The headlines for the Phase III topline results for Lilly‘s solanezumab were pretty much the same as for bapineuzemab: Primary endpoints were missed, and by that definition, the Phase III program was a failure. But there was a very ambiguous subplot which leaves the sola saga looking somewhat different from bapi, whose development as an IV drug was summarily terminated by JNJ/Pfizer. Lilly reported that a secondary analysis of pooled data for mild-moderate patienrs showed cognitive benefit, and a secondary analysis of mild patients showed that the reported benefit was confined to that group. However, a primary endpoint of cognition in mild AD patients had been implemented in the second Phase III trial, and the results there were negative.
Lilly will not unveil details until October, which leaves some important questions yet-to-be-answered. For example, what were the actual p values for these analyses? If the pooled data for mild patients from both trials was significant at, say, p=.049, but the primary endpoint for cognition in the mild AD patients in the second trial ‘failed’ at p=.051, the inconsistency in ‘significance’ might just reflect the effect of having a smaller sample size to work with. This could be another instance where the binary concept of ‘statistical significance’ obscures rather than illuminates. Just as importantly, what was the magnitude of the reported cognitive benefit? From the little information provided thus far, it does not sound like it was accompanied by any measurable functional benefit, and perhaps these cognitive fluctuations are just statistical noise. But it does leave open the door for the hypothesis that in treating Alzheimer’s, earlier is better, and that the differences between sola and bapi (dosing, epitope binding site) that had generally been interpreted as being in bapi’s favor, perhaps worked for solanezumab.
One thing we do know for sure: A couple of Wall Street analysts who speculated that Lilly could potentially file a NDA on the basis of these data must be asked–‘What Have You Been Smoking (WHYBS)?’ Had Lilly hit a primary endpoint in one but not both of the trials, perhaps one could assemble a scenario for filing, but not here.
Had solanezumab struck out as definitively as bapineuzemab did, even in the earliest-stage AD patient cohort, this would have been a heavy blow to the amyloid hypothesis. For now, that judgment hangs in abeyance, at least until a fuller disclosure of the results, and almost certainly beyond.
The upcoming September issue of NeuroPerspective will feature our annual review of Alzheimer’s, where we examine all of the recent developments in the field in context.