In Vitro and Non-human animals

Introduction

To date, researchers continue to publish a wealth of studies examining the pharmacology, behavioral effects, and toxicity of MDMA in non-human animals, chiefly rodents and non-human primates. MDMA neurotoxicity is most frequently addressed, but other areas include MDMA pharmacology, toxicity in other organs or tissues, developmental toxicity studies, acute, short term and long-term behavioral effects of MDMA, self-administration studies. While examinations of MDMA pharmacology and stimulus properties are two useful ways of learning more about the acute effects of the substance and how its subjective effects are experienced by nonhuman animals, these studies are for the most part irrelevant to assessing the safety of MDMA in humans. Hence this review does not review papers addressing the acute effects or pharmacology of MDMA in non-human animals unless those effects are considered relevant to assessing the safety of human trials with MDMA. However, information in these areas of study can be found in other published literature reviews (Cole et al. 2003; Green et al 2003B).

Neurotoxicity

As discussed in the 2002 revision to the IB, the majority of studies of MDMA in non- human animals concern detecting, assessing, and understanding MDMA neurotoxicity. Some publications describe novel or previously unreported means of examining the long- term effects of MDMA on the brain, whereas others look at mediating factors, such as ambient temperature and the presence or availability of antioxidants. Still others seek to establish behavioral indicators of MDMA neurotoxicity. While most of the publications appearing after the completion of the 2002 update of the IB do not offer new information on the topic, a few developments significantly alter conclusions reached in 2002. Perhaps most significantly, one of the studies reviewed in that document has since been retracted, and provocative findings from another study published in early 2004 raises issues concerning possible differences between effects produced by experimenter- administered versus self-administered drug.

Detection and Assessment

While some recent publications continue to confirm signs of damage to serotonin axons and signs of oxidative stress in the brains of rats, mice, and guinea pigs (Fornai et al. 2003; Saadat et al. 2003), other studies quality or call these findings into question (Fantegrossi et al. 2004B; Fornai et al. 2003; Pubill et al. 2003). One study detected damage to neuron bodies in rats (Schmued et al. 2003), though another failed to find signs of cell death in mouse brain (Fornai et al. 2003a). Studies of MDMA neurotoxicity have examined the brains of rats, mice (where dopamine toxicity is expected rather than serotonin toxicity), guinea pigs, and rhesus monkeys.

In the first study reporting signs of damage to neurons themselves, rather than just axons, Schmued and colleagues reported that when brain slices from rats given 20 or 40 mg/kg MDMA showed signs of neuronal degeneration, as assessed with the stain Fluoro-Jade B, while little or no indications of degeneration were seen in rats given 10 mg/kg. The authors noted that hyperthermia was more closely associated with signs of neuronal degeneration than MDMA dose, and they suggest that differences in rat strain, age, and body temperature might explain no previous studies have detected harm to serotonin neuron bodies. Somewhat contradictory findings were reported by Fornai and colleagues (Fornai et al. 2003a; 2003b) in studies in mice. Their studies examined brain slices of GABA-ergic cells in the striatum and substantia nigra of mice given 4 injections of 5 mg/kg every 2 hours, and they found nuclear inclusions, signs of DNA damage or repair, and other indicators of oxidative stress, but no signs of apoptosis, or cell death.

While rodent studies tended to detect signs of MDMA neurotoxicity, a study in rhesus monkeys that had self-administered MDMA approximately three times a week for an 18- month period failed to find signs of frank neurotoxicity (Fantegrossi et al. 2004B). On average, monkeys in this study self-administered cumulative doses of 2 to 4 mg/kg per hour-long session, though doses of up to 15 mg/kg were administered on at least one session. Presence of axonal degeneration was measured in vivo with PET using a radioligand (radioactively labeled drug) that binds to VMAT, a protein associated with axon terminals, and by assessing brain neurotransmitter and VMAT content after monkeys were killed, Fantegrossi and colleagues failed to detect any changes markers of axonal health, or changes in brain neurotransmitters. Brain serotonin levels were lower in monkeys that had self-administered MDMA, but the difference was not statistically significant, and no differences in brain dopamine were found. This study differed from other MDMA neurotoxicity studies in that drug administration was under the immediate control of the subject rather than being administered non-contingently by the experimenter. If the effects of experimenter-contingent (not under the subject's control) MDMA differ from those of self-administered MDMA, then nearly all MDMA neurotoxicity studies in non-human animals may need to be reconsidered, since these studies used experimenter-contingent administration. This study also suggests that when self-administered in doses similar to those used by human Ecstasy users, MDMA produces little or no changes or damage to brain serotonin neurons. However, as discussed in a later section, the study did find that monkeys were similar to humans in that they lost interest in self-administering MDMA over time, in a manner possibly analogous to "loss of magic" reported by some long-time Ecstasy users and discussed in the IB. It is possible that "loss of magic" is an indicator of harm to serotonin axons and lower brain serotonin. It is also possible that reduced interest in self-administering MDMA is instead a sign of non-neurotoxic changes in the brain, such as increased presence (upregulation) or decreased presence (downregulation) of serotonin receptors.

