*4.3. Recommendations*

The global burden of axial symptoms can be considered a proxy for disease stage because of their correlation with disability and death. A fine-grained assessment of each axial symptom, the accurate evaluation of their relationship with dopaminergic therapy, and the integration of technology outcome measures into the clinical practice should favor a better understanding of candidates to DBS, with the potential to predict their disease course and the probability to improve after DBS. Pending clinical trials aiming at the evaluation of the e ffect of DBS on PD-related axial symptoms, we recommend to accurately evaluate in clinical practice the presence, severity, and impact on patient's daily life and independence of axial symptoms before surgery, and discuss the weight of each symptom with the patient, clarifying its poor, good, or indeterminate probability of improvement after DBS. Severe FoG and speech issues, in particular, represent a potential challenge in the managemen<sup>t</sup> of patients undergoing DBS, while camptocormia and Pisa syndrome could have good chances of improvement and should not be considered contraindications for DBS.

### **5. The Need for Patient-Centered Outcome Measures**

### *5.1. The Standard Rule*

From a regulatory point of view, the FDA and European Medicine Agency (EMA) request the presence of motor fluctuations as a mandatory criterion for DBS indication in PD [10]. Reduction in severity and frequency of motor fluctuations represents one of the most relevant achievements obtained by DBS, which translates into the improvement of QoL revealed by randomized controlled trials [8,53].

### *5.2. Pros and Cons*

In CAPSIT-PD, it is recommended that the patients perform the self-reporting diary one week per month during the three preoperative months, indicating the presence of four conditions: complete OFF, partial OFF, complete ON, and ON with dyskinesias [1]. However, these measures are highly subjective, and wrong or missed entries may occur in about one third of cases—also when using electronic motor diaries [54]. That is, objective home-based quantification by wearable sensors of PD motor symptoms [55], including FoG [56], should be explored carefully in patients considered for DBS, also with respect to the predictive value of these measures. Indeed, when it comes to predicting the outcome of DBS in PD and patient-centered outcome measures are employed, some discrepancies arise. Patient-centered outcome measures are represented by QoL, evaluated by the validated Parkinson's Disease Questionnaire 39 (PDQ-39) [57] or by its short form (PDQ-8) [58], and ADL functioning or

independence, typically measured by UPDRS part-II and the Schwab and England (S&E) scale [59]. The importance of measuring patient-centered outcomes relates to the discrepancy between the judgment made in-clinic by the neurologist and the degree of satisfaction [60] and independence obtained by the patient during daily life [54]. Two key factors may account for this discrepancy: (1) motor symptoms observed by clinicians explain only a small part of the complex picture of PD, which encompasses several non-motor symptoms; (2) the standardized tasks assessed during in-clinic visits and the non-quantitative, non-continuous, non-ecologic in-clinic examinations may not represent a comprehensive measure of the patient situation and condition during daily life. This last aspect is true even when limiting the evaluation to motor symptoms, in particular episodic motor symptoms such as FoG [61], which are not adequately captured during in-clinic standard assessments [54].

When it comes to analyzing determinants of improvement in QoL after STN DBS, a post-hoc analysis of the EARLY-STIM trial found smaller QoL improvement at 24-months follow-up in patients with better pre-surgical PDQ-39 scores [6]. Interestingly, patients with pre-surgical PDQ-39 scores ≤ 15 had no significant change in QoL following surgery. This finding is not meant to be caused by a ceiling effect of STN-DBS to improve motor symptoms in the EARLY-STIM cohort, because the change in QoL over the two years was independent of the severity of parkinsonian motor signs assessed by UPDRS-III [8]. Accordingly, in another study analyzing a cohort of 85 PD patients treated with DBS, the magnitude of motor symptom improvement with a pre-surgical LCT was only borderline associated with improvement of QoL after DBS (*p* = 0.053) [62].

A systematic review [63] demonstrated that higher baseline QoL predicted larger QoL changes after surgery in three out of four studies. The analysis of the 18 studies included in this review yielded mixed results with respect to the predictive value of other clinical and demographical features. There are two main explanations for such discrepant findings: (1) most of these studies were not primarily designed to detect predictors of QoL change after DBS and results are influenced by the main a priori hypothesis tested in each study; (2) factors contributing to QoL in PD include not only motor disability but also non-motor symptoms [64], and the interplay between these domains might be individualized and have a different weight in each subject. Moreover, among motor symptoms, axial disability (and its response to therapies) might have a high impact, which has not been ye<sup>t</sup> explored carefully in regard to QoL or ADL outcome after DBS.
