Neurometabolic effects of psilocybin, (MDE) and d-methamphetamine in healthy volunteers

Earlier functional imaging studies with hallucinogens suggested a pattern of right‑hemispheric predominance and increased activity in frontocortical regions, most prominently the anterior cingulate, a region implicated in attention and emotion. Such drug‑induced states have been used as experimental models of acute psychosis because they can produce positive symptom‑like phenomenology; however, neuroimaging findings in naturally occurring schizophrenia are mixed, with chronic patient studies often showing hypofrontality. Evidence on stimulant amphetamine effects is inconsistent, and data on entactogens (for example MDMA or related compounds) are sparse. Given overlapping but distinct psychological effects of hallucinogens, entactogens, and stimulants, a direct comparison under controlled conditions could help disentangle substance‑specific neurometabolic signatures and their relation to cognitive activation capacity. Gouzoulis‑Mayfrank and colleagues designed a double‑blind, placebo‑controlled FDG‑PET study to assess acute neurometabolic effects of a hallucinogen (psilocybin, PSI), an entactogen (3,4‑methylenedioxyethylamphetamine, MDE), and a stimulant (d‑methamphetamine, METH) in healthy volunteers. The investigators aimed to (1) characterise regional cerebral glucose metabolism changes at recreationally relevant doses, (2) explore correlations between metabolic changes and psychopathological signs, and (3) examine how each drug modulated a frontal‑language cognitive activation (word association) versus a control word repetition task.

Thirty‑two healthy volunteers (21 men, 11 women; mean age 34.2 years, range 27–47), mostly physicians or psychologists, were recruited after screening for medical and psychiatric illness (SCID) and provided informed consent. Subjects were allocated to one of four double‑blind groups (placebo, PSI, MDE, METH), each ultimately comprising eight participants; for METH the dose was increased midway through the study and those two subgroups were pooled because no clear differences were observed. Exclusion criteria included current or past axis I/II disorders and first‑degree family history of major psychiatric illness. Drugs were encapsulated identically and given orally. Doses were: psilocybin 0.2 mg/kg (capped at 15 mg), MDE 2 mg/kg (capped at 140 mg), and methamphetamine initially 0.2 mg/kg (later raised to 0.4 mg/kg; capped at 35 mg). Each subject underwent two scanning sessions with the same substance 2–4 weeks apart, performing either a word association (activation) or a word repetition (control) task in counterbalanced order. Cognitive tasks: 640 high‑frequency spoken German nouns were delivered binaurally over 32 minutes at one word every 3 seconds. In the association task subjects overtly produced an associated word; in the control they repeated the stimulus word. Spoken responses were recorded for later analysis. Scans were performed during the peak drug effect: stimuli presentation began 2 minutes before intravenous administration of 132–307 MBq 18‑FDG and lasted 32 minutes; emission imaging started 30 minutes post‑injection and lasted 30 minutes. Imaging and quantification: MRI (1.5 T) provided anatomical reference. PET data were coregistered to MRI and 113 regions of interest (ROIs; 55 per hemisphere plus three brainstem ROIs) were defined. To reduce partial volume effects ROIs exceeded 2.5× the PET FWHM. Absolute pixelwise glucose consumption was calculated using the autoradiographic method with a lumped constant of 0.52 and then normalised to the global metabolic rate to yield regional rMRGlu. For statistical analysis these 113 ROIs were combined into 37 ROI groups and composite regions (frontal neocortex, posterior neocortex, entire neocortex, subcortical, and ratios). Statistical approach: Between‑group comparisons of global and normalised regional rMRGlu (control/repetition scan) versus placebo used Mann‑Whitney U‑tests. Within‑subject changes between association and repetition scans were analysed with Wilcoxon matched‑pairs tests. Correlations between metabolic measures and psychopathology used Spearman coefficients, limited to ROIs showing significant or trend deviations from placebo. No correction for multiple testing was applied; the authors present the analyses as exploratory. Psychopathology scales included PANSS subscales, Bech‑Rafaelsen mania and melancholia scales, STAI state anxiety, and the Hallucinogen Rating Scale (HRS).

