Long-Term Cognitive Profile and Incidence of Dementia After STN-DBS in Parkinson s Disease

Transcription

1 Movement Disorders Vol. 22, No. 7, 2007, pp Movement Disorder Society Long-Term Cognitive Profile and Incidence of Dementia After STN-DBS in Parkinson s Disease Selma Aybek, MD, 1 Aline Gronchi-Perrin, 1 Alexandre Berney, MD, 2 Sabina Catalano Chiuvé, 3 Jean-Guy Villemure, MD, 4 Pierre R. Burkhard, MD, 3 and François J.G.Vingerhoets, MD 1 * 1 Service de Neurologie, CHUV, Lausanne, Switzerland 2 Service de Psychiatrie, CHUV, Lausanne, Switzerland 3 Service de Neurologie, HUG, Genève, Switzerland 4 Service de Neurochirurgie, CHUV, Lausanne, Switzerland Abstract: An effect of subthalamic nucleus deep brain stimulation (STN-DBS) on cognition has been suspected but longterm observations are lacking. The aim of this study was to evaluate the long-term cognitive profile and the incidence of dementia in a cohort of Parkinson s disease (PD) patients treated by STN-DBS. 57 consecutive patients were prospectively assessed by the mean of a neuropsychological battery over 3 years after surgery. Dementia (DSM-IV) and UPDRS I to IV were recorded. 24.5% of patients converted to dementia over 3 years (incidence of 89 of 1,000 per year). This group of patients cognitively continuously worsened over 3 years up to fulfilling dementia criteria (PDD). The rest of the cohort remained cognitively stable (PD) over the whole follow-up. Preoperative differences between PDD and PD included older age ( years; years), presence of hallucinations and poorer executive score ( ; ). The incidence of dementia over 3 years after STN-DBS is similar to the one reported in medically treated patients. The PDD presented preoperative risk factors of developing dementia similar to those described in medically treated patients. These observations suggest dementia being secondary to the natural evolution of PD rather than a direct effect of STN-DBS Movement Disorder Society Key words: dementia; subthalamic; DBS; Parkinson s disease. Subthalamic nucleus (STN) deep brain stimulation (DBS) for Parkinson s disease (PD) has been validated: STN deep brain stimulation (STN-DBS) provides good motor outcome, 1 and allows significant reduction of antiparkinsonian medication. 2 There has been much debate as to whether STN-DBS may cause adverse effects on cognition and behavior 3-7 with suspicion of a negative impact on cognition, particularly executive functions. In medically treated PD, the prevalence of dementia is now considered to be much higher than previously thought, reaching 60% after 10 to 15 years of motor *Correspondence to: Francois Vingerhoets, Service de Neurologie, CHUV, Rue Bugnon 44, 1011 Lausanne, Switzerland. francois.vingerhoets@chuv.ch Received 4 December 2006; Revised 20 February 2007; Accepted 23 February 2007 Published online 18 April 2007 in Wiley InterScience (www. interscience.wiley.com). DOI: /mds *Use of different variant of the same tests at each evaluation (see Method). disability with an almost 6-fold increased risk when compared with normal controls. 8 Prevalence and incidence of dementia in surgically treated patients is unknown, as cognitive data on the long-term impact of STN-DBS are scarce. 9 In this prospective study, we examined the long-term evolution of cognition after STN-DBS, including the incidence of dementia, and we studied potential risk factors for postoperative cognitive decline. METHODS We studied 57 (36 men, 21 women, age years, PD duration years) consecutive nondemented PD patients who underwent bilateral STN-DBS and who potentially reached at least their two years follow-up. Inclusion criteria were the presence of PD following the United Kingdom Parkinson s 1 Disease Society Brain Bank Criteria, with untreatable motor fluctuations and dyskinesia, without any sign atypical for PD at clinical examination and with at least 25% improve- 974

