R- and S-citalopram concentrations have differential effects on neuropsychiatric scores in elders with dementia and agitation

Abstract

Aims: The aim was to determine the relationship between (R) and (S)-citalopram enantiomer exposure (AUC(0,24 h)) and therapeutic response in agitated individuals greater than 60 years old with Alzheimer's dementia (AD).

Methods: Citalopram enantiomer exposures (AUC(0,24 h)) derived from an established population pharmacokinetic analysis were utilized to explore the relationship between (R)- and (S)-citalopram area under the curve (AUC(0,24 )) and Mini-Mental State Examination (MMSE), Neurobehavioural Rating Scale-Agitation Subscale (NBRS-A), modified Alzheimer's Disease Cooperative Study-Clinical Global Impression of Change (mADCS-CGIC) and Neuropsychiatric Inventory Agitation subscale (NPIA) scores. Time dependent changes in these scores (disease progression) were accounted for prior to exploring the exposure effect relationship for each enantiomer. These relationships were evaluated using a non-linear-mixed effects modelling approach as implemented in nonmem v7.3.

Results: (S)-AUC(0,24 h) and (R)-AUC(0,24 h) each contributed to improvement in NBRS-A scores (k3(R) -0.502; k4(S) -0.712) as did time in treatment. However, increasing (R)-AUC(0,24 h) decreased the probability of patient response (maximum Δ -0.182%/AUC(0,24 h)) based on the CGIC while (S)-AUC(0,24 h) improved the probability of response (maximum Δ 0.112%/AUC(0,24 h)). (R)-AUC(0,24 h) was also associated with worsening in MMSE scores (-0.5 points).

Conclusions: Our results suggest that citalopram enantiomers contributed differentially to treatment outcomes. (R)-citalopram accounted for a greater proportion of the adverse consequences associated with racemic citalopram treatment in patients with AD including a decreased probability of treatment response as measured by the CGIC and a reduction in MMSE scores. The S-enantiomer was associated with increased probability of response based on the CGIC.

Keywords: Alzheimer's disease; agitation; citalopram; escitalopram; pharmacodynamics.

Figure 1

Visual predictive checks on (R) and (S) NBRS‐A models. A) (R)‐enantiomer model of NBRS‐A score and B) (S)‐enantiomer model of NBRS‐A score

Figure 2

Response probability of participant mADSC‐CGIC vs. average (R)‐citalopram AUC(0,24 h). (S)‐citalopram of 500 μg l⁻¹ day (blue); (S)‐citalopram of 1000 μg l⁻¹ day (orange); (S)‐citalopram of 1500 μg l⁻¹ day (grey); (S)‐citalopram of 2000 μg l⁻¹ day (yellow); Grey shaded region: Treatment probability < placebo

Figure 3

Model prediction of an average participant response in the placebo (blue line) and citalopram group (orange/grey lines). A) NBRS‐A score improvement with respect to treatment time, B) MMSE score improvement with respect to treatment time, C) NPIA score improvement with respect to treatment time. Average AUC(0,24 h) used in the model is 1500 μg l⁻¹ day for (R)‐citalopram and 1500 μg l⁻¹ day for (S)‐citalopram. placebo, (R)‐citalopram, (S)‐citalopram. placebo, (R)‐citalopram, (S)‐citalopram