Adult-onset temporal lobe epilepsy suspicious for autoimmune pathogenesis: Autoantibody prevalence and clinical correlates.

Temporal lobe adult-onset seizures (TAOS) related to autoimmunity represent an increasingly recognized disease syndrome within the spectrum of epilepsies. In this context, certain autoantibodies (autoABs) were often associated with limbic encephalitis (LE). Here, we aimed to gain insights into (a) the distribution of 'neurological' autoABs (neuroABs, defined as autoABs targeting neuronal surface structures or 'onconeuronal' ABs or anti-glutamate acid decarboxylase 65 (GAD65) autoABs) in a large consecutive TAOS patient cohort, to characterize (b) clinical profiles of seropositive versus seronegative individuals and to find (c) potential evidence for other autoABs. Blood sera/cerebrospinal fluid (CSF) of TAOS patients (n = 800) and healthy donors (n = 27) were analyzed for neuroABs and screened for other autoABs by indirect immunofluorescence on hippocampal/cerebellar sections and immunoblots of whole brain and synaptosome lysates. Serological results were correlated with clinico-neuropsychological features. 13% of TAOS patients (n = 105) were neuroAB+, with anti-GAD65 and anti-N-methyl-D-aspartate receptors (NMDAR) as most frequent autoABs in this group. In our screening tests 25% of neuroAB- patients (n = 199) were positive (screening+), whereas all control samples were negative (n = 27). Intriguingly, key clinico-neuropsychological characteristics including magnetic resonance imaging (MRI) findings, epileptiform electroencephalographic (EEG) activity, and inflammatory cellular infiltrates in CSF were shared to a greater extent by neuroAB+ with neuroAB-/screening+ patients than with neuroAB-/screening- patients. Serological testing in a large consecutive TAOS patient series revealed seropositivity for anti-GAD65 autoABs as the most frequent neuroAB. Intriguingly, neuroAB+ individuals were virtually indistinguishable from neuroAB-/screening+ patients in several major clinical features. In contrast, neuroAB-/screening- TAOS patients differed in many parameters. These data support the potential presence of so far unrecognized autoABs in patients with TAOS.

Blood and cerebrospinal fluid immune cell profiles in patients with temporal lobe epilepsy of different etiologies.

Inflammation plays a role in the pathogenesis of immune-mediated epilepsy, but also in epilepsy of other etiology such as hippocampal sclerosis. This study aimed to characterize immune cell signatures in the peripheral blood (PB) and cerebrospinal fluid (CSF) in temporal lobe epilepsy (TLE) of different etiologies. We retrospectively evaluated CSF routine parameters and immune cell profiles using flow cytometry in a cohort of 51 patients and 45 age-matched controls with functional disorders. Groups were comprised of patients with nonlesional TLE (n = 26), TLE due to hippocampal sclerosis (n = 14), or limbic encephalitis with antibodies against the 65-kDa isoform of glutamic acid decarboxylase (GAD65-LE; n = 11). TLE patients showed increased proportions of human leukocyte antigen-DR isotype (HLA-DR)-expressing CD4+ T lymphocytes in the CSF. Furthermore, they were characterized by a shift in monocyte subsets toward immature CD14low CD16+ cells in the PB and blood/CSF-barrier dysfunction. Whereas TLE patients in general showed similar immune cell profiles, patients with GAD65-LE differed from other TLE patients by increased proportions of HLA-DR-expressing CD8+ T lymphocytes and type 2/3 oligoclonal bands. These findings point to a role of innate and adaptive immunity in TLE. CSF parameters may help to discriminate epilepsy patients from controls and different forms of TLE from each other.

CD19+ B-cells in autoantibody-negative limbic encephalitis.