2 COGNITIVE PROFILE AND INCIDENCE OF DEMENTIA 975 FIG. 1. Follow-up of patients. E, evaluated with detailed neuropsychological testing; NE, nonevaluated with detailed neuropsychological testing; PD, patients without dementia; PDD, patients with dementia. ment of the Unified Parkinson s Disease Rating Scale (UPDRS) motor scale by levodopa (L-dopa) tests. 2 Exclusion criteria included overt ongoing psychiatric disease, severe depression (MADRS 19), general contraindication to surgery, and dementia following DSM-IV criteria (i.e., memory impairment plus decline in one other cognitive field with impact on daily living) or severe impairment (score below the 5th percentile of the healthy population: C5) in two or more tested domains (see below). Clinical evaluations were made prior to surgery (mean months) during the on medication state and at 3, 6, 12, 18, 24, and 36 months and once a year postoperation with STN-DBS turned on and under the current medication. UPDRS I to IV and the total L-dopa equivalent dose (LED) were computed 2 : 100 mg* L-dopa with dopa-decarboxylase inhibitor 130 mg controlled release L-dopa* preparations 83 mg L-dopa with dopadecarboxylase and COMT inhibitors (i.e., the L-dopa dosages marked by * were increased by 20% when an inhibitor of COMT was associated) 1 mg pergolide 1.5 mg cabergoline 1 mg lisuride 1 mg pramipexole 10 mg bromocriptine 3 mg ropinirole. Presence of hallucination was assessed with the UPDRS I (item 2) (scores 0 and 1 absence of hallucinations, scores 2, 3, and 4 presence of hallucination). This dichotomization was done since vivid dreams (score 1) seem less specific for PD and more related to medication. 10 At each follow-up, clinical criteria for dementia were reviewed by the neurologist and the impact of cognition on daily living was carefully discussed with the patient and caregiver. Neuropsychological tests were performed at baseline for the 57 patients, at 6 months ( months, range: months) for 53 patients and at 3 years ( months, range: months) for 50 patients (Fig. 1) by the same trained neuropsychologist. Six specific cognitive domains, i.e., language, praxis, visual perception, calculation, memory, and executive functions, using standardized tests (see Appendix) were assessed. Potential test retest bias was considered minimal because of the large time interval between each assessment. In addition, we used different variant of the same tests at each evaluation, when available. For internationally validated tests and for locally developed tests ( Examen Neuropsychologique du CHUV : French version commercially available at Division Autonome de Neuropsychologie Lausanne, Switzerland), 11,12 the impairment criterion was based on the distribution (percentile) of a corresponding normative sample stratified by age, education, and gender. Standardization of our test battery was performed on 10 healthy subjects for each decade, from 20 to 80 years old and for each education level; level I: 9 years of education, level II: 9 to 12 years, level III: 12 years. To obtain comparable categorical data for each test we defined a0to3score: 0 normal performances ( C25: above the 25th percentile of the healthy population), 1 limited performances ( C10 C25: between the 25th and the 10th percentile of the healthy population); 2 moderate impairment ( C5 C10: between the 10th and the 5th percentile), 3 severe impairment ( C5: below the 5th percentile). Finally, for each cognitive domain, we obtained a Cognitive Index (CI), corresponding to the sum of the related subtests. In 19 PD patients who underwent both assessments the same day (including 11 of the present study), our CI correlated linearly (r 0.89) with the Mattis dementia rating scale (MDRS); based on this correlation, estimated MDRS of all patients at inclusion in the present study were above 120. A trained psychiatrist made a structured clinical evaluation and scored patients on the Hamilton Anxiety Rating Scales (HARS) and the Montgomery Asberg Depression Rating Scale (MADRS). Statistical Analyses Results are presented as mean standard deviation. Student t tests were used for normally distributed parametric values, the Wilcoxon signed-rank test and Mann Whitney U test were used for nonparametric data. Bonferoni correction was applied for multiple comparisons. Significance level was established at P The cumulative proportion of patients becoming demented was estimated using Kaplan Meier survival analysis. The event of interest was the occurrence of dementia. The proportion of patients surviving at each time point was the proportion of patients not demented at that time. The timing variable was the time from baseline to the

