Brain Advance Access originally published online on February 25, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brain, Vol. 127, No. 4, 900-913, 2004
© 2004 Guarantors of Brain
doi: 10.1093/brain/awh109
5-HT1A receptor binding and intracerebral activity in temporal lobe epilepsy: an [18F]MPPF-PET study
1 EA1880, Federal Institute of Neurosciences (IFR19), Lyon, 2 INSERM, 3 Debrousse Hospital, EEG Department, Lyon, 4 CERMEP, PET Center, Lyon, 5 CNRS and 6 Neurological Hospital, EEG Department, Lyon, France
Correspondence to: Isabelle Merlet, CERMEP, 59 boulevard Pinel, 69003 Lyon, FranceE-mail: merlet{at}cermep.fr
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The aim of our study was to assess abnormalities in 5-hydroxytryptamine-1A (5-HT1A) receptor density in patients suffering from refractory temporal lobe epilepsy (TLE). Experimental data in animals show that 5-HT1A receptors are predominantly located in limbic areas, and that serotonin, via these receptors, mediates an antiepileptic and anticonvulsant effect. In TLE patients, we quantified 5-HT1A receptor density in epileptogenic and non-epileptogenic areas, as defined by intracranial recordings with stereo-electroencephalography (SEEG). Nine TLE patients and 53 control subjects were studied by PET using a 5-HT1A receptor antagonist ([18F]MPPF). Anatomical regions of interest (ROIs) were drawn on patient and control MRIs co-registered with PET. PET data were quantified using a simplified model to assess binding potential (BP) values in each ROI, with cerebellum as reference. For each patient, a normalized percentage BP change was calculated as the relative variation of BP in each ROI compared with the corresponding ROI in control subjects. In patients, ROIs explored by SEEG were categorized according to their degree of epileptic activity (ictal onset, ictal spreading, interictal spikes, no epileptic activity) and according to their lesional aspect and volume (lesional with volume loss, lesional without volume loss, non-lesional). Compared with control values, the binding to 5-HT1A receptors in TLE patients was decreased in the epileptogenic temporal lobe. BP decrease was significantly greater in: (i) regions involved in the seizure onset than regions where only interictal paroxysms or no epileptic activity was recorded; and (ii) regions where the discharge propagated than regions where only interictal paroxysms or no epileptic activity was recorded. BP decrease was shown to be significantly influenced by the existence of a lesion on MRI. However, in the group of ROIs with normal quantitative and qualitative MRI aspect, BP decrease remained strongly correlated to the degree of epileptic activity. This study shows that in vivo availability of 5-HT1A receptors is decreased in epileptic patients compared with normal subjects. This decrease is highly correlated to the degree of epileptogenicity of cortical areas explored by intracerebral recordings, and does not reflect only pathological changes or neuronal loss in the epileptic focus.
Key Words: 5-HT1A receptors; intracerebral recordings; PET; serotonin; temporal lobe epilepsy
Abbreviations:
AED = antiepileptic drug; AMT = 
-[11C]methyl-L-tryptophan; BP = binding potential; 5-HT1A = 5-hydroxytryptamine-1A; IS = interictal spiking activity only; NE = no epileptic activity; NI = non-implanted; ROI = region of interest; SEEG = stereo-electroencephalography TLE = temporal lobe epilepsy
Received July 11, 2003. Revised October 23, 2003. Accepted December 14, 2003.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Hyperexcitability, synapse re-organization and imbalance between inhibitory and excitatory synapses are three dysfunctions characterizing the epileptogenic zone. The neurotransmitters that influence the excitatory/inhibitory balance include the excitatory (glutamate and aspartate) and inhibitory (GABA) amino acids (Engelborghs et al., 2000
). In addition, monoaminergic systems, such as the serotoninergic ascending pathway, also play a role in the modulation of the cortical and subcortical activity in the mammalian nervous system (Barnes and Sharp, 1999). These ascending pathways may modify the cortical and subcortical excitatory/inhibitory balance, and therefore contribute to the control of epileptic activity. Among the 17 subtypes of serotonin receptors identified to date, the 5-hydroxytryptamine-1A (5-HT1A) receptor is the most widely studied (Peroutka, 1995). The possibility of an interaction between serotonin and epilepsy was first proposed in the late 1950s by Bonnycastle et al. (1957), who observed that brain serotonin concentration was increased by anticonvulsant agents. Since then, a number of studies have been performed on different types of experimental models of epilepsy. While some authors described a convulsant effect of 5-HT1A agonists in absence-type epilepsies (Gerber et al., 1998; Filakovszky et al., 1999), the majority of studies suggest that serotonin might have, on the contrary, an anticonvulsant and antiepileptic effect via 5-HT1A receptors. Serotonin, via the 5-HT1A receptor, was shown to delay the kindling process (Lerner-Natoli, 1987; Wada et al., 1992, 1997) to decrease the frequency of seizures induced by kainic acid (Gariboldi et al., 1996) or bicuculline (Salgado-Commissariat and Alkadhi, 1997), and to inhibit the epileptiform activity induced by an Mg2+-free medium on rat hippocampal slices (Tokarski et al., 2002). Moreover, agents that increase the concentration of endogenous serotonin, such as the inhibitors of serotonin re-uptake (fluoxetine), were evaluated as anticonvulsant in genetically epilepsy-prone rats (Dailey et al., 1992; Yan et al., 1994, 1995) in partial seizures generated by low-frequency electrical stimulation in rats (Watanabe et al., 1998), as well as in kindled rats (Wada et al., 1993). The anticonvulsant action of fluoxetine was recently described as being mediated by 5-HT1A receptors (Lu and Gean, 1998). Therefore, the predominant effect of serotonin in different experimental models of epilepsy (except in absence-type epilepsy) is a 5-HT1A-mediated inhibition of the epileptic activity.
In humans, immunohistochemical studies have revealed increased levels of serotonin in cortical dysplasia with focal epilepsy (Trottier et al., 1996). PET studies using -[11C]methyl-L-tryptophan (AMT) suggest alterations in serotonin synthesis in lesional and cryptogenic epilepsy (Chugani et al., 1998; Fedi et al., 2001). Most recently, in a PET study using [18F]FCWAY in temporal lobe epilepsy (TLE), Toczek et al. (2003) reported that 5-HT1A receptor binding is reduced in TLE foci.
