Introduction

Deep brain stimulation (DBS) provides symptomatic motor benefit for patients with advanced Parkinson’s disease (PD) [1,2,3,4]. The benefit of symptom control through DBS surpasses that of optimal medical treatment in patients with motor fluctuations and dyskinesias, and it is a relatively safe treatment option for motor complications of idiopathic PD [1,2,3,4,5]. DBS is often performed in relatively early-onset PD, a population in which it has been estimated that at least 5–10% of cases are not sporadic, but may carry genetic mutations [6, 7]. Genetic cases often are phenotypically different compared to sporadic patients, and this factor may influence clinical outcome [6, 8].

Though DBS has demonstrated efficacy, randomized studies have been performed in PD patients without genetic characterization raising questions of suitability of various monogenic forms and their relevance in DBS outcome. It is known that medication effects may vary between different mutations. For example, patients with PRKN mutations generally are particularly prone to levodopa-induced dyskinesias, whereas patients with LRRK2 mutations tend to show a normal sustained benefit for levodopa [8,9,10,11]. The effects of other antiparkinsonian drugs, such as rasagiline, may also be modulated by the genotype [12]. Given the variability in medication effects, it is conceivable that there are also differences in the treatment response to DBS in advanced monogenic PD. There are several case reports and small case series of DBS outcomes in patients with genetic PD, but due to a lack of information synthesis, we performed a systematic review on the effects of DBS in genetic PD.

Methods

Search strategy

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement was followed [13]. We performed a PubMed search from inception to June 26, 2018 with keywords “deep brain stimulation or DBS”, “Parkinson’s or Parkinson or Parkinsonism” and “genetic or gene or GBA or PRKN or PARKIN or LRRK2 or SNCA or PINK1 or VPS35 or DJ-1 or UCHL1 or GIGYF2 or HTRA2 or TMEM230 or CHCHD2 or RIC3 or ATP13A2 or PLA2G6 or FBX07 or SYNJ1 or VPS13C or DNAJC6”. All original English language articles concerning genetic PD patients treated with DBS were included. Animal studies and review articles were excluded.

The initial search identified 220 articles, and we included an additional 16 relevant studies found in the manual search of reference lists (Fig. 1). All abstracts of these studies were screened, and 184 studies were excluded in the first round (no monogenic PD patients or not treated with DBS n = 64, review or commentary article n = 92, animal study n = 28). The remaining 52 studies were assessed fully for eligibility and six more studies were excluded in the second round (genetic test negative n = 2, no genetic testing n = 1, review or commentary article n = 3). Finally, 46 studies of these 236 studies met all selection criteria and were included in the systematic review (Table 1). A summary of the included studies is presented in Table 2. The included studies reported 221 genetic PD patients who were treated with DBS. However, two studies reported partially the same patients [14, 15].

Fig. 1
figure 1

Flow chart of study inclusion and exclusion

Full size image
Table 1 The data extracted from the included studies
Full size table
Table 2 Summary of key findings according to the mutated gene
Full size table

Specific aims

This review of evidence aimed to systematically investigate DBS outcome in monogenic PD compared to the general PD population. The primary aim was to evaluate the motor benefit of the DBS operation in each monogenic PD type. An additional aim was to evaluate effects on non-motor symptoms, including possible cognitive and neuropsychiatric symptoms.

Selection criteria

Search terms and the PubMed search were planned by two authors (T.K. and V.K.). All titles and abstracts were reviewed by one investigator (T.K.). Studies were excluded if the title and/or abstract were not suitable for the aim of the review. Full texts were obtained for appropriate studies or if the relevance of an article was uncertain. The inclusion criteria for the selected studies were as follows: (1) a human study, (2) genetic PD patients treated with DBS, and (3) English language. The data extracted from each study were study year, first author’s family name, number of patients, mutated gene, specific mutation, patient age at disease onset and DBS implantation, target nucleus of DBS, more specific lead positioning, pre- and postoperative UPDRS-III scores, follow-up time and outcome (Table 1). UPDRS-III scores of control cohort's (mutation non-carriers, NC) are also reported in Table 1 if the information was available. In the outcome evaluation, an improvement of 30% or more in the UPDRS-III motor score was considered to indicate favourable outcome; 20–30%, moderate outcome; and < 20%, poor/mild outcome [58,59,60].

