*2.3. Recommendations*

There is evidence for an earlier use of DBS as a treatment option to improve patients' QoL and early levodopa-responsive axial symptoms, while minimizing the psychiatric consequences of overtreatment. Long-term results from the EARLY-STIM trial would allow the better defining of which PD features are associated to a long-term successful outcome. However, there is not enough knowledge on how early into the disease history DBS should be considered, given the paucity of published data and the current lack of knowledge on how to predict disease progression and DBS response in such early stages. We recommend considering each case singularly, according to the patient's phenotype, age, needs, and expectations in patients whose symptoms significantly impact ADL and QoL despite a reasonable number of attempts to provide the best medical therapy.

### **3. The Evolution of the Levodopa Challenge Test**

### *3.1. The Standard Rule*

The second recommendation of CAPSIT-PD is dopaminergic responsiveness confirmed by a levodopa/apomorphine challenge test (LCT). Accordingly, the test has to demonstrate at least a 33% decrease in the Unified Parkinson's Disease Rating Scale (UPDRS) part III score in the "defined-on condition" (best therapeutic e ffect after medication agreed by patient and physician) compared to the "defined-o ff condition" (at least 12 h after receiving the last medication dose).

### *3.2. Pros and Cons*

The threshold of 33% for UPDRS improvement is considered relevant to rule out possible misdiagnoses (i.e., identifying atypical parkinsonism for which DBS is not recommended). The LCT is also important to inform the possible outcome of surgery, showing the likely extent of symptom improvement after surgery, and to establish realistic expectations from DBS [24]. Indeed, it is generally accepted that symptoms improving with levodopa are likely to respond to DBS [25]. However, there are some exceptions and caveats to these widely accepted concepts. Firstly, levodopa-resistant tremor represents one of the indications of DBS, even in the absence of disabling motor fluctuations [26,27], given its excellent e ffect in controlling or even suppressing tremors, regardless of the deep nuclei targeted [28].

Another relevant challenge related to the LCT is the cut-o ff of 33%. This value has been validated by a study based on its ability to predict chronic levodopa responsiveness, with a positive predictive value for the PD diagnosis of 88.6% [29]. Notably, the Movement Disorders Society (MDS)-sponsored UPDRS scale introduced in 2008 has some di fferences in the scoring of the part-III (motor part), and a study analyzing the MDS-UPDRS scores with the old UPDRS [30] ones after an acute LCT found an excellent correlation between the two scales, with the 30% UPDRS score variation used for predicting sustained

long-term levodopa response equivalent to 24% in MDS-UPDRS [31]. However, data from STN-DBS clinical trials seems to indicate that an excellent response to levodopa (i.e., >50% UPDRS part-III improvement) could be associated with a better DBS motor outcome [32]. A meta-analysis published in 2006 on the STN-DBS outcomes supports this hypothesis, demonstrating that the magnitude of decrease in both UPDRS part-II and part-III scores exhibits a dose–response relationship with the presurgical response to the levodopa challenge test [33]. However, it is unknown whether the magnitude of response at the LCT may predict better ADL and Qol after DBS. Moreover, the relevance of axial symptoms as a source of disability, their heterogenous response to dopaminergic therapies, and their influence on the trajectory of the PD course put in light further considerations on the usefulness of the presurgical LCT simplistically considered as a >30% motor response.
