Microdialysis of the lateral and medial temporal lobe during temporal lobe epilepsy surgery

Microdialysis of the lateral and medial temporal lobe during temporal lobe epilepsy surgery

Surgical Neurology 63 (2005) 70 – 79 www.surgicalneurology-online.com Technique Microdialysis of the lateral and medial temporal lobe during tempora...

523KB Sizes 2 Downloads 40 Views

Surgical Neurology 63 (2005) 70 – 79 www.surgicalneurology-online.com

Technique

Microdialysis of the lateral and medial temporal lobe during temporal lobe epilepsy surgery Philip M. Thomas, MRCSI*, Jack P. Phillips, FRCSI, William T. O’Connor, PhD The National Department of Neurosurgery (PMT, JPP), Beaumont Hospital, Dublin 9, Ireland The Department of Human Anatomy and Physiology (PMT, WTO’C), The National Neuroscience Network, Conway Institute of Biomolecular and Biomedical Research, University College, Dublin 9, Ireland Received 15 December 2003; accepted 12 February 2004

Abstract

Background: This study was undertaken to establish whether, in temporal lobe epilepsy (TLE), there are relative differences between the lateral and ipsilateral medial temporal lobe in the extracellular levels of 3 of the human brain’s major neuroactive amino acids. Methods: Seven generally anesthetized patients with TLE undergoing anatomically standardized resective surgery had at operation microdialysis catheters inserted within the middle temporal gyrus (ie, lateral temporal lobe) and anterior hippocampus (ie, medial temporal lobe). Surface electrocorticography (ECoG) recordings were also obtained. Samples of 10-minute dialysate were quantified for glutamate, aspartate, and gamma-aminobutyric acid (GABA) using high-performance liquid chromatography; corresponding ECoG data were assessed for epileptiform activity. Where available, resection tissue was subjected to histopathological analysis. Results: The ratio of mean bsample 3Q dialysate levels of glutamate, aspartate, and GABA was approximately 20:2:1, respectively, in both the minimally epileptiform lateral (n = 7) and medial (n = 5) temporal lobe; between the 2 sets of samples, these levels were not significantly different (P N 0.05 for each amino acid studied). From the vigorously epileptiform medial temporal lobe of 2 patients, sample 3 dialysate levels of the excitatory amino acids glutamate and aspartate were found in considerably greater concentrations (between 15- and 37-fold) with correspondingly less dramatic increases of the inhibitory amino acid GABA (more than 11- and 13-fold). Laterally resected tissue (obtained in 3 cases) did not demonstrate significant cortical or subcortical abnormalities; medial resection tissue from all patients demonstrated, in varying degrees, hippocampal sclerosis. Conclusions: In the absence of significant tissue hyperexcitability, despite known differences in local cellular and/or histopathological architecture, the extracellular relationship among glutamate, aspartate, and GABA is not dissimilar in both the lateral and ipsilateral medial temporal lobe of TLE patients. Considerable disparity in dialysate levels recovered (eg, from the vigorously epileptiform medial temporal lobe) may be related to the functional (ie, hyperexcitable) status of the sampled tissue. D 2005 Elsevier Inc. All rights reserved.

Keywords:

Epilepsy surgery; Hippocampus; Microdialysis; Neuroactive amino acids; Temporal lobe

1. Introduction Recent neuroanatomical and functional studies in primates have suggested an associative interaction between the * Corresponding author. Neurosurgery Administration Office, The National Department of Neurosurgery, Beaumont Hospital, Dublin 9, Ireland. Tel.: +353 1 809 2175; fax: +353 1 809 2302. E-mail address: [email protected] (P.M. Thomas). 0090-3019/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.surneu.2004.02.031

lateral and ipsilateral medial temporal lobe [15], and it is now thought that such connections may explain why clinical features of seizures originating in either region are often similar [25]. However, despite known cytoarchictectural differences between the lateral and medial human temporal lobe, it is not known whether such differences are manifested at the neurochemical (ie, microenvironmental ) level in temporal lobe epilepsy ( TLE ). The revived interest

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79 Table 1 Summary of 7 patients with TLE proceeding to resective surgery Patient no.