Findings from another study in rhesus monkeys raises questions about how doses of MDMA given to non-human primates match doses given to humans (Bowyer et al. 2003). A dose is considered equivalent across species if it produces similar levels of a drug in blood and brain. Most researchers have relied on interspecies scaling to arrive at dose equivalents for their studies, but there has long been controversy over whether interspecies scaling is an appropriate model for calculating MDMA dose equivalents (see the IB, also McCann et al. 2001; Vollenweider et al 2001; Vollenweider et al. 1999). Because of this controversy, some people have questioned the high and repeated doses used in nonhuman animal studies, arguing that these doses are not in fact equivalent to doses used by humans. Recently, Bowyer and colleagues collected blood from rhesus monkeys after a single dose of 10 mg/kg S-(+)-MDMA (Bowyer et al. 2003) in a study that also assessed long-term effects after four days of twice-daily dosing with 10 mg/kg S-(+)-MDMA. Plasma MDMA in these animals after the first 10 mg/kg dose was ten times the levels seen in humans given between 1 and 2 mg/kg MDMA. In agreement with these findings, a study in swine reported that plasma levels of 8 mg/kg MDMA were about eight times higher than those seen in humans after typical recreational doses (Fiege et al. 2003). Peak plasma levels of MDMA were seen 20 minutes after administration, probably as a result of administering the drug by injection rather than orally. In rhesus monkeys, MDMA half-life was 8.3 hours, a value that is very close to its half-life in humans (7 to 9 hours). Plasma MDA levels in these swine were similar to those reported in human Ecstasy-related fatalities. Bowyer and colleagues used only the S-(+) enantiomers of MDMA, while humans almost always use the racemate, and it is possible that the racemate would have produced different results. However, the study in swine used the racemate, and still reached similar findings with respect to plasma MDMA levels. Taken together, these findings suggest that dose regimens used in non-human primates are not equivalent to doses used by humans. The above findings in rhesus monkeys and swine suggest either that current interspecies scaling calculations need to be revised or that interspecies scaling is an inappropriate means of estimating dose equivalence for MDMA. If either case is true, then studies in non-human animals do not provide a basis for estimating a neurotoxic dose of MDMA in humans.

A previous report of dopamine toxicity in non-human primates after 3 injections of 2 mg/kg MDMA given within a six-hour period addressed in the 2002 revision of the IB (Ricaurte et al. 2002) has been retracted (Ricaurte et al. 2003). At the time of its publication, critics remarked on the relatively high mortality rate reported in the study, lack of evidence of dopamine toxicity in Ecstasy users, and whether the dose regimen used was genuinely reflective of doses used by most Ecstasy users (Mithoefer et al. 2003). In line with previous studies in humans (Kish et al. 2000; Reneman et al. 2002A; Semple et al. 1999), subsequent attempts to replicate study data failed, leading to the discovery that Ricaurte and colleagues had administered the psychostimulant methamphetamine to their subjects, and not MDMA. Attempts to replicate the original findings in monkeys failed to detect dopamine neurotoxicity after three oral (intragastric) doses of up to 8.6 mg/kg MDMA to squirrel monkeys within a six-hour period, or after three 4 mg/kg injections (Ricaurte 2004), suggesting that even high doses of MDMA are unlikely to reduce dopamine function in primates. Ricaurte and colleagues also failed to replicate findings in baboons. It is well-known that mice exhibit lower levels of dopamine and the dopamine metabolite DOPAC after MDMA administration, but to date, they are the only species of mammal showing signs of dopamine, and not serotonin, neurotoxicity after repeated doses of MDMA. Even a recent attempt to confirm indications of dopamine toxicity in guinea pigs (Saadat et al. 2003) failed to find it, though the study did find that MDMA produced a lesser degree of hyperthermia in guinea pigs. The retraction of the study by Ricaurte and colleagues reinforces previous studies suggesting that when it appears, MDMA neurotoxicity is specific for serotonin in all mammals studied so far except mice.

Some findings from a study in Dark Agouti rats, a strain lacking an enzyme involved in MDMA metabolism reported that initial dose of 12.5 mg/kg MDMA increased the degree of hyperthermia seen after subsequent, and lower, doses of 2, 4, 5 or 6 mg/kg MDMA, and this was especially true in a warm environment (Green et al. 2003A). While the authors did not report on brain neurotransmitter levels, these findings suggest that illicit Ecstasy users reporting at least one large or "binge" dose of Ecstasy may increase their chances of experiencing elevated body temperature after subsequent uses of MDMA. Bowyer and colleagues found that plasma S-(+)-MDMA levels were higher at the seventh of eight doses in a four-day S-(+)-MDMA regimen (10 mg/kg given twice daily) than after the first dose of S-(+)-MDMA (Bowyer et al. 2003). The significance of this difference is unclear, but again, it is possible that large and repeated doses of MDMA may alter MDMA metabolism or sensitivity. Another study in rats comparing the effects of repeated MDMA (four hourly doses of 5 mg/kg) at lower (16 C) versus higher (28 C) ambient temperatures detected reductions in serotonin and 5-HIAA in both conditions (McGregor et al. 2003a). However, the researchers found greater reduction in serotonin when MDMA was administered in a warm environment. The same dosing regimen produced long-term behavioral changes, such as increased anxiety (discussed below under "Behavioral changes") and apparent impairment in object memory (discussed in "Effects on Learning and Memory.") It appears that elevated body temperature exacerbates degree of reduction in serotonin after MDMA. Note that rats were hypothermic after receiving MDMA in a 16 C environment, but still exhibited some reductions in serotonin and serotonin metabolites. However, if one large dose of MDMA does increase susceptibility to elevated body temperature, and if elevated body temperature in turn plays a role in MDMA neurotoxicity in primates, then taking large doses of MDMA in warm conditions (as in nightclubs) may not be comparable to taking smaller doses of MDMA in less warm locations, as those found in clinical trials, and studies in Ecstasy users are liable to overestimate degree of harm to serotonin function posed by clinical trials with MDMA.