In this paper, we explored the possible relationships between epilepsy and serotonin using PET and [18F]MPPF, a 5-HT1A receptor antagonist, with patients suffering from refractory TLE who, as part of their pre-surgical evaluation, underwent stereotactic implantation of intracerebral electrodes [stereo-electroencephalography (SEEG)] and video-SEEG monitoring of their seizures. Our aim was to correlate changes in 5-HT1A receptor binding with the degree of epileptic activity in the explored regions, as assessed by the paroxysmal ictal and interictal activity recorded by intracerebral electrodes.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Patients and controls
Nine TLE patients, eight males and one female, aged 23–58 years (mean ± SD, 35.9 ± 11.8) were investigated at the Neurological Hospital in Lyon between February 2001 and January 2003. Patients, clinical features and depth electrode investigations are described in Tables 1 and 2.
|
|
All patients suffered from drug resistant temporal lobe epilepsy and all but one were receiving antiepileptic drug (AED) polytherapy at the time of PET investigation (see Table 1 for details). All patients underwent intracranial SEEG in the course of their presurgical investigation. The [18F]MPPF-PET was performed either during the third week of the SEEG investigation with depth electrodes in place, or before or after (7 months) depth electrode implantation. The last seizure occurred 1–15 days before the PET study. Each patient had between seven and 15 depth electrodes implanted unilaterally (in five patients) or bilaterally (in four patients). The cortical structures to be explored were chosen according to interictal EEG, video-scalp-EEG ictal data, interictal [18F]fluorodeoxyglucose PET, interictal and ictal single photon emission tomography (SPECT) and brain MRI data. The video-SEEG monitoring was carried out over 10–16 days.
Fifty-three healthy volunteers consisting of 27 women (age 19–70 years, mean ± SD, 42.6 ± 14.8) and 26 men (age 20–68 years, mean ± SD, 40.8 ± 13.4) also underwent a brain MRI and a [18F]MPPF-PET and were used as a control population. None of the control subjects had psychiatric or neurological illnesses as assessed by medical examination and inspection of the anatomical T1 MRI. Similar to the patient group, control group investigation involved a cerebral MRI and a [18F]MPPF-PET session.
Patients and controls were evaluated for depression using the General Health Questionnaire (GHQ-28) of Goldberg (Goldberg and Hillier, 1979). None of the patients or controls reached the threshold (i.e. score 7) for depression. GHQ-28 depression scores ranged between 0 and 5 in patients (mean ± SD, 2.75 ± 2.1) and between 0 and 4 in controls (0.42 ± 0.8).
All subjects gave their informed consent to the protocol, which was approved by the local ethical committee (CCPPRB, Centre Léon Berard, Lyon) in accordance with the Declaration of Helsinki.
Data acquisition
MRI
MRI acquisition consisted of a 3D anatomical T1-weighted sequence using a 1.5 T Siemens Magnetom scanner (Siemens AG, Erlangen, Germany). The anatomical volume covered the whole brain with mm3 voxels.
PET
[18F]MPPF was obtained by nucleophilic fluoration on a nitro precursor with a radiochemical yield of 20–25% at the end of synthesis and a specific activity of 32–76 GBq/µmol (Le Bars et al., 1998).
PET sessions were performed on a CTI-SIEMENS HR+ (Siemens, Knoxville, TN, USA) during the afternoon. For tracer injections, an intravenous catheter was placed in the radial vein of the left arm. A thermoformable head holder was moulded for each subject in order to limit head movement during acquisition. Prior to the emission acquisition, a 10-min transmission scan was performed using three 68Ge rod sources for the measurement of tissue and head support attenuation.
After intravenous injection of a bolus of 186 ± 30 MBq [18F]MPPF, the dynamic PET scan of emission, consisting of 35 frames of increasing duration (20 s to 5 min) was acquired to evaluate the local radiotracer concentration during 50 min post-injection. The PET scanner was operated in 3D mode. Images were corrected for scatter and attenuation and reconstructed using a filtered back projection (Hamming filter of cut-off 0.5 cycles/pixels), in order to provide a 3D volume of 63 slices (2.42 mm thickness), with 128 x 128 voxels in plane (2.06 x 2.06 mm).
Data analysis
For each dynamic file, we visually controlled the head movements. In two cases where head movements occurred, the static images were taken as reference for an automatic linear spatial alignment (Automated Image Registration package; Woods et al., 1992) of the dynamic acquisition. From the realigned dynamic acquisition, we computed a new static image of equilibrium from 0 to 50 min post-injection. PET static images were re-oriented according to the bi-hippocampal plane. The same transformation matrix was applied to the dynamic data. Transaxial MRIs were co-registered with the static PET and re-sliced with the same sampling as the PET data. This pre-processing resulted in a complete data set (anatomic MRI, static and dynamic PET) with common orientation and size. Using a contour tool (CAPP; CTI/Siemens) we drew 400 regions of interest (ROIs) upon the registered MRI, grouped into 37 volumes of interest representing the actual anatomical regions (see Fig. 1). ROIs were outlined anatomically, following the grey matter ribbon. The identification of key sulci on individual MRI (central, precentral, postcentral, intraparietal, parieto-occipital, temporo-occipital) allowed for the anatomical delineation. Bilateral ROIs were drawn in temporal poles, parahippocampal gyri, amygdala, hippocampi and fronto-parietal opercula, as well as in occipital, inferior middle and superior temporal, inferior and superior parietal, orbito-frontal, inferior, middle and superior frontal, anterior and posterior cingulate, and insular gyri. As a reference for the simplified model, a large ROI was also drawn in the cerebellum. For raphe nuclei, which are difficult to delineate on MRI, contours were first drawn on the static PET image and then displayed on the subject MRI to verify their proper location in the peri-aqueductal grey matter of the brainstem.
|
All ROIs were applied to the dynamic image and, for each ROI, time–activity curves were extracted and used for data quantitation using a simplified reference tissue model developed for [11C]WAY100635, another antagonist of 5-HT1A receptors (Lammertsma and Hume, 1996
; Gunn et al., 1998) and recently validated for [18F]MPPF studies (Costes et al., 2002). This model provides three parameters [k2, R1 = k1ref/k1ROI, and binding potential (BP)] without the need for arterial sampling to define the input function: the free and non-specific ligand kinetic is based on the time–activity curve of a reference region (i.e. cerebellum) that is assumed to be devoid of 5-HT1A-specific binding. Using this model, we previously confirmed that BP was a good index of local receptor concentration (Bmax) in normal subjects (Costes et al., 2002). Of the three parameters extracted, only BP (BP = Bmax/Kd) values were further analysed. Reliability of cerebellum time–activity curves (TAC) was assessed by checking that each cerebellum individual TAC for controls and patients was included in the mean TAC obtained from all 53 controls ± 2 SD.