Sex/age (y)

TLE seizure laterality

Preoperative drug therapy

1

F/19

Right

2

M/38

Right

3 4

F/38 F/20

Left Left

5

M/37

Right

6

F/20

Right

7

M/19

Left

Carbamazepine, sodium valproate Carbamazepine, topiramate Carbamazepine Carbamazepine, lamotrigine Carbamazepine, sodium valproate, vigabatrin Sodium valproate, lamotrigine Phenobarbitone, lamotrigine, tiagabine

in the surgical treatment of pharmaco-resistant TLE, and the parallel advent of novel in vivo techniques [eg, brain microdialysis (MD)], has created unique opportunities for neurosurgeons and neuroscientists to study more closely the chemistry associated with epilepsy. Within the human brain, the amino acids glutamate and aspartate are the principal mediators of excitatory synaptic transmission [6]; gamma-aminobutyric acid (GABA) is the predominant inhibitory neurotransmitter [20]. Various investigators have studied neuroactive amino acids in relation to the electrophysiological status of the medial temporal lobe of conscious patients with TLE [8,9,11,26]; others have reported dialysate levels of various amino acids from the

71

bepileptogenic zoneQ of patients undergoing resective TLE surgery under general anesthesia [4,22]. The application of in vivo human brain MD as a research method in epilepsy has thus provided important ( though not exhaustive ) preliminary information regarding neuroactive amino acids in relation to epileptiform phenomena [23]. In view of the modest accumulation to date of amino acid data regarding the epileptogenic human temporal lobe, we report the first MD results obtained for glutamate, aspartate, and GABA from the intraoperative, spontaneously epileptiform lateral and ipsilateral medial temporal lobe of generally anesthetized TLE surgery patients. We fully describe a relatively straightforward application of MD during TLE surgery (ie, one of the few opportunities to directly perform ethically permissible in vivo study of the human temporal lobe structures). Dialysate levels for the neuroactive amino acids studied were also compared to corresponding surface electrocorticography (ECoG ) measurements of epileptiform activity ( EA ), and where available, resection tissue was subjected to histopathological analysis.

2. Patients and methods The methods used in the present study were approved by the Beaumont Hospital Medical Research Ethics Committee, Dublin, Ireland. Informed consent was obtained preoperatively from all patients. We report lateral and ipsilateral medial temporal lobe MD data obtained from the study of 4 females and 3 males (mean age, 27.3 years ) (Table 1 ) with pharmaco-resistant TLE.

Fig. 1. Intraoperative photograph (patient 2) depicting MD of the lateral temporal lobe. The MD catheter has been placed through a catheter guide-bolt (white arrow). The catheter tip has been inserted 10 mm deep to the cortical surface of the middle temporal gyrus. Electrocorticography electrodes surround the MD catheter insertion point. Dialysate samples (black arrow) were collected at 10-minute intervals.

72

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

Fig. 2. Intraoperative photograph (patient 4) depicting MD of the medial temporal lobe. The MD catheter (white arrow) has been placed through a catheter guide-bolt. The catheter tip has been inserted 10 mm deep to the rostral surface of the anterior hippocampus. An ECoG electrode (black arrow) is within approximately 5 mm of the MD catheter insertion point. Dialysate samples were collected at 10-minute intervals.

2.1. Presurgical assessment The planning of each patient’s operative intervention was based on generally accepted principles of presurgical evaluation for epilepsy surgery [5], which included magnetic resonance imaging (MRI ), continuous scalp electroencephalography (EEG ) with video ictal and interictal telemetry, psychiatric, and neuropsychological assessment, and Wada amylobarbitone sodium internal carotid angiography for evaluation of language dominance and memory.

Once deemed suitable, surgical candidates were maintained on their particular antiepileptic pharmaceutical regimen before operation (Table 1 ). 2.2. Operative anesthesia At the time of surgery, patients had general anesthesia induced with fentanyl (Sublimaze, Jannsen-Cilag, Dublin, Ireland ) and propofol ( Diprivan, Zeneca, Cheshire, UK ). Tracheal intubation and continuous muscle relaxation was

Fig. 3. Schematic diagram (Patient 2) showing the intraoperative time-course of lateral and medial temporal lobe MD and ECoG sampling.

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

73

catheter was connected to a 2.5-mL syringe (CMA Microdialysis ) placed within a CMA 107 microinfusion pump (CMA Microdialysis) and perfused at a flow rate of 5 ll/min with Ringer’s solution (Baxter Healthcare, Dublin, Ireland) containing 4mM K +, 147mM Na+, 2mM Ca 2+, and 155mM Cl . After the perfusate medium reached the outlet tubing of the MD catheter, the catheter was considered safe for insertion ( ie, this confirmed that there would not be a net gain of fluid at the insertion site); the flow rate of the perfusion medium was then altered to 2 ll/min before insertion. Microdialysis catheters were then positioned over the chosen site of study, and an Elger Cortical Mapping Unit (Ad-Tech Medical Instruments, Racine, WI ) was deployed; ECoG electrodes were arranged on the tissue surface surrounding the proposed catheter tip insertion point with at least one electrode within 5 mm of that point (Figs. 1 and 2). Intraoperative ECoG data were then obtained and displayed with a Medelec DG Discovery PR 200 workstation (Vickers Medical, Surrey, UK ) using the following parameters: a bipolar and/or a referential montage with respect to the craniotomy bone margin electrode, sensitivity of 20 to 100 lV/mm, 70 Hz high frequency filter, 0.5 Hz low frequency

