Methylphenidate Differentially Affects Intrinsic Functional Connectivity of the Salience Network in Adult ADHD Treatment Responders and Non-Responders

Simple Summary Methylphenidate is frequently used to treat attention-deficit/hyperactivity disorder (ADHD). It has been reported to correct (aberrant) connectivity between brain regions. However, previous neuroimaging research has not provided a coherent picture. One possible reason is that methylphenidate may not necessarily result in a satisfactory clinical treatment response of every patient. This suggests that neuroimaging research about methylphenidate's mode of action could benefit from subdividing patients into groups of treatment "Responders" and "Non-Responders". However, previous research has not applied this idea so far. Here, we used machine learning techniques to discern Responders from Non-Responders based on clinical symptoms. Fifty-three stimulant-naive adult ADHD patients underwent functional magnetic resonance imaging to determine functional connectivity at rest. Measurements were repeated after 6 weeks' treatment with methylphenidate. We found that, prior to treatment, a specific part of the so-called salience network was hypo-connected with the bilateral putamen in the patient sample as a whole, compared with 50 healthy controls. After methylphenidate treatment, this aberrant connectivity was restored to almost "normal" levels in Responders, but not in Non-Responders. Involvement of the putamen is reminiscent of early positron emission tomography findings, already suggesting modulation of dopaminergic brain regions for the successful treatment of ADHD with methylphenidate. Positron emission tomography (PET) studies have shown involvement of the striatum when treating adult attention-deficit/hyperactivity disorder (ADHD) with methylphenidate (MPH). Results from resting-state functional magnetic resonance imaging (rs-fMRI) for the same issue were less unequivocal. Here, a new analytical framework was set up to investigate medication effects using seed-based rs-fMRI analysis to infer brain regions with alterations in intrinsic functional connectivity (IFC) corresponding with ADHD symptom reduction. In a within-subjects study design, 53 stimulant-naive adult ADHD patients were investigated before and after 6 weeks of MPH treatment, using two major clinical symptom scales and rs-fMRI. The same data were acquired in a sample of 50 age- and sex-matched healthy controls at baseline. A consensual atlas provided seeds for five predefined major resting-state networks. In order to avoid biasing of medication effects due to putative treatment failure, the entire ADHD sample was first categorized into treatment Responders (N = 36) and Non-Responders (N = 17) using machine learning-based classification with the clinical scales as primary data. Imaging data revealed medication effects only in Responders. In that group, IFC of bilateral putamen changed significantly with medication and approached almost normal levels of IFC. Present results align well with results from previous PET studies, with seed-based rs-fMRI as an entirely different neuroimaging method.