Can MRI screen for CSF biomarkers in neurodegenerative disease?

Corey T McMillan1, Brian Avants, David J Irwin

  • 1Department of Neurology, University of Pennsylvania, Philadelphia, USA. mcmillac@mail.med.upenn.edu

Neurology
|December 28, 2012
PubMed
Abstract

Insights

Magnetic Resonance Imaging (MRI) can noninvasively screen for Alzheimer disease (AD) and frontotemporal lobar degeneration (FTLD) pathology. This MRI approach accurately predicts cerebrospinal fluid biomarkers, aiding differential diagnosis.

Area of Science:

  • Neuroimaging
  • Neuropathology
  • Biomarker Discovery

Background:

  • Alzheimer disease (AD) and frontotemporal lobar degeneration (FTLD) present similar clinical symptoms, complicating accurate in vivo diagnosis due to distinct underlying pathologies.
  • Current diagnostic methods often rely on invasive procedures or complex biomarker analysis, highlighting the need for accessible screening tools.

Purpose of the Study:

  • To evaluate the efficacy of Magnetic Resonance Imaging (MRI) as a noninvasive screening tool for differentiating between AD and FTLD.
  • To assess MRI's ability to predict cerebrospinal fluid (CSF) biomarker ratios indicative of AD or FTLD pathology.

Main Methods:

  • Volumetric MRI and lumbar puncture were performed on 185 patients with clinically diagnosed AD or FTLD.
  • Singular value decomposition and linear regression were used to predict the CSF total tau (tt) and β-amyloid (Aβ(1-42)) ratio (tt/Aβ) from MRI data.
  • Accuracy was validated using neuroanatomic visualization and a subset of genetically or autopsy-confirmed cases.

Main Results:

  • MRI-predicted tt/Aβ ratios showed a strong correlation with actual CSF tt/Aβ levels.
  • Distinct neuroanatomic correlates were identified: ventromedial prefrontal regions for low tt/Aβ (FTLD) and posterior cortical regions for high tt/Aβ (AD).
  • The MRI-based prediction achieved 75% accuracy in identifying the underlying diagnosis.

Conclusions:

  • MRI can serve as a noninvasive screening procedure for AD and FTLD.
  • MRI-based predictions of CSF biomarkers offer a surrogate for direct CSF analysis, potentially simplifying differential diagnosis.

Related Concept Videos