Diagnostic and prognostic roles of morphological indices in the characterisation of NPH cohorts

Introduction: External lumbar CSF drainage (ELD) models shunt responsiveness in patients with NPH. However, brain/ventricular changes predictive of responsiveness remain unclear. We used traditional linear and 3-directional measures to characterize morphological differences between cohorts of NPH responders vs. non-responders, Alzheimer’s Disease (AD) and healthy controls (HC). Methods: 21 participants with NPH underwent ELD and pre- and post-intervention imaging, according to a published NPH protocol. T1-weighted MRI scans from 21 age-matched AD and HC cohorts were acquired from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). We used the NIH platform, 3DSlicer, to derive traditional linear indices, [Evans Index (EI), Bicaudate Index (BCI) and Callosal Angle (CA)], and 3-directional measures [z-Evans Index and Brain per Ventricle Ratio (BVR)]. Results: Mean age for NPH study participants was 71.1 ± 6.3 years (18 males, 3 females). There was good intra-rater agreement for all measures (ICC > 0.9). All indices distinguished NPH from AD and HC cohorts (p < 0.001). Within the NPH cohort, nine patients responded to ELD and twelve were non-responders. There were no significant differences in pre-ELD measurements between these groups. Post-drainage, non-responders had a significant decrease in z-Evans Index (p = 0.001) and an increase in BVR at PC (p = 0.024). The increase in BVR was a result of a decrease in the ventricle component of the ratio (p = 0.005). Conclusions: Morphological indices play a role in characterizing NPH vs. non-NPH. Degree of change in ventriculomegaly is not predictive of responsiveness. 3-directional linear measures are superior to traditional indices in differentiating between disease cohorts.