Retraction: "Severe dopaminergic neurotoxicity in primates after a common recreational dose regimen of methylenedioxymethamphetamine (MDMA)" (Science. 2002;297:2260-2263.)
September 5, 2003
With comments by Ilsa Jerome, Ph.D.
We write to retract our report "Severe dopaminergic neurotoxicity in primates after a common recreational dose regimen of methylenedioxymethamphetamine (MDMA)" (1) following our recent discovery that the drug used to treat all but one animal in that report came from a bottle that contained d-methamphetamine instead of the intended drug, racemic MDMA. Notably, d-methamphetamine would be expected to produce the same pattern of combined dopaminergic/serotonergic neurotoxicity (2) as that seen in the animals reported in our paper (1).

The originally published report (1) presented results from multiple studies performed in our laboratory over a span of approximately two years demonstrating that a novel systemic dose regimen of what we believed was MDMA produced severe dopamine neurotoxicity in two species of nonhuman primates, in addition to the previously reported serotonin neurotoxicity (3-6). Subsequent to the publication of those findings, we were unable to extend the dopamine neurotoxicity to orally administered doses. Multiple subsequent attempts to reproduce the original findings with systemically administered doses of MDMA identical to those used in the original study were also unsuccessful, under a variety of laboratory conditions.

We then noted that our studies aimed at extending and replicating the finding of MDMA-induced dopamine neurotoxicity were performed using a new batch of MDMA. This new batch of MDMA was determined to be authentic by several methods, including gas chromatography/mass spectrometry (GC/MS). Upon investigation of our laboratory records, we determined that the studies detailed in our paper (1) utilized a batch of MDMA that had been requested on the same date as a batch of d-methamphetamine, and that both drug requests were for the same amount (10 g), and were processed by the supplier on the same day. Both drugs were delivered to our laboratory on the same day, in the same package. At delivery, the two bottles had different affixed labels, the same delivery reference number, but different batch numbers, as specified in their respective chemical data sheets. Following receipt, both drugs were stored in our laboratory in their original containers, in a locked safe.

When we began to suspect that the two bottles of drug might have borne incorrect labels (i.e., that the putative MDMA was actually d-methamphetamine, and vice versa), we requested that a sample of the drug in the bottle bearing the original and intact label of "d-methamphetamine" be analyzed by various analytical techniques, including GC/MS. Three independent laboratories found the sample to consist of MDMA, with no evidence of even trace amounts of methamphetamine.

Although the drug sample used in our original studies (1) was depleted and the empty bottle labeled MDMA had been discarded, we did have frozen brains from two animals that died shortly after drug treatment during the course ofthe original experiments (1). When these brains were analyzed by GC/MS by three independent laboratories, they were found to contain methamphetamine and its metabolite, amphetamine, neither of which is a metabolite of MDMA(7). Not even trace amounts of MDMA nor its metabolite, MDA, were found inthese brains. Detailed review of our laboratory records revealed that allbut one animal in our study (1) had been treated with the drug in the bottle labeled "()methylenedioxymethamphetamine" (MDMA) processed at the same time as the bottle labeled "d-methamphetamine".

This labeling error does not call into question the results of multiple previous studies demonstrating the serotonin neurotoxic potential of MDMA invarious animal species, including several nonhuman primate species (3-6,8). Regarding the dopamine neurotoxic potential of MDMA in nonhumanprimates, it remains possible that dose regimens in the range of those used by some humans, but different from those thus far tested, produce dopamine neurotoxicity in primates, as they do in rodents (9, 10). Moreover, lasting effects of MDMA on dopaminergic function in humans have recently been reported (11), and some humans with a history of MDMA abuse have developed Parkinsonism (12-14). However, until the dopamine neurotoxic potential ofMDMA in primates can be examined more fully, this possibility remains uncertain.

George A. Ricaurte1, Jie Yuan1, George Hatzidimitriou1, Branden J. Cord2, Una D. McCann3
1 - Department of Neurology, 2 - Department of Neurosciences, 3 - Department of Psychiatry
Johns Hopkins Bayview Medical Center
Johns Hopkins University
School of Medicine, Baltimore, MD 21224, USA.


1. G. Ricaurte, J. Yuan, G. Hatzidimitriou, B. Cord, U. McCann, Science 297, 2260 (2002).

2. V. Villemagne et al., J Neurosci 18, 419 (1998).

3. G. A. Ricaurte, L. E. Delanney, I. Irwin, J. W. Langston, Brain Res. 446, 165 (1988).

4. T. R. Insel, G. Battaglia, J. N. Johannessen, S. Marra, E. B. DeSouza, J. Pharmacol. Exp. Ther. 249, 713 (1989).

5. M. S. Kleven, W. L. Woolverton, L. S. Seiden, Brain Res. 488, 121 (1989).

6. D. L. Frederick et al., Neurotoxicol. Teratol. 17, 531 (1995).

7. A. Cho, Y. Kumagai, in Amphetamine and its Analogs: Neuropsychopharmacology, Toxicology and Abuse, A. Cho, D. Segal, Eds. (Academic Press, New York, 1994), pp. 43-77.

