What evidence would prove the amyloid hypothesis? Towards rational drug treatments for Alzheimer's disease

Abstract

Department of Pharmacology and Biochemistry, University of South Florida College of Medicine, Tampa FL33612-4799, USAAs one of us (DM) listened intently to the debatespresented at the meeting on Challenging Views ofAlzheimer’s Disease, I was struck by the strong desireby some in attendance to find fault with the amyloidhypothesis of Alzheimer’s disease (AD). Argumentswere presented in multiple discussions leading to theconclusion that the relationship between amyloid andAD was so tenuous that we should not even waste timedeveloping drugs directed at reducing amyloid.What was most surprising was the level of proof towhich the amyloid hypothesis was being held by thoseattempting to refute it. I was left with the impressionthat if every symptomin every patient regardless of pa-tient age, gender, state of disease progression or con-current disorders, did not correlate linearly with theamyloid burden, then the assumption was the amyloidhypothesis must be false. For example, some patientsdie with brains loaded with enough amyloid to reachthe criterion for AD, yet they are cognitively fit as fid-dles. Alternatively,patientsdie withfull blowndemen-tia, diagnosed as AD, yet they have little or no amyloidpathology.The animal models are argued to be failures becausenot all aspects of the AD phenotypeare manifest. Theyhavebrainsfilledwithamyloid,yetthereislittleneuronloss or tangle formation. Thus, the amyloid hypothe-sis must be wrong or the mice should recapitulate the