As gender differences in AMT uptake (Nishizawa et al., 1997; Okazawa et al., 2000) and gender-specific age differences in 5-HT1A binding (Meltzer et al., 2001) have previously been reported, BP values in male and female patients were respectively compared with gender-matched control groups (27 females and 26 males). For the ROI of each patient, the percentage BP variation compared with controls (
BP) was calculated with the following formula:
Influence of electrode implantation and other clinical parameters
To test for a possible global effect of electrode implantation, we compared mean global BP and mean BP in raphe nuclei in the group of patients who underwent SEEG during PET versus the group of three patients who had intracerebral exploration either before (patients 1 and 4) or 7 months after the PET session (patient 9).
To test for a possible local effect of electrode implantation, we compared BP variations in: (i) explored areas without epileptic activity (NE); and (ii) in the homologous non-implanted regions on the contralateral side (NI-contra). This comparison was done in the group of patients who had implanted electrodes during PET. Patient 3 was excluded from this group as he had implanted electrodes in the homologous contralateral regions as well.
To test for a possible influence of the delay between last seizure and PET, the duration of epilepsy, the GHQ-28 scores, the carbamazepine, lamotrigine and benzodiazepine daily treatment dose on BP changes, all these factors considered independently were tested with a multiple regression analysis against: (i) global BP changes; (ii) BP changes in raphe; and (iii) BP changes in the hippocampus on the epileptogenic side.
MPPF/SEEG data correlation
In six of nine patients, SEEG was recorded during the PET session. No seizures occurred during the PET acquisition, and no discharge lasting more than 5 s was recorded in these six patients.
All data acquired during the video-SEEG session were reviewed a posteriori and carefully analysed for categorization by three experienced reviewers blinded to PET results (K. Ostrowsky, J. Isnard and P. Ryvlin). The intracerebral activity was classified into: (i) onset: ictal discharge onset (low voltage rhythmic activity or first ictal activity detected); (ii) spread: ictal discharge spreading; (iii) IS: interictal spiking activity only (spikes, polyspikes, spike and waves); and (iv) NE: no epileptic activity. Non-implanted regions (NI) were also categorized according to whether they were ipsilateral (NI-ispsi) or contralateral (NI-contra) to the epileptogenic zone. In one patient (No. 8), an isolated ictal discharge was recorded in the left hippocampus, whereas all other recorded seizures started in the right mesiotemporal structures. For this patient, the left hippocampus was categorized as an ‘onset’ region along with the right mesiotemporal structures. In another patient (No. 9), an asymptomatic discharge was recorded in the right operculum outside the usual regions of seizure onset. Likewise, this region was categorized as an ‘onset’.
MRI analysis
In order to define the nature and extension of lesional areas (see Table 2), patient MRIs were visually analysed by the same observers used for the SEEG data (K. Ostrowsky, J. Isnard and P. Ryvlin). For this analysis, each ROI was categorized as either ‘lesional’ or ‘non-lesional’. As described in Table 2, the visual analysis showed: (i) no abnormality in three patients (Nos 1, 2 and 8); (ii) a unilateral hippocampal sclerosis ipsilateral to the epileptic focus in four patients (Nos 3, 4, 6 and 9), associated, in two of them (patients 4 and 9), with a loss of grey–white matter differentiation in the temporo-polar region ipsilateral to the hippocampal sclerosis; (iii) a temporo-polar porencephalic lesion in one patient (No. 5); and (iv) a cavernous angioma of the left superior temporal gyrus, excised 1 year before SEEG investigation in the remaining patient (No. 7).
In addition, the volumes of ROIs were estimated quantitatively for all subjects (see Table 3). For patients, ROI volumes were considered as significantly reduced when below the mean corresponding ROI volume minus 2 SD in the gender-matched control subgroup. For lesional ROIs (visually defined), where a significant volume decrease was quantitatively measured, the ROI was re-labelled as ‘lesional with volume loss’. None of the non-lesional ROIs (as visually defined) showed significant volume reduction when compared with the corresponding control ROI.
|
Statistical analysis
After verifying that
BP percentages were normally distributed, we performed an analysis of variance (ANOVA) statistical analysis (StatView®; Abacus Concepts) on the full group of patients using BP as the dependent variable, and the type of intracerebral activity (onset, spread, IS, NE, NI-ipsi, NI-contra) as the factor. The same analysis was then performed separately on the group of non-lesional ROIs, lesional ROIs without volume loss and with volume loss. |
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
SEEG results
The results of the SEEG recording for each patient are reported in Table 4. Mesio-temporal seizure onsets were identified in seven patients (Nos 1–4, 6, 8 and 9), whereas in the two remaining patients ictal discharges started outside the mesio-temporal structures. In patient 5, seizures started in the right temporal pole, where a major porencephalic lesion was clearly discernable and spread quickly to mesio-temporal structures as well as to neocortical temporal structures. In the second patient (No. 7), only post-discharges, but no spontaneous seizure, were recorded in the left superior temporal gyrus after electrical stimulation of the anterior or posterior superior temporal gyrus.
|
PET data
Visual analysis
Controls. As described previously, PET static images in controls showed a high uptake of [18F]MPPF in cortical regions recognized to be rich in 5-HT1A receptors, namely limbic (hippocampus, amygdala, parahippocampal gyrus) and paralimbic (temporal pole, insula, anterior and posterior cingulate gyri) regions, and a less intense concentration in neocortical regions. Owing to their high density in 5-HT1A auto-receptors, raphe nuclei were easily identifiable in the brainstem, despite their small volume. [18F]MPPF binding in a control subject is illustrated as an example at the top of Fig. 2.
|
Patients. In six of nine epileptic patients, a clear asymmetry could be visually identified before quantification on radioactivity PET images. Abnormalities were widespread in three patients (Nos 5, 6 and 9) and more focal in three other patients (Nos 1, 3 and 8). Two different examples (patients 3 and 9) are shown in Fig. 2 (bottom). Scans of the three remaining patients (Nos 2, 4 and 7) did not reveal any clear abnormality on visual inspection. These three patients were not different from the others in terms of history, AED therapy, seizure onset or brain MRI.
