Hi BDAZ, SanFran and others.
I probably shouldn't post this perspective particularly after BDAZ but I did it just to give myself some perspective of the required number of dosed in test and control groups to reach 250 MACE events.
So I made the following assumptions.
- Patients - diabetes melletis, with CVD and low HDL.
- Dosing - each patient has a 50/50 chance of being either in the test or control group.
- Blind experiment - double blind - neither the patient nor the doctors nor administrators know whether the drug being given is test or control.
- Test = Crestor + rvx-208 and lipitor + rvx-208 (note my data come from Crestor + rvx-208 only) - i.e. ASSURE/SUSTAIN trial.
- Control group receives Crestor only.
- I used the % of MACE events at 210 days in the test and control groups from the post hoc analysis.
So based on the post hoc analysis at day 210 the test group(Crestor plus rvx-208) had 5.1% MACE events and the control group had 20.8% MACE events. Therefore if the post hoc results are replicated in BETonMACE then
- if 1000 patients were dosed in the test group we could expect 51 MACE events in the test group.
- and if 1000 patients were dosed in the control group at day 210 (7 months) we could expect 208 MACE events in this group.
This would yield 250 MACE events at 7 months from the beginning of dosing.
Of course, as we all know, these results, if replicated, would yield 77% relative risk reduction and 16% points absolute risk reduction.
- The targets for BETonMACE are much lower and therefore a smaller sample would suffice if it meets statistical significance. This in turn could require less time in the trial.
So depending on the speed that RVX is able to start dosing and generating 1000 patients in the test and 1000 in the control there could be very strong results at 7 months - adjusted of course for the staggered dosing starts.
Therefore, it is somewhat unfortunate that Resverlogix said that they will not provide start of dosing progress throughout the trial.
Anyway, I hope I did not mis-interprete the BETonMACE trial design assumptions.
I am sure that the statin BPs must have huge data bases of MACE events by many sub-segments.
FWIW (for what it's worth)
Cheers
Toinv