A study seeking to compare the neurotoxic effects of MDMA with the neurotoxic effects of methamphetamine in rats (Pubill et al. 2003), reported that methamphetamine, but not MDMA, increased astroglial and microglial activity, considered markers of tissue damage and repair. Rats in this study received either two injections of 20 mg/kg MDMA 7 hours apart for four days, or four injections of 10 mg/kg methamphetamine every 2 hours for one day, and were killed three days post-drug (MDMA and methamphetamine) or one week later (MDMA-treated rats only). MDMA did not increase microglial activity in parietal or striatal areas, and there was only a trend for increased microglial activity in hippocampus three days later, with signs of microglial activation no longer visible seven days after MDMA administration. By comparison, MDMA treated rats had signs of lower levels of serotonin transporter sites three and seven days post-MDMA, as assessed via brain paroxetine binding, whereas methamphetamine did not reduce serotonin transporter binding. The reported findings are supportive of changes in serotonin receptor regulation rather than MDMA neurotoxicity, though the researchers note previous studies that have found signs of axonal degeneration after MDMA. It is notable that the researchers did not detect significant hyperthermia in either MDMA or methamphetamine treated rats, raising the possibility that glial activation could have been detected in rats with elevated body temperatures.

Researchers comparing the behavioral and neurochemical effects of repeated doses of MDMA, methamphetamine, or the two combined (4 doses given every 2 hours in rats in a warm environment (28 C) (Clemens et al. 2004) found that the high-dose combination of MDMA and methamphetamine (2 mg/kg) demonstrated reduced brain serotonin and dopamine, though the low-dose (1.25 mg/kg) combination had no effect upon brain serotonin or dopamine. By comparison, MDMA alone only reduced brain serotonin, and methamphetamine alone only reduced brain dopamine. Altered brain neurochemistry is especially notable as the doses used were lower than those given in single-substance conditions (2 mg/kg versus 2 or 5 mg/kg MDMA). If these findings generalize to other species, including humans, then it may be the case that Ecstasy users are affected by the intentional or unintentional coadministration of MDMA and methamphetamine. Consequently, studies of Ecstasy users may not offer an appropriate gauge of extrapolating risk to participants in clinical trials of MDMA.

A study of the effects of delta-tetracannabinol (THC, the active ingredient in cannabis) on MDMA neurotoxicity in rats found that THC and a cannabinoid receptor agonist (activator) at least partially attenuated reductions in brain serotonin (Morley et al. 2004), with THC producing a greater attenuation of serotonin reduction than the cannabinoid receptor agonist. Rats receiving THC showed hypothermic rather than hyperthermic responses to MDMA, but coadministering another compound that prevented hypothermia did not prevent a cannabinoid receptor agonist from attenuating reductions in brain serotonin. In contrast to the findings reported above, these suggest that because they often use cannabis, Ecstasy users may be experience less risk of reduced serotonin function than if they used MDMA only, though Morley and colleagues also note that the doses of THC used in this study were high and probably do not represent doses obtained through smoking cannabis.

A team of researchers in Italy examining the role oxidative stress in MDMA neurotoxicity in mice given four 5 mg/kg injections every 2 hours (Fornai et al. 2003a; 2003b) found strong evidence of oxidative stress and damage to dopamine axons (as expected in mice), but no signs of cell death. They used multiple indicators of oxidative stress, including detection of heat shock proteins (associated with oxidative stress), levels of neurotransmitters, signs of cellular DNA damage and repair, and levels of the axon- associated protein VMAT. In vitro studies described in the same report found ubiquitin- positive inclusions in rat tumor cells incubated with MDMA, considered signs of oxidative stress.

Surprisingly, a study that sought and failed to find indicators of somatic "withdrawal" in mice given twice-daily injections of 10 mg/kg MDMA for five days also failed to find lower levels of dopamine transporter (Robledo et al. 2003). It is unclear why lower brain dopamine levels were not seen in these mice. Findings either indicate a weak association between brain dopamine levels and dopamine transporter abundance or a sign that MDMA neurotoxicity is not consistently produced by this dose regimen in mice.

Most researchers addressing MDMA neurotoxicity continue to be interested in finding evidence in support of specific hypotheses concerning the processes involved in MDMA neurotoxicity, and what means can be used to distinguish potentially neurotoxic compounds from non-neurotoxic ones. Studies used pharmacological challenges either intended to be neuroprotective or exacerbate neurotoxicity (Clemens et al. 2004; Morley et al. 2003; Peng and Simantov 2003; Sanchez et al. 2003; Thompson et al. 2004), altered aspects of the environment, such as ambient temperature (Green et al. 2003A; McGregor et al. 2003a) or stress (Johnson et al. 2003), and examined the effects of neurochemical lesioning (Ferrucci et al. 2003). Studies tended to support findings that oxidative stress is involved in MDMA neurotoxicity, though a team of researchers demonstrated that oxidative stress may play a greater role in MDMA toxicity in mice than in rats (Sanchez et al. 2003). Protective treatments included familiar ones, such as reduced ambient temperature, and unfamiliar ones, such as THC) (Morley et al. 2004). Other findings include the discovery that mice with lesioned norepinephrine systems had greater reductions in brain dopamine than mice with intact norepinephrine systems (Ferrucci et al. 2003), and that administering centrally active GABA-ergic drugs reduced MDMA lethality and attenuated brain dopamine reduction in mice (Peng et al. 2003).

One surprising finding reported by Johnson and associates relates to the neuroprotective effects of stress in female mice injected with four doses 15 mg/kg S-(+)-MDMA (one dose every two hours). Contrary to general expectations, mice exposed to several stressors, including restraint stress, a 16 C (61 F) environment, and exposure to ethanol, actually exhibited greater levels of brain dopamine than unstressed mice (Johnson et al. 2003). Increasing levels of corticosterone, the rodent version of cortisol, so that it matched levels in stressed animals did not attenuate MDMA effects on dopamine levels. Cool ambient temperature is already known to reduce MDMA neurotoxicity, perhaps through reducing oxidative stress or MDMA metabolism, and the effects of ethanol may be explained by direct pharmacological or neurochemical actions, but the beneficial neuroprotective effects of restraint stress remain difficult to explain. Because the authors used only S-(+)-MDMA, it is possible that these findings will not generalize to racemic MDMA. However, such findings at least raise the possibility that environmental stress does not necessarily exacerbate MDMA-induced reductions in brain monoamines.

