*6.3. Recommendations*

More e fforts are needed to understand which NMS are predictors of good or poor outcome, how di fferent targets of DBS (STN or GPi) should be indicated to treat di fferent NMS based on the ability to decrease total LEDD and dopamine-agonists LEDD [17], or directly treat particular symptoms by means of the stimulation of specific networks involved in pain or mood, apathy and attention [73]. We recommend to carefully assess the presence and severity of non-motor symptoms before surgery and explain to the patients that when the disease burden is mainly driven by non-motor symptoms, DBS might not be the best therapeutic option to consider.

### **7. A New Role for Genetics**

### *7.1. The Standard Rule*

One of the most remarkable advances in our understanding of PD pathogenesis in the last 20 years is represented by genetics. The increasing power of genetic analyses led to the identification of several chromosomal loci that cause or modulate the risk for PD [74]. Moreover, specific genetic mutations have been associated to specific clinical features and di fferent disease courses, which could have an impact on the selection for DBS.

That is, only some evidence on the di fferential DBS response in di fferent forms of monogenic PD has been put forward, suggesting some di fferences in the magnitude of response [14,75–77]. However, in this case there are no specific recommendations on the use of genetics in the clinical practice of DBS centers.

### *7.2. Pros and Cons*

The main advantage of knowing the genotype of a PD patient appraised for DBS is related to the knowledge of disease evolution associated to a particular gene variant. However, despite the e ffort of systematic reviews [76,77] and one meta-analysis [14], the small size of the cohorts reported and the paucity of data on di fferent gene variants and brain targets other than STN, do not allow researchers to reach firm conclusions. For example, *LRRK2*-G2019S variant carriers, described in 44 out of 50 *LRRK2*

subjects with DBS reported in the literature, show an excellent response to STN DBS, which is also the most reported target [76]. G2019S is the most frequent *LRRK2* variant and produces a phenotype overlapping to late-onset, non-mutated PD with frequent presence of tremor and good response to dopaminergic medications [78]. However, in three out of four reported cases with *LRRK2*-R1441G variant, a mutation variant rarely found outside northern Spain, poor response to STN-DBS was reported [76]. The paucity of data characterizing the phenotype of *LRRK2*-R1441G variant makes it impossible to assume that the poor DBS outcome was due to more severe disease progression and development of DBS resistant features. A similar issue applies to carriers of glucocerebrosidase (*GBA*) gene variants which have a high prevalence of neuropsychiatric symptoms, especially impulsive compulsive behavior and hallucinations, and a higher risk to develop early over disease course cognitive disturbances [13]. Indeed, *GBA*-associated PD showed worse cognitive and functional performances and lower reductions in dopaminergic medication after surgery [14]. However, it is unknown which variants mostly contribute to this result. Remarkably, the risk of hallucinations and cognitive impairment, as well as survival, di ffers across *GBA* subjects, being higher in subjects carrying complex and severe variants [13,79].
