Diffusion tensor imaging analysis along the perivascular space suggests impaired glymphatic clearance in Lewy body dementia subtypes

BackgroundParkinsons disease dementia (PDD) and dementia with Lewy bodies (DLB) are subtypes of Lewy body dementia (LBD) and overlap in symptoms and pathology. Glymphatic function is implicated in LBD pathophysiology due to reduced clearance of abnormal proteins. We aimed to investigate differences in diffusion tensor imaging along the perivascular space (DTI-ALPS), a candidate in vivo measure of glymphatic function, between LBD sub-types. We compared DTI-ALPS between DLB, PDD, Parkinsons with normal cognition (PD-NC), and control participants. MethodsWe recruited participants with DLB, PDD, PD-NC, and controls from neurology clinics and patient support groups. Participants were aged 50-80, clinically diagnosed. Exclusions were confounding neurological or psychiatric conditions or metal precluding MRI scanning. Participants underwent MRI brain, plasma sampling and clinical and cognitive assessments. DTI-ALPS was calculated for each participant. As DTI-ALPS can be influenced by white matter distribution, we also calculated a metric called "complexity". We tested group differences in DTI-ALPS, and associations with clinical variables across all patient groups including cognition, motor scores, sleep and fluctuations. ResultsFifty-one DLB (43 M, 72.7{+/-}5.5 years), 35 PDD (25 M, 69.5{+/-}7.8 years), 60 PD-NC (27 M, 63.1{+/-}7.33 years), and 26 controls (13 M, 66.7{+/-}9.28 years), were included for analysis. DTI-ALPS significantly differed between groups (F(3, 165) = 22.68, p<.0001), controlling for age and sex. DTI-ALPS did not differ between PD-NC and controls, whilst cognitively impaired groups (PDD, DLB) had reduced DTI-ALPS relative to PD-NC and to controls(pFDR<.05). DTI-ALPS was further reduced in DLB relative to PDD. These findings held when corrected for complexity, which accounts for differences in white matter distribution (F(3, 161) = 28.11, p<.0001). Finally, DTI-ALPS correlated with cognition (MOCA: {beta}=2.95, p<.0001; composite cognitive score: {beta}=1.81, p=.046), REM sleep behaviour disorder ({beta}=-4.36, p=.019), and cognitive fluctuations ({beta}=14.89, p=.014). ConclusionsWe showed that DTI-ALPS, is reduced in LBD compared to PD-NC and controls; and is further reduced in DLB compared to PDD. DTI-ALPS is an easily-extracted metric from widely-used MRI scans which has potential as a useful imaging marker in LBD for clinical trials and in the clinical setting, to identify LBD patients with a more aggressive disease course.