Large-Scale and Local Functional Connectivity Changes Following Psilocybin Administration in Methamphetamine Use Disorder

Psilocybin has been proposed to transiently alter large-scale brain network organisation by acting on cortical 5-HT2A receptors, reducing within-network synchrony and increasing communication between networks. In methamphetamine use disorder, previous resting-state fMRI studies have shown disruptions in default mode, salience, dorsal attention and frontoparietal systems, as well as abnormalities in local neural synchrony, but it remains unclear whether these patterns can be reorganised after a targeted intervention. The paper frames this as an important gap because most earlier work has been cross-sectional and has not examined post-treatment changes in multiscale brain connectivity. Chaganti and colleagues therefore set out to evaluate whether psilocybin administration alongside psychotherapy was associated with changes in large-scale functional connectivity and regional homogeneity, a voxel-wise measure of local synchrony, in people with methamphetamine use disorder. They also aimed to test whether any connectivity changes were associated with clinical measures of methamphetamine use, craving and psychological distress. The study was exploratory and longitudinal, with MRI scans acquired before and after treatment.

The study was a single-arm, open-label trial conducted in Sydney, Australia. The extracted text indicates that it was reported in line with STROBE guidelines and prospectively registered. Participants were adults seeking treatment for methamphetamine use disorder, reporting methamphetamine use on at least 4 days per month at screening, and without serious mental illness, medical conditions, or medications contraindicated with psilocybin. Recruitment took place between February and November 2023. Although a sample size of 15 was selected, 14 participants were recruited, and 10 had usable MRI data at both time points. All participants received three preparatory psychotherapy sessions over two weeks, then a single oral dose of psilocybin (25 mg), followed by two integration psychotherapy sessions over one week. The psychotherapy included motivational enhancement and acceptance and commitment therapy elements. MRI scans were obtained at two time points: before psilocybin dosing and within one week afterwards. Clinical outcomes were assessed at baseline and again 28 days after psilocybin administration. These included timeline follow-back (past-28-day methamphetamine use, verified by urine drug screening), the Depression, Anxiety, Stress Scale (DASS-42), and a visual analogue scale for craving. MRI data were collected on a 3T scanner. Resting-state fMRI and T1-weighted structural images were acquired in the same session. Preprocessing used FSL and included motion correction, slice-timing correction, coregistration, normalisation, brain extraction, band-pass filtering, and regression of white matter and cerebrospinal fluid signals using a CompCor approach. Mean framewise displacement did not differ significantly between sessions, and no scans exceeded the usual motion threshold, so volume censoring was not applied. Functional connectivity was analysed with the CONN toolbox using ROI-to-ROI correlations across 46 regions spanning canonical networks and subcortical structures. The authors highlight a priori regions of interest in frontoparietal, default mode, salience, striatal, hippocampal and cerebellar circuits. ReHo was calculated with AFNI to assess local synchrony, using unsmoothed data, standardisation to whole-brain mean ReHo, and subsequent spatial smoothing. Within-subject pre-post changes in connectivity were tested with general linear models, while voxel-wise ReHo differences used permutation testing with 5,000 permutations in FSL randomise. Clusters were thresholded at voxel-wise p<0.01 with a minimum extent of 30 voxels. Exploratory associations between neuroimaging changes and clinical outcomes were examined with Spearman correlations, without correction for multiple comparisons.

Among the 10 participants with both pre- and post-treatment scans, 80% were male, the mean age was 46 years, and 90% were Caucasian. The sample had long-standing methamphetamine use, with a mean age at first use of 32 years. Most participants smoked methamphetamine only, some injected, and several were taking prescribed stimulants or other psychotropic medications during the treatment period. Comorbid ADHD and anxiety or depression were each reported in 30% of the sample. After psilocybin and psychotherapy, the main functional connectivity changes involved reorganisation across attention, default mode, salience and frontoparietal systems. Connectivity increased between the dorsal attention network and the frontoparietal network, and between the dorsal attention network and the default mode network. There were also increases between the default mode network and salience network, between the visual network and thalamus, between the hippocampus and frontoparietal network, between medial visual cortex and caudate, and between posterior cerebellum and frontal eye field. Some connectivity increases reflected reduced negative correlations rather than stronger positive coupling. In contrast, connectivity decreased between the caudate and both the language network and salience network. ReHo analyses showed a mixed pattern of local synchrony changes. Post-treatment, ReHo decreased in regions including the posterior cingulate gyrus, thalamus and subcallosal cortex, while increased ReHo was observed in the postcentral gyrus and frontal pole. The authors interpret this as evidence of region-specific reorganisation of local neural coherence. Exploratory clinical correlations suggested that greater reductions in methamphetamine use were associated with changes in connectivity in attentional and hippocampal-prefrontal circuits. Specifically, lower methamphetamine use on timeline follow-back correlated with change in connectivity between the right frontal eye field and posterior cingulate cortex (Spearman's r=-0.71, p=0.0205) and between the right hippocampus and left lateral prefrontal cortex (r=-0.83, p=0.0026). Changes in DASS scores correlated positively with change in connectivity between the right frontal eye field and posterior cingulate cortex (r=0.78, p=0.0142) and negatively with connectivity between the left inferior frontal gyrus and right caudate (r=-0.83, p=0.0061). For ReHo, lower methamphetamine use correlated with lower ReHo in the left putamen (r=-0.77, p=0.0085). Changes in DASS scores were negatively correlated with ReHo in the right nucleus accumbens (r=-0.81, p=0.0083), and craving severity was negatively correlated with ReHo in the right anterior cingulate gyrus (r=-0.82, p=0.0108) and right thalamus (r=-0.90, p=0.0020). These analyses were explicitly exploratory.

The authors interpret the findings as preliminary evidence that psilocybin-assisted psychotherapy is associated with measurable reorganisation of intrinsic brain connectivity in methamphetamine use disorder. They emphasise that the strongest changes were seen in networks commonly implicated in stimulant addiction, especially the dorsal attention, default mode and salience systems. In their view, increased coupling between attentional and default mode regions may reflect better integration between externally directed attention and internally focused processing, which could relate to cognitive flexibility and reduced maladaptive self-focus. They also note that persistent reductions in frontostriatal connectivity, particularly involving the caudate, were linked to greater cumulative methamphetamine exposure, which they present as consistent with earlier research suggesting long-lasting corticostriatal alterations in chronic stimulant use. They argue that the thalamocortical and visual-attentional changes support a distributed, network-wide effect rather than a purely cortical one. At the local level, the authors see the ReHo findings as complementary to the network-level results, suggesting a redistribution of local synchrony towards frontal and sensorimotor regions that support executive and behavioural control. Chaganti and colleagues state that the clinical correlations, while exploratory, are in line with prior work linking network integrity to substance use severity, craving and psychological distress. They suggest that local synchrony changes may precede or enable broader shifts in large-scale connectivity, and they place this within hierarchical models of brain organisation and contemporary ideas about network plasticity. They also connect the findings to broader neuroimaging trends, including dynamic connectivity modelling and precision psychiatry approaches that seek MRI-derived biomarkers of treatment response. The authors acknowledge several important limitations. The sample was small, there was no placebo or active control group, and the correlation analyses were not corrected for multiple comparisons. They also caution that the voxel-wise cluster-defining threshold was relatively liberal, which increases the risk of false-positive findings. They therefore describe the study as exploratory and call for larger, controlled studies with longer follow-up to assess durability and clinical significance. Their overall interpretation is that the results provide preliminary neuroradiologic evidence of functional reorganisation after targeted intervention in methamphetamine use disorder and support multimodal rs-fMRI as a potential biomarker of network-level plasticity.