ROI analysis
The comparison of each patient with their gender-matched control group revealed significant BP changes in eight of nine patients. Results are presented in Table 5.
|
Significant BP decreases (i.e. inferior to the mean control values – 2 SD) were focal in three patients (Nos 1, 4 and 6) and located ipsilateral to the seizure onset, in the temporal pole (patient 1), in the hippocampus (patient 4) or in the mesiotemporal structures and temporal pole (patient 6). Two patients (Nos 5 and 9) showed a more widespread significant BP decrease ispilateral to seizure onset in mesiotemporal regions, temporal pole and temporal neocortex, as well as in the ipsilateral insula for patient 5. In two patients (Nos 3 and 8), significant BP decreases were found in mesiotemporal structures and in the temporal pole bilaterally.
In addition, two patients (Nos 4 and 9) showed a significant BP increase in bilateral frontal and parietal regions (patient 4) and in the controlateral mesiotemporal regions, temporal pole, and inferior temporal gyrus (patient 9). Finally, one patient only showed a significant BP increase in raphe nuclei (patient 7), and another one showed no significant BP variation compared with controls (patient 2).
Compared with intracerebral data, the significant decrease of BP was concordant with the ictal onset region in five cases (patients 4–6, 8 and 9), occurred in a propagation area in another case (patient 1), and in the non-explored temporal pole in one patient (No. 3). In this latter case, however, the temporal pole was more than likely involved in the discharge spreading. In the last two patients (Nos 2 and 7), the maximal BP decrease was concordant with the onset zone but did not reach the significant threshold.
When considering BP changes in each patient with MRI abnormalities, the region of significant MPPF binding decrease matched the region of morphological change in one patient (No. 5). In most other cases, significant BP decreases could be found both in lesional and non-lesional regions. Finally, in one patient (No. 7), a BP decrease was found in the lesional region, but did not reach a significant level. In the three patients with no MRI abnormalities, one showed no abnormalities in MPPF binding (patient 2), one showed a focal decrease in the temporal pole ipsilateral to seizure onset (patient 1) and the third showed a bilateral decrease of BP in mesiotemporal and temporopolar regions (patient 8).
Global BP and BP in raphe
Global BP in patients ranged beween 0.47 and 0.74 (mean ± SD, 0.60 ± 0.08) and were not significantly different for female (0.62 ± 0.11) or male (0.62 ± 0.10) controls.
BP in raphe ranged from 0.42 to 0.84 (0.59 ± 0.17) in patients, from 0.17 to 0.81 (0.48 ± 0.15) in male controls and from 0.23 to 0.99 (0.52 ± 0.17) in female controls. In patients 4 and 7, BP values in raphe exceeded mean control values +2 SD, respectively, by 76.4 and 70.8%.
There was no correlation between global BP and BP in raphe.
Influence of electrode implantation and other clinical parameters
Compared with controls, the mean (± SD) variation of global BP was –5.8 ± 12.9% in the group of six patients who had implanted electrodes during PET data acquisition, and 5.4 ± 12.1% in the group of three patients for whom the PET was not performed during the SEEG investigation. This difference was not significant. No significant difference was found in raphe nuclei between the mean BP variation in the group of implanted patients (24.4 ± 31.7%) compared with the group of non-implanted patients (18.3 ± 50.3%).
Similarly, there was no local influence of electrode implantation. Compared with controls, BP change was 3.4 ± 8.0% in implanted structures where no epileptic activity was recorded (either interictal or ictal) and 2.5 ± 16.4% in the homologous regions of the non-implanted contralateral hemisphere. A within-subject paired t-test did not reveal any significant difference between BP changes in the two groups of regions.
The multiple regression analysis did not reveal any significant effect of epilepsy duration, delay between PET and last seizure, depression score, and AED treatment on global BP changes, BP changes in raphe or BP changes in the hippocampus on the epileptogenic side.
Correlation with intracerebral activity
Compared with the control group, BP in patients was decreased in 90, 74, 57 and 27% of ROIs, where, respectively, discharge onsets, spreading, interictal activity only and no epileptic activity had been recorded. These BP reductions encompassed the mean control values – 2 SD in 45, 28 and 7%, and none of the ROIs belonging, respectively, to the onset, spread, IS and NE groups.
BP versus type of intracerebral activity
On average, BP values were decreased by 26.9 ± 20.7% in onset ROIs, by 17.3 ± 28.1% in spread ROIs, and by 2.5 ± 24.0% in regions where only interictal activity occurred (see Fig. 3). Conversely, BP was increased by 8.4 ± 11.1% in NE regions, 3.2 ± 26.4% in NI-ipsi regions and 10.2 ± 21.2% in NI-contra regions. ANOVA revealed a significant effect of the type of intracerebral activity on BP (F = 15.7; P < 0.0001). The post hoc analysis revealed a significantly greater decrease of BP in regions of seizure onsets than in those where interictal activity (P < 0.004) or no epileptic activity had been recorded (P < 0.0001). Likewise, a significantly greater reduction of BP values was also observed in regions where discharges spread compared with regions where interictal activity (P < 0.04) or no epileptic activity (P < 0.0004) was recorded. No significant difference in mean BP variation was identified between onset and spread regions, or IS and NE regions.
|
BP versus type of intracerebral activity in lesional regionsCompared with controls, BP values were on average decreased by 58.2 ± 22.8% in lesional regions with volume loss and by 42.5 ± 15.6% in lesional regions without volume loss, and increased by 3.7 ± 24.0% in non-lesional regions. ANOVA revealed a significant effect of the ‘lesion’ factor on
BP (F = 31.07; P < 0.0001). The post hoc analysis revealed a significantly greater decrease of BP in lesional ROIs with volume loss than in non-lesional ROIs (P < 0.0001), and in lesional ROIs without volume loss than in non-lesional ROIs (P < 0.0001). No significant difference was identified between lesional regions with and without volume loss. As illustrated in Fig. 4, for each type of intracerebral activity considered, the mean percentage of BP decreased gradually in non-lesional, lesional without volume loss and lesional with volume loss ROIs.
|
When considering only lesional regions with volume loss, no significant difference appeared between the BP decrease in onset and spread ROIs.
