The effect of fampridine on working memory: a randomized controlled trial based on a genome-guided repurposing approach

Abstract

Working memory (WM), a key component of cognitive functions, is often impaired in psychiatric disorders such as schizophrenia. Through a genome-guided drug repurposing approach, we identified fampridine, a potassium channel blocker used to improve walking in multiple sclerosis, as a candidate for modulating WM. In a subsequent double-blind, randomized, placebo-controlled, crossover trial in 43 healthy young adults (ClinicalTrials.gov, NCT04652557), we assessed fampridine's impact on WM (3-back d-prime, primary outcome) after 3.5 days of repeated administration (10 mg twice daily). Independently of baseline cognitive performance, no significant main effect was observed (Wilcoxon P = 0.87, r = 0.026). However, lower baseline performance was associated with higher working memory performance after repeated intake of fampridine compared to placebo (r_s = -0.37, P = 0.014, n = 43). Additionally, repeated intake of fampridine lowered resting motor threshold (F(1,37) = 5.31, P = 0.027, R²β = 0.01), the non-behavioral secondary outcome, indicating increased cortical excitability linked to cognitive function. Fampridine's capacity to enhance WM in low-performing individuals and to increase brain excitability points to its potential value for treating WM deficits.

Fig. 1. Drug selection.

The aim of stage I was to identify approved compounds with a putative impact on WM performance, an established endophenotype of schizophrenia. In filtering steps 1 and 2, qualified genes (i.e., genes whose associated products are targets of approved compounds) concurrently implicated in the susceptibility to schizophrenia and in the execution of a validated WM task were identified. These steps provided candidate compounds for stage II, the randomized controlled trial.

Fig. 2. Correlation between individual measures of the primary outcome and individual baseline WM performance.

Y-axis: individual difference (fampridine – placebo) in 3-back d’ after repeated administration; values above 0 represent improvement under fampridine, values below 0 represent decrease under fampridine. X-axis: individual baseline WM performance (3-back d’), the vertical dotted line represents the baseline performance level at which the regression line and the 0 level of the Y-axis intersect. The gray shaded area represents 95% confidence intervals of the linear regression line, drawn for visualization purposes. rs: Spearman’s rank correlation coefficient.

Fig. 3. Post-hoc analysis of the drug effect on the primary outcome per cognitive baseline group, violin plots.

Y-axis: individual difference (fampridine – placebo) in 3-back d’ after repeated administration; values above 0 represent improvement under fampridine, values below 0 represent decrease under fampridine. X-axis: baseline WM performance (3-back d’) stratified in three performance groups (i.e., low, medium, high terciles). Gray dots and vertical lines represent primary outcome measure means and standard errors of the means, respectively.

Fig. 4. Correlation between resting motor threshold and WM performance.

Y-axis: individual baseline WM performance (3-back d’ under placebo). X-axis: individual resting motor threshold under placebo. The gray shaded area represents 95% confidence intervals of the linear regression line, drawn for visualization purposes. rs: Spearman’s rank correlation coefficient.