Mild cognitive impairment and dementia

Dementia is a common problem in PD, occurring in up to 80% of patients [1]. In PD, cognitive decline is a significant contributor to functional impairment and is associated with reduced quality of life and increased caregiver burden. The pathophysiology of PD dementia (PDD) is complex and likely multifactorial. While PDD is considered one of the nondopaminergic, long-term complications of PD [2], some studies have shown that striatal dopamine function correlates with executive function and may be most important in the initial stages of cognitive decline [1]. However, it is the degeneration of cholinergic neurons, present early in PD [1], that has been the biological basis for the majority of clinical trials for dementia to date.

Of the acetylcholinesterase inhibitors studied in PD, only rivastigmine is of demonstrated benefit, thus leading to its approval by the US Food and Drug Administration (FDA) for the treatment of PDD. The main trial to study rivastigmine in PDD [3] was a multicenter, randomized (2:1) parallel-group, double-blinded, placebo-controlled trial of 24 weeks’ duration, with 362 individuals with PDD in the treatment arm, and 172 in the placebo arm. Parkinson’s disease dementia was diagnosed based on the Diagnostic and Statistical Manual of Mental Disorders, 4th Edn, Text Revision (DSM-IV-TR; now DSM-V) criteria for PDD [4]. Primary endpoints were the Alzheimer’s Disease Assessment Scale–Cognitive (ADAS-Cog) subscale and the Alzheimer’s Disease Co-operative Study-Clinical, Global Impression of Change (ADCS-CGIC) scale. A mean (±standard deviation) improvement of 2.1 (± 8.2) on the ADAS-Cog subscale occurred in the treatment group, whereas the placebo group showed a worsening of 0.7 (± 7.5) points. Rivastigmine also provided significant benefits over placebo with respect to all secondary efficacy variables. Premature study discontinuation occurred in 131 patients, mostly due to adverse events (27.3% in treatment group and 17.1% in placebo group). Adverse events leading to withdrawal from the study were nausea and vomiting. Of note, while significant differences in motor function were not observed between the treatment and placebo arm, tremor was severe enough to cause withdrawal from the study in 1.7% of patients in the rivastigmine group (and none of the patients in the placebo group). A post-hoc analysis demonstrated that those in the rivastigmine group had significantly better outcomes in basic and high-level-function activities of daily living (ADLs) compared with the placebo group at 24 weeks [5]. Based on the study by Emre et al. [3], in the Movement Disorders Society (MDS) Evidence-Based Medicine Review of treatments for the nonmotor symptoms of PD [6], rivastigmine was considered efficacious for the treatment of PDD. Other findings from this study, namely improved neuropsychiatric symptoms as measured by the 10-Item Neuropsychiatric Inventory (NPI) and lower rates of hallucination in the rivastigmine arm [3], have prompted consideration of rivastigmine for use in the treatment of psychotic symptoms in PD.

Regarding other acetylcholinesterase inhibitors, evidence for donepezil is inconclusive; while several randomized trials for donepezil in PD have been conducted [7–9], the majority had small sample sizes (fewer than 20 per arm). In the largest trial of donepezil to date for treatment of cognition in PDD (diagnosed based on DSM-IV criteria) [10], participants were randomized 1:1:1 to double-blinded treatment with donepezil 5 mg, donepezil 10 mg or placebo. There were two co-primary efficacy end points: (i) change from baseline to week 24 in total score for the ADAS-Cog; and (ii) overall change score at week 24 for the Clinician’s Interview Based Impression of Change with caregiver/study partner input (CIBIC). Secondary endpoints included measures of specific cognitive domains and the NPI. Overall, 550 patients from 108 sites across 13 countries were randomized. The mean change from baseline to week 24 in the ADAS-Cog was not significant for donepezil in the intent-to-treat population by the predefined statistical model. Post-hoc ADAS-Cog analysis, removing the treatment-by-country interaction term from the model, revealed a significant, dose-dependent benefit with donepezil (difference from placebo −2.08, P = 0.002 for 5 mg; −3.31, P < 0.001, for 10 mg). Significantly better CIBIC scores compared with placebo occurred in the 10 mg group (3.7 vs 3.9, P = 0.113, for 5 mg; 3.6 vs 3.9, P = 0.040, for 10 mg). Significant benefits in secondary endpoints of cognition were present, but there was no significant benefit in ADLs. In terms of adverse events, there was a higher incidence of cholinergic-related events in the donepezil-treated groups (e.g. nausea, vomiting and diarrhea) and higher discontinuation rates related to adverse events. There were higher rates of worsening parkinsonism in the donepezil-treated patients, but the difference was not significant, without apparent dose dependency and no impact on the Unified Parkinson’s Disease Rating Scale (UPDRS) motor scale score.

