data from JELIS to try to determine the likelihood of REDUCE-IT showing statistically significant MACE (major cardiac events) reduction and reduction in death from coronary artery disease and all cause mortality.
First the numbers. JELIS was a Japanese prospective, randomised open-label, blinded endpoint evaluation involving 18,645 patients. Two thirds of the patients were women. All patients had elevated LDL-Cs and were on low dose statins. 3664 patients had a history of CVD and they form a subset which is called the secondary prevention group. The study was divided equally into a statin only group and a group taking statins and 1.8gms EPA/day.
JELIS ran about 4.5 years. The total number of MACE events during this period was 587. 326 events in the statin only group and 261 in the statin plus EPA. These numbers calculate out to a .8% annual event risk in the statin group and a .622% annual rate in the EPA added group. The estimated risk in the REDUCE-IT trial is a 5.2% annual risk in the placebo group.
The first thing that jumps out is the MACE risk in JELIS is more than 6 times lower than the estimated risk in REDUCE-IT. Death from CVD in Japan in 1999 when JELIS completed, was 22-26/100,000 person years. Also Japanese women have a CVD events incidence that is 2-3 times lower than men. THe point here is the information we are looking for is related to the number of events. Even though the JElIS cohort is more than twice the size of REDUCE-IT, if the rate of events is 6 times higher in REDUCE-IT as far as statistics is concerned REDUCE-IT will show meaningful results in less than half the time it takes JELIS.
JELIS showed a reduction of 20% in MACE events across the board. In the secondary indication group this was statistically significant. In this subset the annual risk was 2.3% for the statin only group and 1.9% for the statin plus EPA. The difference being that 20% in the higher risk group was more meaningful. Even though there were 14,981 patients in the primary group and only 3664 patients in the secondary group, there more CVD events in the secondary group, 335, than in the primary where there were only 231 MACEs.
JELIS has been criticized for a number of reasons. First of all it is not a double blinded study. IMHO that would be more troubling in the USA. JELIS was blinded as a probe study so some controls are in place.
JELIS has also been criticized for the fact the statin doses were low by USA standards. The argument goes that the risk rate could have been lowered simply by increasing the statin levels to reach more optimal levels of LDL-C. This aside it is important to understand EPA had little effect on the LDL-C levels, and its reduction in MACE was not correlated with LDL-C levels. The 20% reduction was independent of LDL-C and suggests a different mechanism of action (anti inflammatory).
IMO the most valuable portion of the JELIS data concerns the secondary intervention group (SIG) because they more nearly resemble the REDUCE-IT cohort. The annual event rate in the statin only group, the group which resembles the placebo group in REDUCE-IT was 2.3% which is about 40% the estimated risk orf the placebo group of REDUCE-IT. The SIG is only 3664 patients which is less than half the size of REDUCE-IT. SIG scored stat significant 20% MACE reduction in 4.5 years. That was a total of only 355 total events (197 control, 158 EPA). At 7000 patient REDUCE-IT level we are looking at twice as many patients and more than twice the risk. As far as predicted events you take the product of these two factors which gives you at least four times the event rate which means your will see this many events (355) per year once you hit 7000 patients. Doubling the number of SIG patents should result in a statistically significant reduction in all MACE including mortality studies even at only 20% MACE reduction...
The odds are we may even see a higher reduction in MACE reduction in the REDUCE-IT EPA group, because their EPA levels will be considerably lower than the Japanese. Studies indicate absolute EPA levels are inversely correlated with CVD risk.
The comparison of patients enrolled in REDUCE-IT compared to the patients enrolled in JELIS strongly suggests REDUCE-IT will show statistically significant MACE reduction and most likely lower mortality rates in a period of two years or less beginning in mid 2013...to mid 2015.
":>) JL