Fig. 4. High-performance liquid chromatography chromatograms for glutamate ( GL ) and aspartate ( A ). The chromatogram on the left depicts an external standard for glutamate and aspartate; glutamate and aspartate peaks elute at approximately 1.36 and 0.97 minutes, respectively. The chromatogram in the center represents a water injection or b blank Q reference. The chromatogram on the right represents glutamate and aspartate levels ( patient 5 ) detected in sample 3 medial temporal lobe dialysate.

achieved with vecuronium ( Norcuron, Organon, Dublin, Ireland ). Throughout surgery, anesthesia was maintained to keep within normotensive limits using 1% to 1.5% inhalational isoflurane, an O2/N2O ratio of 40:60, and fentanyl, without the use of additional pharmacological agents to activate potentially dormant epileptic foci. 2.3. Intraoperative techniques Approximation of the temporal lobe was by pterional craniotomy; the approach to the medial temporal lobe was by anatomically standardized surgical methods [14,19,28]. In all cases of MD sampling, a CMA 70 MD catheter (CMA Microdialysis, Stockholm, Sweden) comprising a 60-mm shaft and a 10-mm dialysis membrane ( polyamide; M r 20 000 cut-off) with an outer diameter of 0.62 mm was used. Before insertion of MD catheters into the middle temporal gyrus (ie, the lateral temporal lobe) ( Fig. 1 ) or the anterior hippocampus (ie, the medial temporal lobe) ( Fig. 2), each

Fig. 5. High-performance liquid chromatography chromatograms for GABA ( GA ). The chromatogram on the left depicts an external standard for GABA; GABA elutes at approximately 4.43 minutes. The chromatogram in the center represents a water injection or b blank Q reference. The chromatogram on the right represents the GABA level ( patient 2 ) detected in sample 3 lateral temporal lobe dialysate.

74

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

Table 2 Lateral temporal lobe sample 3 dialysate levels, ECoG, and MRI/histopathology Patient no.

Sample side

Glutamate (lM)

Aspartate (lM)

GABA (lM)

ECoG

1 2 3 4 5 6 7

Right Right Left Left Right Right Left

2.336 1.469 0.236 61.398 2.037 0.327 4.712

0.364 0.000 0.000 3.616 0.305 1.379 2.123

0.389 0.000 0.019 3.274 0.037 0.017 0.032

Minimal Minimal Minimal Minimal Minimal Minimal Minimal

(6.977) (109.030) (2.759) (292.245) (6.706) (12.186) (425.917)

(0.636) (21.030) (0.497) (35.036) (0.943) (3.712) (53.520)

(0.461) (1.503) (0.062) (0.835) (0.083) (0.157) (2.229)

Lateral temporal MRI/histopathology EA EA EA EA EA EA EA

NSA/NSA NSA/NSA NSA/NSA NSAa NSAa NSAa NSAa

NSA indicates no significant abnormalities. Figures in parentheses represent sample 1 dialysate levels. a Lateral temporal lobe MRI result in the absence of tissue resection specimen.

filter, and a paper speed of 30 mm/s. The operative diathermy and suction units were turned off during recordings to avoid ambient electrical interference. To minimize insertion trauma, care was then taken to insert MD catheters under microscopic guidance at a relatively slow rate (ie, approximately 10 mm/min ) within the designated regions of study ( ie, 10 mm deep to the exposed surface of the middle temporal gyrus or anterior hippocampus) . Microdialysis catheters were used in conjunction with a single-channel GMS guide-bolt ( GMS, Kiel-Mielkendorf, Germany ) placed within a clamp affixed to a Budde Halo retractor arm (Ohio Medical Instruments, Cincinnati, OH, USA) (Figs. 1 and 2). The guide-bolt-retractor arm assembly allowed for stability and ease of placement of the catheter. To account for tissue disruption-based amino acid release upon MD catheter insertion and the establishment of steady-state dynamics within the microenvironment [1,2,16], 5 minutes after MD catheter insertion, dialysate samples were collected at 10 -minute intervals for a period of at least 30 minutes ( though not exceeding 50 minutes in any case) into separate

