How can we really improve screening methods for AD prevention trials?

This edition of Alzheimer’s & Dementia: Translational Research & Clinical Interventions includes an important article from Lutz et al. on the potential advantages of a novel genetics-based risk algorithm (GRBA) for selection of participants in Alzheimer’s disease (AD) prevention trials [1]. The GRBA, described previously [2], has been used to identify subjects for the well-known TOMMORROW trial of pioglitazone for prevention of symptomatic AD dementia or MCI due to AD. Efficient screening methods are of paramount importance for AD prevention trials. Studies such as TOMMORROW require years of follow-up, and they cost scores of millions of dollars. These costs accrue in part from the screening methods themselves, but they relate much more strongly to the trials’ size and the duration of labor-intensive followup needed to identify sufficient “events” for adequate statistical power. Certainly, the GBRA described here must be among the lower cost methods of screening. Relying as it does only on age and genetic markers available from peripheral blood, it should have clear cost advantages when compared against more elaborate or invasive screening techniques such as amyloid-PET scanning or CSF biomarker assays. Not surprisingly, therefore, Lutz et al. compare their GBRA with some of the latter, more ambitious methods. But the more salient portion of their argument relates to the comparison of the GBRAwith similarly low cost and minimally invasive methods relying on age and peripheral blood markers. Given that the GRBA is relatively inexpensive per candidate screened, how does it perform otherwise? Two factors determine its efficiency. First is the number of persons who must be screened to provide the needed number of “events”. Second is the efficiency of the method in rejecting persons who will not “convert” from cognitively normal to