"This is your brain (an egg); this is your brain on drugs (an egg frying in a pan)."
This "fried egg/fried brain" analogy has long been used to depict the adverse
consequences of drug use.  But just what are the neurotoxic effects
of drugs used in a recreational or a therapeutic context?  Clearly, in the
field of MDMA research, the term "neurotoxicity" has been very broadly
applied to describe the effects of the drug in both experimental animals and
man.  Unfortunately, there has been very little effort to define what is
meant by MDMA neurotoxicity much less to distinguish MDMA's  "neurotoxic"
actions from its potential to cause neuropathological effects, i.e. effects
associated with degenerative disorders of the nervous system.    In short,
everybody talks about drug-induced neurotoxicity but little attempt is made
to define it in terms meaningful to the human condition.  In this chapter, I
will briefly address specific aspects of the MDMA neurotoxicity issue by
looking at definitions of this term as well as some different types of
effects of MDMA in animals (and humans), and I will talk about functional
outcomes.  I won't address dosage regimen per se but I will touch on some
cross species extrapolation issues.

Brain damage vs. neurotoxicity:

 So just what is neurotoxicity in the context of the effects of MDMA use?
For example, can this drug damage the brain?  For that matter, just what is
brain damage and can chemicals and drugs actually damage the brain?
These all are relevant issues with respect to understanding potential risks associated
with the use of MDMA or any other agent that affects the nervous system.
 When one speaks of brain damage, it is usually equated with neuropathology.
Thus, traumatic injury or neurological diseases such as Alzheimer's,
Parkinson's, Huntington's or multiple sclerosis all have distinct
neuropathological underpinnings defined by changes in brain cells viewed
under a microscope.  These neuroanatomical abnormalities serve as the
structural (brain cell) basis for the functional deficits associated with a
given condition.  No one will argue that the neuropathological effects
underlying the devastating symptoms associated with diseases or trauma of
the nervous system do NOT constitute brain damage.  Likewise, on the
surface, it would seem safe to assume that acute or long-term administration
of MDMA would not be linked to underlying neuropathology (brain damage) in
the absence of some neurological symptoms.  Unfortunately, this assumption
would be erroneous on at least two counts.  First, owing to the functional
reserve of the nervous system, damage or even near complete destruction of a
given brain area is not necessarily associated with loss of brain function.
For example, in victims of Parkinson's disease, loss of upwards of 80% of
the target neurons is required before characteristic symptoms emerge.  A
second consideration is the fact that damage to the putative target of MDMA,
the serotonin-containing neurons, may not be obvious due to our lack of
understanding as to the function of this component of the nervous system.
Thus, it is possible that repeated administration of MDMA over long periods
of time may result in subtle brain pathology (damage) and such damage
eventually may be manifested by subtle changes in behavior, mood,
learning or memory, to name a few effects associated with the serotonergic
nervous system.
 When one encounters the term neurotoxicity, this usually is in reference to the
effects of chemical exposures in the environment or workplace, or with
self-administration of drugs of abuse.  The specter of chemical-induced
carcinogenesis and birth defects has long dominated public perception of
the hazards of exposure to specific chemicals and drugs (the "lumps and stumps" mentality).
There is, however, more than ample evidence in the literature for the propensity
of all classes of chemicals to damage the developing and adult nervous system,
as well as to cause cancer and birth defects.  Viewed in these terms, neurotoxicity
is synonymous with chemical-induced brain damage.  This does not discount the
fact that chemicals and drugs have actions on the nervous system that change its
biochemistry, indeed, by design that is how drugs achieve their therapeutic actions.
Yes, unwanted short to long-term changes in brain chemistry can be viewed as
"neurotoxic" because these changes may represent undesired effects of the compound
(ethanol would certainly fit into this definition).  Viewed in these latter terms,
however, neurotoxicity cannot be equated with brain damage in the absence
of evidence for neuropathology.  This is the crux of the arguments
surrounding the effects of MDMA (see below).  For the purposes of the
arguments put forth here, a chemical will be considered to be neurotoxic if
it causes brain damage.
 Neurotoxic episodes in man have been well documented following the ingestion
of tainted food (e.g. domoic acid), the application of tainted acne
medication (e.g. triethyltin) or the exposure to industrial solvents and metals
(e.g. mercury and carbon disulfide) to name a few.  All of these exposures have
been associated with deaths and were linked to the neuropathological effects
of the offending agent.  It is not a question as to whether some chemicals are
neurotoxic but rather which chemicals preferentially attack the nervous
system and at what exposure levels.  Moreover, owing to the extreme cellular
complexity of the nervous system, one cannot predict which brain area or
cell type will be vulnerable to a given neurotoxic chemical or whether
symptoms of exposure will be overt or hidden (see O'Callaghan, 1995).
Nor can one assume that the neurotoxic effects of a drug are just dose-related
extensions of its pharmacology.
 For example, therapeutic dosages of a drug known as MK-801, an anti-seizure
medication, antagonize the toxic actions of excessive levels of the neurotransmitter
glutamate by blocking its receptors throughout the brain.  At high dosages, MK-801
has been shown to destroy neurons in a small area of cerebral cortex, a brain region
unrelated to the sites of its therapeutic actions.  By analogy, even though the
psychostimulant actions of MDMA may be mediated through the serotonergic
nervous system, there is no reason, a priori, to assume that serotonergic neurons
would be affected at doses that might be toxic to other brain systems (or other organs).
All of this may seem confusing but the facts of the matter are dictated by our neurobiological
make-up which, in turn, predicts the following:  1) chemicals and drugs can damage the brain;
2) areas of the brain that mediate the desired pharmacological effects may not be the areas
vulnerable to toxicity and 3) subtle brain damage can occur in experimental animals and in
man in the absence of overt symptoms.

