RE:
"I am interested in what you say about the secondary endpoints. Is this the one
https://clinicaltrials.gov/ct2/show/study/NCT00721045? I thought it was more of a fact-finding mission which came up with some quite remarkable results, what did they miss?"
That is the CHF phase 2 trial. It was an exploratory trial, but the pre-specified endpoints were stated in slides 7 and 8 of the presentation given at the American Heart Association conference in Nov 2011 (see
http://www.asx.com.au/asx/statistics/displayAnnouncement.do?display=pdf&idsId=01240869).
To go through the endpoints...
1. NYHA Class:
No mention of this in the analysis, I am sure there would have been a mention if a stat sig improvement in NYHA Class was seen.
2. ECG:
LVEF - closest improvement to stat sig was .184 (for pooled MPC).
LVESV - closest improvement to stat sig was .15 (for pooled MPC). 150M MPC at 6 months had stat sig improvement at .015 (but was worse at 3 months). 25M MPC has almost stat sig improvement of .07 after 3 months.
3. 6 minute walk delta:
Pooled MPC - at 12 months the closest improvement to stat sig was .227
150M MPC came close to stat sig improvement over 12 months at .062
4. BNP
Not even mentioned
5. Immunology
No measures, just this text in the wrap-up "Treatment of patients in this study with allogeneic MPCs was not associated with a clinically significant immune response"
So all in all, one dose for one time period showed stat sig improvement in a pre-defined end-point (150M MPC improved LVESV at 6 months).
A couple of final points I want to make:
1. MACE is a much more important end-point than any of those that were shown in the presentation to be end-points. The result for MACE was truly impressive for 150M dose (although quite concerning that the 75M dose fared worse than the control).
2. The number of patients (about 15 in each dose cohort) means that statistical significance was always going to be very difficult to achieve - most of the reported measures improved, with a couple of exceptions over specific time periods. With the larger phase 3 trial these "trends" should become statistically significant.
(20min delay)