Quality control

The quality of the included studies was evaluated according to the Newcastle-Ottawa Scale (NOS) [61]. NOS includes selection, comparability, and exposure or outcome. The scale ranged from 0 to 11 stars, with the highest rating representing the greatest quality. Six months or more was a limit for the adequate follow-up time. Pre- and postoperative evaluation was thought to be accomplished if the outcome was reported properly with percentage improvement of the UPDRS-III score or verbally. A total score of 0–3 was considered to indicate to poor quality; 4–7, moderate quality; and 8–11, good quality. The NOS total score is presented in Table 1 and the scale is presented more accurately in Supplementary Table 1. A summary of the assessed quality of the studies is presented in Supplementary Table 2.

Results

A summary of the primary results is presented in Table 2. Altogether, 46 studies and 221 monogenic PD patients treated with DBS were included in the systematic review (Table 1).

LRRK2

Seventeen studies [9, 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] reported 87 patients (target: subthalamic nucleus (STN) n = 79, not available (NA) n = 8). The outcome was reported in 73 patients (83.9% of patients); with percentage improvement of the UPDRS-III score in 49 patients and verbally in 24 patients. The motor outcome was mostly favourable in patients with LRRK2 mutation. Only five studies with ten patients reported poor/mild/moderate outcomes. Both patients with the p.T2031S (c.6091A > T) mutation (n = 2) developed neuropsychiatric problems 5–7 years after implantation. The outcome appeared poor in patients with p.R1441G (c.4321C > G) mutations whereas it appeared excellent in patients with p.G2019S (c.6055G > A) mutations.

PRKN

Eighteen studies [11, 15, 16, 19, 21, 31,32,33,34,35,36,37,38,39,40,41,42,43] reported 67 patients (STN n = 51, globus pallidus interna (GPi) n = 5, zona incerta n = 1, NA n = 10). The outcome was reported in 57 patients (85.1%); UPDRS-III percentage improvement was reported in 45 patients and the outcome was described verbally in 12 patients. Fifty-one patients (76.1%) had favourable long-term motor outcomes. Six patients in three different studies were reported to have modest or poor outcomes.

GBA

Five studies [14, 15, 19, 44, 45] reported 50 patients (STN n = 33, GPi n = 4, ventral intermediate nucleus (VIM) n = 1, NA n = 12). Samples partially consisted of same patients in two studies [14, 15]. The outcome was reported in 30 patients (60.0%); UPDRS-III percentage improvement in 28 patients and the outcome was described verbally in 2 patients. Eighteen patients were reported to have favourable, three patients moderate and nine patients poor long-term motor outcomes. One study reported better outcomes with STN-DBS and VIM-DBS than with GPi-DBS [15]. GBA mutation carriers developed cognitive impairment faster than patients without mutations.

SNCA

Five patients were reported in five case reports [46,47,48,49,50] (STN n = 4, GPi n = 1). The motor outcome was favourable for all patients in the short-term but 3/5 patients developed cognitive and/or neuropsychiatric problems a few years after implantation. The percentage change in the UPDRS-III score was documented in two patients.

VPS35

Four studies [51,52,53,54] reported five patients (STN n = 3, NA n = 2). Favourable motor outcome was reported in four cases and minor motor benefit complicated by dysarthria in one case. The percentage change in the UPDRS-III score was reported in three patients.

PINK1

Five case reports [21, 32, 38, 55, 56] including one patient in each report (STN n = 4, GPi n = 1) were reported. Favourable motor outcome was observed in three patients and moderate outcome in one case. One patient developed imbalance, gait impairment, dysarthria, and behavioral changes after operation and mental deterioration was documented a few years later.

Exclusion of poorer quality studies

Unfortunately, many studies (Table 1) lacked important information as shown in the Supplementary Table 1. Poorer quality studies have tendency for bias; therefore, in the Supplementary Table 3, data are presented after exclusion of poorer quality studies such as studies lacking the information about DBS target, pre- and postoperative evaluation, adequate follow-up time or outcome information. Furthermore, as Lythe et al. [14] and Angeli et al. [15] reported partly the same patients, we tested the conclusions also when the smaller study was excluded. Nevertheless, after the exclusion of these studies, the results remained essentially the same (Supplementary Table 4).

