6 September 2022
Assessing Cognition in Paediatric Clinical Trials: Benefits, Challenges, and Solutions
Understanding cognition is a growing trend in research, but what benefits can cognitive assessments have for researchers conducting paediatric research, and what are the best practices involved?
Cognitive or neuropsychological assessment has become increasingly common in both adult and paediatric clinical trials, particularly in therapeutic areas where cognitive impairment is a hallmark of disorder, such as mood disorders and neurodevelopmental disorders. Cognition measures are typically patient-reported outcomes, often in the form of a battery of tasks or questionnaires which each target a specific cognitive domain, such as memory or attention. Examples of this include the Cambridge Automated Neuropsychological Test Battery (CANTAB™), and NIH Toolbox Cognition.
Based on a recent search using TrialTrove, a clinical trial database, approximately 20,000 paediatric clinical trials have started since January 2010. Only 458 of these included endpoints assessing cognition. However, 351 of these trials investigated CNS disorders, which represents just over 10% of all paediatric CNS trials included in the search. The most common therapeutic areas to include cognitive assessments include ADHD, epilepsy, and autism, indicating that cognitive endpoints are often used in trials where cognitive impairment is a key feature of the disorder.
Cognitive endpoints have been used successfully in both adult and paediatric clinical trials, and have contributed to regulatory approval of new compounds. However, assessing cognition in children comes with additional challenges, compared to adults. Here, we explore the benefits of cognitive endpoints in paediatric clinical trials, and discuss the challenges and solutions for successful implementation.
The Benefits of Assessing Cognition
Cognitive assessment provides insight into a drug’s effect on neuropsychological processes, and, consequently, brain function. Objective cognitive assessments, such as tasks, can be particularly useful in paediatric populations, as children may be less able to self-report changes in their cognitive functioning than adults.
Safety and Tolerability
In paediatric clinical trials, drugs have the potential to disrupt healthy development, which could cause long-term adverse effects or delayed adverse drug reactions. Cognitive assessments can assess the impact of a drug on brain development, making them an important tool in paediatric safety and tolerability trials.
For example, CANTABTM Reaction Time and Spatial Working Memory tasks were used in a Phase II safety and tolerability trial of guanfacine extended-release in 6-17-year-olds with ADHD (1). These cognitive assessments demonstrated that guanfacine-XR did not cause cognitive impairment, and contributed to successful submission to regulatory bodies.
Cognitive endpoints have also been used to assess efficacy. In paediatric ADHD trials, CANTABTM assessments have been used to assess the efficacy of several cognition-enhancing drugs, including atomoxetine and methylphenidate (2, 3). These cognitive endpoints showed significant improvement in executive function, sustained attention, and memory, following treatment.
Conversely, a randomised controlled trial used CANTABTM assessments to provide Class I evidence that everolimus was not effective at reducing cognitive impairment in children with tuberous sclerosis complex (4).
Cognitive assessments can also inform appropriate dosing. Paediatric dose selection can be challenging due to different PK/PD dynamics in children compared to adults, lack of paediatric kinetics data, and the need to formulate doses according to body weight. Many neurocognitive assessments, including CANTAB, have demonstrated dose-response sensitivity. Thus, cognitive assessments can be used to detect dose-related adverse cognitive effects, as well as indicate under-dosing and accompanying lack of benefit.
A Phase III trial in migraine patients aged 12 to 17 years used CANTABTM assessments to evaluate the neurocognitive effects of two doses of topiramate. The higher dose, but not the lower dose, was associated with significantly slowed psychomotor processing across three cognitive tasks (5). These cognitive endpoints successfully contributed to NDA submissions.
Challenges and Considerations
All patient-reported assessments used in clinical trials, whether adult or paediatric, should demonstrate good reliability and validity. However, assessing cognitive function in children presents additional challenges which need to be considered to ensure valid measurement.
The Developing Brain
Children’s brains and cognitive abilities are still developing throughout childhood and adolescence – in fact, the brain continues to mature until around 25 years old. Additionally, different cognitive skills develop at different developmental stages, and the age at which individual children reach developmental milestones is highly variable. Classifying performance on cognitive assessments as ‘healthy’ becomes challenging, as assessments must account for differences in ability due to developmental stage, and the broad range of abilities considered developmentally normal at a particular age.
This results in two considerations. Firstly, the difficulty of an assessment must capture the cognitive abilities of the youngest and oldest participants with equal validity, to reduce the likelihood of floor or ceiling effects.Secondly, cognitive assessments must facilitate meaningful comparison between children of different ages, such as interpreting scores in terms of expected cognitive ability rather than raw values. In most assessments, a 17-year-old will outperform a 5-yearold, but this does not mean that the 5-year-old is cognitively impaired, or conversely, that the 17-year-old is not.
