Recalling Biobank Participants to Clinical Study of Alzheimer’s Disease – FinnGen Pilot Study

Abstract

AbstractBackgroundPopulation‐based biobanks that comprehensively ascertain health and genetic data hold promise to facilitate identification of Alzheimer’s disease (AD) trial participants. We conducted a pilot study in individuals with mild impairment in cognition or AD to assess the feasibility of recalling biobank participants. We also tested the utility of computer‐ and telephone‐based cognitive assessments as cost‐effective alternatives to in‐person neuropsychological testing and investigated plasma‐based biomarkers.MethodsWe utilized FinnGen, a large population‐based biobank study of up to 500,000 participants representing almost 10% of the Finnish population. Single‐center cross‐sectional study included individuals with ICD‐10 codes for AD (G30) and mild cognitive disorder (MCD; F06.7). Dementia screening instruments were in‐person Mini‐Mental State Examination (MMSE) and telephone interview for dementia (TELE). Episodic memory was assessed with three trial word list tasks in three formats; in‐person, with web‐based tool and with the modified Telephone Interview for Cognitive Status including three learning trials. Each of these yielded immediate and delayed free recall measures. Blood samples were collected to determine plasma‐based biomarkers: amyloid beta (Aβ)1‐40 to Aβ1‐42 ratio, glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL) and phosphorylated‐tau181 (pTau‐181).ResultsA total of 78% (110/140) of a pre‐specified FinnGen sub‐cohort were successfully reached. Of those, 39% (43/110) were willing to participate and 19% (N = 27, mean age = 72y, 13 women) completed the study. FinnGen and hospital records largely matched. For 8/12, MCD was due to other reasons than early AD. MMSE and TELE correlated .818 (p< .001). Immediate and delayed recall measures correlated across three formats: from .584 to .679 (p’s<.003) and from .250 to .424 (p’s: .031 ‐ .218), respectively. AD group had higher levels of GFAP (p = .002) and pTau‐181 (p = .020). TELE was negatively correlated with GFAP (p = .011), NfL (p = .001) and pTau‐181 (p = .006). Moreover, p‐Tau181 had negative correlations with episodic memory measures (all r’s< ‐.202) with telephone‐administered immediate recall reaching statistical significance (r = ‐.394, p = .042).ConclusionsInformative, customized clinical recall studies from FinnGen are feasible. Remote testing is a cost‐effective approach in recall studies. Next phase of our study aims to invite cognitively unimpaired FinnGen participants for screening of high AD risk individuals.