Discussion

We report the following key findings: (1) DBS outcome appears excellent in patients with LRRK2 p.G2019S (c.6055G > A) mutations, good in patients with PRKN mutations and poor in patients with LRRK2 p.R1441G (c.4321C > G) mutations, (2) the overall benefit of DBS in SNCA, GBA and LRRK2 p.T2031S (c.6091A > T) mutations may be decreased due to rapid progression of cognitive and neuropsychiatric symptoms, and (3) in other mutations, the motor outcome in DBS-treated genetic PD patients appears generally comparable to that of sporadic PD patients.

A recent smaller review of 30 studies described the effects of DBS mainly in patients with LRRK2, PRKN and GBA mutations [62]. In the present PRISMA-compliant systematic review of 46 studies and 221 patients, the most comprehensive data were available for patients with LRRK2 and PRKN mutations. The combined evidence suggests that patients with LRRK2 mutations generally have a good response to DBS, and patients with the most common LRRK2 mutation, the p.G2019S mutation [7], may even have better outcome than the general PD population. However, the reported LRRK2 cases of p.R114G, p.T2031S and p.N1437H (c.4309A > C) mutation carriers appeared to have less favourable outcome. This interpretation is limited by the small number of reported DBS-treated cases of rarer LRRK2 mutations. For the PRKN mutations, the literature supports a view that patients with PRKN mutations are optimal candidates for DBS.

Apart from the LRRK2 and PRKN genes, the published literature concerning individual monogenic mutations and DBS is less comprehensive and the data are clearly limited with respect to both the number of patients and duration of follow-up. The available data are limited to five DBS-treated patients with VPS35 mutation, and the patients have shown favourable sustained motor outcome in 4/5 cases. The available literature also suggests that most patients with mutations in GBA tend to achieve favourable long-term motor outcome from STN-DBS. Despite good motor outcome, GBA mutation carriers may develop cognitive impairment after DBS faster than patients without mutations. SNCA patients commonly develop cognitive and neuropsychiatric problems [8]. The literature supported a good motor outcome after DBS also in patients with SNCA mutations; however, 3/5 patients developed cognitive and neuropsychiatric problems a few years after DBS implantation. Indeed, the non-motor features of genetic PD may be a limiting factor in the overall benefit of DBS in some mutations, such as SNCA and LRRK2 p.T2031S. While the motor benefit from DBS may initially be clear, the rapid non-motor progression may lessen the sum value for the quality of life. A recent study in SNCA A53T mutated rodents suggested that DBS may be neuroprotective [63]. Nonetheless, in human PD patients with SNCA mutations, the neuropsychiatric progression appears to be rapid despite DBS. The issue could be the level of damage at the time of implantation, and earlier DBS in these patients might possibly provide different outcomes.

Preoperative response to levodopa is the best single predictor of the postoperative outcome of DBS [64]. This indicator appears useful also in patients with monogenic mutations and the response was reported in practically all included studies. Another relevant predictor is the localization of DBS electrodes [65]. Unfortunately, there were studies, which did not report DBS targets and most studies lacked information about lead positioning. As the literature expands in the future, the effect of targets and lead positioning should be investigated in more detail. In most studies, STN was preferred over GPi as the target. Hence it remains ambiguous whether there are any relevant differences of clinical outcome between STN and GPi stimulation in monogenic PD. One study reported also a patient with VIM stimulation which is an unusual target for PD patients because VIM stimulation improves only tremor, not other PD symptoms [66, 67]. Finally, it is important to note that the genetic status may have a positive as well as a negative influence on outcome of surgery and this issue should be taken into consideration in the interpretation of DBS studies. For example, the EARLYSTIM trial was performed with young-onset PD patients [5] and there could have been an overrepresentation of PRKN patients in the sample.

In conclusion, monogenic PD patients have variable DBS outcomes depending on the mutated gene. Most patients benefit from STN-DBS, at least in the short-term; however, the current evidence does not support or is questionable for DBS implantation for patients with p.T2031S or p.R114G mutations in the LRRK2 gene or mutations in the SNCA or GBA genes. The best outcome from DBS surgery appears to be in patients with LRRK2 p.G2019S or PRKN mutations.