capped microvials (CMA Microdialysis ); sample 3 dialysates acquired (ie, 25-35 minutes after catheter insertion) were therefore taken to represent the most common sample (ie, data-time points) with the least likelihood of reflecting the effects of MD catheter insertion trauma [13,17]. Given the flow rate of 2 ll/min, and the dead volume of the MD catheter outlet tubing (6.8 ll), there was a time lag of 3 minutes 24 seconds from tissue event to dialysate sample collection. At the time of collection, all MD samples were placed in an ice-bath and then frozen at 808C. Microdialysis samples were then thawed at 28C within 24 hours of collection and subjected to high-performance liquid chromatography (HPLC ) analysis. A brief case -in - point schematic diagram of intraoperative methods performed is included in Fig. 3. 2.4. Assessment of electrophysiology Epileptiform activity as sampled by ECoG was considered to have occurred using similar criteria as for scalp EEG [12] (ie, EA confirmed if sharp waves, spikes, or multiple

Fig. 6. Ten-second referential ECoG recording segments depicting (A) minimal EA from the lateral temporal lobe (patient 3), (B) minimal EA from the medial temporal lobe (patient 6), and (C) vigorous EA from the medial temporal lobe (patient 1). Electrocorticography recording segments depicted above were each taken from a 60-second epoch at the midpoint of the final two thirds of respective sample 3 dialysate sampling intervals.

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

75

2.5. Dialysate analysis Glutamate and aspartate absolute recovery was determined by precolumn derivatization of a 10-ll dialysate sample with o-phthaldialdehyde/mercaptoethanol reagent and separation by reversed-phase HPLC on a Biophase ODS 5-lm particle column (Knauer, Berlin, Germany ). The mobile phase contained 0.1M sodium acetate, 6.25% methanol, 1.5% tetrahydrofurane, and pH 6.95, perfused at a flow rate of 1 mL /min. A linear gradient system was used to clean the column after elution of glutamate and aspartate. This involved switching to 100% methanol for 2 minutes and then switching back to the original acetate buffer. The

Fig. 7. Line graphs representing glutamate (A), aspartate (B), and GABA (C) levels detected in consecutive 10-minute lateral temporal lobe dialysate samples acquired from patient 2 (triangular data points), patient 4 (square data points), and patient 7 (circular data points).

spikes were identified). Taking into account the time-lag from tissue event to dialysate acquisition, the overall frequency of EA for each 10-minute dialysate sampling interval was categorized using a spike -per-page schema [24] on the basis of visual analysis of a 60-second epoch at the midpoint of the last two thirds of each sampling period (ie, 6.5-7.5 minutes).

Fig. 8. Histograms representing mean sample 3 dialysate concentration (AM F SEM) for glutamate (A), aspartate (B), and GABA (C) from 7 patients with lateral temporal (LT) lobe minimal EA, 5 patients with medial temporal (MT) lobe minimal EA, and sample 3 dialysate concentrations from the vigorously epileptiform MT lobes of Patients 1 and 7.

76

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

Table 3 Medial temporal lobe sample 3 dialysate levels, ECoG, and histopathology Patient no.

Sample side

Glutamate (lM)

Aspartate (lM)

GABA (lM)

ECoG

Hippocampal sclerosis grading

1 2 3 4 5 6 7

Right Right Left Left Right Right Left

101.099 (89.410) 14.691 (22.156) 7.405 (62.841) 2.514 (38.021) 1.258 (3.582) 4.363 (9.886) 211.861 (471.116)

21.860 0.000 1.374 0.670 0.193 0.764 14.482

4.241 0.724 0.306 0.423 0.190 0.143 4.817

Vigorous EA Minimal EA Minimal EA Minimal EA Minimal EA Minimal EA Vigorous EA

Grade Grade Grade Grade Grade Grade Grade

(9.643) (3.414) (10.044) (5.175) (0.561) (1.876) (41.930)

(2.644) (0.812) (6.521) (0.294) (0.234) (0.245) (6.693)

III I I II I II I

Figures in parentheses represent sample 1 dialysate levels.

excitation wavelength of the Hitachi F1000 fluorescence detector ( Hitachi, Tokyo, Japan ) was set at 330 nm, and the emission cut off filter was set at 440 nm. The limit of detection was 0.5 pmol /sample for glutamate and aspartate [18] (Fig. 4). The GABA assay used was on the basis of precolumn derivatization of a 10-Al dialysate sample with o-phtaldialdehyde /t-butylthiol reagent and separation by reversedphase HPLC on a Nucleosil 3 C18 column ( Knauer ) perfused under isocratic conditions at a flow rate of 0.8 mL /min. The mobile phase was 0.15M sodium acetate, 1mM EDTA, 50 % acetonitrile, and pH 5.4. The BAS LC4B electrochemical detector ( Bioanalytical Systems, West Lafayette, IN, USA ) was set at +0.75V. The limit of detection was 50 fmol / sample for GABA [18] (Fig. 5 ). 2.6. Histopathology assessment Where available, depending on the surgical technique, lateral resection tissue was postoperatively examined by conventional histopathological methods [10]. Medially resected tissue, containing the anterior hippocampal region that had been intraoperatively subjected to MD and ECoG sampling, was also examined and graded for hippocampal sclerosis [27]. 2.7. Statistical analyses Microdialysis data are reported as absolute recovery levels ( lM ) detected in 10-minute dialysate samples acquired at a perfusion flow rate of 2 ll/min. Where average values have been calculated, these have been expressed as the mean F standard error of the mean. Statistical analysis between groups was assessed using unpaired two-tailed Student t test (Microsoft Excel, 2002); P values of b 0.05 were considered significant.