Some researchers performed in vitro studies of serotonin and dopamine transport to see whether they could better understand aspects of neuronal transport associated with neurotoxicity (Bogen et al. 2003; Saldana et al. 2003). In vitro research using rat synaptosomes that compared synaptosomal uptake of radioactively labeled transmitter with vesicular (cellular, active) uptake conclude that vesicular uptake, and not synaptosomal uptake, is associated with the neurotoxic properties of MDMA. A comparison of the effects of 10 nM MDMA on serotonin and dopamine transport by tumor cells expressing human serotonin transporter (Saldana et al. 2003), with transport assessed at temperatures intended to mimic ordinary (37 C) and hyperthermic (40 C) body temperatures reported increasing transport of serotonin, but not dopamine, at higher temperatures (Saldana et al. 2003). Higher ambient temperatures seemed to increase rate of serotonin transport and inhibit dopamine transport, leading the researchers to propose that as serotonin levels decline after MDMA, dopamine may still enter into synapses.

In general, reports addressing MDMA neurotoxicity in non-human animals support previously reported conclusions. These include findings of reduced brain serotonin in rats and monkeys and reduced dopamine in mice, the impact of hyperthermia on MDMA neurotoxicity, evidence of a role for oxidative stress, and signs of damaged serotonin axons. However, findings reported since 2002 may alter some conclusions reported then. There is no longer any basis to believe MDMA produces dopamine neurotoxicity in non- human primates, and a study in rodents reported evidence of harm to neuron bodies, though these findings were dependent on dose and presence of hyperthermia. An examination of the brains of monkeys that had self-administered MDMA over an eighteen-month period found no indicators of reduced dopamine function and non- significant declines in brain serotonin, and no signs of axonal damage as assessed through measuring levels of VMAT. Though some findings point to additional routes of neurotoxicity and other studies question the degree or type of neurotoxicity seen after MDMA, an overall evaluation of these findings does not yet change the assessment of safety reached in the IB and the 2002 update to the IB, except with respect to dopamine toxicity. Human participants receiving no more than a cumulative dose of about 2.8 mg/kg on two separate occasions in a normothermic environment should be at minimal risk for potential serotonin neurotoxicity after MDMA, and so far it appears that they do not face any risk of dopamine toxicity.

Effects on Learning and Memory

Non-human animals appear to be less susceptible to impaired learning and memory after MDMA than might be estimated from studies of Ecstasy users (see for example Frederick et al. 1998; and further discussion in the IB). There are only a few reports of impaired learning in rodents (Marston et al. 1999), and none in monkeys, with differences in monkeys only seen after pharmacological challenge (Taffe et al. 2002). Even reports of apparent learning or memory deficits in rodents reviewed in the 2002 update of the IB (Morley et al .2001; Pompei et al. 2002) may be interpreted as reflections of increased anxiety or sociability rather than signs of a learning or memory impairment. While neonatal or young rats may be more susceptible to the effects of MDMA on learning (see Williams et al. 2002), previous research has not detected learning or memory impairments after MDMA, even at doses that reduced brain serotonin. At the time of this review, still remains true for non-human primates (Bowyer et al. 2003; Taffe et al. 2003), but recent reports suggest that this may no longer be true for rats (McGregor et al. 2003a; Sprague et al. 2003A, but see also Timar et al. 2003).

Eleven months after receiving daily doses of 10 mg/kg MDMA or vehicle (no MDMA) for four consecutive days, MDMA-treated rhesus monkeys failed to perform any differently than controls on a test battery designed to measure learning, memory and motivation in non-human primates (Taffe et al. 2003). In controls, tryptophan depletion improved performance on a spatial sequential search task, while MDMA-treated monkeys did not benefit from tryptophan depletion. Neurochemical analyses were not reported in this study, so presence or degree of reduced brain serotonin in MDMA-treated monkeys is not known, but the doses used appear to be comparable to doses that lower serotonin in the same species. These findings suggest that long-term functional effects in monkeys may only emerge in specific circumstances, as under transient alterations in serotonin function (Taffe et al. 2002). The relevance of these findings to studies of cognitive function in Ecstasy users remains unclear, since an examination of research findings suggests a dissociation between impaired cognitive function and signs of reduced serotonin function (see discussion in "Studies of Ecstasy Users.") These results also indicate that when they can be found, changes in cognitive function appearing after repeated MDMA in nonhuman primates are subtle.

Bowyer and colleagues (2003) assessed performance on a learning and behavioral test battery in male rhesus monkeys given 10 mg/kg S-(+)-MDMA twice daily for four consecutive days. Monkeys with high plasma levels of MDMA took longer to perform the tasks a week after the last MDMA administration, but their performance had returned to normal a month after the final dose of MDMA. The learning task consisted of making the correct sequence of lever presses to earn a reward on an array of six levers. It would appear in this case that the changes in performance were temporary. By contrast, studies in Ecstasy users have not found that impaired memory or executive function dissipates after a period of abstinence, and some studies even found impaired cognitive function only in abstinent Ecstasy users.

