*3.3. Recommendations*

We recommend using LCT as a key tool to obtain relevant presurgical information on the patient's status and the possibility of improvement after DBS. However, the rule of UPDRS part-III improvement >30% should not be strictly applied. Although patients a ffected by disabling dopa-resistant tremor could represent an exception to this rule, improvement >50% can be associated with greater overall benefit in most patients. The LCT response of disabling axial symptoms, such as FoG, is important and should be weighted independently from the percentage UPDRS part-III total score improvement.

### **4. The Relevance of Axial Symptoms: How Sensitive Is Current Clinical Assessment?**

### *4.1. The Standard Rule*

The term 'axial symptoms' is commonly referred to as a group of PD motor features encompassing gait impairment, postural instability, postural abnormalities, and speech disorders, especially dysarthria and stuttering. These are a major source of disability because they are associated with reduced mobility, communication di fficulties, recurrent falls, and subsequent injuries [34]. Moreover, they are markers of advanced disease and are often resistant to dopaminergic therapies or exhibit an heterogenous pattern of response to levodopa [35]. There are no precise indications on how to consider these symptoms and their pre-surgical response to levodopa in the clinical practice.

### *4.2. Pros and Cons*

The evidence on the e ffect of DBS on axial symptoms is controversial. A meta-analysis published in 2004 showed that one year after surgery, STN-DBS or globus pallidus pars interna (GPi) DBS can improve gait and balance symptoms, with an e ffect size similar to the preoperative e ffects of dopaminergic medication [36]. However, the improvement provided by DBS seems not sustained over the years. Evidence for axial symptom progression, despite a good control of PD appendicular motor symptoms, has been shown in open-label, long-term follow-up studies, although some extent of axial improvement related to stimulation is reported in the first years after surgery [37–40].

The relevance of axial symptoms as a marker of disease progression was disclosed in a cohort of 143 PD patients treated with STN-DBS [5], in whom axial disability during the follow-up period was strongly associated with an increased risk of death (hazard ratio of 4.3), proving to be the most accurate mortality predictor, even superior to the cognitive status.

Axial symptoms track disease progression and disability, therefore an accurate presurgical evaluation of levodopa-responsiveness would be necessary for estimating the extent of response after DBS. Indeed, worsening or amelioration after DBS of speech, posture and gait disorders is multifactorial and depends upon clinical variables such as disease duration [16,41], the type of axial symptom (gait often improves after DBS, speech may worsen as a stimulus-related side e ffect), their interplay with dopaminergic medications [42,43], the brain target employed for DBS, the frequency [44,45] and distribution of stimulation [46], and placement of active electrode contact [41].

Levodopa-resistant axial symptoms are considered a relative contraindication for surgery [47]; however, the current pre-surgical clinical examination is unable to detect early axial signs which may foster worse DBS outcomes. In particular, FoG evaluation poses grea<sup>t</sup> challenges, given the episodic nature of this phenomenon and the complex relationship with dopaminergic medications. Trunk postural abnormalities also represent a source of di fficulty at the time of DBS selection (including on which target to choose), because camptocormia and Pisa syndrome might be responsive to STN-DBS, even with poor or no amelioration after LCT [48,49].

These di fficulties which impact on the selection process might be overcome by integrating objective evaluation involving kinematic analysis and wearable sensors to the pre-operative clinical examination. Specifically, novel technological developments on wearable sensors for home-monitoring have the potential to provide a measure of axial symptoms and their relation to levodopa intake in a naturalistic way [50,51]. These technologies should be employed to detect early axial signs which might predict worsening after DBS and assist the identification of the best candidates. Indeed, a study employing kinematic assessment of gait demonstrated a correlation between presurgical levodopa response of stride length and range of motion and FoG outcome after DBS [52].