bA Q EA [24] (ie, minimal EA, b 6 spikes /min ) ( Fig. 6A ). Comparative examples of dialysate recovery-time profiles of lateral temporal lobe MD during minimal EA are included in Fig. 7. Lateral temporal lobe mean sample 3 dialysate recovery of glutamate, aspartate, and GABA ( n = 7 ) was ( lM F SEM ): 10.359 F 7.893, 1.112 F 0.475, and 0.538 F 0.425, respectively (Fig. 8). Microdialysis of the medial temporal lobe of 5 patients (patients 2- 6, inclusive ) ( Table 3 ) also occurred during periods of spontaneously minimal EA ( Figure 6B ); mean sample 3 dialysate recovery of glutamate, aspartate, and GABA (n = 5 ) was (lM F SEM) 6.046 F 2.143, 0.600 F 0.215, and 0.357 F 0.093, respectively (Fig. 8). Between the minimally epileptiform lateral and medial temporal lobe, dialysate levels were not significantly different ( P N 0.05 for each amino acid studied). In contrast, MD of the medial temporal lobe of patients 1 and 7 (Table 3) occurred during periods of spontaneous category bD Q EA [24] (ie, vigorous EA, N 24 spikes /min) (Fig. 6C) during the sample 3 sampling period and throughout all sampling periods, respectively. Comparative examples of dialysate recovery-time profiles of medial temporal lobe MD during spontaneously minimal (patient 5) vs vigorous ( patient 1) EA are included in Fig. 9. For patients 1 and 7, medial temporal lobe sample 3 dialysate recovery of glutamate, aspartate, and GABA was (lM) 101.099 and 211.861, 21.860 and 14.482, and 4.241 and 4.817, respectively (Fig. 8). In all cases, lateral temporal lobe MRI was unremarkable for significant pathology ( Table 2 ); histopathological examination of laterally resected tissue (where available) also did not reveal fundamental abnormalities (Table 2). Medially resected tissue, in all cases, revealed hippocampal sclerosis, though the degree to which this pattern was observed varied ranging from grades I to III (mild to marked or b classic Q Ammon’s horn sclerosis ) [27] (Table 3).

3. Results There were no complications attributed to our methodology of intraoperative MD performed in this study. We present absolute recovery MD data for lateral and ipsilateral medial temporal lobe glutamate, aspartate, and GABA from 7 patients undergoing anatomically standardized TLE surgery. Microdialysis of the lateral temporal lobe of all patients ( Table 2) occurred during periods of spontaneous category

4. Discussion In an attempt to control for certain variables among patients, and thus allow valid comparisons to be made, universally applied methods were performed pre -, intra -, and postoperatively, including the presurgical work- up, anesthetic protocol, type of MD catheter used, MD catheter insertion depth, MD catheter perfusion protocol, recording

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

Fig. 9. Line graphs representing glutamate (A), aspartate (B), and GABA (C) levels detected in consecutive 10-minute medial temporal lobe dialysate samples acquired from patient 1 (triangular data points) and patient 5 (square data points). In the case of patient 1, concomitant to the acquisition of sample 3 dialysate, there occurred a period of spontaneously vigorous EA.

and interpretation of electrophysiological measurements, storage and analysis of dialysate samples, and assessment of histopathological specimens.