When considering only lesional regions without volume loss, BP was decreased by 41.2 ± 20.1% in onset ROIs, by 46 ± 4.1% in spread ROIs, by –61.5% (one value) in IS ROIs and by –38.4 ± 34% in NE regions. ANOVA did not reveal any difference of BP between these groups.
BP versus type of intracerebral activity in non-lesional regions
When considering only non-lesional ROIs, BP values were decreased by 19.1 ± 14.8% in onset ROIs and by 12.6 ± 26.1% in spread ROIs (see Fig. 4). BP was increased by 2.0 ± 17.6% in regions where only interictal activity occurred (IS), by 8.4 ± 11.1% in NE regions, by 4.3 ± 26.0% in NI-ipsi regions and by 10.2 ± 21.2% in NI-contra regions. ANOVA revealed a significant effect of the type of intracerebral activity on BP (F = 9.4; P < 0.0001). The post hoc analysis showed a greater decrease of BP in regions of seizure onsets than in those where interictal activity (P < 0.02) or no epileptic activity (P < 0.001) had been recorded. Likewise, a significantly greater reduction of BP values was also observed in regions where discharges spread compared with regions where interictal activity (P < 0.05) or no epileptic activity (P < 0.003) was recorded. No significant difference in mean BP variation was identified between onset and spread regions, or IS and NE regions.
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
By combining [18F]MPPF-PET with intracerebral recordings, we show in this study that decreased 5-HT1A receptor binding, recently reported by Toczek et al. (2003)
, is related to the paroxysmal activity in the epileptogenic temporal lobe, and does not just reflect lesion-related histological changes. Before discussing these conclusions, the influence of electrode implantation and clinical parameters on PET findings deserves some attention.
Parameters possibly influencing our results
Studies in a number of animal species have revealed that implantation of electrodes per se may induce neurochemical, histological or vascular alterations (Boast et al., 1976; Ben Attia et al., 1992; Loscher et al., 1995). However, we found no significant effect of electrode implantation on global brain or raphe nuclei BP. Moreover, regarding a possible local effect of electrode implantation, we found no BP differences between non-epileptic areas on the focus side and contralateral non-implanted homologous regions. This suggests that, in our group of patients, global or focal changes in 5-HT1A binding due to electrode implantation are unlikely.
Symptoms of depression and anxiety are frequent in patients with epilepsy (Kanner and Palac, 2000), and might interact with our BP measures, as 5-HT1A receptors have been shown to be involved in major depression (Cowen, 2000). Indeed a post mortem study on suicide victims with major depressive disorders showed reduced 5-HT1A receptor mRNA in the hippocampus (Lopez et al., 1998). A PET study using [11C]WAY100635 showed an abnormal reduction in 5-HT1A receptor BP in patients compared with controls located mainly in the raphe, and in mesiotemporal cortex, but extending also to occipital and postcentral regions (Drevets et al., 1999). In our group of patients, although individual GHQ-28 did not reach the threshold for depression in any patient (or control), the mean depression score was different in patients from that in controls. However, no significant effect of this score could be identified on BP changes in raphe or in hippocampus on either side.
The delay between the occurrence of last seizure and the PET might also possibly influence [18F]MPPF binding. Studies in animal models of epilepsy show a trend towards a binding increase to 5-HT1A receptors appearing 1 week after stage 5 in the kindling model (Cagnotto et al., 1998), or at day 6 in the kainate model (Van Bogaert et al., 2001). These changes were shown to be long-lasting. In our population of patients, we did not evidence any effect of the delay between last seizure and PET on BP changes in raphe or hippocampus, or on global BP modifications.
Finally, no significant effect of the AED treatment could be shown on BP modifications, although AED are known to interact with the serotoninergic system and to be effective in depression disorders (Harden, 2002).
Nevertheless, given the very small number of measures, and the high SDs, this does not allow for definitely ruling out the influence of all these factors on our results.
Individual data
Given the small group of patients, individual results are difficult to discuss, especially in terms of relationships between the epileptic activity and the BP changes.
Individual results are compatible with a close relationship between the significant decrease in MPPF binding and the location of the area where seizures started and/or spread. Results are, however, very variable across patients, as some of them showed very focal areas of BP decrease, whereas in other cases the PET functional abnormality was more widespread. It is, however, interesting to note that in the case where bilateral discharges were recorded, significant bilateral changes in MPPF binding occurred.
Individual data also show a clear relationship between the existence of a lesion on MRI and a significant BP decrease. Indeed, a significant BP decrease was measured in 11 of 13 lesional regions (with or without volume loss). However, it is noteworthy that significant and major BP decreases also occurred in non-lesional regions.
Interestingly, in some patients, a significant increase in MPPF binding was measured in regions located on the contralateral side. In rats, an increase in binding to 5-HT1A receptors has been identified during the late chronic phase, bilaterally in the hippocampus, after kainic acid injection (Van Bogaert et al., 2001) or kindling (Clark et al., 1993; Cagnotto et al., 1998). Contrary to these experimental results, the BP increase we observed in our group of TLE patients was never located within the epileptogenic zone, but involved contralateral regions. Given a possible antiepileptic effect of serotonin through 5-HT1A receptors, this increase could be interpreted as a regulation process in response to the epileptic discharges occurring, part of which could contribute to the modulation of neuronal hyperexcitability.
In two patients we observed a significant binding increase compared with controls in raphe nuclei. In these two patients, we failed to find a relationship between this increase and other clinical factors such as TLE syndrome (mesiotemporal versus neocortical seizure onsets), MRI, treatment, past history, or GQH-28 score for depression or anxiety. On the contrary, in their study, Toczek et al. (2003) described a decrease of [18F]FCWAY binding in raphe nuclei of epileptic patients. These contradictory results might be explained by the different binding properties of [18F]MPPF and [18F]FCWAY used in these two studies, as [18F]MPPF might be more sensitive to variation in the concentration of endogenous serotonin than [18F]FCWAY.
