Cognition endpoints for Parkinson's disease: A roadmap for developing patient-focused assessment tools to accelerate drug development

Abstract

Cognitive dysfunction is a prevalent and impactful feature of Parkinson's disease (PD), with substantial heterogeneity in onset, severity, and progression, and a high lifetime risk of dementia. Despite its clinical importance, effective treatments and fit-for-purpose outcome measures remain limited. This article reports outcomes from a multi-stakeholder roundtable convened by The Michael J. Fox Foundation for Parkinson's Research, outlining priorities to advance patient-focused endpoints and improve clinical trial design to accelerate therapeutic development for cognitive dysfunction. In the context of the current understanding of cognitive impairment across the PD/dementia with Lewy bodies (DLB) spectrum and existing neuropsychological assessment approaches, recommendations are made regarding necessary literature reviews, including qualitative data and related conceptual models, disease models including pathophysiology and neuropsychological profile, and clinical outcome assessment (COA) reviews, including content validity, psychometrics, and fitness-for-purpose. Additionally, novel conceptual models of cognitive impairment across the normal, mild cognitive impairment (MCI) and dementia spectrum are needed and may require further qualitative evidence.

Plain language summary

Many people with Parkinson's disease experience problems with thinking, memory, and decision-making. These problems can start at different times, vary in how severe they are, and may worsen over time, with many people eventually developing dementia. Despite how much these symptoms affect people in their daily life; there are few effective treatments and limited ways to measure changes that truly matter to patients. This article explains what is currently known about thinking problems in Parkinson's disease and related conditions, considers tests used to assess them when developing new treatments, and identifies important gaps in understanding. It also summarizes recommendations from a meeting led by The Michael J. Fox Foundation to improve patient-focused measures and make clinical trials more meaningful and effective.

The importance of cognitive dysfunction in Parkinson's disease

Cognitive changes are among the most worrisome symptoms associated with Parkinson's disease (PD). ¹ Treatments to address them remains an area of high unmet need because of their significant impact on daily function, interpersonal relationships, and quality of life for people with PD (PWPD) and their care partners.^2–4 PD dementia (PDD) represents a severe and disabling stage of the disease in which functional independence is lost.^5,6 Importantly, there is substantial heterogeneity in the prevalence, severity, cognitive profile, and rate of progression of cognitive dysfunction in PD. Estimates indicate 20–40% of PWPD have mild cognitive impairment (MCI) at an early stage or within two years of symptom onset, whilst overall prevalence has been estimated at 40%, with a life-long prevalence of 80% for dementia. ⁷ Additionally, cognitive differences have also been observed in prodromal or at-risk groups e.g., isolated REM sleep behavior disorder (iRBD). ⁸ Rates of progression are variable, but a majority of PWPD surviving greater than 10 years disease duration will develop dementia, and those with GBA-PD are at risk of more rapid progression, while many of those in the early stages of PD may show little to no impairment or worsening in the near term.^9–11 With the recent advances in biological definition and clinical staging of PD and dementia with Lewy bodies (DLB), DLB is also recognized as sharing both underlying neuropathology and aspects of clinical presentation, including prominent cognitive changes. ¹²

Due to the importance of cognitive dysfunction in PD, DLB and related disorders, since 2024 The Michael J. Fox Foundation for Parkinson's Research (MJFF) has brought diverse stakeholders together, including academic, industry, and nonprofit representatives, to discuss the challenges and opportunities of outcome assessment and trial design for cognitive dysfunction in PD. This article reports on some of the proceedings of this dialogue, summarizes the current state of outcome measurement for cognitive impairment and outlines the importance of multi-stakeholder collaboration to promote patient-focused tools to accelerate drug development.

Current understanding

Several qualitative studies have gathered preliminary evidence on cognitive signs and symptoms of importance to PWPD. As noted, cognitive differences may be present in prodromal; or at-risk groups including RBD and hyposmia, whilst in early PD (i.e., Stage 1 or 2 on the Hoehn and Yahr (H&Y) scale), cognitive symptoms are already reported as common non-motor symptoms, with a broad range including difficulty with memory, concentration, and word finding. In more advanced PD (i.e., H&Y ≥ 2.5), cognitive impairment was the sixth and fourth most bothersome non-motor manifestation reported by PWPD and caregivers, respectively, and clinicians identified it as the most important to treat. ¹³ In other work in GBA-PD and idiopathic PD (iPD) (H&Y 1-3) memory loss was commonly reported as either the most bothersome or most important to treat symptoms. ¹⁴