The rationale for examining the utility of N-methyl-d-aspartate (NMDA) receptor antagonists such as memantine in PDD stems from in vitro and animal data suggesting that this class of agents may be neuroprotective, perhaps by interfering with glutamate-mediated excitotoxity [11], as well as the relative success of this agent in the treatment of Alzheimer’s disease [12]. However, evidence for memantine was felt to be insufficient in the MDS Evidence-Based Medicine Review [6], particularly in light of the conflicting evidence found in the randomized control trials (RCTs) investigating it [13, 14]. A possible benefit of memantine on psychiatric symptoms (as assessed by the NPI) has been suggested in subgroup analyses of patients with dementia with Lewy bodies [14] and warrants further consideration.

In terms of other pharmacological classes that may be of utility in PDD, in a controlled study [15], the selective norepinephrine reuptake inhibitor atomoxetine, while not efficacious for the treatment of clinically significant depressive symptoms, was associated with improvements in global cognitive performance and daytime sleepiness. The biological plausibility of this is based on several lines of evidence. There is a putative role of norepinephrine in cognition [16], particularly in relation to prefrontal cortex functioning, including attention and executive abilities. The latter likely accounts in large part for the efficacy of atomexetine for attention-deficit disorder (an FDA-approved indication). Early involvement of the brainstem noradrenergic nucleus (the locus coeruleus) in PD is well established [17]. Thus, this agent may be of utility for treating cognitive dysfunction, particularly of the frontal/executive type [17], in PD, and examination of this question in a RCT may be warranted.

Approximately 25–30% of nondemented PD patients have mild cognitive impairment (PD-MCI), and accumulating evidence suggests that a substantial subset of PD-MCI patients go on to develop PDD [18]. At the time of writing, no large randomized trials for treatment of PD-MCI have been published, although an exploratory, 12-week placebo-controlled RCT showed benefits of the monoamine oxidase B (MAO-B) inhibitor rasagiline on attention and executive functions [19]. There is an ongoing trial examining the impact of rasagiline on PD-MCI [20]. Results of an open-label trial of atomoxetine in 12 nondemented PD patients with executive dysfunction [21] suggest that consideration for study of this drug for the treatment of nonamnestic MCI in PD is warranted as well.

Regarding nonpharmacological therapies, preliminary data suggest that cognitive training or rehabilitation can improve executive abilities [22] and speed of processing [23] in nondemented patients, but it remains unknown whether there are any acute or long-term cognitive benefits to treating PD-MCI patients with these modalities.

