Well the ALXN trial results are out, and I listened to some of the CC.
These are early results and they are fairly complex. So far they have only analyzed the MB-CK and death numbers, not the other endpoints.
First the bad news. They missed some endpoints and only the bolus plus infusion arm showed any efficacy at all. So if this was a Phase III they would be in some trouble. However this was a Phase II, and the good news is that they showed some fairly clear efficacy where it counts.
What's happened is that the surgery apparently released a bunch of CK-MB in everyone (surprise!). Not surprisingly, the drug didn't do anything to prevent this. They showed a plot of _peak_ CK-MB readings, drug vs. placebo. There's a sharp peak at low levels (from the surgery) where there is no differentiation at all between placebo and drug. As you go out further (higher peak readings), however, the drug and placebo clearly diverge. At say 100 ng/ml (20x normal) you get a significant divergence between drug and placebo. (The divergence is also there significantly at somewhat lower levels). This is a level that is apparently historically associated with a 4-5x increase in mortality.
From the point of view of their predefined endpoints on CK-MB, it looks like these were set at too low a level, so no significance in some of their pre-defined endpoint because everybody's level was raised some by the surgery. Essentially they were trying to be too cautious (It's easier to show a moderate effect than a big one, right?) and shot themselves in their foot.
Ironically the shorts will now change their tune. They had been grumbling about how nobody knows how significant moderately raised levels of CK-MB are, and now the company will start agreeing with them. The shorts will doubtless (legitimately) complain also about missed endpoints and sub-group analysis.
Bottom line though is that this seems to be a pretty good (if somewhat confusing) Phase II result. Clearly the Phase III will proceed only on the bolus+infusion arm, and also only look at severely elevated CK-MB levels plus death.
Note also that we are still to find out about many secondary endpoints here, like cognitive function. The drug was safe (setting to rest some worries raised by the original small Phase II) and also clearly effectively blocked terminal complement activation. There is still some chance in my view that we will see a statistically significant reduction in death in this trial if we wait awhile for delayed complications to set in amongst the placebo patients.
Altogether a somewhat messy result, although you wouldn't know it yet from the cheering. I sold part of my holdings and hope to buy back at a somewhat lower price.
Peter |
