Background

Detecting the progression from amnestic mild cognitive impairment (aMCI) to Alzheimer’s dementia remains a key research area; particularly when recruiting the right patients into clinical trials.

A cognitive function which is compromised early in the development of Alzheimer’s disease, more than six years prior diagnosis in fact, is episodic memory.

A reliable assessment of episodic memory is the CANTAB Paired Associate Learning (PAL) task. Indeed, this test has shown to be sensitive to different stages of disease progression, to drug effects and to different biomarkers of Alzheimer’s disease.

Objectives

To explore whether baseline performance on CANTAB tests, including PAL, can predict which patients with aMCI will show declines in their functional abilities, over the course of two years.

Methods

The study sample comprised 147 patients with aMCI, aged 55 to 90 years, who were enrolled in 13 European memory clinics as part of the IMI PharmaCog study,

When recruited, the patients all showed subjective memory complaints: a Mini-mental State Examination score ≥24 and Clinical Dementia Rating score of 0.5. The CANTAB cognitive test battery and the Functional Activities Questionnaire (FAQ) were administered throughout the study follow-up (24 months).

Cerebrospinal Fluid (CSF) samples were also collected, and levels of Amyloid-β 42, phosphorylated tau 181 (p-tau) and tau proteins were measured.  

Results

For the analysis, prediction models were developed to classify individuals by change in FAQ.

CANTAB PAL in combination with age, gender and CANTAB Spatial Working Memory (SWM) tasks were able to predict those subjects with the greatest decline in FAQ with excellent performance:

• Accuracy 0.83
• Specificity 0.75
• Sensitivity 1.0

​Figure 1 illustrates how the participants with the fastest decline on FAQ have the poorest performance on the CANTAB episodic memory Task PAL (PALTEA) over 24 months, supporting the ability to predict decliners using CANTAB task performance.