Effect of ecstasy [3,4-methylenedioxymethamphetamine (MDMA)] on cerebral blood flow: a co-registered SPECT and MRI study

Chang and colleagues frame their study around the neurotoxic profile of 3,4-methylenedioxymethamphetamine (MDMA, 'Ecstasy'), which in preclinical work damages serotonergic (5-HT) nerve endings. Because serotonergic pathways contribute to regulation of the cerebral vasculature, the investigators highlight a plausible mechanism by which MDMA could alter regional cerebral blood flow (rCBF). Previous human data on MDMA’s biological effects were limited at the time, with mixed findings from cerebrospinal fluid, PET and spectroscopy studies and reports of neuropsychiatric and neuropsychological abnormalities in users. The study set out to assess possible chronic (long-term, abstinent users) and subacute (weeks after controlled administration) effects of MDMA on brain perfusion. To do so, the researchers used co-registered single photon emission computed tomography (SPECT) and magnetic resonance imaging (MRI) to compare rCBF and brain structure in abstinent recreational MDMA users versus matched drug-naive controls, and to examine rCBF changes in a subset of users who received controlled MDMA doses and were rescanned afterwards. The emphasis was on detecting regional perfusion changes that might relate to serotonergic dysfunction.

This was a case–control study with an embedded open administration component. Twenty-one recreational MDMA users (17 males, 4 females) and 21 age-, gender- and socioeconomically-matched drug-naive control subjects underwent MRI and SPECT. Inclusion criteria for MDMA subjects required recreational use at low doses at least twice per year for at least 1 year and abstinence for a minimum of 2 weeks before baseline imaging. Participants completed psychiatric interviews, physical and neurological examinations, and urine toxicology screens; exclusions covered recent positive screens for several drugs, current psychiatric or medical medication, alcohol dependence or other substance dependence (except nicotine), significant head trauma, pregnancy and contraindications to MRI. Ten of the MDMA users subsequently participated in a Phase I inpatient component in which they received MDMA and underwent repeat imaging; of those, eight were rescanned within 1.5–3 weeks after the second MDMA dose and two were scanned more than 2 months later. Subjects were required to abstain from other drug use and had repeat urine toxicology prior to the second scan. The extracted text does not clearly report the exact dosing regimen for each subject, although dose–response analyses were reported and some individual doses (mg/kg) appear in the results. MRI was performed on a 1.5-T scanner using sequences optimised for co-registration with SPECT (inversion recovery and spin-echo sequences providing good white/grey/CSF contrast). SPECT acquisitions combined 133-xenon inhalation for absolute rCBF quantification and 99mTc-HMPAO perfusion imaging for higher spatial resolution; the HMPAO scans were calibrated to the 133Xe-derived absolute flow using a non-linear regression model. Images were co-registered intra-subject and partial volume correction was applied to account for CSF/atrophy-related dilution. Primary outcomes were global and regional CBF and measures of global brain volume and percentage CSF. Statistical analyses used mixed-model analyses of variance (ANOVA) comparing MDMA users and controls with hemisphere and brain region as within-subject factors, and repeated-measures ANOVA for the pre–post comparisons in the administered subgroup. Multiple linear regression assessed associations of cumulative lifetime MDMA exposure and time since last use with imaging measures. Corrections for multiple comparisons included the Huynh–Feldt correction for ANOVA sphericity and post-hoc Fisher’s PLSD with Bonferroni adjustment; the paper reports conventional P-value thresholds after correction.

Baseline comparisons between the 21 abstinent MDMA users and 21 matched controls showed normal MRI scans in all participants. After co-registration and partial volume correction, global CBF in abstinent MDMA users was 2.3% lower than in controls, a difference that was not reported as statistically significant. No significant correlations were found between global or regional CBF and duration, frequency or recency of prior MDMA use. In contrast, global brain volume correlated negatively with duration of MDMA use (reported correlation r = -0.57, P = 0.02), suggesting an association between longer use and reduced brain volume. In the subgroup that received MDMA and were rescanned, eight subjects imaged between 1.5 and 3 weeks after the second dose demonstrated detectable decreases in rCBF. A statistically significant interaction between drug status and hemisphere was reported (F = 6.6; df 2,14; P = 0.01). Compared with eight matched control subjects, these post-MDMA scans showed significantly reduced rCBF across multiple brain regions. The largest regional reductions cited were in the caudate (approximately -9.1% and -9.2% for left and right measurements reported), the superior parietal cortices (about -10.0% and -9.0%), and the right dorsolateral prefrontal cortex (about -10%). Global CSF volume and %CSF did not change significantly following MDMA administration. The magnitude of post-administration rCBF decreases tended to be greater in subjects who received higher total MDMA doses and in those scanned sooner after dosing. Correlations between total MDMA dose and rCBF change were reported for the left globus pallidus (r^2 ≈ 0.66, P = 0.04) and right caudate (r^2 ≈ 0.78, P = 0.009), but these associations did not remain significant after correction for multiple comparisons. Two subjects scanned more than 2 months after MDMA administration showed increased global CBF (+13.6% and +8.6%), indicative of temporal variability in the perfusion response. The extracted text does not provide full statistical tables or exact P-values for all regionwise comparisons.

Chang and colleagues interpret their findings as showing relatively normal long-term rCBF in abstinent low-dose recreational MDMA users, alongside evidence of subacute regional hypoperfusion after controlled MDMA administration. They propose two non-exclusive explanations for the near-normal baseline rCBF: either chronic, low-dose MDMA does not produce sustained alteration in 5-HT-mediated cerebrovascular regulation sufficient to change perfusion, or neuronal networks adapt over time through recovery or compensatory mechanisms such as axonal reinnervation or input from other neurotransmitter systems. The investigators link the pattern of subacute decreases in rCBF—most pronounced in parietal and dorsolateral frontal regions and in basal ganglia structures—to serotonergic mechanisms. They discuss how MDMA or its metabolites might provoke 5-HT-mediated vasoconstriction or transient increases in extracellular 5-HT secondary to terminal damage, producing regional hypoperfusion in the weeks after dosing. The authors also consider that indirect effects on dopaminergic systems could contribute, particularly in basal ganglia regions rich in dopaminergic receptors, and they note parallels with animal studies showing focal perfusion and metabolic changes after related compounds. Limitations acknowledged in the paper include the small number of subjects who received MDMA (limiting statistical power and the generalisability of the dose–response observations), the small number of female participants (precluding robust assessment of gender effects), and limited longitudinal follow-up to define the temporal course of perfusion changes. The authors emphasise that the subacute hypoperfusion findings are preliminary and call for larger, longitudinal studies with acute and longer-term follow-up to determine whether the observed regional hypoperfusion is transient and to clarify underlying mechanisms. They also note that their co-registration and partial volume correction procedure reduces the likelihood that perfusion measures were confounded by CSF-related dilution due to atrophy.