77

In the lateral temporal lobe, given that all MD catheters were inserted 10 mm deep to the surface of the middle temporal gyrus, the catheter’s 10 -mm dialysis membrane tip likely interacted with the extracellular fluid of gray matter for 2 to 3 mm of its vertical span superficially, and the remainder (7- 8 mm ) was likely within the subcortical white matter. In the medial temporal lobe, all MD catheters were inserted 10 mm deep to the rostral surface of the anterior hippocampus; catheter tips therefore likely interacted distally with various regions including the extracellular fluid of the hippocampal pyramidal cell layer and dentate fascia. Thus, dialysate levels for all neuroactive amino acids studied probably reflect an amalgam of the microenvironment of, laterally, neuronal and glial cortical and subcortical cells and, medially, both neuronal and glial hippocampal cells. It is likely that initial neuroactive amino acid dialysate levels from both the lateral and medial temporal lobe of all patients also represent, in part, the effects of insertion trauma and the establishment of a new steady-state proximal to the catheter membrane tip [1,2,16]. In the wake of minimal EA, we were not surprised to find that lateral and medial temporal glutamate, aspartate, and GABA dialysate levels appeared to reach low levels within 25 to 45 minutes after MD catheter insertion, broadly in keeping with the previous observations of some researchers [17] and other investigators who inserted MD catheters intraoperatively into cortical brain regions of generally anesthetized TLE patients [4,22]. It is worth noting that in one case (patient 4), during spontaneously minimal EA throughout lateral temporal lobe MD sampling, although relatively low amino acid dialysate levels were reached within 35 to 45 minutes, there was a slightly prolonged trend to these levels (Fig. 7). It is known that both glutamate and aspartate are zwitterionic molecules and are unable to diffuse across membranes; it is also known that uptake mechanisms have an important role in regulating the extracellular concentrations of glutamate and aspartate in the brain [6]. Although speculative, in the case of this patient, it is possible that the lateral temporal lobe MD data reflect an aberrant or perhaps pathological alteration in the mechanisms of uptake by neuronal or glial cortical and/or subcortical transporters [3]. Unfortunately, because of the operative approach taken, lateral resection tissue was not obtained and thus the question of whether, for example, cortical and /or subcortical gliosis may have led to reduced rates of reuptake could not be clarified. Nevertheless, the time profile of dialysate glutamate levels detected in the lateral temporal lobe of all patients was correspondingly similar to that observed for aspartate; observations that lend some support to the notion that both glutamate and aspartate are co-stored and/or co-released (to different degrees) [7], and that their mechanisms of reuptake are functionally similar. In contrast, in the case of patient 4, lateral temporal dialysate GABA levels demonstrated a striking brise and fallQ trend (Fig. 7C ). It is known that

78

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79

GABA release is stimulated by depolarization of presynaptic GABAergic neurons (probably via the depolarizing action of glutamate and/or aspartate on glutamate receptors present on these cells) [21]. Thus, the fact that the initial level (ie, sample 1) for GABA in the dialysate acquired from the lateral temporal lobe of patient 4 was lower than that observed in sample 2 and sample 3 suggests that even if insertion trauma was primarily responsible for the initial level, it could not have accounted for the subsequent levels detected (Fig. 7C). Again, although speculative, the dialysate GABA levels detected in this case could be explained by excitatory neuroactive amino acid –stimulated GABA neuronal release (a notion supported by the observation that dialysate GABA levels only returned to lower levels in association with the decline to lower levels of glutamate and aspartate) (Fig. 7). The finding that the ratio of mean sample 3 dialysate levels of glutamate, aspartate, and GABA was approximately 20:2:1, respectively, in both the minimally epileptiform lateral and medial temporal lobe of the patients with TLE studied does suggest that despite known differences in local cellular and/or histopathological architecture, the relationship between these neuroactive amino acids may not be regionally specific. In particular, the medial temporal lobe data are similar to results reported from the intrahippocampal study of conscious epileptic patients [8,26]. These investigators also reported a dialysate hippocampal glutamate -GABA ratio (in MD basal samples ) of approximately 20:1. As previously noted, MD of the medial temporal lobe of patients 1 and 7 (Table 3) occurred during periods of vigorous EA (Fig. 6C) during the sample 3 sampling period and throughout all sampling periods, respectively. Glutamate and aspartate levels detected in sample 3 dialysates from these patients (Table 3) were between 15- and 37fold higher than mean sample 3 dialysate levels from the minimally epileptiform medial temporal lobe (Fig. 8A and B); there were correspondingly less dramatic increases of the inhibitory amino acid GABA (more than 11- and 13fold) (Table 3; Fig. 8C). These findings (for glutamate and aspartate) are similar to those reported from the first intraoperative MD study of focally epileptiform human brain [4] and (for glutamate and GABA) from the intrahippocampal MD study of conscious patients with epilepsy [8,26]. Clearly, the functional chemical and electrical pathways and mechanisms involved in all forms of TLE are complex. Regarding the vigorously epileptiform medial temporal lobe (given the relatively small number of cases, the lack of temporal and spatial resolution in the MD method used, and the limited variability in EA), then the present findings do not permit conclusions differentiating b causal Q relationships among dialysate levels of glutamate, aspartate, and GABA in relation to the grade of medial temporal sclerosis and/or interictal or ictal EA. These findings also do not permit definitive statements to be made concerning medial temporal lobe vesicular and non-