How do you detect subtle damage of the brain?

 If one accepts the notion that chemicals can damage the brain and that
damage constitutes neurotoxicity, then all that is needed to assess neurotoxicity
are the appropriate techniques.  As with the evaluation of traumatized or diseased brains,
neuroanatomical methods have remained the dominant means for detection and characterization
of neurotoxicity.  Where cells are killed outright by the offending agent, it is possible to visualize
the damaged areas with tissue stains that have been in common usage for more than a century.
Of course, under these circumstances, the functional deficits associated with the loss of cells
already may have provided the clues to point to a neurotoxic exposure.  This situation is unlikely
to describe the real-world situation.  Here, the greatest concern is directed toward detecting and
preventing the cumulative damage that occurs following protracted exposures to chemicals or drugs
whose damage is not initially detected using traditional neuroanatomical stains.  Examples would be
drugs or chemicals that kill only a few cells, where the surviving cells would be in far greater
numbers than those that were destroyed (i.e. like looking for the "needle in the haystack").  Perhaps
an even more likely situation would be chemical destruction of parts of neurons with sparing of the nerve cell itself.
This is the case put forth for MDMA neurotoxicity (see below).  Under both of these scenarios,
the selective and discrete nature of neurotoxic effects dictates the need for special
techniques/indicators to identify cells damaged (but not necessarily killed)
by a given neurotoxic agent.  This is not an easy task because, as mentioned
above, while the targets of neurotoxic insults may be limited to a very
small area of the nervous system, any area of the nervous system may be
affected.  For the small drop of damaged brain to be detected within the sea
of unaffected tissue, it requires an indicator of neurotoxicity that possess
several features, as follows: 1) it must reveal diverse types of injuries to
any area of the nervous system, 2) it must be sensitive to low levels of
damage, 3) it must be specific to the damage (neurotoxic) condition so that
therapeutic effects of drugs are not falsely attributed to adverse effects.  Succinctly stated,
the ideal neurotoxicity endpoint would be an indicator of damage at any level anywhere in
the nervous system that would not pick up therapeutic actions of drugs.
 The propensity of the damaged brain to cause enlargement of a specific cell type known as the
astrocyte and for damaged neurons to become impregnated with silver (argyrophilia) are
two of only a handful of generic indicators of brain damage, regardless of the causative agent.
Astrocyte enlargement, known as astrogliosis, refers to the reaction of this brain cell type to all
types of brain injury.   The hallmark of this response is the accumulation of a protein within astrocytes
known as glial fibrillary acidic protein (GFAP).  Increases in GFAP, therefore, serve as an indicator
of astrogliosis and, by extension, of neurotoxicity.  Increased GFAP expression can be examined in
slides of brain tissue with antibodies that recognize this protein.  Alternatively GFAP levels in samples of
brain tissue can be measured by sensitive immunoassays. Elevations in GFAP are widely accepted as indicators of brain damage associated with neurological diseases such as Alzheimer's and multiple sclerosis.  More recently, enhanced expression of GFAP has been validated as an indicator of  neurotoxicity by using a wide variety of prototype chemical neurotoxicants.  These include agents that damage many regions of the brain and many different cell types within a brain region, as would be expected to occur under "real-world" conditions.  Moreover, increases in GFAP reveal subtle damage to neurons, such as loss of nerve endings, under conditions where traditional neuropathological stains fail to reveal the damage. Importantly, GFAP levels do not change with pharmacological agents administered at therapeutic dosages.  Thus, GFAP assessments fulfill the desired requirements for an indicator of neurotoxicity.
 As with the increases in GFAP associated with chemical-induced
neurotoxicity, staining of brain cells with special silver degeneration