Diagnostic criteria to identify PD-MCI via neuropsychological testing have defined a two-level approach. ¹⁵ Level I (Abbreviated Assessment, Possible PD-MCI) requires impairment on a scale of global cognitive ability validated for use in PD or on a limited battery of neuropsychological tests (e.g., only one test per cognitive domain or fewer than five cognitive domains). Level II (Comprehensive Assessment) requires formal, comprehensive neuropsychological testing that includes at least two tests for each of the five cognitive domains (attention and working memory, executive, language, memory, and visuospatial). These domains were selected based on literature evaluating the presence of impairment on cognitive/neuropsychological tasks in PD patients and their ability to predict progression to PDD. ¹⁶

Items assessing cognition are commonly included within clinical outcome assessments (COAs, Table 1) such as the MDS-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) item 1.1, and health related quality of life assessments (e.g., the Parkinson's Disease Questionnaire), but these may not be appropriate for clinical trials targeting cognition, particularly longitudinal trials where the ability for participants to reliably self-report may change along with cognitive decline. Thus, approaches to the measurement of cognition for clinical trials have tended to focus on neuropsychological tests, i.e., performance-based tasks or ‘PerfO’ assessment; and the International Parkinson and Movement Disorder Society (MDS) is developing guidelines for assessing cognitive abilities, including recommendations for COAs based on PerfOs. ¹⁷ This has included work to evaluate both global assessments (e.g., MoCA, DRS-2, and PD-CRS) and individual neuropsychological tasks.^17–19 Composites of well-established, neuropsychological tests have recently been proposed to measure cognitive change in PD. The Parkinson's Disease Composite of Executive Functioning (PaCEF) has detected change over time within a cohort of PD-MCI and PDD patients. ²⁰ A second composite, the Cognitive Summary Score (PD-CSS) has good discriminant validity and measures a broader range of cognitive functions by including tests of episodic memory, visuospatial function, information processing speed, executive function, working memory, and language. ²¹ Whilst assessments developed for Alzheimer's disease (AD) have been employed in some trials, the distinct cognitive as well as clinical profile of PD suggests they may be unsuitable due to the confounding impact of other PD symptoms (e.g., motor) on activities of daily living (ADL).^22,23 Other assessments have been developed specifically for PD and which seek to measure cognition dependent function. The Parkinson's disease – cognitive functional rating scale (PD-CFRS) is a patient-reported (PRO) or informant-reported (ObsRO) instrument focused on how cognitive difficulties have hindered daily life activities, specifically avoiding motor problems, with items derived via an expert clinical panel. ²⁴ Evidence for reliability and psychometric validity in PD patients with normal cognition and PD-MCI have been reported. The Penn Parkinson's Daily Activities Questionnaire (PDAQ) is an ObsRO assessment intended to measure cognition dependent instrumental activities of daily living in PD. Developed via literature review and focus groups of patients and informants, the assessment has also been psychometrically validated, including a short-form (PDAQ-15) in PD patients in normal, PD-MCI, and dementia stages of cognition.^25,26 Importantly, for both PD-CFRS and PDAQ, gaps may exist against FDA Patient-Focused Drug Development guidance, particularly in respect of content validity, and the field may be lacking the comprehensive and representative input needed to establish what is important to patients, which must underpin the item content of clinical trial assessments.

Table 1.

Clinical outcome assessment glossary.

Glossary	Abbreviation	Definition
clinical outcome assessment	COA	A clinical outcome assessment describes or reflects how a person feels, functions, or survives and can be reported by a patient (through PROs), a non-clinical observer (such as a care partner) (ObsROs), a clinician (ClinROs), or through performance of an activity or task (PerfOs).
patient-reported outcome	PRO	A measure of a patient's health status as reported directly from the patient without added interpretation of the patient's response by a clinician or anyone else. A PRO can be measured by self-report or by interview provided that the interviewer records only the patient's response. Symptoms (e.g., functional slowness, fatigue) or other unobservable concepts known only to the patient (e.g., pain) can only be measured by PRO measures. PROs can also assess the patient perspective on functioning or activities that may also be observable by others.
observer-reported outcome	ObsRO	An assessment of observable signs, events or behaviors related to a patient's health condition as reported by individuals who observe the patient in daily life, like family members or care partners. ObsROs are particularly useful for patients who cannot report for themselves (e.g., individuals who are cognitively impaired). An ObsRO measure does not include medical judgment or interpretation.
clinician-reported outcome	ClinRO	A report coming from observable signs, behaviors, or other manifestations, which may be gathered from clinical interviews and or examinations related to a disease or condition (e.g., limb rigidity). ClinRO measures involve a clinical judgment, or interpretation and cannot directly assess symptoms that are known only to the patient.
performance outcome	PerfO	A measurement based on standardized task(s) actively undertaken by a patient according to a set of instructions (e.g., performing a neuropsychological test). A PerfO assessment may be administered by an appropriately trained individual or completed by the patient independently.

