Blockade of 5-HT 2 receptor selectively prevents MDMA-induced verbal memory impairment

The study used a double-blind, placebo-controlled, within-subject design with 17 healthy recreational MDMA users (9 male, 8 female; age range 19–27, mean (SD) 22.76 (2.75)). Lifetime MDMA use varied widely (mean 72.4 lifetime occasions), with most participants classified as mild to moderate users; when heavy users were excluded the mean lifetime use fell to 13 occasions. Inclusion criteria included age 18–35, history of MDMA use, good physical health and BMI 18–28; exclusion criteria included DSM-IV addiction, psychiatric/neurological disorder, pregnancy and cardiovascular abnormalities. Medical screening and standard blood chemistry/haematology were completed. Ethical approval and regulatory permits for MDMA were obtained. Each subject completed six experimental sessions separated by at least seven days. Sessions followed a two-drug scheme: a pretreatment (T1) given 30 minutes before treatment (T2). The six T1–T2 combinations were placebo–placebo, pindolol 20 mg–placebo, ketanserin 50 mg–placebo, placebo–MDMA 75 mg, pindolol 20 mg–MDMA 75 mg, and ketanserin 50 mg–MDMA 75 mg. Doses were chosen to approximate therapeutic receptor blockade: ketanserin reportedly blocks about 91% of 5-HT2 receptors and pindolol blocks about 40% of 5-HT1A receptors. Conditions were balanced across subjects and a double-dummy technique synchronised Tmax. Subjects refrained from other drugs for at least a week prior to sessions; urine drug screens and pregnancy tests were performed before dosing. Memory was assessed at the expected Tmax of MDMA (1.5–2 h after T2). The test battery comprised three tasks sensitive to MDMA effects: a Word-Learning Task (WLT; Dutch clinical version) measuring immediate recall over three trials and recognition after a 30-min delay (primary parameter: sum score of correct recalls across three trials, plus recognition accuracy and reaction time); a Spatial Memory Task assessing short-term non-verbal localisation memory (75 trials with 0, 2 or 4 s delays; main outcomes: mean localisation error and reaction time); and an event-based Prospective Memory Task (240 trials of a letter discrimination foreground task with occasional embedded prospective cues; main outcome: prospective memory failures). Vital signs (blood pressure, body temperature) were recorded as safety measures. Blood samples for pharmacokinetic assays (MDMA, pindolol, ketanserin) were taken before the memory tasks at 1.5 h post-T2 and analysed by GC–MS. Statistical analysis focused on testing interactions between pretreatments and MDMA using two separate repeated-measures General Linear Model (GLM) ANOVAs. GLM 1 examined factors MDMA (present/absent) and ketanserin (present/absent); GLM 2 examined MDMA (present/absent) and pindolol (present/absent). Significant interactions were followed by drug–placebo contrasts to clarify the nature of effects. The alpha level was set at P = 0.05 and analyses were performed in SPSS version 15.0. Safety measures were analysed using the same GLM approach.

Seventeen complete data sets were analysed. Across the memory battery, single doses of MDMA 75 mg produced impairments on all tasks: MDMA reduced immediate recall on the Word-Learning Task (WLT), increased prospective memory failures on the prospective task, and increased localisation error on the spatial memory task. These task-level effects are reported in the GLM analyses and are consistent with prior intoxication studies. Prospective memory: MDMA significantly increased the number of prospective memory failures in the No-Go trials in one GLM comparison (F1,16 = 7.8; p = 0.013) and showed a trend in the other GLM (F1,16 = 4; p = 0.06). Neither ketanserin nor pindolol, nor their interactions with MDMA, significantly affected prospective memory performance. Spatial memory: MDMA significantly increased localisation error in both GLM comparisons; the extracted text truncates the detailed statistics for this outcome. Pretreatment with ketanserin or pindolol did not prevent the MDMA-induced increase in localisation error. Verbal memory (WLT): The GLM analyses revealed a significant interaction between ketanserin pretreatment and MDMA on verbal learning. Pretreatment with ketanserin prevented the MDMA-induced reduction in immediate recall on the WLT. Subsequent drug–placebo contrasts indicated that placebo–MDMA and pindolol–MDMA combinations both produced significant reductions in immediate recall, whereas the ketanserin–MDMA combination did not differ from the placebo–placebo condition. In other words, ketanserin selectively blocked MDMA-induced impairment of verbal learning, whereas pindolol did not alter MDMA’s effect on verbal memory. Pharmacokinetic data: blood concentrations of MDMA, ketanserin and pindolol were measured and mean concentrations are reported in a table in the original text; the extracted text does not provide those numerical values. Safety measures (blood pressure, body temperature) were assessed and showed no interactions between pretreatments and MDMA in the analyses reported.

Van Wel and colleagues interpret the findings as evidence that acute MDMA impairs multiple memory domains during intoxication and that stimulation of 5-HT2A receptors plays a central role in MDMA-induced verbal memory impairment. The selective prevention of WLT impairment by ketanserin supports the hypothesis that direct or indirect activation of 5-HT2A receptors mediates the verbal memory decrement observed with MDMA, whereas lack of effect of pindolol argues against a crucial role for 5-HT1A receptors in these acute memory effects. The authors place their results in the context of prior imaging and receptor studies, noting that alterations in 5-HT2A receptor densities have been observed in recent and ex-MDMA users and suggesting that acute 5-HT2A stimulation during intoxication may contribute to observed performance decrements. They also acknowledge that involvement of 5-HT2A receptors may be task-specific: ketanserin selectively protected verbal (WLT) performance but did not ameliorate MDMA-induced impairments in spatial or prospective memory, which could reflect differences in the neural networks and cortical layer distributions of 5-HT1A and 5-HT2A receptors. Important limitations noted by the investigators include the pharmacological profile of pindolol, which is not a selective 5-HT1A antagonist but primarily a beta-blocker with approximately 40% 5-HT1A blockade; selective and full 5-HT1A antagonists for human use were not available. Consequently, negative pindolol findings do not definitively exclude any contribution of 5-HT1A receptors. The authors also caution that the present data do not resolve whether acute receptor-mediated effects and persistent memory deficits seen in chronic users share the same underlying mechanisms. Finally, they suggest future work examining genetic variation in 5-HT2A (for example the His452Tyr polymorphism) as a potential moderator of MDMA-induced verbal memory impairment. The paper concludes that their data demonstrate MDMA-induced verbal memory impairment is mediated by 5-HT2A receptor stimulation, while noting the selectivity and limitations of this conclusion.