stains can be used to show regions and cell types damaged by neurotoxic
exposures.  Silver stains are not as extensively validated as GFAP as indicators of neurotoxicity.  Where silver stains have been used, however, they show at least equal sensitivity to GFAP, they reveal sites of damage in the absence of overt cytopathology as assessed by traditional neuroanatomical methods and drug effects do not screen positive.  When analysis of GFAP is coupled with mapping of argyrophila using silver stains, there is a remarkable correspondence between the regional and cellular patterns of neurotoxicity revealed by the two techniques.  Thus, it is likely that the two approaches to neurotoxicity assessment represent specific and sensitive methods for the assessment of all types of neurotoxic exposures.

What is serotonin neurotoxicity?

 In light of the points made above, it would seem sensible to apply GFAP
assays and silver degeneration stains to determine whether MDMA is
neurotoxic.  This has been done in a number of laboratories and I will
elaborate on some of the findings below.  First, however, it is useful to
review the context within which MDMA is considered to be neurotoxic.  In
almost all studies using experimental animals and humans, MDMA is described
as a serotonin neurotoxin.  What does this mean?  Serotonin neurotoxicity
implies that MDMA damages the serotonergic nervous system.   Because MDMA
was known to release serotonin from the nerve endings of serotonin
containing neurons in experimental animals, these serotonergic neurons were viewed as the presumed targets of any neurotoxic effects of the drug.
Indeed, subsequent measurements of serotonin levels after administration of
high dosages of MDMA to rats showed  weeks-long decreases in this
neurotransmitter.  Further, measurements of the enzyme that catalyzes the
synthesis of serotonin (tryptophan hydroxylase) and of the protein that
transports the released serotonin back into the nerve endings (serotonin
transporter) also showed reductions as a result of high doses of MDMA.
Because these three constituents of serotonin nerve endings all were reduced
for long periods of time (weeks to months) as a result of large doses
of MDMA, these changes were viewed as evidence of serotonin neurotoxicty,
i.e. MDMA-induced brain damage.
 There is little argument that the protracted decreases in the constituents of serotonergic neurons resulting from the acute or chronic administration of MDMA are not drug-like (subjective) actions of the compound that serve as the basis for its self-administration.  Moreover, the persistent nature of the decreases in these serotonergic endpoints could be considered manifestations of toxicity, at least at a metabolic level within serotonergic neurons.  Over the past decade, however, there has been broad recognition of the malleability of the adult nervous system.  This "plasticity" certainly extends to the serotonergic nervous system and to its components affected by MDMA.  For example, it is now known that treatment with antidepressants such as paroxetine (Paxil) or fluoxetine (Prozac) can decrease the numbers of serotonin transporters as can a condition that does not even involve exposure to a drug: a food restriction diet. Pharmacotherapy with antidepressants such as fluoxetine or dieting are not conditions one often associates with neurotoxicity.   Because MDMA and Prozac share the propensity to decrease the serotonin transporter suggests that MDMA can be viewed as much as an antidepressant agent as a "serotonin neurotoxin."  Taken together, these observations indicate that changes (even long-term changes) in markers of serotonergic neurons are likely a reflection of neuronal plasticity, i.e. adaptive changes that occur in response to drug therapy in an otherwise intact neuron.   Thus, alterations in parameters associated with the functioning of serotonin neurons can not be taken as evidence of neurotoxicity, in the absence of evidence for serotonin neuron pathology.  Not only are such "markers" of serotonergic neurons not useful as stand-alone measures of neurotoxicity, it also is quite likely that current medications may induce changes in these markers and that such changes would be the expected effects of the long-term therapeutic actions of  these drugs.