8. W. Slikker Jr. et al., Toxicol. Appl. Pharmacol. 94, 448 (1988).

9. D.L. Commins et al., J. Pharmacol. Exp. Ther. 241, 338(1987).

10. E. O'Shea, B. Esteban, J. Camarero, A. R. Green, M. I. Colado, Neuropharmacology. 40, 65 (2001).

11 G. Gerra et al. Behav. Brain Res. 134, 403 (2002).

12. S. Mintzer, S. Hickenbottom, S. Gilman, N. Engl. J. Med. 340, 1443 (1999).

13. S.M. Kuniyoshi and J. Jankovic N Engl J Med. 349, 96 (2003).

14. G. Ricaurte, unpublished data.

15. We gratefully acknowledge helpful discussions with Dr. Jonathan Katz of the NIDA Intramural Research Program, Baltimore, MD, Dr. W. Lee Hearn, Laboratory Director of the Miami-Dade County Medical Examiner Department, Miami, FL and Dr. Nancy Ator, Johns Hopkins University School of Medicine, Baltimore, MD. We are also grateful for the expert chemical analytical assistance of Ms. Rebecca Fernandez, Toxicologist I, Miami-Dade County Medical Examiner Department, Miami, FL, Terry L. DalCason of the DEA North Central Laboratory, Chicago, IL., Dr. Max Courtney, Forensic Consultant Services, Fort Worth, TX., Dr. Michael Daggett, Quest Diagnostics-Nichols Institute, Chantilly, VA, Dr. Ivy Carroll and associates at RTI International, Research Triangle Park, NC, and Dr. Roger Foltz at the University of Utah, Salt Lake City, UT.

Ilsa's comments about the retraction:

"The Gerra piece they are citing as evidence of dopaminergic toxicity in humans is flimsy stuff. It is a neuroendocrine challenge study, and this generally raises my suspicion, as neuroendocrine parameters are multiply determined to start with. Second of all, since (at least to my understanding) 5HT2A receptors are linked with dopaminergic tone, and Reneman has found transient changes in 5HT2A receptors, it is possible that what we are seeing is a result of that serotonin change. Or it could even be related to pre-existing conditions, since the 12 male Ecstasy users had greater prevalence of personality disorders. Neuroendocrine challneges are even less telling than transporter imaging studies. And, as you noted, this was in men only. The only virtue of Gerra's work is that he is very diligent about abstinence from drug use, as everyone is tested twice a week throughout the three-week period of abstinence. Naturally Ricaurte doesn't mention this in his retraction; you might be led to think it is an actual imaging study or a diagnosis of movement disorders. (At least he didn't cite Parrott's on-line study).

You can find out more about this study here:

Here is what I said about the study in Overall Effects:

Overall Effects: Male ecstasy users and non-drug users both exhibited an increase in prolactin release after bromocryptine challenge, but only non-drug user controls exhibited a significant increase in growth hormone post-challenge. There was only a trend for increased growth hormone after bromocryptine challenge in ecstasy users. Ecstasy users scored higher on the TPQ Novelty Seeking scale, HDRS depression rating scale and MMPI D (depression) scale than non-drug users, and they also had higher scores on some (but not other) BDHI scale scores. A greater number of ecstasy users were diagnosed with personality disorders than were non-user controls, including borderline, avoidant and (partial diagnosis) anti-social disorder. While higher prolactin AUC after bromocryptine was associated with lower novelty seeking in members of both groups, higher growth hormone AUCs were associated with greater novelty seeking only in non-drug users. Lower growth hormone AUCs were associated with a higher total (lifetime) number of exposures to ecstasy, an indication that blunted growth hormone response to the D2 agonist bromocryptine was associated with greater use of ecstasy. The authors contend that blunted growth hormone response without blunted prolactin response has precedence in studies of people with other conditions.

Comments: To date, this is the first comparison of ecstasy users with non-drug using controls on bromocryptine challenge. As with most reports from the same team (Gerra et al. 2000; Gerra et al. 2001), this paper is notable for employing continual urinary monitoring of drug intake, and offering a stronger verification of abstinence from drug use than reports relying on self-reports or urinary analysis conducted on the study day only. The authors interpret their findings as reflecting a direct effect of MDMA on dopamine function, but acknowledge that changes in 5HT2A receptors or pre-existing differences in the dopamine system might also account for different responses to bromocryptine challenge. Others (Meltzer and Reynolds 1999) have questioned the usefulness of neuroendocrine challenge studies in reflecting brain function, since changes in neuroendocrine response to a test drug is multiply determined. Greater frequency of personality disorders in ecstasy users suggests that at least some participants in this conditions suffered from pre-existing conditions, and it is not impossible for pre-existing conditions to be associated with changes in neuroendocrine function. Study limitations include small sample size, retrospective study design, unusually self-selected participants in the ecstasy user group (individuals seeking information or treatment for substance abuse), and restriction of study participation to men. Since time since last use of ecstasy was reported at three weeks before the study day, it is possible that changes in neuroendocrine response to bromocryptine reflect a transient phenomenon, possibly related to reduced number of 5HT2A receptors. As noted in a study by Reneman et al. (2002), MDMA-associated reductions in 5HT2A receptors increase again after a period of abstinence.

Also, we learn in the retraction that studies occurred for about 2 years. This means we could double the estimates I made of grant money spent on the studies to $800,000 to $2 million, though I know little enough about animal research to tell if these are credible numbers, underestimates or overestimates.

Related Links:  Scientific Community Responds to Ricaurte Study  |  MAPS MDMA Research Page  |  MAPS in the Media