Semple et al (Reference Semple, Ebmeier and Glabus1999) report a reduction in vivo of ¹²³I-labelled 2β-carbomethoxy-3β-(4-iodophenyl)tropane (β-CIT) uptake in the cerebral cortex of 3,4-methylenedioxymethamphetamine (MDMA, ‘ecstasy’) users. They interpret this observation to be an indication of a decrease in serotonin transporters in the cortex of MDMA users. However, there are serious methodological concerns with this interpretation of their data.

It has been demonstrated that the radioligand [¹²³I]β-CIT binds with high affinity to dopamine, serotonin and noradrenaline transporters in human brain (Reference Farde, Halldin and MuellerFarde et al, 1994; Reference Laruelle, Wallace and SeibylLaruelle et al, 1994). In the case of the serotonin transporter, in vivo displacement of β-CIT binding by selective serotonin reuptake inhibitors (SSRIs) has established that specific (displaceable) binding occurs in brainstem and thalamus (Reference Laruelle, Baldwin and MalisonLaruelle et al, 1993; Reference Pirker, Asenbaum and KasperPirker et al, 1995; Reference Tauscher, Pirker and de ZwaanTauscher et al, 1999). However, this is not true for the cerebral cortex. Indeed, Laruelle et al (Reference Laruelle, Baldwin and Malison1993) observed that [¹²³I]β-CIT uptake in cortical areas was unaffected by citalopram administration in non-human primates. Similarly, recent SSRI displacement studies of [¹¹C]McN-5652, a selective serotonin transporter radioligand for positron emission tomography (PET) imaging, failed to observe specific binding in the cerebral cortex (Reference Parsey, Simpson and HwangParsey et al, 1999). The lack of evidence for specific binding to serotonin transporters in the cerebral cortex in vivo is not surprising when one considers the paucity of these transporters in primate cortex (Reference Jagust, Eberling and BiegonJagust et al, 1996). We are aware of only a single report of apparent displacement of β-CIT by citalopram in primate cortex: a PET study of [¹¹C]β-CIT uptake in two cynomolgus monkeys (Reference Farde, Halldin and MuellerFarde et al, 1994). However, the shape and time-scale of the binding curves for the cynomolgus monkeys are strikingly different from those observed in other non-human primate species (Reference Laruelle, Baldwin and MalisonLaruelle et al, 1993) and in humans (Reference Farde, Halldin and MuellerFarde et al, 1994; Reference Laruelle, Wallace and SeibylLaruelle et al, 1994; Reference Pirker, Asenbaum and KasperPirker et al, 1995). This discrepancy is particularly pronounced for the cortical curve, and one wonders to what extent these data may be relevant to human studies. Be that as it may, the bulk of the evidence indicates that serotonin transporters are present in sufficient density to be measured reliably with [¹²³I]β-CIT only in the thalamus and brainstem, and not the cerebral cortex. The region of choice is the raphe area of the brainstem because the thalamus may have a substantial admixture of noradrenaline transporters (Reference Farde, Halldin and MuellerFarde et al, 1994) and because it is difficult to avoid scattered radiation from the much greater accumulation of activity in the striatum in a thalamic region of interest. We have found this to be true in our studies of serotonin transporters with [¹²³I]β-CIT, and we have observed that uptake in cortical regions does not differ significantly from the non-displaceable (non-specific) uptake seen in the cerebellum (Reference Heinz, Ragan and JonesHeinz et al, 1998). At extended times (>4 hours post-injection in humans), when specific binding to serotonin transporters in the brainstem approaches a near-equilibrium plateau and non-specific uptake continues to washout throughout the brain, it becomes clear that cortical uptake is ‘tracking’ that of the cerebellum.

This latter point raises a further methodological concern. Semple et al (Reference Semple, Ebmeier and Glabus1999) imaged [¹²³I]β-CIT uptake at 90 minutes post-injection hoping to assess radioligand binding to serotonin transporters. However, near-equilibrium conditions for β-CIT at serotonin transporters are not established in human brain earlier than about four hours post-injection (Reference Laruelle, Wallace and SeibylLaruelle et al, 1994; Reference Pirker, Asenbaum and KasperPirker et al, 1995). Once near-equilibrium has been established, [¹²³I]β-CIT binding to serotonin transporters in the brainstem is quite stable and persists well into the following day (Reference Laruelle, Wallace and SeibylLaruelle et al, 1994; Reference Pirker, Asenbaum and KasperPirker et al, 1995). Measurements at extended times of [¹²³I]β-CIT activity in human brainstem (following decay correction and subtraction of non-specific uptake) are simply proportional to the density of serotonin transporters (Reference Laruelle, Wallace and SeibylLaruelle et al, 1994). Unfortunately, this is not the case for the measurements of Semple et al (Reference Semple, Ebmeier and Glabus1999) at 90 minutes post-injection. At times this early, the system is not near equilibrium, and factors related to radioligand delivery and washout, rather than transporter binding per se, play a prevalent role in determining the appearance of [¹²³I]β-CIT images. Thus, it seems likely that factors such as blood flow, blood-brain barrier integrity, tissue permeability, etc. have confounded the cortical measurements that Semple et al (Reference Semple, Ebmeier and Glabus1999) have assumed to be due to serotonin transporters.

In summary, although Semple et al (Reference Semple, Ebmeier and Glabus1999) report an interesting reduction in β-CIT uptake in the cerebral cortex of MDMA users, there is no scientifically sound basis for ascribing this observation to a decrease in cortical serotonin transporters.