Using cerebrospinal fluid ATN quantification to classify research subjects in a diverse clinical population

Background The ATN Framework was proposed to better classify subjects on the Alzheimer’s Disease (AD) spectrum for research purposes. The role of the MassGeneral Institute for Neurodegenerative Disease (MIND) Tissue bank is to collect cerebrospinal fluid (CSF), plasma, and associated clinical data from a diverse neurological subject population. Ab40, Ab42, p‐tau 181, and t‐tau were quantified in CSF samples from the MIND Bank and classified according to the ATN framework to assess the utility of this research framework in a clinical trial relevant, non‐cohort setting. Method 750 CSF samples were collected from clinically indicated lumbar punctures at the Massachusetts General Hospital Outpatient and Inpatient Neurology Services. The CSF samples were analyzed using Eurolmmun ELISA assays on a semi‐automated TECAN FreedomEVO liquid handler to obtain ATN values. To optimize sensitivity and specificity we plotted these value across varying thresholds, and selected the value at which both were maximized. Assessment of cognitive impairment and neurological diagnosis was performed by chart review by an experienced neurologist. Result After sliding threshold analysis of clear cognitively unimpaired versus AD‐Dementia cases, a threshold value for Ab42/40 Ratio (0.082, 92% sensitivity, 91% specificity), pTau (41.8 pg/mL, 91% sensitivity, 91% specificity), and tTau (280 pg/mL, 88% sensitivity, 88% specificity) was determined. 25% of the MIND bank was classified as likely AD from chart review and a similar proportion was biofluid A+T+N+. The overlap of these two subgroups was extensive but not complete. Out of the 188 A+T+N+ subjects, 152 presented clinically with a form of AD, 6 had other neurodegenerative diseases, 23 had other forms of cognitive impairment, and 7 were cognitively unimpaired. 8 of the 419 A‐T‐N‐ subjects had a clinical diagnosis of AD or mixed AD. Conclusion The ATN framework can be used to provide an extra level of certainty with regards to diagnosis of likely AD in a clinical trial relevant subject population. Even in this population with clinical indication for LP, there is a small number of resilient cognitively unimpaired individuals that have AD positive biomarkers but no cognitive impairment.