3 976 S. AYBEK ET AL. visit at which dementia was diagnosed and to the last available visit for nondemented PD. RESULTS At baseline, except for poor executive performance as usually encountered in advanced PD, cognitive performances were normal (Table 1), and, by definition, no patient was demented according to DSM-IV, in particular their cognitive status did not interfere with daily living activities and they did not present significant impairment in more than one cognitive field. At 6-months evaluation, 5 of 57 patients (prevalence 8.7%) fulfilled dementia criteria when examined by the neurologist. This represents an incidence of 202 of 1,000 per year. At 3-years, fourteen patients (prevalence of 24.5%) were demented, representing an incidence of 89 of 1,000 per year from baseline. Between 6 months and 3 years, 5 patients died without having the long-term detailed neuropsychological evaluation. They were clinically evaluated in average 8.5 months before their death: none presented DSM-IV criteria of dementia at that time. They did neither differ from PD for the cognitive profile at baseline and at 6 months, nor for its evolution between these time points. Four of the nondemented patients were not examined with detailed neuropsychological testing at the 6-months follow-up (3 refused, 1 had systemic infection): all were examined at 3-year and were not demented at that point. At 3-years, 7 patients were not examined in detail (5 died, 1 refused, 1 moved to another country), representing a drop out for detailed neuropsychological testing of 12.3% (Fig. 1). Incidences of dementia in the 50 patients who underwent serial detailed neuropsychological examination are 214 of 1,000 per year at 6 months and 98 of 1,000 per year (Fig. 2) at 3-years with prevalences of 9.4 and 28%, respectively. At 6 months, the cognitive profile for the 53 patients who underwent neuropsychological testing declined in executive functions (especially for initiation, inhibition, and reasoning), for encoding and free recall capacities (Table 1). The 5 patients who converted to dementia worsened in almost all cognitive domains, significantly for the global memory score ( ; ; P 0.05). They ended up significantly worse than nondemented PD in all fields (Table 2) except for the direct digit span and ideomotor praxis. The 48 nondemented patients worsened in executive functions only ( ; ; P 0.05) [particularly initiation ( , ; P 0.05)]. At 3-years, the 50 patients who underwent neuropsychological testing worsened significantly in memory, particularly encoding capacities and inverse digit span, executive functions, inhibition, attention, and in constructive praxis (Table 1). The 14 patients who converted to dementia significantly worsened from baseline in all tested functions (Table 2) except for the language, direct digit span, and calculation; they performed significantly worse than PD in all fields (Table 2). After excluding these 14 patients, the remaining 36 PD patients who did not develop dementia had a neuropsychological profile (Table 2) similar to the preoperative evaluation except CI normal range TABLE 1. Cognitive profile of the whole cohort Baseline number of patient C10 CI baseline CI 6 mo CI 3 yr Language (71%) Constructive praxis (95%) a,b Ideomotor praxis (88%) Visual perception (97%) Memory (91%) c,d Direct digit span (95%) Inverse digit span (90%) a,d Rey: encoding/recall (76%) a a Rey: recognition (88%) b Executive functions (34%) a c Initiation (62%) c Inhibition/planning (34%) a a,d Attention (88%) a Reasoning (93%) a Calculation (76%) b a P 0.05 compared to baseline. b P 0.05 comparing 6 mo to 3 yr. c P 0.01 compared to baseline. d P 0.01 comparing 6 mo to 3 yr. CI, cognitive index; C10, with scores above the 10th percentile.

