Dissociable and common effects of methylphenidate, atomoxetine and citalopram on response inhibition neural networks

Response inhibition is an executive function that allows the detection and modification of unwanted actions. Its underlying neurochemistry and neurobiology have been explored by combining classic neuropsychological paradigms, such as the go/no-go task (GNG), with targeted pharmacology and functional neuroimaging. We sought to further this literature by using single doses of methylphenidate (30 mg), atomoxetine (60 mg), citalopram (30 mg) and placebo to probe dopaminergic, noradrenergic and serotonergic aspects of response inhibition. Twenty-seven (27) healthy, right-handed males participated in a randomised, double blind, placebo-controlled, within subject, crossover fMRI study to examine stop-related BOLD activation correlates of a modified GNG task. Methylphenidate demonstrated activation versus placebo in the pregenual cingulate (dorsal anterior cingulate), right inferior frontal, left middle frontal, left angular and right superior temporal gyri and right caudate. Atomoxetine demonstrated activation versus placebo across a broad network of cortical regions. Both methylphenidate and atomoxetine, but not citalopram, activated superior temporal, right inferior frontal and left middle frontal clusters. Citalopram only activated the left inferior occipital lobe. Taking the above as functionally defined regions of interest, we examined the specificity of stop-related drug activity by comparing mean activations across the four conditions. Only methylphenidate demonstrated drug-specific effects with increased activation of the pregenual cingulate and decreased activation of the caudate. Direct comparison of methylphenidate and atomoxetine showed broad recruitment of prefrontal regions but specific effects of methylphenidate in the pregenual cingulate and caudate revealing dissociable modulations of response inhibition networks.