AbstractBackgroundBrain lesions causing amnesia are in a memory circuit centered on the subiculum. Deep brain stimulation (DBS) sites connected to this circuit are linked to cognitive decline in Parkinson’s disease but cognitive improvement in Alzheimer’s Disease (AD). Insight into this paradox may come from directly comparing the topography of lesion and DBS‐induced effects.MethodsWe studied 53 amnesia‐causing lesion locations and 46 patients (92 DBS sites) from Phase 1 and Phase 2 randomized controlled trials of fornix DBS for AD. Connectivity between lesions and DBS sites was computed using a normative atlas of functional connectivity data from 1000 healthy subjects. Baseline cognitive scores were compared to 1‐year post‐DBS activation scores.ResultsConnectivity between AD patients' DBS sites and a subiculum region derived from amnesia‐causing brain lesions correlated with cognitive improvement (Spearman’s rho = 0.33, p = 0.027). There was a difference between sham‐first (Spearman’s rho = 0.33, p = 0.027) and stimulation‐first (Spearman’s rho = 0.25, p = 0.27) cohorts. Whole‐brain connectivity maps of the subiculum region and AD response fingerprints were aligned: 17/19 local maxima appeared in both maps, dice coefficient 0.71 (p<0.00001). Both maps showed data‐driven peaks in an overlapping subiculum region. However, DBS data showed a significant interaction effect between patient age and subiculum connectivity upon outcome in a general linear model (p = 0.031). Randomization status was controlled as a covariate and was not significant. Specifically, higher subiculum connectivity in older patients led to improved cognition, while higher connectivity in younger patients resulted in worsening cognition. The data‐driven inflection point for this interaction effect was at age 65 at any subiculum connectivity value.ConclusionOur findings suggest that lesions causing amnesia and DBS sites modulating cognition converge on common neuroanatomy. However, patient age appears to be critical variable in determining whether DBS sites connected to the subiculum result in cognitive improvement or decline. Our results provide insight into the paradox of mixed outcomes across DBS cohorts and support restricting ADvance II enrollment of fornix‐targeted DBS to patients over 65. The underlying mechanisms remain unclear, and this study must be replicated in additional datasets.
Read full abstract