Depression

Depression is common in PD patients and has significant consequences including functional disability, higher rates of caregiver stress, decreased quality of life and poorer outcomes [24]. Several RCTs for depression in PD have been published. Selective serotonin reuptake inhibitors (SSRIs) and bupropion are commonly used clinically. While bupropion is used to treat depression in PD and has the advantage of minimal adverse sexual side effects and good tolerability from a motor standpoint, evidence for its efficacy is lacking. On the other hand, several RCTs for SSRIs in PD depression have been conducted, although the size of the trials and other limitations did not provide sufficient evidence for conclusions to be drawn in the last MDS Evidence-Based Medicine Review [6], and an early meta-analysis suggested that the benefit of SSRI treatment is less in PD than in non-PD depression [25]. However, since that time, the RCT Study of Antidepressants in Parkinson’s Disease (SAD-PD) has provided class I evidence for use of one SSRI, paroxetine, and the serotonin and norepinephrine reuptake inhibitor venlafaxine (extended release), in the treatment of depression in PD [26]. This was a 12-week, multicenter study in which 115 PD patients were randomized; of note, the sample size was only half of the projected enrollment. The final analyses included 42, 34 and 39 patients in the paroxetine, extended-release venlafaxine and placebo groups. The mean (± standard deviation) paroxetine dosage at week 12 was 24 ± 11 mg/day, and that of extended-release venlafaxine was 121 ± 75 mg/day. Depression was diagnosed based on DSM-IV criteria, and subsyndromal depression wasbased on the 17-item Hamilton Depression Rating Scale (HDRS-17) score, which was also the primary outcome measure. Secondary outcomes included the Montgomery–Åsberg Depression Rating Scale (MADRS), Beck Depression Inventory (BDI-II), Geriatric Depression Scale (GDS) and NIMH Clinical Global Impression (CGI) scores. Significant improvements in HRDS-17 scores were noted in the paroxetine and venlafaxine group (−13.0 and −11.0 points difference between baseline and 12 weeks, respectively) compared with the placebo group (−6.0 points). Significant beneficial effects of paroxetine and venlafaxine were also noted for the secondary depression outcomes examined. While over 85% of patients in each group reported adverse events, there were no significant differences in overall frequency of adverse events between groups. Insomnia was less common in the paroxetine group compared with the venlafaxine group. Venlafaxine was associated with an 8.5 mmHg increase in seated systolic blood pressure, and paroxetine with a 1.3 kg increase in weight. Importantly, significant differences in UPDRS motor scores were not different between the treatment arms and placebo.

While tricyclic antidepressants (TCAs) are often avoided in older adults because of concerns about side effects (particularly related to anticholinergic activity), several RCTs have investigated the utility of TCAs in PD depression, and both nortriptyline and desipramine were considered likely efficacious in the MDS Evidence-Based Medicine Review [6]. Menza et al. [27] compared nortriptyline with paroxetine in a double-blinded, randomized, placebo-controlled trial. Eighteen PD patients with depression or dysthymia as diagnosed by DSM-IV criteria received paroxetine, 17 received nortriptyline and 17 received placebo. The primary endpoint was the HDRS-17 score. Overall, 65% of patients completed the study, with no between-group differences in discontinuation rates. Nortriptyline was superior to placebo regarding change in HDRS-17 score, while paroxetine was not. Both medications were well tolerated, with no differences in study discontinuation rates and cognitive outcomes. However, paroxetine, but not nortriptyline, was associated with a higher average number of side effects than placebo, including fatigue and orthostatic hypotension. In another RCT comparing citalopram and despiramine in 48 PD patients [28], the study completion rate was 94%. The primary endpoint was MADRS scores, and both treatment arms showed significant improvement in overall MADRS scores compared with placebo. Adverse events were twice as common in the desipramine group.

Among other considerations, the occurrence of “off”-time depressive and other NPSs as manifestations of the PD “off” period prompted investigation of dopamine agonists for the treatment of depression in PD, and efficacy has been demonstrated. In a 12-week, randomized, placebo-controlled study of pramipexole [29], the 139 patients randomized to the pramipexole group who were included in the final analysis had significant improvements in BDI (the primary endpoint) and UPDRS scores compared with the 148 in the placebo group. Importantly, secondary analyses (i.e. path analysis) showed that approximately 80% of the treatment effect was due to the effect of pramipexole on depressive symptoms rather than on motor symptoms. Adverse events were not significantly more common in the treatment group. Pramipexole was also compared with sertraline in a 14-week, randomized trial using the HDRS-17 as the primary endpoint, with similar results [30]. While there are some data to suggest that ropinirole improves depression in PD [31], no RCTs with depression as a primary endpoint have been conducted for this agent, and there are insufficient data to ascertain whether the antidepressant effect of dopamine agonists in PD is a class effect or is specific to pramipexole.