Mechanisms of 5-HT1A receptor binding decrease in the epileptogenic temporal lobe
The reduction of 5-HT1A binding that we observe on the side of the epileptogenic zone is in agreement with data obtained in some epilepsy models. Indeed, in genetically epilepsy-prone rats, a decrease of binding of the agonist [3H]8-OH-DPAT [8-hydroxy-2-(di-n-propylamino) tetralin] to 5-HT1A receptors was observed in the hippocampus (Statnick et al., 1996). It also is in agreement with results of a study using another antagonist of 5-HT1A receptors, [18F]FCWAY, showing reduced 5-HT1A binding in mesial and lateral temporal regions on the side of the epileptic focus (Toczek et al., 2003).
[18F]MPPF has an affinity (Ki = 3.3 nM) close to that of serotonin (Ki = 4.7 nM) for the 5-HT1A receptor (Zhuang et al., 1994), and was shown to be sensitive to endogenous variations of serotonin concentration (Zimmer et al., 2002). Thus, a decrease of MPPF binding can be interpreted as reflecting either a decrease in receptor density or an increase of endogenous serotonin, resulting in a competition for receptor binding by the radioligand. In the first hypothesis, a causal process could be involved in which depletion in serotonin could lead to a down-regulation of 5-HT1A receptors and hyperexcitability. In the second hypothesis, a reactional process might be involved, in which epilepsy itself leads to an increase in serotonin concentration in order to modulate the neuronal hyperexcitability.
In the absence of a concomitant investigation of serotonin synthesis it is impossible to decide whether the binding decrease we observed in discharge onset and spreading zones reflects a decreased density in 5-HT1A receptors, or a higher occupancy of these receptors by endogenous serotonin. Studies in human epilepsy on resected tissues have indicated an increase of serotonin concentration and of its metabolite, 5-hydroxyindolacetic acid (5-HIAA), in the epileptogenic zone, this increase being greater in regions where frequent interictal spikes had been recorded (Louw et al., 1989; Pintor et al., 1990). These data suggest that the decrease in receptor binding observed in our patients could be explained by an increase of serotonin concentration resulting in a competition with the radioligand for receptor binding.
To that extent, the increase in [11C]-methyl-tryptophan ([11C]AMT) uptake in tuberous sclerosis (Chugani et al., 1998; Fedi et al., 2003), cortical dysplasia and cryptogenic epilepsy (Fedi et al., 2001), including TLE with no hippocampal volume loss (Natsume et al., 2003), supports this hypothesis, as far as [11C]AMT uptake is interpreted as reflecting 5-HT synthesis and not quinolinic acid synthesis, as recently suggested (Chugani and Muzik, 2000).
On the other hand, as in our study, Toczek et al. (2003) also reported a decrease in 5-HT1A binding in their patients, using a PET ligand with a much higher affinity for 5-HT1A receptors than MPPF, and being therefore potentially less displaceable. This therefore would be more in favour of a decrease in 5-HT1A receptor density than of a higher occupancy of these receptors by endogenous serotonin.
Binding modifications in relation to intracerebral activity
The analysis of our data using anatomical ROI allowed investigation of the relationship between regional BP decreases of 5-HT1A receptors and the type of epileptic activity recorded by intracerebral electrodes. Independently of the existence of a lesion, BP variations in patients compared with controls gradually decreased from regions showing no epileptic activity to regions with interictal activity, seizure propagation and seizure onsets. Therefore, these data suggest that the reduction of BP is correlated to the degree of epileptic activity in the explored areas.
However, when considering only lesional regions, the BP decrease no longer correlated with the degree of epileptic activity, so that part of the above observed changes might reflect partial volume effect. Indeed, the limited spatial resolution of PET results in a partial volume effect, that in particular affects the quantification of signals in small structures or heterotopic nodules (Hoffman et al., 1979). Thus, when structural abnormalities are present, such as atrophy or inhomogeneity in grey matter content, the BP can appear smaller than it actually is, and a correction for partial volume effect is necessary in order to accurately quantify changes in the binding to receptors (Labbe et al., 1996; Koepp et al., 1998; Rousset et al., 1998; Hammers et al., 2001).
Accordingly, in our group of patients, the significantly greater BP decrease in lesional ROIs (either with or without volume loss) than in non-lesional regions can be attributed to partial volume effects. Therefore, to avoid any influence of partial volume effects, lesional regions were eliminated from the further statistical analysis.
Conversely, in non-lesional regions the differences in BP decrease for onset, spreading, interictal spiking and non-epileptic regions are unlikely to reflect partial volume effects, since the volume of homologous regions was not significantly different in patients and controls. In these non-lesional regions, we found that the BP decrease still correlated to the degree of epileptic activity. Therefore, these results show that MPPF binding not only reflects structural changes or neuronal loss, but it also can be considered as a marker of the epileptogenic zone.
|
Acknowledgements |
|---|
We wish to thank V. Berthet, C. Vighi and M. Lionnet for careful handling of patients and controls during PET investigations, C. Pierre for informatics assistance, the Chemistry Department of CERMEP for [18F]MPPF synthesis and PET acquisitions, A. Paquin for advice on the manuscript, and D. Comar, J. F. Pujol and G. Gimenez for providing encouragement and support for this study. This work was financially supported by the Claude Bernard University of Lyon (BQR, 2001) and by a grant from Glaxo-Smith-Kline (awarded to K.O.).
|
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Barnes NM, Sharp T. A review of central 5-HT receptors and their function. Neuropharmacology 1999; 38: 1083–152.[CrossRef][ISI][Medline]
BenAttia M, N’Gouemo P, Belaidi M, Rondouin G, Chicheportiche R. Kindling and electrode effects on the benzodiazepine receptors density of olfactory bulb and hippocampus after olfactory bulb kindling. Neurosci Lett 1992; 143: 74–8.[CrossRef][ISI][Medline]
Boast CA, Reid SA, Johnson P, Zornetzer SF. A caution to brain scientists: unsuspected hemorrhagic vascular damage resulting from mere electrode implantation. Brain Res 1976; 103: 527–34.[CrossRef][ISI][Medline]
Bonnycastle DD, Giarman NJ, Paasonen MK. Anticonvulsivant compounds and 5-hydroxytryptamine in rat brain. Br J Pharmacol 1957; 12: 228–31.
