The brain changes of Alzheimer disease (AD) start many years before the onset of cognitive symptoms. Most models of AD propose a stepwise progression of brain pathology starting with amyloid plaque deposition, then tau tangle formation, then neurodegeneration. Cerebrospinal fluid (CSF) and imaging biomarkers have been developed that allow detection of AD brain pathology in living individuals, but these modalities have significant drawbacks that limit their widespread use. Over the last three years, there has been rapid development of blood-based biomarkers that accurately detect AD brain pathology. In this study, we propose to study some of the most promising blood-based biomarkers for three types of AD brain pathology: amyloid (A?42/A?40), tau (phosphorylated tau [pTau] isoforms), and neurodegeneration (neurofilament light chain protein [NfL]). The research team has developed immunoprecipitation-mass spectrometry assays for plasma A?42/A?40 and pTau isoforms that will be further optimized and automated as part of the proposed project. The Knight Alzheimer Disease Research Center cohort will be studied, and has available data on plasma and CSF NfL, clinical dementia diagnosis, performance on cognitive tests, health history, amyloid PET, tau PET, structural brain volumes by MRI, genetic markers, numerous CSF biomarker measures, discovery proteomics data, and autopsy reports. Approximately 1,700 matched pairs of banked plasma and CSF samples from ~1,000 individuals will be examined, which is similar in size to recent major studies of cognitive outcomes as a function of biomarker combinations. The correlation of the blood-based measures with better established CSF and imaging measures will be evaluated. Biomarkers of amyloid, tau and neurodegeneration will be used independently and in combination within a modality (blood-based, CSF or imaging) to predict the risk for current or future symptomatic AD. For all analyses, the effects of individual characteristics (including age, sex, years of education, APOE ?4 genotype, polygenic risk score, race, and medical comorbidities) will be evaluated to identify factors that modify the expression of symptoms associated with biomarker levels. We hypothesize that the combination of plasma A?42/A?40, pTau isoforms and NfL will perform better than amyloid PET in predicting risk for current or future symptomatic AD. Because blood tests are well-accepted by patients, physicians, and researchers, an accurate blood test for symptomatic AD would likely be widely used and could be a game-changer in improving AD research, accelerating clinical trials and enabling more accurate diagnoses in the clinic.
We will examine whether a combination of blood-based measures that reflect three different aspects of Alzheimer disease (AD) pathology better predict current or future AD symptoms than amyloid PET, a gold standard for AD brain pathology. If a blood test can provide better diagnostic information than more logistically complex tests, this could accelerate efforts to find drugs that prevent AD symptoms.