vesicular neurotransmitter release, the local distribution of synapses, specific uptake mechanisms, or substrate diffusion properties. 5. Conclusion Microdialysis during TLE surgery represents one of the few opportunities to safely investigate the in vivo extracellular microenvironment of temporal lobe structures. In the absence of significant tissue hyperexcitability, even in the wake of known differences in local cellular and /or histopathological architecture, the extracellular relationship among glutamate, aspartate, and GABA is not dissimilar in both the lateral and ipsilateral medial temporal lobe of patients with TLE. Considerable disparity in dialysate levels (eg, recovered from the vigorously epileptiform medial temporal lobe) may be related to the functional (ie, hyperexcitable) status of the sampled tissue. It is hoped that our experience will encourage the continued, and perhaps routine, use of MD during TLE surgery towards the establishment of b expectedQ regional values for various biochemical markers and corresponding electrophysiological data. Acknowledgments This work was supported in part by funding assistance from the Health Research Board (Dublin, Ireland), the Charitable Infirmary Charitable Trust (Dublin, Ireland), the Irish Brain Research Foundation (Dublin, Ireland), and the Stanley Foundation (Bethesda, MD, USA). For their helpful assistance, we would like to thank the following: Dr Norman Delanty, Consultant Neurologist and Epileptologist; Professor Michael Farrell, Consultant Neuropathologist; Dr Kevin Murphy, Epilepsy Fellow; the technical staff of the EEG Department and the operating theatre nursing and ancillary staff, Beaumont Hospital, Dublin, Ireland. This work is dedicated in memory of Mr Alan E. Synnott, a close friend ( PMT ), an eminent Irish legal professional, an author, and a philanthropist who suddenly and tragically passed away during the final stages of his endeavor. Although not a clinician or a scientist, Mr Synnott nevertheless had a special interest in and high regard for clinical neuroscience research; deeply missed by his wife, children, and many others, the spirit of this work shall remain a tribute to his ethos. References [1] Amberg G, Lindefors N. Intracerebral microdialysis: II. Mathematical studies of diffusion kinetics. J Pharmacol 1989;22:157 - 83. [2] Benveniste H. Brain microdialysis. J Neurochem 1989;52:1667 - 79. [3] Bradford HF, Young AMJ, Crowder JM. Continuous glutamate leakage from brain cells is balanced by compensatory high-affinity reuptake transport. Neurosci Lett 1987;81:296 - 302.

P.M. Thomas et al. / Surgical Neurology 63 (2005) 70 –79 [4] Carlson H, Ronne-Engstrfm E, Ungerstedt U, Hillered L. Seizure related elevations of extracellular amino acids in human focal epilepsy. Neurosci Lett 1992;140:30 - 2. [5] Carren˜o M, Lqders HO. General principles of presurgical evaluation. In: Lqders HO, Comair YG, editors. Epilepsy surgery. Philadelphia7 Lippincott Williams and Wilkins; 2001. p. 185 - 99. [6] Dingledine R, McBain CJ. Glutamate and aspartate. In: Siegel GJ, Agranoff BW, Albers RW, Fisher SK, Uhler MD, editors. Basic neurochemistry: molecular, cellular and medical aspects. Philadelphia7 Lippincott-Raven; 1999. p. 315 - 33. [7] Docherty M, Bradford HF, Wu JY. Co-release of glutamate and aspartate from cholinergic and GABAergic synaptosomes. Nature 1987;330:64 - 6. [8] During MJ, Spencer DD. Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet 1993;341: 1607 - 10. [9] During MJ, Ryder KM, Spencer DD. Hippocampal GABA transporter function in temporal-lobe epilepsy. Nature 1995;376:174 - 7. [10] Farrell MA, DeRosa MJ, Curran JG, Secor DL, Cornford ME, Comair Y, Peacock WJ, Shields WD, Vinters HV. Neuropathologic findings in cortical resections (including hemispherectomies) performed for the treatment of intractable childhood epilepsy. Acta Neuropathol 1992;83:246 - 59. [11] Fried I, Wilson C, Maidment NT, Engel Jr J, Behnke E, Fields TA, MacDonald KA, Morrow JW, Ackerson L. Cerebral microdialysis combined with single-neuron and electroencephalographic recording in neurosurgical patients. J Neurosurg 1999;91:697 - 705. [12] Gloor P. Contributions of electroencephalography and electrocorticography to the neurosurgical treatment of the epilepsies. In: Purpura D, Penry J, Walter R, editors. Neurosurgical management of the epilepsies. New York7 Raven Press; 1975. p. 59 - 105. [13] Hillered L, Persson L. Neurochemical monitoring of the acutely injured human brain. Scand J Clin Lab Invest 1999;59(Suppl. 229):9 - 18. [14] Kim R, Spencer D. Surgery for mesial temporal sclerosis. In: Lqders HO, Comair YG, editors. Epilepsy surgery. Philadelphia7 Lippincott Williams and Wilkins; 2001. p. 643 - 52. [15] Lavenex P, Amaral DG. Hippocampal-neocortical interaction: a hierarchy of associativity. Hippocampus 2000;10:420 - 30. [16] Lazarewicz JW, Hagberg H, Hamberger A. Extracellular calcium in the hippocampus of unanesthetized rabbits monitored with dialysisperfusion. J Neurosci Methods 1986;15:317 - 28. [17] Meyerson BA, Linderoth B, Karlsson H, Ungerstedt U. Extracellular measurements in the thalamus of parkinsonian patients. Life Sci 1990;46:301 - 8. [18] Morari M, O’Connor WT, Ungerstedt U, Fuxe K. Dopamine D1 and D2 receptor antagonism differentially modulates stimulation of striatal neurotransmitter levels by N-methyl-d-aspartic acid. Eur J Pharmacol 1994;256:23 - 30. [19] Niemeyer P. The transventricular amygdala-hippocampectomy in temporal lobe epilepsy. In: Baldwin M, Bailey P, editors. Temporal lobe epilepsy. Springfield7 Charles C. Thomas; 1958. p. 461 - 82. [20] Olsen RW, DeLorey TM. GABA and glycine. In: Siegel GJ, Agranoff BW, Albers RW, Fisher SK, Uhler MD, editors. Basic neurochemistry: molecular, cellular and medical aspects. Philadelphia7 LippincottRaven; 1999. p. 335 - 46. [21] Ong HY, Leong SK, Garey LJ. Distribution of glutamate receptor subunit GluR1 and GABA in human cerebral neocortex: a double