Application of silver stains and GFAP analysis for the assessment of
MDMA-induced neurotoxicity

        As noted above, one way to resolve the controversy as to whether
MDMA is neurotoxic to serotonin neurons would be to use sensitive and
selective indicators of neurotoxicity such as silver stains and GFAP
analysis.  When a dose of MDMA (20 mg/kg) that caused 50% decreases in brain
serotonin was administered to the rat, it failed to increase GFAP or result
in silver staining at any point after dosing.  Daily dosages of up to
30 mg/kg for a week also did not increase GFAP.  Only when given at fairly
enormous dosages to the rat (4 x 50 mg/kg over 24 hours)  was evidence of
damage obtained.  Even under these circumstances,  however, increases in
GFAP and silver staining were observed in the cortex but the damaged areas
were not those associated with serotonin neurons.  These findings indicated
that only massive doses of MDMA can cause damage to the brain of the rat and
that the damage that occurs is not related to the serotonin nervous system.
The implication of these findings is two-fold: 1) changes in markers of
serotonin neurons can occur independent of damage to these neurons and 2)
large doses of MDMA are required to damage the nervous system of the rat,
approximately 100 times the human dosage taken in a recreational context.
One easy explanation for the failure to see damage to serotonin neurons
after MDMA, as assessed by assaying GFAP or using silver stains, was that
the techniques were not sensitive enough.  To address this issue, the known
serotonin neurotoxin, 5,7-dihydroxytryptamine, was administered to the rat
at a dosage that produced decreases of serotonin equivalent to those seen
with MDMA.   This resulted in large increases in GFAP (40-100%) in the areas
of the brain where serotonin was decreased and these effects were
accompanied by silver staining.  These findings indicate the sensitivity of
GFAP and silver staining as indices of chemical-induced damage to serotonin
neurons.  Thus, hallmarks of brain damage occur after damage to serotonin
neurons but are absent following administration of high dosages of
MDMA.
 