A study assessing spatial memory seven days after rats received two 20 mg/kg injections of MDMA given 12 hours apart found that MDMA-treated rats learned to navigate the Morris water maze (test of spatial memory) as well as saline-treated rats, but that they had difficulty recalling or locating the spot where the escape platform had been when it was removed (Sprague et al. 2003A). MDMA-treated rats in this study also had lower levels of serotonin in the hippocampus when assessed 15 days post-drug. It should be noted that since memory was assessed seven days post-drug, and given the transience of the effects found in monkeys, it is not clear if this impairment reflects a long-term effect. An examination of the impact of ambient temperature on long-term effects of MDMA in rats (McGregor et al. 2003a) reported that MDMA (four hourly doses of 5 mg/kg) impaired performance on a test of object memory, assessed by noting time spent exploring a novel versus familiar object, and that performance on this task was further reduced in rats given MDMA in a warm environment (28 C) versus a cooler one (16 C). . In contrast with studies finding impaired learning or memory in rats, a comparison of the effects of single versus multiple doses of S-(+)-MDMA and amphetamine in rats failed to find impaired active or passive learning after four injections of 10 mg/kg S-(-)-MDMA given every two hours (Timar et al. 2003). Timar and associates did not assess brain serotonin or serotonin function, so it is not known whether rats had lower levels of brain serotonin, though the dose the researchers used is known to reduce brain serotonin in rats. Both of the learning tasks in this study rely on fear-based learning, so it is possible that MDMA-associated increases in anxiety made it more likely that MDMA-treated rats would perform well at these tasks.

As was true of research previously reviewed in the 2002 update of the IB, published findings appearing in 2003 concerning the long-term effects of MDMA on learning and memory in non-human animals are inconclusive. Studies in rodents sometimes reported signs of impaired learning or memory after MDMA, while studies in monkeys found impairment or changes in performance that was either temporary or that only appeared under pharmacological challenge.

Behavioral Effects and their Association with Neurotoxicity

Researchers continue to study the behavioral effects of MDMA in rodents, including assessments of anxiety (Clemens et al. 2004; McGregor et al. 2003a; 2003b; Robledo et al. 2003; Thompson et al. 2004), sensitization to other drugs (Cole et al. 2003) or locomotor activity (Itzhak et al. 2003). Interest in the long-term effects of MDMA on anxiety and "depression-like" behavior (performance on the forced swim test, used to assess potential antidepressants) continues to interest researchers, both as a result of findings of reduced psychological well-being in Ecstasy users and because of a presumed relationship between mood or affect and the serotonin system. So far, most studies have found increased anxiety after repeated doses of MDMA (Fone et al. 2002; Gurtman et al. 2002; Morley et al. 2001). However, reduced anxiety has been found on occasion (Mechan et al. 2002), and may be an alternative explanation for some findings described as measures of social memory (Pompei et al. 2002), as discussed in the 2002 update to the IB. Researchers also reported that MDMA increased sensitivity to other stimulants (Fone et al. 2002).

Most recent publications reported increased anxiety in rodents after MDMA, as measured in the emergence test (Clemens et al. 2003; McGregor et al. 2003a; 2003b; Morley et al. 2003; Thompson et al. 2004) and in the open field test (McGregor et al. 2003a). Studies found increased anxiety after dose regimens not known to reduce brain serotonin, and not just after doses that usually reduce brain serotonin, suggesting that changes in anxiety do not represent a marker for MDMA neurotoxicity. Repeated doses of MDMA also reduced social interactions between pairs of novel rats (Clemens et al. 2003; Morley et al. 2003; McGregor et al. 2003a; 2003b; Thompson et al. 2004). It is notable that in some studies, neuroprotective treatments, such as fluoxetine (Thompson et al. 2004) or THC, or a cannabinoid CB1 agonist (Morley et al. 2003) did not attenuate MDMA-associated reductions in social interaction, even when (in the case of THC), the treatment attenuated anxiety after MDMA (Morley et al. 2003).

Anxiety-increasing doses of MDMA in studies ranged from four 5 mg/kg injections every 2 hours (Clemens et al. 2003) to four 10 mg/kg every two hours for one day (Timar et al. 2003), with most studies administering four hourly doses of 5 mg/kg for two consecutive days (McGregor et al. 2003a; 2003b; Morley et al. 2003; Thompson et al. 2004). Researchers assessed anxiety as early as four weeks post-MDMA (Clemens et al. 2003), and as late as 12 to 15 weeks post-MDMA (Thompson et al. 2004). In most cases, MDMA increased both social and non-social anxiety. Administering MDMA in a cooler environment (16 degrees C versus 28 C) did not reduce the appearance of anxiety afterwards (McGregor et al. 2003a).

While four injections of 2.5 mg/kg MDMA did not reduce brain serotonin and 5HIAA to the same degree as 5 mg/kg (Clemens et al. 2003), the lower dose increased anxiety as assessed via emergence test, while there was only a trend for increased anxiety, measured 11 weeks post-drug, after the higher dose. By contrast, the higher dose, and combinations of MDMA and methamphetamine, reduced social interactions, whereas the low dose did not. Another study found that even a single 5 mg/kg injection increased anxiety without any significant changes in brain serotonin or 5HIAA (McGregor et al. 2003b). Taken together, examining all studies assessing anxiety after MDMA suggest MDMA is more likely to increase anxiety in rodents, and that increased anxiety after MDMA is not caused or produced by lower brain serotonin, and may instead be the result of some other as yet unmeasured change in brain function or structure. One possibility is that changes in receptor activity, such as changes in 5HT2A receptors, result in these and other behavioral changes.

Two studies reported on the effects of MDMA on the "forced swim" test, considered a measure of depression-like behavior in rats (McGregor et al. 2003a; Thompson et al. 2004). Sixteen to 18 weeks after receiving four hourly doses of 5 mg/kg MDMA for two days in a row in either 16 or 28 C environments, rats showed increased immobility and reduced escape attempts during the forced swim test (McGregor et al. 2003). Rats in another study that received the same MDMA regimen (four hourly doses of 5 mg/kg given for two days) also showed increased immobility and decreased active response on the forced swim test (Thompson et al. 2004). A five week course of fluoxetine administered in drinking water (with rats ingesting approximately 6.2 mg/kg fluoxetine per day) attenuated MDMA-associated reductions in forced swim performance, even though this dose of fluoxetine surprisingly did not improve forced swim performance in saline-treated rats. Furthermore, fluoxetine had very little effect on serotonin levels in MDMA-treated rats, though it did decrease 5HIAA levels. The authors propose that fluoxetine ameliorated anxiety and depression-like effects by acting on serotonin receptors and not through neuroprotection.