Nonselective monoamine oxidase inhibitors (MAOIs) were previously widely used for treatment of depression in the general population, but their used decreased dramatically as safer drugs were developed. Subsequently, rasagline, a selective MAOI that specifically targets the B-isoform, was developed and is widely used in PD as an adjunct to levodopa (l-DOPA) therapy [32]. Preliminary data suggest that rasagiline may have antidepressant effects that are independent of its motor benefit [33]. The ADAGIO (Attenuation of Disease Progression with Azilect Given Once-daily) study [34] was a double-blind, placebo-controlled, delayed-start trial of rasagiline in de novo PD. In an analysis of patients taking an antidepressant any time during the 36-week phase 1, in which patients were randomized to rasagiline (1 or 2 mg/day) or placebo, NPSs were assessed by the MDS-UPDRS-experiences of daily living (EDLs), original UPDRS and Parkinson’s Fatigue Scale (PFS). Sixteen percent (191/1174) of patients were treated with antidepressants, approximately 75% with an SSRI, during phase 1. The EDL depression and cognition item scores improved significantly in the rasagiline group compared with the placebo group (P = 0.045 and P = 0.002, respectively). The PFS (P < 0.001) and daytime sleepiness (P = 0.006) scores also improved significantly in the rasagiline group compared with placebo. The effect on depression remained significant after controlling for improvement in motor symptoms (P = 0.009). There were no serious adverse events in the combined rasagiline and antidepressant group suggestive of serotonin syndrome. Thus, rasagiline augmentation of antidepressant treatment was associated with improvement in mood, cognition, fatigue and daytime sleepiness in de novo PD. These findings suggest a role for the dopamine system in clinically significant NPSs in early PD (D. Weintraub, unpublished data).

Nonpharmacological treatments for depression, including cognitive behavioral therapy (CBT), may be as efficacious and preferred by patients [35]. Given the relative burden of frequent in-person visits, studies on telephone- and web-based administration of CBT are also being pursued. Another nonpharmacological interventions for depression for which sufficient data are lacking but which is a promising avenue of active investigation is transcranial magnetic stimulation [6]. Regarding the effects of deep-brain stimulation (DBS) on depression in PD, in a randomized trial examining the effects of subthalamic nucleus (STN)- versus globus pallidus interna (GPi)-DBS, the primary outcome was motor improvement, but the BDI was examined as a secondary outcome. The BDI scores [36] improved by 0.6 points and increased by 1.3 points in the GPi and STN arms (P = 0.02). This has led some to wonder whether the NPS outcomes might be better with GPi versus STN lead placement for DBS surgery.

Psychosis

Psychosis is uncommon in early and untreated PD but may occur in up to as many as 60% of patients over the disease course [37]. It is particularly associated with older age and dementia [37]. As with other NPSs, psychosis in PD causes significant caregiver burden and increases the probability of nursing home admission [38, 39]. The most common manifestations are visual hallucinations, but other symptoms include illusions, hallucinations in other sensory modalities and delusions. The pathophysiology of psychosis in PD is complex and multifactorial, but a contribution from both the underlying neurodegenerative disease process as well as exogenous dopaminergic medication exposure is probable. To this end, the majority of treatment trials for PD psychosis have targeted the dopaminergic system alone or in combination with the serotonergic system, although agents targeted selectively at the serotonergic system are being investigated as well, as detailed below.