4 COGNITIVE PROFILE AND INCIDENCE OF DEMENTIA 977 FIG. 2. Kaplan-Meier curve of the incidence of dementia. y-axis: Proportion of PD without dementia; x-axis: Time course after the operation (months). Each step down represents one patient who developed dementia and each cross represents the time point of the last follow-up for each patient who did not develop dementia. for inhibition, which significantly worsened ( ; ; P 0.05). Over the months between the two postoperative follow-ups, the whole cohort had a worsening of inhibition, inverse digit span, memory (particularly recognition), constructive praxis, and calculation (Table 1). The 14 patients developing dementia worsened in all cognitive fields, significantly for constructive praxis, ideomotor praxis, calculation, inverse digit span, memory (particularly encoding). The rest of the cohort showed only a significant worsening for executive functions (inhibition) and inverse digit span. Patients who developed dementia were similar to the rest of the cohort for gender, duration of PD, baseline UPDRS III, LED, psychiatric profiles (except hallucinations) and the preoperative cognitive evaluation except executive functions (Table 2). The two groups had similar evolution for anxiety (HARS), depression (MADRS) and motor outcomes (Table 3). Correct electrodes positioning were confirmed in the 12 patients who developed dementia and who were controlled by postoperative MRI (performed routinely since 2000). No persistent complication occurred in either group. Transient complications comprised 4 confusions (1 in PDD, 3 in PD), one hyponatremia and one acute paranoid state, both in PD. At baseline, patients who subsequently developed dementia differed from the rest of the cohort only by an older age, presence of more hallucinations, and poorer executive functions especially inhibition and initiation (Tables 2 and 3). However, there was important overlaps between PDD and PD groups for these scores preventing them to be good preoperative predictors (Fig. 3). Even score combining these three risk factors did not reach 40% sensitivity for prediction of postoperative dementia. We analyzed the risk profile of the 5 patients who died before the long-term neuropsychological testing; they were older (mean 68.6 years), none of them had hallucinations. TABLE 2. Cognitive differences between PD and PDD at 6 mo and at 3 yr Baseline 6 mo 3 yr PDD PD PDD PD PDD PD N Language a b Constructive praxis a c b Ideomotor praxis c b Visual perception a c b Memory c a c b Direct digit span a Inverse digit span a c b Rey: encoding/recall b c b Rey: recognition b c b Executive functions a b,c c b Initiation a b,c c b Inhibition/planning a a c b,c Attention a c b Reasoning b c b Calculation a a b a P 0.05 between PD and PDD. b P 0.01 between PD and PDD. c P 0.05 compared to baseline.

5 978 S. AYBEK ET AL. TABLE 3. Neurological and psychiatric data Baseline 3 yr postop stimulation ON PDD PD PDD PD Age (yr) a Age 65 yr old 12/14 (86%) 18/43 (42%) Gender (Nb female) 4/14 (28%) 17/43 (39%) LED UPDRS III OFF UPDRS III ON UPDRS hallucinations 5/14 0/43 a HARS MADRS a P 0.05 comparing PD and PDD. DISCUSSION We found that, in spite of current selection criteria, excluding demented patients from STN-DBS, 28% developed dementia over three-year follow-up, leading to a postoperative dementia incidence between 89 and 98 of 1,000 per year (clinical versus detailed neuropsychological evaluation). Potential confounders, such as depression or anxiety, 13 were excluded by psychiatric examinations and mood scores. These estimated incidences are higher than those reported in other studies with a range between 24 and 46 of 1,000 per year over 3 to 5 years. 1,9,14 An explanation for such differences relies in the average age of these cohorts (55 to 59 years), younger than our (64 year): age being one of the main determinant for the future development of dementia. 15,16 Another possible factor involves the use of different cognitive tests potentially leading to different exclusion criteria for cognition. However, CAPSIT 17 criteria, published after we started our long-term prospective followup, proposed a cut-off score of 120 on the MDRS. This corresponds to a clinical state of dementia that is very similar to what we obtain by applying the DMS-IV criteria for dementia in addition to our extensive battery of neuropsychological tests (patients with a severe deficit ( C5) in more than one domain being excluded). Such variability in the estimates of dementia incidence in PD has also been observed, with similar range, in studies dealing with medically treated patients (42.6 to 112 of 1,000 per year). 8 Even if comparison with community-based studies should be regarded cautiously, such similarities of incidence estimates may suggest that the occurrence of dementia after STN-DBS corresponds FIG. 3. Baseline findings associated with postoperative dementia. Patients who did not develop dementia (PD: black circles) are compared with those who developed it (PDD), at 6 months (white circles) or thereafter (crosses). (A) Baseline Executive function score; (B) Baseline Hallucination score; (C) Age at operation. Although means (black squares) and SD (error bars) lead to significant differences, there is a substantial overlap, preventing discrimination between the two groups. In addition, there is no difference between PDD developing dementia at 6 months and those developing it at 3 years.