Since many studies now find that changes in psychological well-being in Ecstasy users are liable to be related to variables other than Ecstasy use (see discussion in "Studies in Ecstasy Users" and in the 2002 update to the IB), the relevance of increased anxiety in MDMA-treated rodents remains unclear, but supports the possibility of an association between MDMA and anxiousness. However, because these effects are seen both after neurotoxic and non-neurotoxic dose regimens, increased anxiety cannot be treated as an indicator of reduced serotonin function. Given that findings from human studies do not assign a strong or unique association between MDMA and anxiety or depression, and given that humans, these findings in no way increase the estimation of potential risk of increased anxiety in human participants receiving MDMA.

Developmental Toxicity

In the IB and especially in the 2002 update to the IB, it was noted that exposing rats to MDMA either prior to birth or shortly after birth altered brain chemistry and sometimes produced impaired learning and memory. There has been controversy concerning studies describing developmental toxicity in humans (see for example McElhatton et al. 1999), but researchers continue to examine the effects of prenatal (Kelly et al. 2002; Koprich et al. 2003A), neonatal (Kelly et al. 2002; Koprich et al. 2003B; Williams et al. 2003), or early "youth" (Bull et al. 2004) exposure to MDMA in rats. So far, these studies have detected changes in brain neurochemistry and behavior.

Rat pups whose mothers had received twice-daily injections of 15 mg/kg MDMA on gestational days 14 through 20 were more active in a novel cage three days after birth compared to pups born to saline-treated rats (Koprich et al. 2003A), but fine motor activity was unaffected by MDMA exposure. Twenty-one days after birth, MDMA- treated rats had lower levels of serotonin and dopamine metabolites, possibly implying slower neurotransmitter turnover, since there were no differences in brain serotonin or dopamine content. By contrast, another research team (Kelly et al. 2002) reported that brain serotonin levels were not lower in rats born to mothers given 20 mg/kg MDMA twice daily over four consecutive days and starting on gestational day 15. Differences in research findings may be due to different dosing regimens, or to using different means of assessing brain serotonin levels, as Koprich and colleagues used high performance liquid chromatography (HPLC) whereas Kelly and colleagues assessed serotonin transporter binding with radioactively labeled paroxetine.

Neonatal rats exposed to high, frequent doses of MDMA (twice-daily injections 20 mg/kg for 10 days, beginning on postnatal day 11) performed less well than controls on water maze tasks that assessed spatial memory (Williams et al. 2003). Another study employing the same dosing regimen and time of dosing in rats found reduced serotonin and dopamine content in specific brain areas (Koprich et al. 2003B). Contradictory findings are reported in a study that varied time of neonatal exposure, and only administered twice-daily injections of 20 mg/kg MDMA for four days (Kelly et al. 2002). No differences were seen in serotonin levels (measured by paroxetine binding) when MDMA exposure occurred prior to postnatal day 25, but brain serotonin levels were presumably reduced in rats given a first dose of MDMA on postnatal day 25 or 30. Kelly and colleagues found even greater reduction in brain serotonin when MDMA was begun on postnatal day 90. Differences in findings may be the result of differences in length of dose regimen (four versus ten days) and differences in assessment of brain serotonin. In reviewing these studies, it is important to note that most researchers believe that rat neonates are analogous to human fetuses in the third trimester of pregnancy, and so these findings would reflect a window of vulnerability occurring at the third trimester.

Young rats given four hourly doses of 5 mg/kg MDMA on two consecutive days exhibited increased anxiety and decreased social interaction (Bull et al. 2004), and brain serotonin and 5-hydroxyindoleacetic acid (5-HIAA) was lower. However, serotonin transporter density was apparently unchanged, leading the authors to conclude that behavioral changes were associated with changes in number or sensitivity of receptors, though changes in serotonin function could also be involved.

As was the case with previously reviewed research, these studies support the existence of one or more critical periods when exposure to high or repeated doses of MDMA could alter brain development, impair specific forms of learning and increase anxiety, though it may be the case that MDMA must be administered for longer than four days for this effect to occur. Though doses used in these studies remain high, the findings continue to support the exclusion of pregnant women and women who are not using an effective means of contraception. Because MDMA-assisted therapy is not treating an acutely life- threatening illness, this restriction seems appropriate even if the high and repeated doses used in rodent studies may be overestimating the risk of developmental toxicity.

Other Possibly Related Effects

An investigation into the effects of MDMA on regulation of an area of the brain related to circadian rhythm (Dafters and Biello 2003) found that when the dopamine synthesis inhibitor AMPT or the D2 antagonist haloperidol were administered along with daily doses of 20 mg/kg MDMA injected on three consecutive days, they reduced MDMA- associated hyperthermia. However, these drugs did not alter the effects of MDMA on suprachiasmatic nucleus (SCN) neuron response to the 5HT1A receptor agonist 5- OHDPAT. The authors conclude that alterations in the SCN must be the result of MDMA metabolites, and not body temperature or dopamine presence per se, but they did not assess brain serotonin levels. The significance of these findings for understanding MDMA effects in humans is not clear.

Toxicity

Cardiotoxicity

Unpublished research finding that MDMA was an agonist at the 5HT2B receptor was reported to institutional review boards in the 2002 update to the IB. After the recent discovery that fenfluramine, an anorectant taken off the market in part because of its association with valvular heart disease (VHD), acted on the newly discovered 5HT2B receptor, it was suggested that drugs activating this receptor may increase the likelihood of VHD (Rothman and Baumann 2002). A receptor activity assay demonstrated that MDMA is a 5HT2B agonist, and incubation with the fairly large dose of 10 mM MDMA stimulated heart valve cell growth, as did identical concentrations of fenfluramine, dexfenfluramine and MDA. However, only fenfluramine and dexfenfluramine produced statistically significant increases in heart valve cell growth. VHD is even rare after chronic fenfluramine use (Davidoff et al. 2001; Rothman and Baumann 2002), and there are no medical reports of VHD appearing in an Ecstasy user. These findings do not suggest any increased risk of VHD after intermittent administration of MDMA.

