3600 patient years was the estimate needed to acheive 250 MACE events based upon projections at the start/design of BETonMACE. This was based upon projections of 2400 patients in the trial for an average length of 18 months (2400 patients X 1.5 years = 3600 patient years). Yes, there were assumptions and unknowns (i.e. event rate in control patients and effect size of apabetalone). However, for let's just assume for now that 3600 patient years is still the magic number.

If 3600 patient years equals 250 MACE events, then 2700 patient years equals 188 MACE events (75% of 250; trigger of sample size re-estimate analysis).

Based upon patient enrollment figures provided in DSMB reports or news releases, we can arrive at the approximate number of patient years. Between two sequential enrollment updates, we know how many more patients were added. For example, around 12/15/2016 there were 800 patients and on 3/17/2017 there were 1200 patients. In this example, 400 patients were added sometime between 12/15/2016 and 3/17/2017. However, we do not know if most of these 400 were added closer to 12/15/2016 or closer to 3/17/2017. Therefore, an approximation of patient years will be a range considering the two extremes.

Do this exercise for all of the DSMB/new release enrollment updates and I get the following:

By end of Q1 2018, there will have been between 2046 and 2688 patient years.

By end of Q2 2018 there will have been between 2596 and 3088 patient years.

By end of Q3 2018 there will have been between 3146 and 3488 patient years.

By end of Q4 2018 there will have been between 3546 and 3838 patient years.

Note 1: I only account for 2200 patients as stated on 1/11/2018 and do not account for any additional patients that have/may be added after 1/11/2018. Therefore, the above are underestimating the contributions from these additional patients.

Note 2: I assume patients stop once they get to 2 years. 

Note 3: A majority of the uncertainty in the min/max range is due to not knowing how many of the first 600 patients were enrolled closer to the 11/11/2015 first patient announcement and the 9/12/2016 first update. 

Note 4: I fully admit that I may have made calculation errors, so do your own due diligence and make your own conclusions/interpretations from the above.

BearDownAZ