[22]

[23] [24]

[25]

[26]

[27]

[28]

79

immunolabelling and electron microscopic study. Exp Brain Res 1996;112:147 - 57. Ronne-Engstrfm E, Hillered L, Flink R, Sp7nnare B, Ungerstedt U, Carlson H. Intracerebral microdialysis of extracellular amino acids in the human epileptic focus. J Cereb Blood Flow Metab 1992;12:873 - 6. Sherwin AL. Neuroactive amino acids in focally epileptic human brain: a review. Neurochem Res 1999;24:1385 - 95. Wennberg R, Quesney F, Villemure J-G. Epileptiform and nonepileptiform paroxysmal activity from isolated cortex after functional hemispherectomy. Electroencephalogr Clin Neurophysiol 1997;102: 437 - 42. Williamson PD, Engel Jr J, Munari C. Anatomic classification of localization-related epilepsies. In: Engel Jr J, Pedley TA, editors. Epilepsy: a comprehensive textbook. Philadelphia7 Lippincott-Raven; 1997. p. 2405 - 16. Wilson CL, Maidment NT, Shomer MH, Behnke EJ, Ackerson L, Fried I, Engel Jr J. Comparison of seizure related amino acid release in human epileptic hippocampus versus a chronic kainite rat model of hippocampal epilepsy. Epilepsy Res 1996;26:245 - 54. Wyler AR, Dohan Jr FC, Schweitzer JB, Berry AD. A grading system for mesial temporal pathology (hippocampal sclerosis) from anterior temporal lobectomy. J Epilepsy 1992;5:220 - 5. YaYargil MG, Teddy PJ, Roth P. Selective amygdalo-hippocampectomy: operative anatomy and surgical technique. Adv Tech Stand Neurosurg 1985;12:93 - 123.

Commentary In this report, Thomas et al evaluate the relationship between surface ECoG and tissue levels of glutamate aspartate and GABA obtained by microdialysis from middle temporal gyrus and anterior hippocampus for 7 temporal lobe epileptics during temporal lobectomy under general anesthesia. They found that the relationship between glutamate aspartate and GABA b is not dissimilar in both the lateral and ipsilateral medial temporal lobe of patients with TLE.Q This in itself is not surprising because it is suspected that any such differences would be minimal, if at all present. It may be that clinically significant differences are well below the resolution of the present technique and/or sample size. One of the real points of this paper is that elegant research can be accomplished in human subjects during the course of providing routine clinical care. For this, the authors should be commended. This, of course, is not a new concept. It is, however, a concept in need of constant reinforcement in this era of economic pressure to produce large clinical output with declining reimbursement. Allen R. Wyler, MD Seattle, WA 98122, USA