Thermoregulation

Researchers continue to be interested in MDMA effects on body temperature, its mechanisms of action and the relationship between body temperature and neurotoxicity. To date, all recently performed studies have been performed in rodents. Recent publications examined possible neurochemical mechanisms (Fantegrossi et al. 2004B; Fantegrossi et al. 2003) and involvement of skeletal muscle thermoregulation, or "non- shivering" heat production, in MDMA-induced hyperthermia (Fiege et al. 2004; Mills et al. 2003; Sprague et al. 2003B).

Both racemic and S-(+)-MDMA produced dose-dependent hyperthermia in mice, while R-(-)-MDMA never produced hyperthermia, even at reliably lethal doses (Fantegrossi et al. 2003). This study found that group housing (six or twelve mice per cage) increased MDMA lethality, and that cold ambient temperature (a room kept at 4 C, or 39 F) reduced lethality for group-housed mice, suggesting a role for hyperthermia in aggregate (crowd or group related) toxicity of MDMA. Further investigations by the same author in mice found that both alpha1 adrenergic and 5HT2A antagonists reduced hyperthermia (Fantegrossi et al. 2004A), but only nantenine, a novel plant-based compound bearing some structural resemblance to MDMA, reduced MDMA-induced hyperthermia 30 minutes after MDMA administration.

Two studies in mice published by the same research team found relationships between MDMA-induced changes in body temperature and non-shivering heat production (Mills et al. 2003; Sprague et al. 2003B). One study found that thyroid hormones and presence of intact thyroid or hypophyseal area were associated with elevated body temperature after a single dose of 40 mg/kg MDMA in rats. Intact rats had elevated core and skeletal temperature after MDMA (Sprague et al. 2003B), but hypothermia was seen after the same dose of MDMA in hypophysiectomized and thyroparathyroidectomized rats. Transgenic mice lacking the gene for the uncoupling protein 3 (or UNCP3) had lower elevations in body temperature than normal mice after 10 to 40 mg/kg MDMA, and were more likely to survive after receiving 50 mg/kg MDMA (Mills et al. 2003). It is notable that doses of MDMA in these studies were high; lower doses may not trigger non- shivering heat production in rodents or humans. While the significance of these findings to humans is somewhat unclear, especially given the high doses used to elicit hyperthermia in these studies, these findings do offer another explanation for elevated body temperature after MDMA.

Swine susceptible to malignant hyperthermia, a genetic disorder probably related to abnormal calcium metabolism in skeletal muscles, had higher body temperatures than non-susceptible swine after receiving 8 mg/kg and 12 mg/kg MDMA, though non- susceptible animals showed temperature increases after 12 mg/kg MDMA (Fiege et al. 2003). The researchers successfully reduced hyperthermia in malignant hyperthermia susceptible swine by administering the muscle relaxant dantrolene. These findings suggest that humans susceptible to the same disorder may respond similarly to MDMA. Though findings suggest that people possessing the genetic disorder that causes malignant hyperthermia would experience hyperthermia after MDMA, the rarity of the condition makes it an unlikely explanation for most cases of hyperthermia in Ecstasy users.

No cases of significantly elevated body temperature or clinically significant hyperthermia have occurred during any human trials of MDMA. The lack of any such occurrences continues to demonstrate the minimal risk of administering 1 to 2.5 mg/kg MDMA to humans in controlled, normothermic settings. As discussed earlier in this review, human volunteers given MDMA in clinical trials sometimes experienced a slight elevation in body temperature, and even this elevation was not always detected. The effects of MDMA on body temperature appear to be dose dependent (Greene et al. 2003), but also appear to be related to ambient temperature and activity. These studies in non-human animals do not alter our original estimation that human volunteers taking part in clinical trials of MDMA are very unlikely to experience hyperthermia.

Liver Toxicity

As discussed in the IB, there have been sporadic reports of liver problems in Ecstasy users. While their etiology is not fully understood, high ambient or body temperature may be involved (see for example Andreu et al. 1998; Caballero et al. 2002; Henry et al. 1992). One researcher posited that liver problems may arise from impurities in illicit Ecstasy (Green et al. 2003B), though without presenting supportive evidence for the claim. In vivo and in vitro investigations of MDMA effects on the liver have been discussed in the IB and in the 2002 update of the IB. These investigations have found that MDMA seemed to impair liver function in cells when given at very high doses, and that these effects appear to be temperature-dependent (see Carvalho et al. 2002; Montiel- Duarte et al. 2002), As noted in the previous review of the literature, the doses used are probably not relevant either to doses taken by Ecstasy users or to doses administered in clinical trials.

One of the research teams responsible for several previous investigations performed another study investigating the effects of the MDMA metabolite MDA, and a putative metabolite alpha-methyl-dopamine (or Alpha-MeDA) on rat hepatocytes (Carvalho et al. 2003). 0.2, 0.4, 0.8, and 1.6 mM MDA, incubated with cells for 3 hours at 37 deg C (human body temperature), failed to impair rat liver cell function or reduce viability at any dose tested, while the highest dose of Alpha-MeDA reduced cell viability and increased some signs of impaired liver function (such as reduced glutathione content, without increasing signs of lipid peroxidation, a marker of oxidative stress. MDA is only a minor metabolite of MDMA in humans (de la Torre et al. 2001; Pizarro et al. 2002; Segura et al. 2001), and the concentration of alpha-MeDA that altered liver function is probably much greater than that expected in humans, assuming that levels of alpha- MeDA will always be lower than those of MDMA. Study findings do not alter earlier risk estimations related to liver toxicity. To date, no liver problems have been reported in human volunteers enrolled in clinical trials of MDMA.