The strongest evidence for treatment of psychosis in PD exists for clozapine, which has activity at both dopaminergic and serotonergic receptors. The PSYCLOPS (PSychosis and CLOzapine in PD Study) [40] was a 4-week, placebo-controlled, double-blinded RCT that examined clozapine in 60 PD patients. Primary outcome measures were the CGI for psychosis and the UPDRS for parkinsonism; the Brief Psychiatric Rating Scale (BPRS) and Scale for the Assessment of Positive Symptoms (SAPS) were also examined. In total, 54 participants completed the trial. At a mean dose of 24.7 mg/day for those in the clozapine group resulted in significantly more improvement in CGI, BPRS and SAPS scores. There was no significant worsening of motor symptoms in either group and, in fact, a significant beneficial effect on tremor was noted in the clozapine group. In one patient, clozapine was discontinued due to leucopenia. A 4 beats/min increase in heart rate and a 0.7 kg increase in weight were noted in the clozapine group, but otherwise side effects and adverse events were no different between the groups. Based on the results of the PSYCLOPS and two reports based on 4-week, open-label extensions of it, the MDS Evidence-Based Medicine Review [6] determined that clozapine is efficacious for the treatment of psychosis in PD. On the other hand, the evidence for quetiapine was considered insufficient based on the presence of conflicting data and several methodological concerns of the studies published, including small sample sizes and low-quality ratings [6]. Despite this, quetiapine is the most widely used antipsychotic in PD, prescribed in over two-thirds of cases [41], whereas clozapine accounts for less than 2% of prescriptions, likely due to the burden of blood monitoring that is required [41]. Of note, regarding other atypical antipsychotics, based on available evidence, olanzapine was deemed unlikely to be efficacious [6] and is thus recommended against. An open-label study [42] of aripiprazole in 14 PD patients with psychosis has raised concerns that the motor side effects of this drug may make it unsuitable for use in PD as well.

Given the significant safety concerns regarding the use of atypical antipsychotics in PD patients, including worsening of parkinsonism and increased mortality [43], alternative treatments are clearly needed. A promising agent in this regard is the selective serotonin 5-HT2A inverse agonist pimavanserin [44]. In a 6-week, placebo-controlled RCT of pimavanserin, the primary outcome was change in the total Parkinson’s Disease-adapted Scale for Assessment of Positive Symptoms (SAPS-PD) score from baseline to day 43. Importantly, the assessment was conducted by a centralized blinded rater. In an attempt to minimize confounding by the placebo response, after screening and before baseline, participants entered a 2-week lead-in period during which nonpharmacological brief psychosocial therapy adapted for PD was administered. Overall, 199 patients were randomized and 185 were included in the final analysis. The 95 randomized to the treatment arm had significant improvements in SAPS-PD scores compared with the 90 in the placebo arm (difference of −3.06). Significant improvements in CGI and caregiver burden were also seen. Of note, an improvement in nighttime sleep (a secondary outcome assessed by the Scale for Outcomes of Parkinson’s disease [SCOPA]-sleep scale), without worsening in daytime sleepiness, was found. No significant differences in motor function were observed between the treatment groups. Serious adverse events occurred in 11% of the pimavanserin arm and in 4% of the placebo arm.

As mentioned above, consideration of rivastigmine for use in the treatment of psychotic symptoms in PD has been prompted by lower rates of hallucinations among patients randomized to a rivastigmine treatment arm in a trial of that agent for treatment of PDD [3].

Others

Impulse control disorders, including compulsive gambling, buying, sexual behavior and eating, occur in 13.6% of patients with PD [45], and are the result of medications (and not the underlying disease itself) [46]. Thus, while a reduction in the dosage or cessation of the inciting medication, usually a dopamine agonist, is the cornerstone of therapy for ICDs, given the symptomatic benefits seen from dopaminergic medications, pharmacologic treatments of ICDs have been pursued. A 17-week, double-blind, placebo-controlled, crossover RCT of amantadine was considered by the MDS Evidence-Based Medicine Review [6] to provide insufficient evidence based on the small sample size (n = 17, with five patients dropping out) and other methodological issues. The other pharmacological agent being investigated in an RCT for treatment of ICDs in PD is naltrexone. In this study, PD patients (n = 50) with an ICD were enrolled in an 8-week, randomized (1:1), double-blinded, placebo-controlled study of naltrexone 50–100 mg/day (flexible dosing). The primary outcome measure was remission based on the Clinical Global Impression of Change (CGIC) score, and the secondary outcome was change in symptom severity using the Questionnaire for Impulsive–Compulsive Disorders in Parkinson’s Disease – Rating Scale (QUIP-RS) ICD score. Forty-five patients (90%) completed the study. There was no significant between-group difference in remission status (odds ratio = 1.6, 95% confidence interval [CI] = 0.5–5.2, Wald χ² [d.f.] = 0.5 [1], P = 0.5), but naltrexone treatment led to a significantly greater decrease in QUIP-RS ICD score over time compared with placebo (regression coefficient for interaction term in linear mixed-effects model = −0.9, F [d.f.] = 4.3 [1.49], P = 0.04). Cohen’s effect size (d = 0.61) indicated a medium to large treatment effect on the QUIP-RS ICD score (D. Weintraub, unpublished data). An RCT of CBT for impulse control behaviors in PD [47] shows the promise of this modality, and larger trials are needed.

