Cholinesterase Inhibitors to Slow Progression of Visual Hallucinations in Parkinson&Apos;s Disease

Overview

Rationale: Visual hallucinations (VH) are the most common non-motor symptoms in Parkinson's disease (PD). As an independent predictor for cognitive decline and nursing home placement they form an important disability milestone in the course of PD. According to current clinical guidelines minor VH do not require treatment per se. But as minor VH precede the stage of major VH without insight and PD associated psychosis (PDP) they offer an opportunity for early intervention. Neuroleptic drugs delay the transition into PDP but are unsuitable for early treatment of VH due to their side effects. We hypothesize that cholinesterase inhibitors (ChEI) are a well-tolerated alternative for the early treatment of minor VH to delay the progression to PDP, and that brain network analysis is suitable to predict treatment response. Objective: Investigate whether early treatment with ChEI delays the progression of minor VH to major VH without insight or PDP. In addition, we will measure motor control, psychotic symptoms, cognitive impairment, mood disorders, daytime sleepiness, adverse events and compliance, disability, caregiver burden and care use. We assess the cost-effectiveness of early chronic treatment of VH with ChEI. Finally, we analyse changes of functional brain networks before and during treatment. Study design: A randomized, double blind, placebo-controlled, multi-center trial with an economic evaluation. Study population: 168 patients with PD and VH after fulfilling the in-and exclusion criteria. Intervention: Rivastigmine capsule 6 mg BID or placebo BID for 24 months. Main study parameters/endpoints: The primary outcome measure is the median time until PD patients with minor VH progress to major VH without insight. The clinical endpoint is defined as the start with antipsychotic treatment. Secondary outcome measures are changes in motor control, psychotic symptoms, cognitive impairment, mood disorders, daytime sleepiness, cholinergic deficiency, the number of adverse events, compliance, disability and caregiver burden. The median time until PD patients with minor VH progress to PD dementia is measured by means of changes in cognitive function. The secondary neurophysiological outcome measures are peak frequency, functional connectivity, topological network organisation and the direction of information flow. All relevant costs will be measured and valued. Nature and extent of the burden and risks associated with participation, benefit and group relatedness: The burden of participation consists of a total of 5 clinical visits (every 6 months), 5 telephone interviews on adverse events during the escalation phase and 9 questionnaires on health related costs (every 3 months). In a subgroup 3 additional visits for EEG recording are needed. There is a risk for adverse reactions with rivastigmine treatment; the most common are nausea and vomiting.