6 COGNITIVE PROFILE AND INCIDENCE OF DEMENTIA 979 to the natural evolution of PD rather than as a consequence of the surgery. This is also supported by the stability over 3 years of the neuropsychological profile in patients not developing dementia with only minimal impairment of executive functions as can be expected from the natural evolution of medically treated nondemented PD In contrast to the long term follow-up results, we found that 9.4% of patients developed dementia within 6 months from implantation (36% developing postoperative dementia) representing an incidence of 214 of 1,000 per year, twice as high as the biggest rate ever reported in medically treated patients. Those early-demented patients did not suffer from specific surgery complication (transient confusion or infection) and this raises the question of a direct impact of STN-DBS on cognition, at least in some patients. Such early postoperative increase of dementia incidence has been reported in other series with figures between 5.8% at 13 weeks postop 21 and 30% at 3 months. 6 Direct comparison to medically treated PD has also recently confirmed a larger decline in executive function at 6 months in DBS treated patients, 22 in keeping with our observation. In this latter study, three cases of dementia resulted from surgical complications (2 electrode misplacements, 1 hemorrhage) but the larger decline in DBS treated patients was maintained after exclusion of these patients. These results suggest that the operation itself may trigger the occurrence of dementia in at-risk patients. Such effect might be nonspecific as cognition has been showed to be altered in all types of surgery (not only brain surgery) 23 leading to a transient cognitive decline followed by a recovery at 2 years in healthy elderly populations. With PD leading to progressive cognitive decline in the vast majority of patients, such return to the baseline of the cognitive status cannot be expected in our patients but rather a return to the natural evolution, including persistence of dementia. In addition to such potential effect of the intervention, we found a slow worsening of executive functions in the whole cohort over the 3 years, in agreement with other authors. Saint-Cyr et al. evaluated eleven patients between 3 and 12 months postoperation 5 and demonstrated a negative impact of STN-DBS on executive functions, mostly letter fluency (initiation) and Trail making test (TMT) B (inhibition). Those two functions have been reported to be sensitive to STN-DBS implantation since it might interfere with frontal cortico-subcortical loops 24 as shown by PET studies. 25 However a direct effect of the stimulation itself is still a matter of controversy 22,26 since no differences were seen when stimulation was turned OFF to ON, 3 and 12 months postoperatively for verbal fluency and even an improvement was seen in Stroop test. Here again the interpretation of this executive decline in operated PD should be cautious as medically treated patients may also display such a decline and the operation may have only precipitated it. With such potential precipitating effect of STN-DBS on dementia in at-risk patients, it might be desirable to refine selection criteria to avoid operating patients at risk of such poor cognitive outcome. We were not able to identify predictors of early dementia (6 months) and this might be due to the small number of patients involved (Fig. 3). For long-term dementia (3 years) the patients who will develop dementia were significantly older, scored worse in executive functions and presented more frequently hallucinations preoperatively than those who did not develop dementia. These risk factors are similar to those predicting dementia in medically treated PD. 20,27 However none of them allows discriminating between pre-pdd and PD patients, as substantial overlap exists (Fig. 3). Even the combination of 3 risk factors allowed prediction of only part of the occurrence of dementia when excluding any nondemented PD was avoided. As we found hallucinations, based only on UPDRS definition, having the better discriminating potential, an expanded evaluation of this risk factor seems desirable in future studies. Such