Anxiety is common in PD, occurring in approximately 40% of patients. While there are no RCTs for treatment of anxiety in PD, anxiety was a secondary outcome in four RCTs examining treatments for depression in PD (two of the antidepressants discussed above [27, 28], one of atomoxetine [15] and one of CBT [35]). A pooled analysis of these trials [48] showed that the pooled effect size for antidepressants on anxiety in PD was large (d = 1.13) but nonsignificant (95% CI = −0.67 to 2.94). Subgroup analyses were conducted, and a large and significant effect size of TCAs (d = 1.40, 95% CI = 0.09–2.70) but not SSRIs (d = 0.85, 95% CI = −0.40 to 2.09) was found. Overall, TCAs, atomoxetine and CBT all demonstrated a significant and large secondary effect on anxiety outcomes [48].

Apathy is also a common NPS, with approximately 40% of patients affected. In a 6-month, double-blinded, placebo-controlled RCT [49], the effect of rivastigmine on apathy in 30 nondemented, nondepressed patients with PD was assessed. The primary end point was change in the Lille Apathy Rating Scale (LARS), and measures of functional impact and quality of life were secondary endpoints. There was a significant reduction in apathy in the treatment group: 37% of the patients in the rivastigmine group and 83% of the patients in the placebo group were classified as apathetic at 6 months. Significant improvements in functional ability (but not quality of life) were also noted. This was independent of effects of rivastigmine on cognition. Regarding other classes of drugs that are of promise in the treatment of apathy in PD, the selective D2/D3 agonist piribedil has also been preliminarily investigated in a small, placebo-controlled, randomized study [50] of PD patients with apathy following withdrawal of dopamine agonist medications post-DBS and shows promise.

Patient population

Increasingly, studies are using clinical characteristics, biomarkers or genetic profiles to subtype patients who may have a differential response to a specific treatment, potentially leading to an increased likelihood of demonstrating an active-treatment effect. This attempt to homogenize the study population in some way, and minimize the variability in response, has been applied in studies of psychiatric conditions such as alcohol dependence. In terms of study design, patients can be enrolled simply on the basis of having a particular characteristic, or the study population can be stratified on that characteristic. A potential example of this for NPSs in PD relates to the finding from secondary analyses that patients in the PDD rivastigmine placebo-controlled study [3] who experience psychosis at baseline had a more robust response to rivastigmine treatment compared with patients without psychosis. Thus, future cognitive studies in PDD patients might consider focusing the study population on those patients with psychosis in addition to having PDD, or stratifying the study population on the basis of having psychosis.

Table 30.1

Key challenges and possible solutions in clinical trials for neuropsychiatric symptoms (NPSs) in Parkinson’s disease (PD)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Oral dopamine agonists in the management of Parkinson’s disease Management of disease-related behavioral disturbances in Parkinson’s disease A neurobehavioralist approach to the management of cognitive impairment in Parkinson’s disease Lessons learned: symptomatic trials in early Parkinson’s disease Lessons and challenges of trials for other nonmotor complications of Parkinson’s disease Lessons and challenges of trials involving ancillary therapies for the management of Parkinson’s disease

Stay updated, free articles. Join our Telegram channel

Join