Performance of Plasma Biomarkers Combined with Structural MRI to Identify Candidate Participants for Alzheimer's Disease-Modifying Therapy

M Manjavong,J M Kang,A Diaz,M T Ashford,J Eichenbaum,A Aaronson,M J Miller,S Mackin,R Tank,M Weiner,R Nosheny
DOI: https://doi.org/10.14283/jpad.2024.110
Abstract:Background: Recently, two monoclonal antibodies that lower amyloid plaques have shown promising results for the treatment of Mild Cognitive Impairment (MCI) and mild dementia due to Alzheimer's disease (AD). These treatments require the identification of cognitively impaired older adults with biomarker evidence of AD pathology using CSF biomarkers or amyloid-PET. Previous studies showed plasma biomarkers (plasma Aβ42/Aβ40 and p-tau181) and hippocampal volume from structural MRI correlated with brain amyloid pathology. We hypothesized plasma biomarkers with hippocampal volume would identify patients who are suitable candidates for disease-modifying therapy. Objectives: To evaluate the performance of plasma AD biomarkers and hippocampal atrophy to detect MCI or AD with amyloid pathology confirmed by amyloid-PET or CSF biomarkers in ADNI. Design: A cross-sectional and longitudinal study. Setting and participants: Data were from the Alzheimer's Disease Neuroimaging Initiative. Participants were aged 55-90 years old with plasma biomarker and structural MRI brain data. Measurements: The optimum cut-off point for plasma Aβ42/Aβ40, p-tau181, and NFL and the performance of combined biomarkers and hippocampal atrophy for detecting cognitive impairment with brain amyloid pathology were evaluated. The association between baseline plasma biomarkers and clinical progression, defined by CDR-Sum of Boxes (CDR-SB) and diagnostic conversion over two years, was evaluated using a Weibull time-to-event analysis. Results: A total of 428 participants were included; 167 had normal cognition, 245 had MCI, and 16 had mild AD. Among MCI and AD, 140 participants had elevated amyloid levels by PET or CSF. Plasma Aβ42/Aβ40 provided the best accuracy (sensitivity 79%, specificity 66%, AUC 0.73, 95% CI 0.68-0.77) to detect drug candidate participants at baseline. Combined plasma Aβ42/40, p-tau181, and hippocampal atrophy increased the specificity for diagnosis (96%), but had lower sensitivity (34%), and AUC (0.65). Hippocampal atrophy combined with the abnormal plasma p-tau181 or hippocampal atrophy alone showed high sensitivity to detect clinical progression (by CDR-SB worsening) of the drug-candidate participants within the next 2 years (sensitivity 93% and 89%, respectively). Conclusion: Plasma biomarkers and structural MRI can help identify patients who are currently eligible for anti-amyloid treatment and are likely to progress clinically, in cases where amyloid-PET or CSF biomarkers are not available.
What problem does this paper attempt to address?