Tada's original question was: "Does the trial finish with 250 events with less than 2400 patients dosed or will they have to continue through 2400 patients in order to get statistically significant numbers?"

According to the trial design with the original 2400 patients, patients will be treated for a minimum of one year and maximum of two years. First patients in will get 2 years of dosing; last patients in will only be dosed for 1 year. 1:1 ratio of placebo- to RVX-208- treated patients. In theory, the last of the 2400 patients should be enrolled/start dosing sometime H2 2017. So unless BETonMACE reaches 250 MACE events by 2017 and ends the trial, which is very unlikely, then yes 2400 patients will have started dosing prior to reaching 250 MACE events.

Their goal is 3600 patient years (2400 patients with avg treatment for 18 months) to achieve 250 MACE events based based on the event rate of the placebo group and anticipated reduction in event rate in RVX-208 group. Back in October 2015, Mike Sweeney and Kausik Ray broke this down even further and I summarized it in this post:

"BETonMACE is modeled very similar to the EXAMINE trial, in which the 3-point MACE event rate in their patients was right around 10.5% at 18 months. BETonMACE is an event driven trial, with a goal of 250 events and a minimum of 2400 patients. 10.5% X 2400 patients gives you your 250 events at 18 months. Granted, the RVX-208 treatement arms are expected to have lower event rates. But as Michael Sweeney and Kausik Ray pointed out, BETonMACE will additionally have a low HDL (<45 mg/dL female, <40 mg/dl male) requirement that was not there for EXAMINE, so the average event rate for BETonMACE should actually be higher."

So maybe the 3-point MACE event rate in the placebo group will be greater than 10.5%. I think that is what Tada is implying. However, this may or may not be counterbalanced by a reduced 3-point MACE event rate in RVX-208-treated group. Just for fun, let's say the 3-point MACE event rate is 15% for the placebo and 10.5% for the RVX-208 treated (30% RRR). The average event rate for the combined group would be 12.75%, which is about 20% greater than the original 10.5% estimate. So yes, Tada, I can see how there may be a greater event rate than originally planned. I'm still not holding my breath for the 125 event futility analysis to occur in H1 2017. However, when it does occur, this will surely be informative as to refining the bookends for the 250 events.

On a side note, it is really difficult to extrapolate the %RRR in 3-point MACE from the ASSURE/SUSTAIN post-hoc analysis. There were very few 3-point MACE events in those relatively short 6-month trials. Most of the events in the post-hoc analysis of the 5-point MACE were not the strict 3-point MACE events (death, non-fatal MI, non-fatal stroke). From the published data (Atherosclerosis 2016) on the ASSURE/SUSTAIN trials for the combined diabetics and non-diabetics, there was 1 death and 4 MI in the 331 RVX-208 patients and 3 deaths and 0 MI in the 168 placebo patients. No strokes were reported. The other 13 and 14 5-point MACE events in the RVX-208 and placebo groups, respectively, were reported as coronary revascularization or hospitilization for unstable angina or heart failure. If you limit to just the diabetic patients in SUSTAIN/ASSURE, there were even fewer 3-point MACE events to extrapolate from. There were 0 deaths, 1 MI and 0 stroke in the RVX-208 group and 2 deaths, 0 MI and 0 stroke in the placebo group. Long story short, the 55% and 77% RRR in 5-point MACE is largely based on incidence of coronary revascularization or hospitilization for unstable angina or heart failure, which aren't a part of the 3-point MACE metric that BETonMACE demands.

BearDownAZ