Genetic Profile and Symptom Pattern Explain Variability of Deep Brain Stimulation Effect in Dystonia

Objective: Bilateral globus pallidus pars interna deep brain stimulation (GPi-DBS) is a recognized and effective treatment option for drug-resistant dystonia patients. However, the clinical GPi-DBS outcomes vary significantly. Herein, we explored the pre-implant factors affecting GPi-DBS effectiveness. Methods: Genetic profiles, symptom distribution, age at onset, disease duration, and severity of a cohort of 31 GPi-DBS dystonia patients were collected. Dystonia motor severity was evaluated before and after GPi-DBS using the Burke–Fahn–Marsden Dystonia-Rating-Scale (BFMDRS-M). We assessed the interplay of the aforementioned factors in determining the BFMDRS-M improvement through a multilinear regression analysis. Results: BFMDRS-M score showed a significant improvement (47.8%) since the first year, remaining stable at 5 years (54.3%). Lower limb (0.20), upper limb (0.16) and trunk (0.24) symptoms showed a significantly larger improvement compared to cranial symptoms (0.07, p < 0.05). Consequently, patients with more pronounced lower limb motor symptoms displayed a greater GPi-DBS effect (p < 0.01). However, pre-treatment localization of motor symptoms accounted only for 31% of the Inter-Patient Variability (IPV) in post-GPi-DBS improvement. Amelioration varied also across genetic profiles, with the largest improvement reported for DYT-TOR1A patients (n = 9, 64.2% in the first year), predicting 36% of IPV. Interestingly, combining motor and genetic profiles predicted 73% of the IPV. Including the clinical profile of the patient (age at onset, disease severity and duration) increased prediction accuracy to 81%. Interpretation: Our results suggest that motor and genetic profiles contribute independently to the efficacy of the GPi-DBS treatment. These results may support a personalized prediction of DBS outcomes in dystonia patients.