Drug development landscape

At present, the only approved treatment for cognitive impairment associated with PD is rivastigmine for PDD, a cholinesterase inhibitor approved in 2006 and there are no approved disease modifying treatments, or symptomatic therapies for cognition prior to dementia. Consequently, there is interest in the ability of disease modifying treatment to slow or prevent cognitive impairment along with motor and other non-motor symptoms, and in additional symptomatic treatments with mechanisms targeting cognition. Multiple mechanisms of cognitive decline in PD have been proposed, including α-synuclein aggregation, mitochondrial function, synaptic plasticity, gut–brain axis, neuroinflammation, neurotrophic factors, and impaired neurotransmitter systems (e.g., cholinergic, dopaminergic, serotoninergic and noradrenergic). ²⁷ According to the latest review of the clinical trial pipeline, of the 53 symptomatic trials evaluating targeted symptoms (i.e., other than general motor and/or non-motor symptoms), 9 were evaluating cognitive function, making it the third most common target. ²⁸

Gaps in understanding

Despite the interest in cognition, aspects of the patient experience are not well understood. Key gaps for drug developers include: an understanding of both severity and frequency of cognitive difficulties; their evolution over time; the impact on patients in respect to how they feel or function in daily life; the influence of additional factors such as use of aids or assistance, fatigue or stress; and the interplay between cognitive and motor function and its impact on activities of daily living, including falls. Further work may help to differentiate these aspects, guide decisions regarding COA type, and support modification or development of functional (e.g., activities of daily living) or health-related quality of life (HRQoL) assessments.

Important too may be the use of neuropsychological assessment data to guide and support qualitative research. Patient experience may differ from findings generated via neuropsychological assessment to establish presence and severity of cognitive impairment. PWPD and care partners may have difficulty identifying cognitive deficits and quantifying their severity, as well as potentially misidentifying deficits (i.e., exhibiting a tendency to report deficits as memory impairment that upon neuropsychological testing may result from executive function, or attention/working memory impairment).^29,30 This understanding needs to be generated across the course of the disease to account for progression that might alter the meaningful aspects of health that are targeted for measurement, as the approach to measurement must consider both the severity of cognitive impairment of recruited patients as well as trial duration/expected rate of progression. Additionally, related psychometric evidence and control or normative data will be needed from appropriate comparison groups.

As noted, substantial focus to date has been placed on the development and validation of neuropsychological test based (PerfO assessment) COA approaches. Whilst these may be critical for the understanding of severity, and some key concepts, there is an absence of well-developed COAs (i.e., PRO, ObsRO, or ClinRO) that might assess frequency and impact, particularly as it relates to relevant ADLs, rather than the tasks that are typical of PerfO assessments and which are not activities undertaken in daily life. Further to this, wider considerations regarding whether COAs are fit-for-purpose, as well as related elements of trial design must be addressed, including heterogeneity of clinical presentation and demographic characteristics such as age, concomitant medication including PD treatments, selection of the best reporter/COA type(s), impact of premorbid level of function, and specific confounds such as measurement of bradyphrenia Vs motor slowing or aphasia Vs dysarthria. It should be recognized that development of novel assessments is lengthy and in preparing the field for both short and long-term needs, careful consideration must first be given to use and adaptation of existing instruments, potentially including novel composites that may be validated with existing data.

As with clinical drug development for AD, endpoint models likely need to support the demonstration of treatment benefit for both cognition and daily function, recognizing regulatory concerns regarding the inability of indirect neuropsychological assessment to measure ADLs, challenges with PROs where there may be loss of ability to reliably self-report, and the presence of early cognitive changes without measurable functional impact (Table 2). ³¹

Table 2.

Illustrative endpoint models for cognition focused clinical trials.

