Abstract
INTRODUCTION
Parkinson’s disease (PD) is a progressive neurodegenerative disease that affects dopaminergic neurotransmission, resulting in bradykinesia, rigidity, and rest tremor. Advanced PD typically involves the development of dopamine medication-related motor complications as well as dopamine-resistant features of axial motor impairments and progressive cognitive impairment [1].
Subthalamic nucleus deep brain stimulation (STN-DBS) is one of the most effective treatments for advanced levodopa-responsive forms of PD that rapidly improves motor disability and reduces levodopa-related complications [2, 3]. Clinical pre-operative predictive factors of optimal STN-DBS motor results have been previously reported [4]. Thus, it has been demonstrated that the outcome of neurosurgery was influenced partly by the age of the patients and the duration of the disease and was markedly dependent upon whether the parkinsonian motor symptoms responded to levodopa treatment [4]. Nevertheless, age of patient at surgery as a clinical predictive factor is still a debated issue [5–9]. Previous studies focusing on elderly PD patients suggest that motor complications dramatically improve in this population whereas activities of daily living and axial subscores tend to worsen [5–7]. Moreover, quality of life assessed with the PDQ-39 is significantly improved in young patient compared to old patients [5]. On the other hand, several factors are thought to predict cognitive decline following STN-DBS, such as advanced age, axial signs, and poor baseline cognitive performance [10, 11]. Indeed, further study shows that negative impact of surgery on frontal executive performance was more consistently observed in patients who were older than 69 years [12].
This report analyses the motor and cognitive outcomes outcome in 100 consecutive PD patients treated with STN-DBS, according to their age (<65 year-old or ≥65). Moreover, it focuses on the post-operative outcome in a sub-group patients 70 years or older.
MATERIALS AND METHODS
Patients
One hundred consecutive patients (42 women, 58 men; mean age, 60.7 ± 8.1 years) with advanced levodopa-responsive PD (Hoehn and Yahr off score ≥ 3; mean disease duration, 12.4 ± 4.3 years) referred for STN stimulation to our centre between 2004 and 2012 were enrolled. All patients fulfilled the same selection criteria, i.e. clinically diagnosed PD, severe levodopa-related complications despite optimal adjustment of antiparkinsonian medication, no surgical contraindications, and no dementia or major ongoing psychiatric illness. Since we did not use an age cut-off for surgery, eight patients over 70 (range: 70 to 78 years) were operated on in our centre. The surgical procedure has been previously described [13]. All the patients gave their informed consent to their participation in the study and the protocol was approved by our local ethic committee.
Two groups of patients were stratified according to age at time surgery: patients younger than 65 years old (group 1, n = 57) and patients 65 years or older (group 2, n = 34). Moreover, we focused on the post-operative outcome in a sub-group patients 70 years or older (group 3, n = 8).
Clinical evaluation
A baseline clinical evaluation was performed 1 month before surgery based on the CAPIT protocol [13]. The motor state was assessed by the UPDRS part III. Axial motor features, i.e. speech, arising from chair, posture and gait, were assessed separately using the corresponding UPDRS III sub-scores (items 18, 27, 28, and 29). Motor performance assessments were performed 6 months after surgery, under two conditions (off-drug/on-stimulation, on-drug/on-stimulation) as previously described [15]. The percentage improvement in motor disability was determined in respect of the preoperative off-drug condition. Complications of levodopa therapy, i.e., dyskinesias (UPDRS part IVA score) and clinical fluctuations (UPDRS part IVB score), were also assessed pre- and postoperatively. All the patients underwent standardized cognitive and psychiatric assessments, i.e. the Mattis Dementia Rating Scale (MDRS) score, the GREFEX version of semantic and phonemic verbal fluency tasks and of the Stroop Color Word test, the Wisconsin card-sorting Test (WCST), Buschke and Grober Free and Cued Selective Reminding Test–Immediate Recall (FCSRT-IR) and the Montgomery & Åsberg Depression scale (MADRS), that were performed before the intervention in the on-drug condition, and at 6 months in the on-drug/on-stimulation condition.
All dopaminergic drugs, expressed in levodopa equivalent daily dose (LED), were recorded before and after surgery. To compare the effects of changes in antiparkinsonian medications, we calculated that a 100-mg daily dose of standard levodopa was equivalent to the following doses of other medications: 133 mg of controlled-release levodopa; 75 mg of levodopa plus entacapone; 1 mg of pramipexole; 5 mg of ropinirole; and 10 mg of apomorphine [2].