Lessons from other compounds and from man

      Lessons learned using experimental animals don't always apply to man,
therefore, the absence of evidence (negative data) cited above is not
evidence for absence of neurotoxic effects of MDMA in man.  In no small
measure this often is why sub-human primates are used in an attempt to model
more closely the effects presumed to occur in humans.  Unfortunately,
different species of sub-human primates also provide different responses to
drugs, including MDMA.  There are, however, lessons that can be learned from
human exposures to other compounds that can be applied to MDMA.  These
compounds are MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine),
methamphetamine and dexfenfluramine.  In the early 1980's a group of
individuals self-administered what they assumed was an analogue of
meperidine (a synthetic narcotic).  The compound administered turned out to
be MPTP, an unintended contaminant that had devastating consequences: many of the individuals exposed to MPTP developed symptoms of Parkinson's
disease.  Subsequent research clearly demonstrated that MPTP damaged
dopamine-containing neurons, the same ones affected in Parkinson's disease.
The MPTP episode raised the specter that similar damage would occur in
another cohort of humans, methamphetamine users, because methamphetamine acts on the same dopamine neurons damaged by MPTP and Parkinson's disease.
 Given the widespread usage of methamphetamine in the 1960's and 1970's, the potential for this drug to have damaged dopamine neurons should now be manifested with age, as is the case for Parkinson's disease.   Because the prevalence of  Parkinson's disease has not increased over the intervening decades suggests that methamphetamine is unlikely to have had a neurotoxic effect on human dopamine neurons.   A recent study of humans exposed to methamphetamine  bears more directly on this issue and provides data more relevant to human MDMA users.  This study involved post-mortem examination of brains from verified methamphetamine users.  Marked decreases in markers of dopamine neurons were found in these brains including dopamine, the enzyme that catalyzes its formation (tyrosine hydroxylase) and the dopamine transporter.  If the decreases in these markers were a reflection of damage to dopamine neurons, then this would have been manifested as symptoms of Parkinson's disease prior to death.  None of these individuals had such symptoms.  Thus, as was the case for markers of serotonin neurons in rats (and, recently, humans) exposed to MDMA, the data for dopamine markers in human methamphetamine users was indicative of an adaptive change in response to the drug rather than a neurotoxic action on the neuron.
 The anorectic agent, dexfenfluramine, is the final example of a human exposure relevant to MDMA.  Although this compound recently was taken off the market due to reports of abnormal heart function, it was also the subject of a controversy involving neurotoxicity.  This stems from the fact that in rats, MDMA and dexfenfluramine have nearly identical actions on serotonin neurons.  As a consequence, dexfenfluramine, like MDMA, became a suspect "serotonin neurotoxin."  Unlike MDMA, dexfenfluramine or its racemate, fenfluramine, have been taken by millions of patients worldwide for 20 years.  Extensive post-marketing patient surveillance has yet to reveal any repercussions of dexfenfluramine use that can be linked to adverse effects on the nervous system.  The data for human exposures to methamphetamine and fenfluramine are consistent with results of research on these agents using experimental animals.  They indicate that these agents have the potential to alter biochemical markers of dopamine and serotonin neurons without causing neurotoxicity.  Given the similarities between the effects of these
compounds and MDMA, it is likely that we can infer that MDMA shares similar
actions in man.
        Despite nearly two decades of research on the neurotoxic properties
of substituted amphetamines, including MDMA, no definitive data have been
obtained to indicate that these compounds are in fact neurotoxic to man.
This interpretation of existing data does not constitute an endorsement of
the recreational or therapeutic use of these compounds.  Rather, it is a
call for continued research on the adaptive responses that these compounds
engender in their target neurons.  If we are to understand the potential for
these drugs to cause neurotoxicity, then we must understand the significance
of their long-term effects in relation to their potential to alter the nervous system.
 
 

References

Adams, J.H. and Duchen, L.W.: Greenfield's Neuropathology, 5th Edition,
Oxford University Press,  Oxford, 1992.

Costa, L.G. and Manzo, L.: Occupational Neurotoxicology, CRC Press, Boca
Raton, 1998

Curzon, G.: Serotonin and appetite.  Ann. NY Acad. Sci. 600: 521-531, 1990

Erinoff, L.: Assessing neurotoxicity of drugs of abuse. NIDA Res. Monograph
136, NIH publication No. 93-3644, U.S. Government Printing Office, 1993.