potential baseline predictors may help us in the evaluation of the risk for further development of dementia after STN-DBS to better inform patients and caregivers. They should not be understood yet as exclusion criteria, since such criteria should also take into account global evaluation of quality of life and daily function of patients and balance these risks with the benefits from motor improvement. Our study does not include such systematic analysis but 5 of the patients who developed dementia where included in our recently published prospective study 28 ; in contrast to the improvement of quality of life as measured by the Parkinson s disease Questionnaire (39) 29 reported by the other PD patients, the 5 patients who developed dementia reported a trend to worsening of quality of life. In conclusion, the 3-year cognitive follow-up of our longitudinal cohort revealed an incidence of dementia after STN-DBS similar to those reported in medically treated patients. However, 36% of patients developing dementia did so within 6 months from implantation suggesting a precipitating effect of the stimulation either nonspecifically or by interference with the frontal cognitive loops. The identification of predictors (older age, presence of hallucination, and poorer baseline executive functions) may help a better patient selection and information concerning the risk of poor cognitive outcome. However the low sensitivity of the prediction calls for

7 980 S. AYBEK ET AL. further studies to better delineate risk factors for developing dementia after STN-DBS. APPENDIX Cognitive Assessment Language Part I (reading) and Part II (naming) of the animal Stroop test (time recorded) Boston Naming Test (number of items correctly stated)* Praxis Constructive praxis: copy from four ADAS figures (number of items correctly drawn) Ideomotor praxis: four pantomime of object use (number of items correctly done) Visual perception Discrimination of four overlapping figures (Poppelreuter) (number of items correctly stated)* Recognition of five famous faces (number of items correctly stated)* Pointing at the location of five main cities on the map of Switzerland (number of items correctly stated) Calculation Four elementary oral calculations and one written multiplication (number of items correctly stated)* Memory Immediate memory: direct digit span (number of maximum numbers remembered)* Working memory: inverse digit span (number of maximum numbers remembered)* Encoding, retrieval capacities and recall memory capacities: Rey s auditory-verbal learning test (number of items encoded after repeated trials, number of items recognized, number of items spontaneously recalled)* Executive functions Initiation: Verbal fluency: literal (M words) and categorical (animals) (number of items stated in 1 min), Nonverbal fluency: drawing of different figures (number of items carried out in 3 min) Inhibition and planning: Stroop naming part III (time scored), Visual antisaccades (10 stimuli on each side), TMT (part B) (time and error scored), Luria graphic sequences (number of items consecutively correctly carried out). Attention: TMT part A (time scored), Code subtest of the WAIS-III (items produced in 90 ) Reasoning: Similarities subtest of the WAIS-III, Piaget s reasoning task (patient is asked to draw the liquid level of four bottles in different positions). REFERENCES 1. Krack P, Batir A, Van Blercom N, et al. Five-year follow-up of bilateral stimulation of the subthalamic nucleus in advanced Parkinson s disease. N Engl J Med 2003;349: Vingerhoets FJ, Villemure JG, Temperli P, Pollo C, Pralong E, Ghika J. Subthalamic DBS replaces levodopa in Parkinson s disease: two-year follow-up. Neurology 2002;58: Alegret M, Junque C, Valldeoriola F, et al. Effects of bilateral subthalamic stimulation on cognitive function in Parkinson disease. Arch Neurol 2001;58: Ardouin C, Pillon B, Peiffer E, et al. Bilateral subthalamic or pallidal stimulation for Parkinson s disease affects neither memory nor executive functions: a consecutive series of 62 patients. Ann Neurol 1999;46: Saint-Cyr JA, Trepanier LL, Kumar R, Lozano AM, Lang AE. Neuropsychological consequences of chronic bilateral stimulation of the subthalamic