Statistical analysis
Statistical analysis was performed with the STATISTICA software version 7.1, StatSoft1, France. The different scores were expressed as mean value ± SD. The cognitive and motor effects of DBS were evaluated by comparison of different scores obtained before (baseline) and 6 months after surgery using a repeated measures analysis of variance. Quantitative and qualitative data were compared between group 1 and 2 using respectively a Mann-Whitney and Kruskal-Wallis test. p < 0.05 was considered statistically significant.
Descriptive statistics were used to describe pre- and post-operative clinical characteristics ingroup 3.
RESULTS
Young (group 1) versus old (group 2)
One hundred PD patients who underwent STN-DBS in our centre were enrolled in the study. Nine patients were excluded from the analysis because of lack of pre-operative data in 5 and lost of follow-up in 4. Group 1 (<65 year old) was composed of 57 patients (24 women; mean age, 56.1 ± 6.4 years; mean disease duration, 11.9 ± 4.2 years). Group 2 (≥65 year old) was composed of 34 patients (14 women; mean age, 68.4 ± 3.0 years; mean disease duration, 13.3 ± 4.4 years). Gender, disease duration, LED, levodopa responsiveness, axial and Hoehn and Yahr scores assessed in the “off” state, and motor fluctuations (UPDRS part IVB) score were not statistically different between groups. Preoperative off-UPDRS III and dyskinesias (UPDRS part IVB) scores were significantly more severe in the group 1 whereas on-axial sub-score was significantly more severe in group 2.
Post-operative motor outcome. Since three patients had perioperative intracerebral haemorrhage (2 in the group 1 and 1 in group 2), post-operative motor outcomes were assessed in 88 patients (Table 1). Six months after surgery, the UPDRS part III score improved from baseline value (off-drug condition), respectively by 66% and 84% in group 1, and by 60% and 78% in group 2, in the off-drug/on-stimulation and on-drug/on-stimulation conditions (P < 0.05). Similarly, axial motor sub-score improved by 54% and 71%, and by 25% and 60%, respectively in group 1 and group 2, in the off-drug/on-stimulation and on-drug/on-stimulation conditions. LED was significantly reduced by 56% in group 1 and by 40% in group 2. The severity of levodopa-related complications (UPDRS IV score) was significantly reduced by 76% in group 1 and by 54% in group 2 (P < 0.05).
Post-operative motor improvement (UPDRS part III and part IV scores) and LED reduction were not statistically different between groups (P > 0.05). The axial motor sub-score which was significantly worse in group 2 in the on-drug condition before surgery was also significantly worse both in off-drug/on-stimulation and on-drug/on-stimulation conditions (P < 0.05).
Pre- and post-operative cognitive and psychiatric assessments. Pre and post-operative cognitive and psychiatric assessments are summarized in Table 2. There were no significant differences between groups with respect to pre-operative cognitive performances and mood disorders. Six months after surgery, mean scores on global cognitive scale, i.e., MDRS, remained unchanged whereas a significant decrease in semantic verbal fluency tasks performance and an increase of the Stroop test interference time were observed in both groups. Total time to complete color word naming in the Stroop test (P = 0.03) and FCSRT-IR free recall score (P = 0.04) significantly worsened after surgery in group 2. Finally, scores in the WSCT in group 2, tended to be worse in all modalities (categories, total errors and perseverative errors) although without statistical significance.
Further analyses revealed that deterioration of scores in the WSCT in all modalities, i.e. number of categories (P = 0.02), errors (P = 0.006) and perseverative errors (P = 0.002), was significantly higher in group 2 compared to group 1. Moreover, both Stroop test interference time and FCSRT-IR free recall were significantly worsened in group 2 compared to group 1.
Surgery- and device-related complications. Overall, serious surgery- and device-related complications included (Table 3) symptomatic intracerebral haemorrhage (n = 3), subdural hematoma (n = 3), thromboembolism (n = 4), infection around the implanted pulse generator and the extension cable (n = 3), pneumoencephalus (n = 1), and post-operative confusion (n = 6). There were no significant differences between groups.
