After full data release, experts say failed Alzheimer’s trial EXPEDITION 3 offers hopeful signals

This new phase III trial of solanezumab reveals that the drug is not effective for mild Alzheimer’s disease patients, despite the hint that it was possibly effective based on post-hoc analyses of earlier studies with this drug.

The findings expose the hazards of such post-hoc analyses, typically done when the desired results are not observed, in the hope of squeezing lemonade from lemons. Although the subanalysis of mild AD patients in the earlier studies suggested a 34% slowing of cognitive decline as assessed by ADAS-Cog, an incremental slowing of 11% was seen in the new study that was not even statistically significant. While some secondary endpoints reached statistical significance, the slowing was so modest as to make no practical difference clinically.

I cannot emphasize enough that such equivocal results as seen in EXPEDITION 3 do absolutely nothing to either confirm or deny the amyloid hypothesis. By now, there have been so many of these studies with antiamyloid agents, with little or no hint of efficacy, that we’ve long passed the definition of insanity: Doing the same thing over and over in the hope of getting a different result.

The combination of all these clinical trial failures with the result of imaging studies that have shown amyloid deposition some 20 years before the expected onset of symptoms clearly tells us that antiamyloid agents should only be considered as potential prophylactics. By the time symptoms appear, disease progression is largely independent of amyloid and may be primarily tau-driven, spreading from neuron to neuron even when amyloid is effectively targeted by therapeutics. Even the A4 and DIAN studies are likely initiating treatment too late to make anything more than a modest effect with little practical value clinically. I am not suggesting that we drop amyloid as a target, only that we stop making these incremental changes in clinical trial design in the hope of getting a different result.

Michael S. Wolfe, PhD, is the Mathias P. Mertes Professor of Medicinal Chemistry at the University of Kansas, Lawrence. He has no financial disclosures.