nucleus in Parkinson s disease. Brain 2000;123 (Part 10): Dujardin K, Defebvre L, Krystkowiak P, Blond S, Destee A. Influence of chronic bilateral stimulation of the subthalamic nucleus on cognitive function in Parkinson s disease. J Neurol 2001; 248: Trepanier LL, Kumar R, Lozano AM, Lang AE, Saint-Cyr JA. Neuropsychological outcome of GPi pallidotomy and GPi or STN deep brain stimulation in Parkinson s disease. Brain Cogn 2000; 42: Aarsland D, Andersen K, Larsen JP, Lolk A, Nielsen H, Kragh- Sorensen P. Risk of dementia in Parkinson s disease: a communitybased, prospective study. Neurology 2001;56: Funkiewiez A, Ardouin C, Caputo E, et al. Long term effects of bilateral subthalamic nucleus stimulation on cognitive function, mood, and behaviour in Parkinson s disease. J Neurol Neurosurg Psychiatry 2004;75: Aarsland D, Larsen JP, Cummins JL, Laake K. Prevalence and clinical correlates of psychotic symptoms in Parkinson disease: a community-based study. Arch Neurol 1999;56: Assal G. Batterie des examens Neuropsychologiques du CHUV. Division Autonome de Neuropsychologie, Bindschaedler C, Assal G, de Tribolet N. Cognitive sequelae following rupture of aneurysms of the anterior communicating artery and the anterior cerebral artery. Retrospective study of 56 cases. Rev Neurol (Paris) 1997;153: Troster AI, Stalp LD, Paolo AM, Fields JA, Koller WC. Neuropsychological impairment in Parkinson s disease with and without depression. Arch Neurol 1995;52: Rodriguez-Oroz MC, Obeso JA, Lang AE, et al. Bilateral deep brain stimulation in Parkinson s disease: a multicentre study with 4 years follow-up. Brain 2005;128 (Part 10): Athey RJ, Porter RW, Walker RW. Cognitive assessment of a representative community population with Parkinson s disease (PD) using the Cambridge Cognitive Assessment-Revised (CAM- COG-R). Age Ageing 2005;34: Hobson P, Meara J. Risk and incidence of dementia in a cohort of older subjects with Parkinson s disease in the United Kingdom. Mov Disord 2004;19: Defer GL, Widner H, Marie RM, Remy P, Levivier M. Core assessment program for surgical interventional therapies in Parkinson s disease (CAPSIT-PD). Mov Disord 1999;14: Marinus J, Visser M, Verwey NA, et al. Assessment of cognition in Parkinson s disease. Neurology 2003;61: Jacobs DM, Marder K, Cote LJ, Sano M, Stern Y, Mayeux R. Neuropsychological characteristics of preclinical dementia in Parkinson s disease. Neurology 1995;45:

8 COGNITIVE PROFILE AND INCIDENCE OF DEMENTIA Stern Y, Tang MX, Jacobs DM, et al. Prospective comparative study of the evolution of probable Alzheimer s disease and Parkinson s disease dementia. J Int Neuropsychol Soc 1998;4: Morrison CE, Borod JC, Perrine K, et al. Neuropsychological functioning following bilateral subthalamic nucleus stimulation in Parkinson s disease. Arch Clin Neuropsychol 2004;19: Smeding HM, Speelman JD, Koning-Haanstra M, et al. Neuropsychological effects of bilateral STN stimulation in Parkinson disease: a controlled study. Neurology 2006;66: Abildstrom H, Rasmussen LS, Rentowl P, et al. Cognitive dysfunction 1-2 years after non-cardiac surgery in the elderly. IS- POCD group. International Study of Post-Operative Cognitive Dysfunction. Acta Anaesthesiol Scand 2000;44: Schroeder U, Kuehler A, Lange KW, et al. Subthalamic nucleus stimulation affects a frontotemporal network: a PET study. Ann Neurol 2003;54: Schroeder U, Kuehler A, Haslinger B, et al. Subthalamic nucleus stimulation affects striato-anterior cingulate cortex circuit in a response conflict task: a PET study. Brain 2002;125 (Part 9): Pillon B, Ardouin C, Damier P, et al. Neuropsychological changes between off and on STN or GPi stimulation in Parkinson s disease. Neurology 2000;55: Mahieux F, Fenelon G, Flahault A, Manifacier MJ, Michelet D, Boller F. Neuropsychological prediction of dementia in Parkinson s disease. J Neurol Neurosurg Psychiatry 1998;64: Gronchi-Perrin A, Viollier S, Ghika J, et al. Does subthalamic nucleus deep brain stimulation really improve quality of life in Parkinson s disease? Mov Disord 2006;21: Auquier P, Sapin C, Ziegler M, et al. Validation of the French language version of the Parkinson s Disease Questionnaire - PDQ- 39. Rev Neurol (Paris) 2002;158:41-50.