The study is performed in four regional study centers: i.e. VUmc-AMC Amsterdam, Atrium MC Heerlen, UMC Groningen en Radboudumc Nijmegen. Each regional study center has a participating neurologist and will station a research nurse - of which some part-time. VUmc-AMC is also national coordinating center and their research nurse is also national trial manager. Because the logistics are complex and the assessments are cumbersome, it is vital that research nurses perform these to facilitate inclusion and to assure proper follow-up assessments. All research nurses will be trained and examined in performing all aspects of the Visits, i.e. explaining the study and using the questionnaires. Every twelve months, the research nurses will follow additional training to keep their expertise up to date and to minimize intra-rater and inter-rater variability. In addition to the four study centers, approximately 30 hospitals are requested to participate as local recruitment center. Eligible patients are recruited by the treating neurologist in these recruitment centers during a regular follow-up appointment at the outpatient clinic. In addition, the treating neurologist can send a recruitment letter to a selected group of potentially eligible patients based on data from the patient files or a clinical database. The recruitment letter will give a short introduction on the topic of visual hallucinations and will underline the importance to participate in this study. Patients that are interested are encouraged to contact the research nurse for information or to make an appointment. The treating neurologist will check the inclusion and -exclusion criteria. He asks the patient for permission to send contact information to the research nurse. If the patient is eligible and agrees, the neurologist will complete the patient registration form with name, gender, date of birth, telephone number and in- and exclusion criteria. After registration, a research nurse will contact the patient by phone within 10 working days. She will introduce the study to the patient, inform the patient on the gross outline of the trial and discuss the possible benefits and disadvantages. She will answer any additional question about the study. If the patient agrees an appointment will be made (visit 1). All study visits will take place at home or in the outpatient clinic of the regional study center. The study medication is prescribed by the participating neurologist in one of the four regional study centers. During Visit 1 Informed Consent will be obtained. The baseline characteristics will be recorded, as well as medication history, year and type of first PD symptoms (e.g. tremor, bradykinesia) and year of PD diagnosis. Current medication is noted and the Levodopa Equivalent Daily Dose (LEDD) will be calculated. Visual acuity is tested with a digital version of the Snellen chart (iPad). During Visit 1 patients will undergo several assessments by the research nurse. During Visit 1 or any other visits (2,3,4 or 5) two 6 mL EDTA blood samples for DNA analysis will be drawn and stored for future genetic research; including possible genetic modifiers of response to medication. At the end of the visit the research nurse will randomize the patient by a central web-based computer program. Medication is sent by mail through a central pharmacy during the complete study period. The treatment should start as soon as possible after randomization and at least within 6 weeks after the baseline assessment. The research nurse inquires the patient about the medication received, medication intake and side effects by telephone after 2,3,5,8 and 11 weeks of treatment. The interviews will be taken according to schedule or as close as possible. Only missing interviews after a dose escalation step will be considered a protocol violation. Four specified assessment visits will take place after Visit 1: at 6 months, 12 months, 18 months and 24 months (respectively Visits 2, 3, 4 and 5; see table 1). The visits will be completed according to schedule or as close as possible, but at least within 6 weeks after the planned date. The long term outcome measures are unlikely to change significantly within this period of time. The assessments of Visit 1 are repeated by the research nurse and completed with a monitor for side effects, any change in (non-parkinsonian) medication and calculation of the LEDD. Patients are asked to bring the remaining number of capsules. These will be counted as a measure of compliance. Patients that have agreed to participate in the additional CHEVAL-in study will visit the outpatient clinic of VU University Medical Center to perform and EEG within approximately 4-6 weeks after baseline assessment and before start on study medication. The EEG recording will be repeated after 6 and if possible 12 months. The recordings will be completed according to schedule or as close as possible, but at least within 6 weeks after the planned date. If the criteria for PD psychosis (primary outcome) or PD dementia (secondary outcome) are met, the study medication and the assessment scheme are altered. The treating neurologist reports the event to the local research coordinator in one of the four main study centers, using the trial mutation form. This form includes the date and the primary outcome measure; that is the UPDRS-1 MDS score (3 or 4 on the hallucinations item, indicating that insight is lost) and the reason for withdrawal of study medication (PD psychosis or PD dementia). If the condition of PD psychosis is reached (primary endpoint), the experimental treatment with rivastigmine or placebo is discontinued. Because this is a medical emergency, the research nurse will plan the next clinical visit as soon as possible, but not before the subject's condition allows participation and not later than 6 weeks from the end of treatment. During this clinical visit (visit '9') the research nurse will follow the same protocol as with the regular visits. Planned visits are cancelled after the primary endpoint is reached and experimental treatment is discontinued but if possible, patients are visited once more after 24 months. Costs will continued to be measured every 3 months until 24 months from the start of treatment. If the condition of PD dementia is reached (secondary endpoint), treatment with rivastigmine is advised according to current PD treatment guidelines. The patient is instructed to stop taking the blinded study medication and will start rivastigmine that is prescribed by the treating neurologist. Using the trial mutation form, the treating neurologist can motivate his decision to switch to open label rivastigmine. For safety, a new dose escalation phase is necessary. The consequences of any important event are discussed with the participating neurologist of the regional study center and if necessary with the project coordinator or project leader of the national coordinating center to decide if a patients needs to withdraw from the study and whether a final clinical visit must be completed. Both the treating neurologist of the recruitment center as the participating neurologist of the participating center are informed about this decision. Withdrawal of individual subjects: Subjects can leave the study at any time for any reason if they wish to do so without any consequences. Specific criteria for withdrawal are not predefined. The investigator can decide to withdraw a subject from the study if the treating neurologist points out urgent medical reasons, which make it necessary to withdraw the patient from the study. Subjects can also be withdrawn in case of protocol violations. The datasets from withdrawn patients will be kept in the study database to facilitate analysis according to the intention-to-treat principle. Replacement of individual subjects after withdrawal: Subjects will not be replaced. Drop out has been accounted for in our power calculation. Moreover the analysis will be done according to the intention-to-treat principle. Follow-up of subjects withdrawn from treatment: If a patient violates the study medication protocol this will be registered. All further study procedures and measurements will be conducted according to the study protocol. Randomization, blinding and treatment allocation: After inclusion, the patient will be randomized by the web based database. The server of the website will operate from the VU University Medical Center. Eligible patients will be randomized by a computer in a 1:1 ratio to the rivastigmine group or the placebo group in a double-blind design. Until the computerized randomization is fully operational, patients will be randomized by the research nurse or, in her absence, by T.J.M. van Mierlo or E.M.J. Foncke. Randomization will be stratified by type of hospital (University Medical Center versus Non-University Medical Center), age (below 65 years or 65 years and older), and disease duration (\<10 years, \>10 years) using variable permuted blocks. Code-breaking sheets for emergency use will be kept in a safe. Study personnel, research nurses, neurologists, and the patients are blinded to the treatment allocation at all times. All data will be entered in the central database before the treatment codes are broken. The randomization code can be broken in case of an emergency.