Posters

Gait impairment is an important diagnostic criterion for Progressive Supranuclear Palsy (PSP) and may be modulated by specific cognitive domains such as executive function. This study aims to investigate the relationship between executive function and gait in PSP, using verbal fluency tasks as a cognitive measure alongside clinical assessments and objective gait analysis with wearable sensors. We explore whether these relationships are sustained longitudinally. Twenty-six participants with PSP were recruited as part of the Oxford Quantification in Parkinsonism (OxQUIP) study, and 17 of them were followed up longitudinally over five visits at three-month intervals. Clinical symptoms were assessed using the PSPRS, and executive function was tested with phonemic (letter F) and semantic (animal names) verbal fluency tasks. Spatio-temporal gait characteristics and gait variability were measured during a two-minute walk task using body-worn sensors. We found that higher semantic fluency was distinctly associated with lower clinical symptom severity (p=0.008) and lower variability in gait speed (p=0.027) and stride length (p=0.041). The association between semantic fluency and gait speed variability (p=0.042) was confirmed by the longitudinal analysis. Higher semantic fluency is linked with reduced gait variability (i.e., more controlled movements), and this link persists over time, potentially reflecting a mechanism to prevent falls. Our findings highlight the clinical importance of executive function for gait stability, emphasizing its role in PSP assessment and interventions.

Magnetic resonance–guided focused ultrasound (MRgFUS) thalamotomy targeting the ventral intermediate nucleus (VIM) is an established treatment for medication-refractory essential tremor (ET), but its efficacy in dystonic tremor (DT) remains unclear. Evidence suggests that extending lesions toward the zona incerta (ZI) may enhance tremor control. We report a single-centre experience of unilateral VIM/ZI MRgFUS in DT patients and compare outcomes with an age-matched ET cohort. Consecutive patients treated between 2020 and 2024 were analysed from a prospectively maintained database. Tremor severity was assessed using the Clinical Rating Scale for Tremor (CRST) and Bain & Findley spirals at baseline and 12 months. ANCOVA analyses were performed for CRST-A, B, C, and spiral scores, with clinical diagnosis as a fixed factor and age and skull density ratio as covariates. Linear regression assessed the effect of age on tremor improvement. Nine DT and 44 ET patients were included. DT patients were younger (64.3±14.9 vs 73.0±7.7 years; t=2.58, p=0.013). At 12 months, ET patients improved more than DT in overall tremor severity (CRST-A: F1,49=5.91, p=0.019) and functional disability (CRST-C: F1,49=6.29, p=0.016), while improvements in task-specific performance (CRST-B: F1,49=0.13, p=0.716) and spirals (F1,49=0.06, p=0.805) were similar. Age did not interact with diagnosis (β=−0.049, 95% CI −0.342 to 0.243, t=−0.339, p=0.736). Adverse events were mild to moderate and transient. In summary, dual MRgFUS targeting VIM/ZI reduces tremor amplitude in both ET and DT, but overall functional gains are greater in ET. Residual dystonic features in DT likely limit improvements in global severity and disability.

In Parkinson’s disease (PD), α-synuclein-mediated damage to the autonomic nervous system (ANS) means failure to regulate blood pressure (BP) is a common comorbidity that is hard to treat, worsens quality-of-life and often precedes diagnosis. Central ANS dysfunction is still poorly understood in-vivo, but likely is crucial as α-synuclein accumulates in these regions long before overt symptoms. Here, we characterised function of a brainstem-cortical ANS network in relation to cardiovascular autonomic failure. 81 individuals with early-PD and 65 aged-matched healthy controls (HC) performed resting-state 3T fMRI, neurological assessments and a 3-minute standing test: lied down for 5-minutes, had BP measured, then stood-up and had BP measured every minute for 3-minutes. We used Dynamic Causal Modelling to investigate how drops in BP linked to effective connectivity in medullary and relay nuclei, hypothalamus, and cortical autonomic regions. Drops in BP during the standing task, in patients, were associated with self-inhibition of the rostro-ventrolateral medulla (RVLM), a nucleus that regulates sympathetic increase of BP. Conversely, in HC, drops were associated with higher RVLM activity. Additionally, the patient model was dominated by bottom-up connectivity from medullary regions to relay and cortical ones, potentially pertaining to state-signalling to higher-order regions, given the decreased ability of medullary regions to regulate BP in patients. Our findings are compatible with known PD pathophysiology and elucidate how in-vivo a baseline RVLM dysfunction may cause poor central ANS reactivity to orthostatic stimuli and predispose patients to BP drops. Finally, this study paves the way to further characterise central autonomic function in-vivo.

Ultra-low-field (ULF) MRI (0.064T) shows promise as a low-cost tool for research and diagnostics for neurodegenerative disorders in lower-middle-income countries (LMICs). We assessed the feasibility of ULF MRI in patients with Parkinson's disease (PD), Parkinson's disease (PDD), Alzheimer's disease (AD), and healthy older adults (HC) as controls. We scanned a total of 16 participants, including PD, PDD, AD, and HC, using axial T1 and FLAIR MRI sequences. Images were assessed qualitatively by a neuroradiologist for image quality and Fazekas scoring for White Matter Hyperintensities (WMH). The MRI images were super-resolved using the Freesurfer tool "SynthSR" to produce 3D images for the evaluation of medial temporal lobe atrophy (MTA). Images were also preprocessed using Freesurfer for segmentation and calculation of cortical thickness and subcortical gray matter using the recon-all-clinical pipeline and white matter hyperintensities (WMH) using WMH-Synthseg. We obtained a reasonable quality of images from the scanner. We were able to demonstrate disease-specific changes in dementia, such as hippocampal atrophy, widening of sulci, and thinning of gyri in the lobes, as well as the presence of deep WMH. Fazekas scores were calculated using axial FLAIR MRI. Further super-resolution to 1mm MPRAGE images helped calculate MTA scores from T1-weighted images. We calculated cortical thickness and subcortical gray matter, and WMH for all the images using automated software. We found the imaging feasible for this age group, and the participants were able to complete the imaging procedure comfortably. This study is the first to demonstrate the feasibility of ULF MRI in patients with neurodegenerative disorders and healthy older adults in an LMIC.