Inflammatory Biomarkers Aid in Diagnosis of Dementia

Area of Science:

• Neurology and Neuroscience
• Biomarker Discovery
• Computational Biology

Background:

• Mixed dementia (MX), combining Alzheimer's disease (AD) and vascular cognitive impairment and dementia (VCID), presents diagnostic challenges due to the lack of established biological criteria for vascular disease.
• Current diagnostic approaches for MX are limited, particularly in differentiating it from single-pathology conditions during a patient's lifetime.
• Autopsy studies frequently reveal co-occurring AD and VCID pathologies, highlighting the clinical relevance of MX.

Purpose of the Study:

• To establish biological criteria for diagnosing mixed dementia (MX) by integrating AD biological markers with MRI-based white matter injury assessments.
• To explore the underlying pathophysiology of MX, AD, and VCID by measuring a panel of CSF and blood biomarkers.
• To identify key biomarkers predictive of MX using machine learning for improved diagnostic accuracy.

Main Methods:

• Mixed dementia (MX) was diagnosed using established biological criteria for Alzheimer's disease (AD) combined with MRI evidence of white matter injury.
• Multiplex biomarker kits were employed to measure proteases, angiogenic factors, and cytokines in cerebrospinal fluid (CSF) and blood samples.
• The Random Forest machine learning algorithm was utilized to identify the most important biomarkers for diagnosing MX.

Main Results:

• Machine learning analysis identified three key proteases (MMP-10, MMP-3, MMP-1), three angiogenic factors (VEGF-C, Tie-2, PLGF), and three cytokines (IL-2, IL-6, IL-13) as highly important.
• These selected biomarkers demonstrated significant correlations with neuropsychological testing results, confirming their clinical relevance.
• The study successfully defined a set of biomarkers potentially useful for diagnosing mixed dementia.

Conclusions:

• The integration of AD biological criteria and MRI-based white matter injury assessment provides a framework for diagnosing mixed dementia (MX).
• Specific proteases, angiogenic factors, and cytokines identified through machine learning hold potential as diagnostic biomarkers for MX.
• These findings advance the understanding of MX pathophysiology and offer tools for improved clinical diagnosis and management.