Cagnotto A, Crespi D, Mancini L, Manzoni C, Presti ML, Gariboldi M, et al. Lasting increase in serotonin 5-HT1A but not 5-HT4 receptor subtypes in the kindled rat dentate gyrus: dissociation from local presynaptic effects. J Neurochem 1998; 70: 850–7.[ISI][Medline]
Chugani DC, Muzik O. Alpha[C-11]methyl-L-tryptophan PET maps brain serotonin synthesis and kynurenine pathway metabolism. J Cereb Blood Flow Metab 2000; 20: 2–9.[CrossRef][ISI][Medline]
Chugani DC, Chugani HT, Muzik O, Shah JR, Shah AK, Canady A, et al. Imaging epileptogenic tubers in children with tuberous sclerosis complex using alpha-[11C]methyl-L-tryptophan positron emission tomography. Ann Neurol 1998; 44: 858–66.[CrossRef][ISI][Medline]
Clark M, Weiss SR, Post RM. Autoradiographic analysis of serotonin receptors and transporter in kindled rat brain. Neurosci Lett 1993; 161: 21–6.[CrossRef][ISI][Medline]
Costes N, Merlet I, Zimmer L, Lavenne F, Cinotti L, Delforge J, et al. Modeling [18 F]MPPF positron emission tomography kinetics for the determination of 5-hydroxytryptamine(1A) receptor concentration with multiinjection. J Cereb Blood Flow Metab 2002; 22: 753–65.[ISI][Medline]
Cowen PJ. Psychopharmacology of 5-HT(1A) receptors. Nucl Med Biol 2000; 27: 437–9.[CrossRef][ISI][Medline]
Dailey JW, Yan QS, Mishra PK, Burger RL, Jobe PC. Effects of fluoxetine on convulsions and on brain serotonin as detected by microdialysis in genetically epilepsy-prone rats. J Pharmacol Exp Ther 1992; 260: 533–40.
Drevets WC, Frank E, Price JC, Kupfer DJ, Holt D, Greer PJ, et al. PET imaging of serotonin 1A receptor binding in depression. Biol Psychiatry 1999; 46: 1375–87.[CrossRef][ISI][Medline]
Engelborghs S, D’Hooge R, De Deyn PP. Pathophysiology of epilepsy. Acta Neurol Belg 2000; 100: 201–13.[ISI][Medline]
Fedi M, Reutens D, Okazawa H, Andermann F, Boling W, Dubeau F, et al. Localizing value of alpha-methyl-L-tryptophan PET in intractable epilepsy of neocortical origin. Neurology 2001; 57: 1629–36.
Fedi M, Reutens DC, Andermann F, Okazawa H, Boling W, White C, et al. alpha-[11C]-Methyl-L-tryptophan PET identifies the epileptogenic tuber and correlates with interictal spike frequency. Epilepsy Res 2003; 52: 203–13.[CrossRef][ISI][Medline]
Filakovszky J, Gerber K, Bagdy G. A serotonin-1A receptor agonist and an N-methyl-D-aspartate receptor antagonist oppose each others effects in a genetic rat epilepsy model. Neurosci Lett 1999; 261: 89–92.[CrossRef][ISI][Medline]
Gariboldi M, Tutka P, Samanin R, Vezzani A. Stimulation of 5-HT1A receptors in the dorsal hippocampus and inhibition of limbic seizures induced by kainic acid in rats. Br J Pharmacol 1996; 119: 813–8.[ISI][Medline]
Gerber K, Filakovszky J, Halasz P, Bagdy G. The 5-HT1A agonist 8-OH-DPAT increases the number of spike-wave discharges in a genetic rat model of absence epilepsy. Brain Res 1998; 807: 243–5.[CrossRef][ISI][Medline]
Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychol Med 1979; 9: 139–45.[ISI][Medline]
Gunn RN, Sargent PA, Bench CJ, Rabiner EA, Osman S, Pike VW, et al. Tracer kinetic modeling of the 5-HT1A receptor ligand [carbonyl-11C]WAY-100635 for PET. Neuroimage 1998; 8: 426–40.[CrossRef][ISI][Medline]
Hammers A, Koepp MJ, Richardson MP, Labbe C, Brooks DJ, Cunningham VJ, et al. Central benzodiazepine receptors in malformations of cortical development: a quantitative study. Brain 2001; 124: 1555–65.
Harden CL. The co-morbidity of depression and epilepsy: epidemiology, etiology, and treatment. Neurology 2002; 59 (Suppl 4): S48–55.
Hoffman EJ, Huang SC, Phelps ME. Quantitation in positron emission computed tomography: 1. Effect of object size. J Comput Assist Tomogr 1979; 3: 299–308.[ISI][Medline]
Kanner AM, Palac S. Depression in epilepsy: a common but often unrecognized comorbid malady. Epilepsy Behav 2000; 1: 37–51.[CrossRef][Medline]
Koepp MJ, Hand KS, Labbe C, Richardson MP, Van Paesschen W, Baird VH, et al. In vivo [11C]flumazenil-PET correlates with ex vivo [3H]flumazenil autoradiography in hippocampal sclerosis. Ann Neurol 1998; 43: 618–26.[CrossRef][ISI][Medline]
Labbe C, Froment JC, Kennedy A, Ashburner J, Cinotti L. Positron emission tomography metabolic data corrected for cortical atrophy using magnetic resonance imaging. Alzheimer Dis Assoc Disord 1996; 10: 141–70.[ISI][Medline]
Lammertsma AA, Hume SP. Simplified reference tissue model for PET receptor studies. Neuroimage 1996; 4: 153–8.[CrossRef][ISI][Medline]
Le Bars D, Lemaire C, Ginovart N, Plenevaux A, Aerts J, Brihaye C, et al. High-yield radiosynthesis and preliminary in vivo evaluation of p-[18F]MPPF, a fluoro analog of WAY-100635. Nucl Med Biol 1998; 25: 343–50.[CrossRef][ISI][Medline]
Lerner-Natoli M. Serotonin and kindling development. Int J Neurosci 1987; 36: 139–51.[ISI][Medline]
Lopez JF, Chalmers DT, Little KY, Watson SJ. A.E. Bennett Research Award. Regulation of serotonin1A, glucocorticoid, and mineralocorticoid receptor in rat and human hippocampus: implications for the neurobiology of depression. Biol Psychiatry 1998; 43: 547–73.[CrossRef][ISI][Medline]
Loscher W, Wahnschaffe U, Honack D, Rundfeldt C. Does prolonged implantation of depth electrodes predispose the brain to kindling? Brain Res 1995; 697: 197–204.[CrossRef][ISI][Medline]
Louw D, Sutherland GR, Glavin GB, Girvin J. A study of monoamine metabolism in human epilepsy. Can J Neurol Sci 1989; 16: 394–7.[ISI][Medline]
Lu KT, Gean PW. Endogenous serotonin inhibits epileptiform activity in rat hippocampal CA1 neurons via 5-hydroxytryptamine1A receptor activation. Neuroscience 1998; 86: 729–37.[CrossRef][ISI][Medline]
Meltzer CC, Drevets WC, Price JC, Mathis CA, Lopresti B, Greer PJ, et al. Gender-specific aging effects on the serotonin 1A receptor. Brain Res 2001; 895: 9–17.[CrossRef][ISI][Medline]
Natsume J, Kumakura Y, Bernasconi N, Soucy JP, Nakai A, Rosa P, et al. Alpha-[11C] methyl-L-tryptophan and glucose metabolism in patients with temporal lobe epilepsy. Neurology 2003; 60: 756–61.