Usability, Feasibility, and Acceptability
To successfully deploy cognitive assessments and collect high-quality data, assessments must be usable, feasible, and acceptable in the target population. In paediatric patients, there may be several potential barriers that can affect this:
- Pre-requisite knowledge or abilities: Children need to have already developed the skills needed to understand and complete a task. This includes: the target cognitive domain (there is little point in trying to assess a skill that a child does not have), knowledge related to the context of the task (for example, typical items on a grocery list for a memory task), and any abilities required to execute the task (such as motor coordination)
- Task comprehension: To complete an assessment to the best of their abilities, children need to fully understand task instructions. This requires use of age-appropriate vocabulary and reading difficulty
- Attention: Children have shorter attention spans than adults, and may quickly become bored during an assessment. Therefore, the duration of a cognitive assessment battery should be within the limits of a participant’s attentional capacity
- Motivation: If a cognitive assessment becomes frustrating or boring, children may lack the intrinsic motivation to continue completing the task to the best of their abilities
Overcoming Challenges Through Study Design
While including cognitive endpoints in paediatric trials is challenging, careful selection of appropriate assessments and study designs can ensure high quality data collection.
Using a series of varied, short duration cognitive assessments can mitigate the impact of children’s shorter attention spans. However, shorter assessments also reduce the amount of data that can be collected.
Computerised adaptive testing adjusts an assessment’s contents according to an individual’s abilities, maximising both brevity and precision. For long test batteries, adaptive testing can reduce the overall duration of a testing session without reducing the number of individual assessments included.
Additionally, standardised rest breaks between tasks, or questionnaires, can maintain focus by facilitating multiple short bursts of attention, rather than one continuous session.
Suitability Across Ages
The suitability of a cognitive assessment will depend on the characteristics of the population, including patient age and therapeutic area. In paediatric clinical trials, patient ages can range from infancy to late adolescence, encompassing a wide range of cognitive abilities.To account for this, studies may use multiple variations of an assessment, with easier versions for younger children and challenging versions for adolescents. However, data collected using different variants of an assessment may not be comparable, which may require further consideration at the statistical analysis stage, including use of relative measures, such as change from baseline. Alternatively, trials may select assessments which gradually increase in difficulty, and terminate once a participant finds trials too challenging. This approach ensures participants of all abilities can complete a mix of easier and more challenging trials, without having to complete many trials beyond their ability level.
Regardless of the assessment used, normative data is an essential tool for assessing cognition in children. Transforming scores on a cognitive assessment to z-scores using age- and IQ-matched normative data models allows performance to be interpreted in terms of expected ability for a child of a given age. This also facilitates direct comparison of performance between children of different ages.
There are three ways to foster task comprehension in paediatric populations: content, delivery,and supporting material. In terms of content, the vocabulary and grammar structures used to explain the task must be accessible and developmentally appropriate. The delivery of these instructions should be engaging and avoid written text. Child-friendly delivery could include use of a sing-song voice, or animated or simplistic video. Comprehension can be further supported by using visual or auditory cues to guide participants’ responses. Similarly, completing a practice trial or familiarisation session can allow children to get used to an assessment before completing it ‘for real’.
Maintaining children’s engagement during cognitive assessments is crucial to collecting valid data. Gamification (applying game-like elements, such as point-scoring, to an assessment), may increase engagement and motivation. However, gamification may be a double-edged sword, as too much focus on obtaining a reward may lead to reduced focus on accuracy or task rules.
Motivation can also be improved through method of task delivery. A pencil-and-paper exam is less engaging than interactive play, so children are likely to be more motivated during assessments which resemble the latter. Technology can also enhance motivation: children are familiar with computer-games, iPads, and smartphones, and many perceive time using these devices as a treat or reward. Computerised cognitive assessment, delivered via iPads or smartphones, may be inherently rewarding and could increase patient motivation.
In conclusion, cognitive assessments can be a valuable addition to paediatric clinical trials, particularly when cognitive impairment is a key feature of the target disorder. Cognitive endpoints have supported paediatric safety, tolerability, and efficacy trials, notably in ADHD and epilepsy. The use of developmentally appropriate, short, and engaging cognitive assessments, supported by normative comparisons, can provide valid, reliable, and informative endpoints in paediatric clinical trials.
1. Kollins SH et al, Psychomotor functioning and alertness with guanfacine extended release in subjects with attention-deficit/hyperactivity disorder, J Child Adolesc Psychopharmacol, 21(2): pp111-120, 2011, doi:10.1089/cap.2010.0064
2. Wu CS et al, Differential Treatment Effects of Methylphenidate and Atomoxetine on Executive Functions in Children with Attention-Deficit/ Hyperactivity Disorder, J Child Adolesc Psychopharmacol, 31(3): pp187-196, 2021, doi:10.1089/cap.2020.0146,
3. Shang CY, Gau SS, Improving visual memory, attention, and school function with atomoxetine in boys with attention-deficit/hyperactivity disorder, J Child Adolesc Psychopharmacol, 22(5): pp353-363, 2012, doi:10.1089/cap.2011.0149,
4. Overwater IE et al, A randomized controlled trial with everolimus for IQ and autism in tuberous sclerosis complex, Neurology, 93(2): e200-e209, 2019, doi:10.1212/WNL.0000000000007749,
5. Pandina GJ et al, Cognitive effects of topiramate in migraine patients aged 12 through 17 years, Pediatr Neurol, 42(3): pp187-195, 2010, doi:10.1016/j.pediatrneurol.2009.10.001
This is a repost of an article that first appeared in International Clinical Trials, August 2022, pages 17-19. © Samedan Ltd.
Tags : No tags found.
Iona Pickett, Operational Scientist