ILoF: a blood‐based optical fingerprinting tool for Alzheimer's disease screening. (31st December 2021)

• Record Type: Journal Article
• Authors: Mumtaz, Mehak; Faria, Simão; Santos, Paulo; Carpinteiro, Cristiana; Pinto, Vanessa; Rodrigues, Sandra; Marques, Filipe; Rocha, Sara; Sampaio, Paula; Paiva, Joana S.
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-12-31
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.054501
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: The high cost and invasiveness of state‐of‐the‐art ATN diagnostic methods underscore the need for alternative approaches for Alzheimer's Disease (AD) patient screening. Blood‐based biomarkers have opened unrivalled opportunities to make AD diagnostics cost‐ and time‐efficient. However, single biomarkers are failing to provide accurate diagnosis/prognosis information for patients suffering from this heterogenous disease, hence the need to combine multiple biomarkers to generate disease‐specific signatures. Here, we present a blood‐based optical fingerprinting tool as a unique 'holistic' cost and time efficient solution for frontline AD diagnostics. Method: Plasma samples from 96 subjects with AD (n=33), mild cognitive impairment due to AD (MCI, n=29), and nondemented controls (n=33) were analyzed with the iLoF technology to generate an optical fingerprint of each sample. Briefly, samples were exposed to a fixed wavelength light‐beam guided through a single mode optical fibre and their back‐scattered signal was acquired and processed to calculate signal‐associated features. The combination of optical features was computed by supervised machine learning algorithms, and four‐fold validation was done to fit the best model. Spike‐and‐recovery experiments for amyloid‐β42, total/phospho‐tau were performed to derive their specific optical fingerprints in blood samples. Result: The iLoF method classified AD, MCI, and nondemented subjects with an AUC of 0.83. The iLoF measurements correlated with amyloid‐β42 ( r =0.783, p >0.0001), total tau ( r =0.815, p >0.0001), and phospho‐tau ( r =0.782, p >0.0001) concentration in CSF, MMSE score ( r =0.759, p >0.0001), and amyloid‐PET scan results ( p >0.0001). Additionally, spike‐recovery experiments showed that iLoF technology can identify and quantify relevant AD biomarkers in blood at pathophysiological concentrations. A coefficient of determination between predicted and real concentrations of r 2 =0.968, r 2 =0.996 and r 2 =0.984 was obtained for amyloid‐β42, total tau and phosphor‐tau, respectively. These biomarker‐directed optical fingerprints can be used to further improve the classification model. Conclusion: In conclusion, blood optical fingerprinting provides a rapid, low‐cost, and convenient tool to enable agnostic screening of AD patients. This disruptive and agnostic technology is expected to become a front‐line screening tool to empower AD diagnosis and may deliver simultaneous screening for multiple diseases underlying dementia. Further work will focus on expanding testing to larger cohorts and addition of other common forms of dementia.