Self-administration and Reward Value

Studies in rats (Cole et al. 20003; Cornish et al. 2003; Schenk et al. 2003; Wakonigg et al. 2003) and monkeys (Fantegrossi et al. 2004B) continue to find that non-human animals find MDMA rewarding. Two of four studies examined self-administration or reward value in animals already familiar with self-administering cocaine (Fantegrossi et al. 2004A), two others involved drug-naïve animals (Cornish et al. 2003; Wakonigg et al. 2003), and one study compared both cocaine-trained and drug-naïve rats (Cole et al. 2003; Schenk et al. 2003).

One study assessed reward value by first associating arrival at the end of a "runway" (an alley) with an injection of 1 mg/kg MDMA or saline (Wakonigg et al. 2003). The speed at which the rat arrived at the alley was considered an indicator of reward value. Researchers found more rapid arrival times after MDMA than after saline in two strains of rat (Long Evans and Sprague Dawley). While cocaine-trained rats learned to self- administer MDMA more readily than drug-naïve rats (Schenk et al. 2003), rats in both conditions worked to receive 0.1 to 2 mg/kg MDMA. Rats made more frequent lever presses when the dose of MDMA they received was low (e.g. 0.5 mg/kg), and stopped responding when only saline was available. Cornish and colleagues (2003) found that rats learned to self-administer a range of doses of MDMA (0.1 to 1 mg/kg), and that rats worked harder for all doses of MDMA when in a warm room (30 C, or 86 F) versus in a 21 C (70 F) "room temperature" setting, suggesting that the rewarding properties of MDMA increase with increasing ambient temperature, at least in rats. Cornish and colleagues also found that rats given MDMA in a warm environment were more likely to engage in social interactions than rats given MDMA in the cooler environment. A study in rhesus monkeys experienced in self-administering other drugs found that they also self-administered MDMA (Fantegrossi et al. 2004A). This study probably examined the same sample first described in a previous paper by the same team (Fantegrossi et al. 2002) reviewed in the 2002 update to the IB. Monkeys already maintaining self- administration of cocaine were presented with MDMA, as racemate and as R-(-)- and S- (+) forms, approximately three times a week, and number of self-administrations was catalogued over an eighteen-month period. Monkeys generally self-administered between 2 and 4 mg/kg MDMA per session, as discussed above in "Neurotoxicity." All forms of MDMA were self-administered at the start of the study, but by the end of the study, monkeys self-administered less racemic and R-(-)-MDMA, suggesting a use pattern analogous to the "loss of magic" reported by Ecstasy users. (Self-administration of S-(+)-MDMA declined over time in one monkey, while another self-administered more of this enantiomer.)

Conditioned place preference, or the tendency to remain in a place associated with a previous drug administration, was used in a study in rats (Cole et al. 2003) and a study in mice (Robledo et al. 2003), though neither of the two studies were focused on the reinforcing effects of MDMA. In one case, the authors investigated the possibility that rats previously given three injections of 10 mg/kg MDMA every 2 hours increased the reinforcing effects of other drugs, including a second dose of MDMA (Cole et al. 2003). They did not find any increased conditioned place preference for any of the drugs tested, MDMA amongst them, and instead found that ethanol produced less conditioned place preference in rats previously given MDMA. These results stand in contrast to an earlier report of increased conditioned place preference for cocaine after previous exposure to MDMA (Fone et al. 2002). The other study sought to establish whether MDMA produced a "withdrawal syndrome" or any signs of an aversive experience after MDMA discontinuation in mice (Robledo et al. 2003). Though the authors reported that administering a serotonin antagonist (metergoline) and a beta adrenergic antagonist (timolol) a day after MDMA administration produced some somatic effects in MDMA treated mice only, metergoline did not reduce MDMA-associated conditioned place preference, and only the highest dose of timolol reduced conditioned place preference. The researchers also failed to find any signs of reduced efforts in working for food reward or any signs of conditioned place aversion for a location previously associated with time after a 10 mg/kg injection of MDMA, indicating that the period when all MDMA effects are gone is not experienced as aversive. How these findings relate to reported sub-acute effects in humans and their relation to drug dependence remains unclear. Humans experience sub-acute effects of MDMA, but as discussed in the IB, most Ecstasy users take the drug on a weekly to monthly basis rather than taking more Ecstasy to avoid sub-acute effects.

Altogether, study findings suggest that rodents and monkeys will self-administer MDMA at doses that are similar to or not much higher than those taken human Ecstasy users. Findings of self-administration and reward value in non-human animals confirm earlier reports of the reinforcing effects of MDMA in non-human animals, and findings concerning rewarding dose range also suggest that dose equivalents calculated through interspecies scaling do not appropriately reflect doses used by humans. It also appears that warm ambient temperature may increase the reinforcing effects of MDMA. None of the findings provides a basis for either increasing or decreasing estimated risk of abuse liability for clinical trials with MDMA.

Conclusion

Recent examinations of MDMA toxicity in heart, liver, and thermoregulatory system report findings similar to those reviewed in the previous update to the IB. Likewise, evidence of MDMA self-administration has already been noted in previous reviews of the literature. Subjects in clinical trials of MDMA are carefully screened for the presence of cardiovascular and hepatic (liver) disease, and to date, clinical trials have not reported heart or liver problems in volunteers. Volunteers are also screened for recent substance abuse or dependence disorders, thus reducing the chance that MDMA will be given to people who will then seek to self-administer Ecstasy. None of the research findings reviewed have given any cause for revising risk estimates associated with participating in clinical trials of MDMA.