Electrical parameters at 6 months. The mean stimulation parameters were: group 1, 147 and 146 Hz, 2.9 and 3 V, and 62 and 64 microseconds; group 2, 146 Hz and 146 Hz, 3.1 and 3.1 V, and 62 and 63 microseconds, respectively for the right and the left sides. There were no significant differences between groups.
Patients≥70 year-old (group 3)
Post-operative motor outcome. Eight patients (3 women; median age, 72.8 ± 2.5 years, range 70 to 78 years; mean disease duration, 15.9 ± 6.6 years, range 10 to 30) were more than 70 year-old at the time of surgery. Six months after surgery, the UPDRS part III score improved from baseline value (off-drug condition), in the off-drug/on-stimulation and on-drug/on-stimulation conditions (P < 0.05) in group 3 (Table 4). Similarly, axial motor sub-score improved significantly in the on-drug/on-stimulation condition. The severity of levodopa-related complications (UPDRS IV score) tended to improve although without statistical significance. LED was significantly reduced by 41%.
Pre- and post-operative cognitive and psychiatric assessments. Pre and post-operative cognitive and psychiatric assessments are summarized in Table 5. Six months after surgery, WSCT scores remained stable whereas MDRS and FCSRT-IR free recall scores significantly worsened in group 3. In the meanwhile, a decrease in verbal fluency tasks and Stroop test performances were observed.
DISCUSSION
In the current study, we show that younger PD patients (<65 years old) and older PD patients (65 years or older) improve to the same extent by continuous bilateral high-frequency stimulation of the STN. However, we find that older PD patients, who had a significantly worse axial motor sub-score before surgery, experience less improvement of axial motor features, both in off-drug/on-stimulation and on-drug/on-stimulation conditions, 6 months after STN-DBS (P < 0.05). Moreover, we report a significant worsening of post-operative global cognitive performances in the older subgroup. Finally, we found no significant differences between groups with respect to the frequency of serious surgery- and device-related complications.
These results are consistent with published studies assessing the effect of age at time of surgery on the motor outcomes in PD patients [5–7]. Derost et al. [5], Vesper et al. [6], and Shalash et al. [7], compared the clinical effect of continuous bilateral high-frequency stimulation of the STN in parkinsonian patients younger than 65 years old versus 65 years old or older. Similarly, they found that UPDRS part III score improved in both groups in off-drug/on-stimulation without significant difference whereas the axial motor sub-score improvement was higher in the young group. The axial response to STN stimulation is usually influenced by age and most of all by L-dopa responsiveness during pre-operative testing [16, 17]. Long-term follow up after STN-DBS has showed that stimulation-induced motor improvement wears off over times with a progressive loss of benefit on axial signs [18]. Gait and balance disturbances typically emerge in advanced PD with a generally limited response to dopaminergic medication and STN-DBS. In our study, all patients shared the same pre-operative selection criteria according to previous studies concerning STN-DBS efficacy predictive factors [4]. In particular, they all showed excellent improvement of their pre-operative axial sub-score after acute levodopa challenge. Therefore, the axial performance worsening in the older PD patients is likely related to degenerative disease progression related to non-dopaminergic lesions.
While global cognitive performances are preserved after chronic STN-DBS, several studies have shown that executive functions can be altered [8, 19]. Decline in verbal fluency is the most consistent neuropsychological adverse effect reported, both in phonemic and semantic tasks [20, 21]. Moreover, several studies showed an accelerated cognitive decline few months after surgery, suggesting that a sub-group of patients may have a predisposition towards PD dementia following STN stimulation [11]. Thus, age, attention, and levodopa response at baseline have been identified to be predictors for cognitive decline at 12 months in STN DBS [19, 22]. Accordingly, we found that global cognitive performances were preserved whereas semantic verbal fluency tasks performances declined in both groups after STN-DBS. In addition, we reported an increase of the Stroop test interference time whatever the age. Conversely, we observed that WSCT, Stroop test interference and FCSRT-IR free recall, were significantly worsened post-operatively in the group ≥65. These results are in accordance with some studies assessing cognitive outcome in older PD patients after STN-DBS [12] but not other [5]. Indeed, Saint Cyr et al, found that STN-DBS has a negative impact on cognitive functions especially in elderly patients over 69 years old [12]. On the contrary, Derost et al. assumed that lower pre-operative performances in older patients may explain post-operative differences between groups [5]. Thereafter, they concluded to an absence of global cognitive deterioration whatever the age at time of surgery. In the present study, there were no significant differences between groups with respect to pre-operative cognitive performances. Moreover, we found impaired episodic memory in older subjects compared to younger after surgery. Cognitive dysfunction has been noted after all types of surgery in older patients, and early post-operative cognitive decline was identified as an independent risk factor for persistent cognitive decline in one study [23]. Yet, there is little evidence that post-operative cognitive dysfunction leads to dementia or that surgery and/or anesthesia accelerate cognitive decline [24]. Moreover, impaired episodic memory is generally observed in patients with DBS implanted for a long time and reflects the natural evolution of PD [11, 25]. Therefore, we assumed that both PD disease progression and surgical procedure may explain cognitive decline inolder patients.