Global cerebral glucose metabolism did not differ significantly between groups during the repetition (control) scan; mean global rMRGlu values (µmol/100g/min) were reported as PLA 37.58 ± 14.27, METH 32.59 ± 3.12, MDE 32.06 ± 10.23, PSI 31.49 ± 5.30 (some missing values due to technical problems). Regional normalised rMRGlu (control/repetition scan): Psilocybin induced focal relative hypermetabolism in right‑hemispheric frontotemporal cortex, most notably the right anterior cingulate (+9.82%, p = .046) and right frontal operculum (+7.74%, p = .046), with a trend in right inferior temporal (+4.16%, p = .093). Concomitant decreases were seen in the right thalamus (−8.73%, p = .016) and left precentral region (−6.13%, p = .016) with a left thalamic trend (−5.62%, p = .093). Composite measures suggested a trend toward reduced subcortical metabolism and an increased right cortical/subcortical ratio. MDE produced a pattern of cortical hypometabolism—most significantly reduced frontal neocortex (−3.19%, p = .006) and entire neocortex (−1.58%, p = .037)—with particular ROI decreases in left precentral (−7.14%, p = .021) and right superior prefrontal (−4.58%, p = .037) regions. At the same time, MDE increased cerebellar metabolism bilaterally (right +10.12%, p = .002; left +7.59%, p = .028) and the right putamen (+6.21%, p = .049). The right anterior cingulate showed a trend increase (+7.26%, p = .064). Methamphetamine showed widespread tendencies toward cortical hypometabolism, with significant bilateral cerebellar hypermetabolism (left +10.52%, p = .039; right +9.91%, p = .020). The composite entire neocortex was slightly reduced (−1.69%, p = .045); several right‑hemisphere ROIs showed trend decreases (parietal, temporoparietal, precentral, superior prefrontal). Correlations with psychopathology: In the PSI group, higher right anterior cingulate activity correlated positively with stereotyped thoughts (PANSS N7) and negatively with state anxiety (STAI‑X1). Right frontal operculum increases correlated with lower general activation (BRMAS). Left thalamic decreases associated with higher general psychopathology, tension/anxiety (PANSS G4, G2), and depressive mood (BRMES). In the MDE group, cerebellar hypermetabolism correlated with reduced volition (HRS volition), depressive mood, cognitive deficits (HRS cognition), and anxiety; right anterior cingulate increases correlated with attentional deficits. In the METH group, cerebellar hypermetabolism correlated with anxiety and inversely with feelings of happiness; cortical hypometabolism related to cognitive deficits and anxiety, while posterior cortical hypometabolism correlated with measures of drug intensity and several PANSS items. Activation versus control (association vs repetition) scans: Under placebo the association task increased metabolism in left frontal operculum (Broca homolog) (+7.13% ± 6.27%, p = .035), left insula (+5.30% ± 6.16%, p = .025), and left middle prefrontal (+3.51% ± 4.11%, p = .05) and decreased left posterior cingulate (−5.30% ± 4.37%, p = .017). Psilocybin markedly attenuated task‑related frontal activation: only a trend increase in left frontal operculum (+4.95% ± 6.05%, p = .075) was observed, whereas decreases occurred in left posterior cingulate (−5.63%, p = .028), right putamen (−3.58%, p = .046), right anterior cingulate (−3.24%, p = .046), and right occipital cortex (−2.83%, p = .028). MDE preserved some activation: significant increases in left middle prefrontal (+4.48% ± 1.85%, p = .018), left superior temporal (+4.55% ± 2.60%, p = .018), and right caudate (+5.51% ± 3.36%, p = .018), with a decrease in right posterior cingulate (−7.09% ± 5.66%, p = .018). Under methamphetamine frontal activation did not reach significance; significant activation occurred in left thalamus (+4.57% ± 4.58%, p = .046) and a decrease in right amygdala/hippocampal region (−8.03% ± 5.15%, p = .028). Cognitive performance: In the repetition control task subjects performed near ceiling in all groups. In the association task psilocybin subjects produced significantly fewer words (PSI 69.4% ± 17.8% of possible words) than placebo (PLA 90.2% ± 8.9%, p = .016). MDE (87.4% ± 11.3%) and METH (95.6% ± 6.5%) did not differ significantly from placebo.

Gouzoulis‑Mayfrank and colleagues interpret the findings as demonstrating distinct neurometabolic signatures for the three drug classes and as broadly consistent with a model‑psychosis framework. Psilocybin produced focal right‑hemispheric frontal hypermetabolism—most markedly a nearly 10% increase in the right anterior cingulate—together with thalamic hypoactivity and an elevated right cortical/subcortical ratio. Despite this baseline frontal hyperactivity, subjects on psilocybin showed an impaired ability to further activate frontal language/association regions during the cognitive task and performed worse on the association task, suggesting a dissociation between resting hyperactivity and task‑related recruitment. By contrast, methamphetamine and MDE shared a pattern of cortical hypometabolism with concomitant cerebellar hypermetabolism; MDE effects were intermediate, combining methamphetamine‑like cortical decreases (especially frontal) with a tendency for right anterior cingulate hyperactivity reminiscent of psilocybin. The authors propose that anterior cingulate increases under PSI and MDE may reflect compensatory engagement of executive–emotional control processes during intense sensory or emotional experiences, as suggested by correlations between cingulate activity and stereotyped thoughts or attentional deficits. Thalamic hypometabolism under psilocybin is discussed in terms of the thalamus’s role as a sensory filter, although in these data thalamic changes correlated more with general psychopathology than with perceptual reports. The cerebellar hypermetabolism seen with MDE and METH is noted as unexpected; the authors point to accumulating evidence that the cerebellum contributes to cognitive and emotional operations and may participate in drug‑induced and psychotic states. The investigators emphasise several limitations: small group sizes (n = 6–8), lack of correction for multiple comparisons, single‑dose design precluding dose–response inferences, and a mid‑study methamphetamine dose escalation. They also note the absence of a true resting baseline scan—the control was a word repetition task—so effects attributable to the control condition versus drug cannot be fully separated. These caveats lead the authors to characterise the study as exploratory. Finally, they suggest that the data support the concept of entactogens as a distinct class occupying an intermediate position between stimulants and hallucinogens, and they propose that the pattern of right‑hemispheric frontal overactivity coupled with impaired task‑related frontal recruitment may provide a model for certain features of acute psychotic episodes; further research is required to clarify mechanisms and clinical relevance.