Evaluation of cerebrospinal fluid levels of VAMP-2 and SNAP-25 in a dementia with Lewy bodies clinical cohort stratified by Alzheimer's pathophysiological biomarkers.

Background: Synaptic protein levels in cerebrospinal fluid (CSF) may represent much-needed objective biomarkers of cognitive impairment, disease progression and drug efficacy in patients with dementia with Lewy bodies (DLB). Soluble N-ethylmaleimide-sensitive factor attachment proteins receptors (SNARE) proteins, such as VAMP-2 and SNAP-25, are implicated in α-synuclein pathophysiology and CSF levels of these proteins are associated with pathophysiological biomarkers and cognitive decline in Alzheimer's disease (AD). The aim of the study was to compare CSF levels of VAMP-2 and SNAP-25 in patients with DLB to cognitively unimpaired controls and AD patients and study their association with cognitive performance and AD and neurodegeneration biomarkers.

Methods: VAMP-2 and SNAP-25 were quantified in CSF from cognitively normal controls (n = 62), DLB (n = 44) and AD (n = 114) patients from the Sant Pau Initiative for Neurodegeneration (SPIN) cohort using homebrew Single Molecule Array assays (Simoa). The DLB group was stratified into two groups with ("DLB + AD", n = 28) or without AD co-pathology ("pure DLB", n = 16) using our validated cut-off for the CSF phosphorylated tau (p-tau)/Aβ42 ratio. We used linear regression to test for group differences (adjusting for age) and association with AD biomarkers. We used standardized w-scores of the cognitive tests to analyze the association of the synaptic markers with cognitive performance.

Results: CSF VAMP-2 and SNAP-25 levels correlated across all groups (r = 0.71-0.9, p < 0.001). Both proteins were decreased in pure DLB (p < 0.001, p = 0.01) but increased in DLB + AD (p = 0.01, p = 0.02) compared to controls and showed good accuracy to discriminate pure DLB from DLB + AD (AUC = 0.84, 0.85). Both proteins were associated with CSF p-tau and total tau (t-tau) across all groups (r = 0.49-0.88, p < 0.001), with the Aβ42/40 ratio in DLB + AD (r = 0.29-0.36, p < 0.001) and in AD (r = 0.12-0.23, p < 0.001) and with CSF neurofilament-light chain (NfL) in controls (r²=0.10-0.11, p < 0.001-0.01) and AD patients (r²=0.01-0.08, p = 0.01 - 0.001). SNAP-25 was associated with CSF NfL in the DLB + AD group (r²=0.15, p = 0.02). CSF VAMP-2 and SNAP-25 were associated with phonemic fluency in pure DLB (r = 0.39 - 0.28, p = 0.01-0.03) and SNAP-25 with the Clock drawing test and the MMSE in DLB + AD (adj.r = 0.15 - 0.14, p = 0.03-0.03) and DLB (adj.r = 0.12 - 0.08, p = 0.02-0.04) groups.

Conclusions: CSF VAMP-2 and SNAP-25 are promising surrogate markers of synapse degeneration in DLB. However, care should be taken when interpreting CSF levels of these synaptic markers in DLB in light of the confounding effect of AD pathophysiological markers.