Fix, A.S., Long, G.G., Wozniak, D.F., Olney, J.W.: Pathomorphologic effects
of N-methyl-D-aspartate antagonists in the rat posterior cingulate/retrosplenial cerebral cortex: a review.  Drug. Dev. Res. 32: 147-152, 1994.

Fix, A.S., Wrightman, K.A. and O'Callaghan, J.P.:  Reactive gliosis induced
by MK-801 in the rat posterior cingulate/retrosplenial cortex:  GFAP
evaluation by sandwich ELISA and immunocytochemistry.  Neurotoxicology 16: 229-238, 1995.

Gorzalka, B.B., Mendelson, S.D., Watson, N.V.: Serotonin receptor subtypes
and sexual behavior.  Ann. NY Acad. Sci. 600: 435-446, 1990.

Green, A.R., Cross, A.J. and Goodwin, G.M.:  Review of the pharmacology and
clinical pharmacology of 3,4-methylenedixymethamphetamine (MDMA or
"ecstasy"). Psychopharmacology 119: 247-260, 1995

Guy-Grand, B., Apfelbaum, M., Crepaldi, G., Gries, A., Lefebvre, P., and
Turner, P.:  International trial of long-term dexfenfluramine in obesity.
Lancet, Nov. 11: 1142- 1145, 1989

Huether, G., Zhou, D., Schmidt, S., Wiltfang, J., Ruther, E:  Long-term food
restriction down-regulates the density of serotonin transporters in the rat
frontal cortex.  Biol. Psychiatry 41: 1174-1180, 1997).

Insel, T.R.: 3,4-Methylenedioxymethamphetamine ("ecstasy") selectively
destroys brain serotonin terminals in Rhesus monkeys.  J. Pharmacol. Exp. Ther.  249: 713-720, 1989.

Jellinger, K.:  The pathology of parkinsonism.  In: Marsden, C.D. and Fahn,
S. (eds.)  Movement Disorders, Vol. 2, pp. 124-165, Butterworth Press, London,
1987.

Jensen, K.F., O'Callaghan, J.P., Miller, D.B. and Olin, J.K.: Mapping toxicant-induced nervous system damage with a cupric silver stain: A quantitative
analysis of neuronal degeneration induced by 3,4 methylenedioxymethamphetamine (MDMA).  In: Assessing Neurotoxicity of Drugs of Abuse, NIDA Research Monograph 136, NIH  Publication No. 93-3644, ed. By L. Erinoff, pp. 133-154, U.S. Government Printing Office, 1993.

Kimbrough, R.D.: Toxicity and health effects of selected organotin
compounds:  a review.  Environmental Health Perspectives 14: 51-65, 1976.
Langston, J.W., Koller, W.C., and Giron, L.T.: Etiology of Parkinson's
disease. In: The Scientific Basis for the Treatment of Parkinson's Disease. , The
Parthenon Publishing Group, Park Ridge, N.J., pp. 33-58, 18992.

Luthman, J., Anderson, H. and Olson, L.: Environmental neurotoxicology: A
review on identification and action of neurotoxicants.  Technology: J. of the
Franklin Institute 332A: 151-182, 1995.

McCann, U.D., Seiden, L.S., Rubin, L.J., Ricaurte, G.A.:  Brain serotonin
neurotoxicity and primary pulmonary hypertension from fenfluramine and
dexfenfluramine: a  systematic review of the evidence. JAMA 278: 666-672, 1997.

McCann, U.D., Szabo, Z., Scheffel, U., Dannals, R.F., and Ricaurte, G.A.:
Positron emission tomographic evidence of toxic effect of MDMA ("ecstasy")
on brain   serotonin neurons in human beings.  Lancet 352: 1433-1437, 1998.

McKenna, D.J. and Peroutka, S.J.: Neurochemistry and neurotoxicity of 3,4-
methylenedioxymethamphetamine (MDMA, "Ecstasy").  J. Neurochem. 54:
14-22, 1990.

Meltzer, H.Y.: Role of serotonin in depression.  Ann. NY Acad. Sci. 600:
486-500, 1990.