Few studies have evaluated the motor and cognitive outcome after STN-DBS in a limited case of PD patients who were older than 70 years [7, 12]. Saint Cyr et al, first reported a series of 6 patients older than 69 years (range 70 to 75 years) [12]. In general, they underlined that older patients did not improve their motor performances (on medication, on stimulation) as much as the younger patients did. Moreover, they demonstrated that cognitive processes involving executive functioning, such as working memory, phonemic fluency, encoding efficiency, susceptibility to interference, associative learning, speed of processing and switching of mental sets, as well as bimanual co-ordination under conditions of divided attention, were impaired in older patients following STN-DBS. Russman et al. then reported a series of 9 patients older than 70 years [7]. They found that patients over age 70 treated by STN-DBS had similar motor complications improvement than younger ones. Nevertheless, they emphasized that UPDRS motor scores, especially axial signs, improved less in the older patients leading to worsening of activities of daily living and less medication reduction. Interestingly, they found that these reduced motor responses to STN-DBS were not predicted by the pre-operative L-dopa responsiveness, suggesting that age is an independent predictive factor. In the present study, we reported the clinical outcome in a limited series of 8 PD patients over 70 years (range 70 to 78 years) who fulfilled the classic selection criteria for STN-DBS. In the majority of these patients, preoperative UPDRS motor score improved under stimulation. In the meanwhile, the severity of levodopa-related complications (UPDRS part IV score) tended to improve and LED were reduced. Moreover, we reported a decline of their cognitive performance. However, these results should be interpreted with caution because of the limited population, the lack of statistical analyses, and the short term 6 months follow-up.
Finally, the incidence of permanent side effects related to the procedure such as intracerebral hemorrhage, seizure, pulmonary embolism, and transitory confusion is considered to be less than 4% [18]. Older age (>70 years) was associated with a tendency towards intra-operative and postoperative confusion without significant increase of serious surgery- and device-related complications [7, 12]. Accordingly, recent studies suggest that incidence of complications related to surgery (including postoperative hemorrhage or infection) is not increased in the elderly population [26, 27]. Conversely, Vesper et al. found that the death rate within the 24 months following surgery and the incidence of infection were higher in elderly patients over 65 years old [6]. Here, we found no significant differences between groups with respect to the frequency of surgical complications. Then, we concluded that DBS-STN is safe in a carefully selected population even over 65 years old.
The main limitation of our study is that we did not include the item 30, i.e. postural stability, in the axial subscore definition. In fact, most clinicians usually regard postural stability as an important axial symptom. According to previous studies, we believed nevertheless that our results would be approximately the same with this item.
CONCLUSIONS
Although effective and safe, our study suggests that STN-DBS has a more negative impact on cognitive functions in elderly patients. A recently published review concluded that there might be a possible advantage of unilateral and bilateral GPi-DBS over STN-DBS when comparing cognitive outcome [28]. Further recent studies have shown that lead trajectories intersecting with caudate nuclei increased the risk of global cognitive decline and working memory performance after STN implantation [29]. In addition, it has been suggested that multiple tracks could increase the risk of memory function deterioration [30]. Together, these findings suggest that surgical procedure should be optimized to minimize post-operative cognitive impairment, especially in older PD patients.
CONFLICT OF INTEREST
Jean Paul Bouwyn reports no disclosure.
Stéphane Derrey reports no disclosures.
Romain Lefaucheur reports no disclosures.
Damien Fetter reports no disclosures.
Audrey Rouille reports no disclosures.
Floriane Le Goff reports no disclosures.
David Maltête reports no disclosures.