Predicting Hippocampal Atrophy More Accurate With Cognitive Status Than CSF Aß

Sara Becker, PhD

Researchers from a study have uncovered data that suggests that cognitive status predicts hippocampal volumes with greater certainty than cerebrospinal fluid (CSF) amyloid-ß 1-42 (Aß42) levels. The Cornu Ammonis 1 (CA1) region and hippocampal-amygdaloid transition area (HATA) were specifically associated with various cognitive domains.

Led by Sara Becker, PhD, department of neurodegenerative diseases, Hertie Institute for Clinical Brain Research, analysis from linear regression models revealed that cognitive group—either cognitively normal (PD-CN) or with mild cognitive impairment (PD-MCI)—is a statistically significant predictor for both the HATA (ß = –.23; 95% CI, –.44 to –.02) and the CA1 region (ß = –.28; 95% CI, –.56 to –.02). Patients with PD-MCI showed significantly smaller volumes than those with PD-CN independent of disease duration and intracranial volume (ICV).

"Our results suggest that the HATA has the potential to differentiate cognitive status in PD, and that CSF total tau levels are associated with hippocampal atrophy. These pilot results should be confirmed in future prospective studies, preferably with a longitudinal design,” Becker and colleagues noted.

Becker and colleagues aimed to examine whether hippocampal volume loss is primarily associated with cognitive status or pathological Aß42 levels by including 45 non-demented PD patients who underwent CSF analyses, magnetic resonance imaging (MRI), and comprehensive motor and neuropsychological examinations.

READ MORE: Early Treatment in Parkinson Disease May Have Small Benefit

The hippocampal subfield volumes were compared between patients with PD-MCI and PD-CN and between patients with low and high Aß42 levels. Investigators also explored the relationship between hippocampal subfield volumes, CSF biomarkers, and activities of daily living (ADL).

For all other hippocampal subfields not including CA1 or HATA, there were no significant effects of either cognitive or Aß42 status found. Additionally, no association to hippocampal subfields reaching significance were found when Aß42 burden was used as a continuous variable.

Correlation analyses showed significant correlations between hippocampal subfields and Cambridge Neurological Test Automated Battery (CANTAB) subtests assessing memory, most notably the Paired Associated Learning (PAL) test, as well as the Stockings of Cambridge (SOC) subtest which assessed executive functions and information sampling independent of ICV.

There were partial correlations which revealed significant associations between worse performance on Repeated Battery for the Assessment of Neuropsychological Status (RBANS) memory subtests and smaller volumes of the hippocampal tail, subiculum, CA1, parasubiculum, and whole hippocampus. The analyses also showed significant associations between lowered RBANS language tests scores and smaller hippocampal tail, subiculum, presubiculum, parasubiculum and whole hippocampus volumes.

"Our results also indicate that computerized tests are more sensitive than paper-and-pencil tests in relation to hippocampal volume, as CANTAB subtests are potentially less reliant on motor functions than the RBANS. Both computerized and paper-and-pencil tests may also reflect AD pathology differently, although more research is necessary to examine sensitivity of both types of tests for detecting cognitive impairment and hippocampal atrophy in PD,” Becker et al noted.

Patients who performed better on the Functional Activities Questionnaire (FAQ), which indicated more severe ADL impairment, were associated with larger volumes of hippocampal status in both the CA1 (r = .31; 95% CI, 0.02–0.55; P = .04) and CA2-3 regions (r = .38; 95% CI, 0.10–0.61; P = .01).

Indications of more severe impairment in motor-driven ADL function, demonstrated by higher motor scores of the FAQ, were correlated with larger volumes of the hippocampal tail (r = .30; 95% CI, 0.007–0.55; P = .04), CA1 (r = .33; 95% CI, 0.04–0.57; P = .03) CA2-3 (r = .38; 95% CI, 0.10–0.61; P = .01), and whole hippocampus (r = .33; 95% CI, 0.04–0.57; P = .03).

REFERENCE

Becker S, Granert O, Timmers M, et al. Association of hippocampal subfields, CSF biomarkers and cognition in non-demented Parkinson’s disease patients. Neurology. Published online November 20, 2020. doi: 10.1212/WNL.0000000000011224