Monaghan, D.T., Bridges, R.J., Cotman, C.W.: The excitatory amino acid
receptors: Their classes, pharmacology, and distinct properties in the function
of the central   nervous system.  Annual. Rev. Pharmacol. Toxicol. 29: 365-402, 1989.

O'Callaghan, J.P.: Quantitative features of reactive gliosis following
toxicant-induced damage of the CNS.  Ann. NY. Acad. Sci. 679: 195-210, 1993.

O'Callaghan, J.P. and Miller, D.B.: Assessment of chemically-induced
alterations in  brain development using assays of neruon- and glia-localized
proteins. Neurotoxicology 10: 393-406, 1989.

O'Callaghan, J.P., and Jensen, K.F.:  Enhanced expression of glial
fibrillary acidic protein and the cupric silver degeneration reaction can be used as sensitive and   early indicators of neurotoxicity.  Neurotoxicology 13: 113-122, 1992.

O'Callaghan, J.P. and Miller, D.B.: Quantification of reactive gliosis as an
approach to neurotoxicity assessment. In Assessing Neurotoxicity of Drugs of Abuse. NIDA Research Monograph 136, NIH Publication No. 93-1944, ed. By L.
Erinoff, pp. 188- 212, U.S. Government Printing Office, Washington, D.C., 1993

O'Callaghan, J.P.,  Jensen, K.F., and Miller, D.B.:  Quantitative aspects of
drug and toxicant-induced astrogliosis.  Neurochem. Int. 26: 115-124, 1995.
Olanow, C.W. and Lieberman, A.N.:  The Scientific Basis for the Treatment of
Parkinson's Disease, The Parthenon Publishing Group, Park Ridge,
N.J., 1992.

Piñeyro, G., Bilier, P., Dennis, T., and de Montigny, C.:  Desensitization
of the neuronal 5-HT carrier following its long-term blockade.  J. Neurosci. 14:
3036-3047, 1994.

Perl, T.M., Bédard, L., Kosatsky, T., Hockin, J.C., Todd, E.C.D. and Remis,
R.S.: An outbreak of toxic encephalopathy caused by eating mussels
contaminated with domoic acid, N. Engl. J. Med. 322: 1775-1780, 1990.

Ricaurte, G.A., Forno, L.S., Wilson, M.A., DeLanney, L.E., Irwin, I.,
Molliver, M.E., and Langston, J.W.:  (+)3,4-Methylenedioxymethamphetamine selectively damages central serotonergic neurons in nonhuman primates.  JAMA 260: 51-55,1988.

Sayre, L.M.: Biochemical mechanism of action of the dopaminergic neurotoxin
1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP).  Toxicol. Lett.
48: 121-149,  1989.

Sprague, J.E., Everman, S.L., and Nichols, D.E.: An integrated hypothesis
for the  serotonergic axonal loss induced by 3,4-methylenedioxymethamphetamine. Neurotoxicology 19: 427-442, 1998.

Steele, T.D., McCann, U.D., Ricaurte, G.A.: 3,4-Methylenedioxymethamphetamine
(MDMA, "Ecstasy"): pharmacology and toxicology in animals and humans.
Addiction 89: 539-551, 1994.

Wilson, J.M., Kalasinsky, K.S., Levey, A.I., Bergeron, C., Reiber, G.,
Anthony, R.M., Schmunk, G.A., Shannak, K., Haycock, J.W. and Kish, S.J.:  Striatal dopamine nerve terminal markers in human, chronic methamphetamine users.  Nature Medicine 2: 699-703, 1996.

Zaczek, R., Battaglia, G., Culp, S., Appel, N.M., Contrera, J.F. and
DeSouza, E.B.: Effects of repeated fenfluramine administration on indexes of
monoamine function in rat brain:  Pharmacokinetic, dose response, regional specificity and time course data. J. Pharmacol. Exp. Ther. 253: 104-112, 1990.