Nishizawa S, Benkelfat C, Young SN, Leyton M, Mzengeza S, de Montigny C, et al. Differences between males and females in rates of serotonin synthesis in human brain. Proc Natl Acad Sci USA 1997; 94: 5308–13.
Okazawa H, Leyton M, Benkelfat C, Mzengeza S, Diksic M. Statistical mapping analysis of serotonin synthesis images generated in healthy volunteers using positron-emission tomography and alpha-[11C]methyl-L-tryptophan. J Psychiatry Neurosci 2000; 25: 359–70.[ISI][Medline]
Peroutka SJ. 5-HT receptors: past, present and future. Trends Neurosci 1995; 18: 68–9.[CrossRef][ISI][Medline]
Pintor M, Mefford IN, Hutter I, Pocotte SL, Wyler AR, Nadi NS. Levels of biogenic amines, their metabolites, and tyrosine hydroxylase activity in the human epileptic temporal cortex. Synapse 1990; 5: 152–6.[CrossRef][ISI][Medline]
Rousset OG, Ma Y, Evans AC. Correction for partial volume effects in PET: principle and validation. J Nucl Med 1998; 39: 904–11.
Salgado-Commissariat D, Alkadhi KA. Serotonin inhibits epileptiform discharge by activation of 5-HT1A receptors in CA1 pyramidal neurons. Neuropharmacology 1997; 36: 1705–12.[CrossRef][ISI][Medline]
Statnick MA, Dailey JW, Jobe PC, Browning RA. Abnormalities in 5-HT1A and 5-HT1B receptor binding in severe-seizure genetically epilepsy-prone rats (GEPR-9s). Neuropharmacology 1996; 35: 111–8.[CrossRef][ISI][Medline]
Toczek MT, Carson RE, Lang L, Ma Y, Spanaki MV, Der MG, et al. PET imaging of 5-HT1A receptor binding in patients with temporal lobe epilepsy. Neurology 2003; 60: 749–56.
Tokarski K, Zahorodna A, Bobula B, Hess G. Comparison of the effects of 5-HT1A and 5-HT4 receptor activation on field potentials and epileptiform activity in rat hippocampus. Exp Brain Res 2002; 147: 505–10.[CrossRef][ISI][Medline]
Trottier S, Evrard B, Vignal JP, Scarabin JM, Chauvel P. The serotonergic innervation of the cerebral cortex in man and its changes in focal cortical dysplasia. Epilepsy Res 1996; 25: 79–106.[CrossRef][ISI][Medline]
VanBogaert P, De Tiege X, Vanderwinden JM, Damhaut P, Schiffmann SN, Goldman S. Comparative study of hippocampal neuronal loss and in vivo binding of 5-HT1a receptors in the KA model of limbic epilepsy in the rat. Epilepsy Res 2001; 47: 127–39.[CrossRef][ISI][Medline]
Wada Y, Nakamura M, Hasegawa H, Yamaguchi N. Role of serotonin receptor subtype in seizures kindled from the feline hippocampus. Neurosci Lett 1992; 141: 21–4.[CrossRef][ISI][Medline]
Wada Y, Nakamura M, Hasegawa H, Shiraishi J. Microinjection of the serotonin uptake inhibitor fluoxetine elevates hippocampal seizure threshold in rats. Neurosci Res Commun 1993; 13: 143–8.
Wada Y, Shiraishi J, Nakamura M, Koshino Y. Role of serotonin receptor subtypes in the development of amygdaloid kindling in rats. Brain Res 1997; 747: 338–42.[CrossRef][ISI][Medline]
Watanabe K, Minabe Y, Ashby CR Jr, Katsumori H. Effect of acute administration of various 5-HT receptor agonists on focal hippocampal seizures in freely moving rats. Eur J Pharmacol 1998; 350: 181–8.[CrossRef][ISI][Medline]
Woods RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992; 16: 620–33.[ISI][Medline]
Yan QS, Jobe PC, Dailey JW. Evidence that a serotonergic mechanism is involved in the anticonvulsant effect of fluoxetine in genetically epilepsy-prone rats. Eur J Pharmacol 1994; 252: 105–12.[CrossRef][ISI][Medline]
Yan QS, Jobe PC, Dailey JW. Further evidence of anticonvulsant role for 5-hydroxytryptamine in genetically epilepsy-prone rats. Br J Pharmacol 1995; 115: 1314–8.[ISI][Medline]
Zhuang ZP, Kung MP, Kung HF. Synthesis and evaluation of 4-(2'-methoxyphenyl)-1-[2'-[N-(2"-pyridinyl)-p-iodobenzamido]ethyl] piperazine (p-MPPI): a new iodinated 5-HT1A ligand. J Med Chem 1994; 37: 1406–7.[CrossRef][ISI][Medline]
Zimmer L, Mauger G, Le Bars D, Bonmarchand G, Luxen A, Pujol JF. Effect of endogenous serotonin on the binding of the 5-HT1A PET ligand 18F-MPPF in the rat hippocampus: kinetic beta measurements combined with microdialysis. J Neurochem 2002; 80: 278–86.[CrossRef][ISI][Medline]