Circulating microglia-derived extracellular vesicles predict recovery after rehabilitation in stroke survivors

BackgroundTimely intensive rehabilitation is crucial to contrast the negative escalation of events that follow a stroke injury, to promote tissue regeneration, and to restore the physiological function. This study aimed to identify measurable blood biomarkers that could be predictive of functional recovery in stroke survivors. Methods30 sub-acute stroke subjects were enrolled at Fondazione Don Carlo Gnocchi Onlus (Italy) during the observation prospective cohort study EXO4STROKE (NCT05370105). Among the clinical scales used for patient profiling, modified Barthel Index (MBI) was considered as primary outcome for the evaluation of rehabilitation recovery. Cytokines, Neurofilaments and circulating Extracellular Vesicles (EVs) were evaluated in the serum of patients at admission (T0) and at discharge (T1) in intensive rehabilitation unit. The Surface Plasmon Resonance imaging (SPRi) technique was exploited to obtain a multiplexed analysis of circulating EVs from brain (neurons, astrocytes, microglia) and non-brain (endothelium, skeletal muscle and platelets) cells, and for the relative quantification of markers of pathological or regenerative processes. Machine learning-based analysis complemented the study by integrating statistically significant features in a cross-validated prediction model targeting functional recovery at T1 from T0 data. ResultsHigh serum levels of IL-6 and Neurofilament Light Chain at T0 were associated with higher residual disability (lower outcome) at T1. In contrast, elevated circulating levels of EVs from neurons (CD171+) and microglia (CD11b+) were associated with better recovery after rehabilitaion. Moreover, the overexpression of the VEGF receptor on microglial EVs (IB4+ and CD11b+), was associated with higher functional improvement. Alternatively, the overexpression of TGF-{beta} receptor on IB4+ EVs during the subacute phase post-stroke was found associated with increased stroke severity. The MBI at T1 was predicted with a median [iqr] absolute error of 5.7[8.7] points. ConclusionThis study identifies a panel of few circulating biomarkers associated to Evs able to objectively assess post-stroke status and predict functional recovery after rehabilitation.