Tese de Doutoramento Medicina 2024 Faculdade de Ciências Médicas, Universidade NOVA de Lisboa

Axial motor features, including gait impairment, freezing of gait (FOG), and postural instability, are common symptoms in Parkinson’s disease (PD) patients. They are found in the early stages of the disease, and their frequency and severity increase with disease progression, leading to loss of mobility, independence, an increased risk of falls, and a decrease in quality of life. Although FOG is typically responsive to levodopa (LD) therapy in the early stages, axial symptoms are generally less sensitive to LD compared to appendicular symptoms such as rigidity, bradykinesia, and tremor. Deep Brain Stimulation of the Subthalamic Nucleus (STN-DBS) is a well-established treatment for PD patients with motor complications (MC). While STN-DBS has shown longterm benefits for appendicular signs, motor fluctuations, dyskinesias, and quality of life, its efficacy in addressing axial signs, including FOG, is more controversial. Some studies have indicated that stimulation can improve gait and FOG, especially in the short-term follow-up, but these benefits may diminish in the long term. Other studies have shown that axial symptoms either remained unchanged or worsened after STN-DBS surgery. Additionally, the pre-surgery clinical and demographic factors predictive of gait and FOG outcomes after STN-DBS are not well established. Managing patients who present FOG in the postoperative phases of DBS remains challenging. Alternative stimulation paradigms such as low-frequency stimulation (LFS) have been attempted to improve post-surgery FOG, but consistent results have not been found across different studies. In this thesis, we aimed to establish the frequency and identify risk factors for the development of post-surgery FOG and gait impairment. Data from 109 PD STN-DBS patients with an 8-year follow-up were retrospectively collected. Survival at 8 years and the frequency of disability milestones were assessed. We found that falls (73% of patients) and freezing (47% of patients) were the earliest and most frequently observed milestones, occurring at 40.4 ± 25.4 and 39.6 ± 28.4 months, respectively. Dementia, hallucinations, and institutionalization emerged later and were less frequent, with a temporal relationship between the presence of disability milestones and death being ascertained. Higher Activities of Daily Living (ADL) scores in the OFF state and older age at surgery were associated with a higher risk of falls and freezing. Eighteen STN-DBS PD patients were evaluated at baseline (pre-surgery) and followed up at 18 months post-surgery. Gait impairment and FOG were assessed under different conditions. Stimulation showed a non-significant improvement in the severity of gait impairment in medication-off conditions but a non-significant worsening in the medication-on condition. Stimulation reduced the severity of medication-OFF FOG, but it had no impact on severity. An increase in the frequency of both gait impairment and FOG in the best functional condition was also observed. FOG severity at 18 months post-surgery was better correlated with the pre-operative response to LD of specific kinematic metrics than with the MDS-UDRS-III response. Regarding the effects of stimulation and LD on gait biomechanics, post-surgical assessment revealed that stimulation was associated with a reduction in speed, step and stride length, and an increase in gait variability. Given the limitations of clinical assessment in PD patients, we studied the motor response to LD using inertial-based 3D kinematic analysis. Seventeen PD patients at the stage of motor complications undergoing pre-surgery evaluation were studied. IMUassisted 3D kinematics detected motor alterations as early as 20 minutes after levodopa administration, earlier than the gold-standard clinical evaluation using the MDS-UPDRS part III, suggesting that kinematic analysis using wearable devices can detect motor alterations related to treatment more comprehensively than the MDS-UPDRS III. After establishing the frequency and risk factors for FOG, we selected PD STN-DBS patients presenting FOG in the best functional (score item 3.11 ≥2 on MedicationON/Stimulation ON condition) to assess the response of FOG LD and different stimulation frequencies (130 Hz and 60 Hz). 17 patients were included in this crosssectional study, and evaluated in five different therapeutic conditions: MedicationOFF/StimulationOFF; MedicationOFF/StimulationON; MedicationON/StimulationOFF and MedicationON/StimulationON. Besides the clinical and kinematic evaluation of gait a model for automatic FOG detection was applied to these patients. At a cohort level, compared to MedOFF/StimOFF, the number of FOG episodes was significantly reduced in the MedONStimON 130Hz condition. A high variability in individual responses to LD or stimulation was observed. While approximately 29% of patients worsened their FOG with LD and were rescued by STN-DBS, about 18% presented the reverse pattern. No significant differences were observed in the number of FOG episodes between different stimulation frequencies, but gait variability emerged as the strongest kinematic dimension associated with FOG. Automatic FOG detection showed a good correlation with clinical FOG metrics across all conditions tested. The present work provides evidence that FOG is a common symptom after STNDBS, with its prevalence increasing over time. Patients with higher disease severity and older age at the time of surgery appear to have a higher risk of developing negative axial outcomes. With the role of a levodopa challenge test (LCT) being recently questioned as a predictor of axial outcomes, we show that the response to LD of specific kinematic gait metrics may be better correlated with FOG outcomes than the overall motor response. We also demonstrated that stimulation and LD may impact gait biomechanics differently. In patients with post-surgery FOG in the best-functional condition, we have shown that FOG is mostly therapy-resistant but partially improved by stimulation and medication. Nonetheless, this response is inferior to the pre-surgery response to LD, suggesting that disease progression may play a role in the emergence of this post-surgery FOG. The clinical and kinematic heterogeneity in FOG responses to LD and stimulation (including frequency) should be clinically considered, and an extensive evaluation with LD and stimulation testing should be offered to patients with post-surgery FOG. 3D-kinematic gait analysis can be a powerful instrument to identify FOG episodes, study gait phenotypes, and clarify the circuit mechanisms of FOG