Cognitive profile as a predictor of the long-term outcome after deep brain stimulation in Parkinson's disease

Abstract

Background: Although dementia is a contraindication for deep brain stimulation (DBS) in patients with Parkinson's disease (PD), the concept is supported by little scientific evidence. Moreover, it is unclear whether PD with mild cognitive impairment (PD-MCI) or domain-specific cognitive impairments affect the outcome of DBS in non-demented PD patients. Objective: To investigate the influence of baseline cognitive profiles of PD on the outcome of DBS. Methods: Baseline cognitive levels of patients with PD who underwent DBS were classified into PD with dementia (PDD) (n = 15), PD-MCI (n = 210), and normal cognition (PD-NC) (n = 79). The impact of the cognitive level on key DBS outcome measures [mortality, nursing home admission, progression to Hoehn &Yahr (HY) stage 5 and progression to PDD] were analyzed using Cox regression models. We also investigated whether impairment of a specific cognitive domain could predict these outcomes in non-demented patients. Results: Patients with PDD showed a substantially higher risk of nursing home admission and progression to HY stage 5 compared with patients with PD-MCI [hazard ratio (HR) 4.20, P = .002; HR = 5.29, P < .001] and PDNC (HR 7.50, P < .001; HR = 7.93, P < .001). MCI did not alter the prognosis in patients without dementia, but those with visuospatial impairment showed poorer outcomes for nursing home admission (P = .015), progression to HY stage 5 (P = .027) and PDD (P = .006). Conclusions: Cognitive profiles may stratify the pre-operative risk and predict long-term outcomes of DBS in PD.