Context of use	Domain	Endpoint hierarchy	Clinical interpretation
Treatment of early/prodromal cognitive changes associated with PD	Cognition (e.g., PerfO)	Primary	Evidence of specific effect on core clinical feature
	Global Clinical Assessment or Functional Assessment (e.g., ObsRO, ClinRO)	Secondary	Evidence of clinical meaningfulness
Treatment of mild cognitive impairment in PD (PD-MCI)	Cognition (e.g., PerfO)	Co-primary or integrated	Evidence of specific effect on core clinical feature
	Global Clinical Assessment or Functional Assessment (e.g., ObsRO, ClinRO)		Evidence of clinical meaningfulness
	Secondary		Wider clinical benefit and/or influence of processes/concepts other than cognition (e.g., falls, other motor, autonomic, neuropsychiatric, sleep, sensory)
Treatment of mild to moderate PD dementia (PDD)	Cognition (e.g., PerfO)	Co-primary	Evidence of specific effect on core clinical feature
	Global Clinical Assessment or Functional Assessment (e.g., ObsRO, ClinRO)		Evidence of clinical meaningfulness
	Secondary		Wider clinical benefit and/or influence of processes/concepts other than cognition (e.g., falls, other motor, autonomic, neuropsychiatric, sleep, sensory)

Readiness for short-term and long-term clinical trial needs

To spur the use or adaptation of existing cognition-focused measures and the development of new measures as potential clinical trial endpoints in PD, DLB, and related disorders, The Michael J. Fox Foundation for Parkinson's Research (MJFF), Shake It Up Australia Foundation, Parkinson's UK, Parkinson Canada, Lewy Body Dementia Association, Cure Parkinson's, and the Critical Path for Parkinson's Consortium co-hosted the PD/DLB Cognition Endpoints Roundtable. This January 2024 roundtable brought together representatives from academia and industry, as well as representatives of regulatory agencies, community partners, patient advocates, and research funders, to build consensus and collaborate on the outcome assessment and trial design methods needed to support the development of new treatments for initial or mild cognitive changes in disorders on the PD/DLB spectrum. ³² This has led to the formation of a cognition working group, with several research efforts being prioritized (Table 3). Short-term goals will include literature reviews and COA reviews to support identification and selection of suitable COAs to support clinical trial endpoints. Longer-term, there is a need to address gaps in understanding of the patient experience to support more detailed COA reviews, modification of assessments, or novel instrument development (Table 4). Furthermore, as understanding increases regarding the potential influence of co-pathology (e.g., amyloid burden) or gene variants (e.g., the GBA gene), revisions of models and recommendations may be necessary.^14,33,34

Table 3.

Priorities for the working group.

Priority area	Workstreams
Literature reviews	Review existing qualitative data and conceptual models incorporating cognition concepts
	Review existing disease models including pathophysiology and neuropsychological profile
COA reviews	Review existing cognition instruments Content validity Psychometrics Fit-for-purpose in context of use
Evidence summaries / Additional evidence generation	Cognition COA recommendations
	Additional qualitative research
	Revised conceptual model(s) of cognition across the spectrum of cognitive functional ability
	Modification of existing or development of novel COAs

Table 4.

Completed and ongoing research efforts to support trial readiness.

MJFF research efforts	Objectives	Findings
State-of-the-field of qualitative research in DLB	A literature review of qualitative studies of the lived experience of DLB to understand whether sufficient data exists to inform the development of a DLB-specific conceptual model.	The review included 27 studies and concluded that there were substantial gaps in understanding the direct experiences of DLB, as most studies reported caregiver experience with limited focus on the experiences of people with DLB. 34
Review of cognition measures for PD/DLB trials	A multistakeholder taskforce was convened to review cognition measures used in PD/DLB trials. The first phase of this initiative involves a research synthesis of existing measures to understand measure performance and limitations. The next phase will identify and/or generate data to evaluate the longitudinal performance of the prioritized instruments.	Ongoing 32
Patient-centered conceptual model of cognition in PD	Work will be undertaken to identify the patient experience of cognition across PD stages, in terms of frequency, severity, and bothersomeness to patients. The relationship between cognitive impairment and changes in daily functioning will be investigated. Input from patients, care partners/knowledgeable informants, and experts will fill gaps in existing research and be used to develop and refine a conceptual model. This model will ultimately be used to inform the selection/adaptation/development of patient-centered COAs and/or digital health technologies (DHTs), and to define clinical trial endpoints.	Ongoing

Through these research efforts, the patient experience of cognitive changes in PD will be better understood so that appropriate trials can be designed, meaningful measures selected or developed, and evidence-based clinical trial endpoints chosen. This new body of evidence will be made available to drug developers pursuing new therapies in PD to help support their regulatory interactions, allowing for the appropriate information about the impacts of new PD therapies on cognition to be generated for regulatory decision making and to inform treatment decisions made by clinicians and patients. It is recognized too that wider efforts will be needed and the working group seeks to encourage the field to consider these issues.