PURPOSE: Flow cytometry helps to elucidate the cellular immune repertoire's mechanisms in patients with temporal lobe epilepsy (TLE) due to limbic encephalitis (LE) subcategories and carries potential significance for subtype-specific treatment. METHODS: We enrolled 62 patients with TLE due to LE associated with no autoantibodies (n = 40), neural autoantibodies (n = 22), as well as autoantibodies against intracellular antigens (n = 15/22). All patients underwent neuropsychological testing, brain magnetic resonance imaging (MRI), electroencephalography (EEG) recordings, and peripheral blood (PB) and cerebrospinal fluid (CSF) investigations including flow cytometry. RESULTS: CD19+ B-cells were increased in the PB and CSF of patients with antibody-negative LE compared with those associated with antibodies against intracellular antigens (Kruskal-Wallis one way analysis of variance (ANOVA) on ranks with Dunn's test, p < 0.05). There were no differences in CD138+ B-cells, CD4+ T-cells, human leukocyte antigen - DR isotype (HLA-DR+) CD4+ T-cells, CD8+ T-cells, and HLA-DR+ CD8+ T-cells in the CSF between groups with LE. The blood-brain barrier is more often impaired in patients with antibody-negative LE than in LE with antibodies against intracellular antigens (chi-square test, p < 0.05). In addition, we detected no correlations between immune cell subsets and clinical or paraclinical parameters in patients with antibody-negative and intracellular antibody-positive LE. CONCLUSIONS: The increase of CD19+ B-cells in the CSF and frequent signs of dysfunctional blood-brain barrier in patients with antibody-negative rather than intracellular antibody-positive LE suggest that CD19+ B-cells play a role in antibody-negative encephalitis although their pathogenic role in the central nervous system (CNS) immunity because of missing correlations between immune cells and clinical and paraclinical parameters remains unknown. Further studies are required to evaluate the usefulness of these B-cells as a biomarker for the stratification of treatment strategies.

Low CSF CD4/CD8+ T-cell proportions are associated with blood-CSF barrier dysfunction in limbic encephalitis.

PURPOSE: Investigating immune cells in autoimmune limbic encephalitis (LE) will contribute to our understanding of its pathophysiology and may help to develop appropriate therapies. The aim of the present study was to analyze immune cells to reveal underlying immune signatures in patients with temporal lobe epilepsy (TLE) with LE. METHODS: We investigated 68 patients with TLE with LE compared with 7 control patients with TLE with no signs of LE screened from 154 patients with suspected LE. From the patients with TLE-LE, we differentiated early seizure onset (<20 years, n = 9) and late seizure onset group (≥20 years, n = 59) of patients. Patients underwent neuropsychological assessment, electroencephalography (EEG), brain magnetic resonance imaging (MRI), and peripheral blood (PB) and cerebrospinal fluid (CSF) analysis including flow cytometry. RESULTS: We identified a higher CD4/8+ T-cell ratio in the PB in all patients with TLE-LE and in patients with late-onset TLE-LE each compared with controls (Kruskal-Wallis one-way ANOVA (analysis of variance) with Dunn's test, p < 0.05). Moreover, a lower CD4/CD8+ T-cell ratio is detected in all patients with TLE-LE with blood-CSF barrier dysfunction, unlike in those with none (Kruskal-Wallis one-way ANOVA with Dunn's test, p < 0.05). CONCLUSIONS: These findings suggest that the proportion of CD4+ and CD8+ T-cells in the CSF of patients with LE associated with blood-CSF barrier dysfunction plays a potential role in CNS (central nervous system) inflammation in these patients. Thus, flow cytometry as a methodology reveals novel insights into LE's genesis and symptomatology. The CD4/8+ T-cell ratio in PB as a biomarker for LE requires further investigation.

Association of cerebrospinal fluid zinc-α2-glycoprotein and tau protein with temporal lobe epilepsy and related white matter impairment.

This study aimed to identify the change of cerebrospinal fluid (CSF) zinc-α2-glycoprotein (ZAG), tau protein and phosphorylated tau 181 (P-Tau 181), and analyze their association with white matter impairment (WMI) in patients with temporal lobe epilepsy (TLE). Levels of ZAG, tau, and P-Tau 181 in CSF were measured by enzyme-linked immunosorbent assay. WMI was scored based on Scheltens and Ylikoski scales. ZAG and total tau (T-Tau) were decreased in CSF of patients with TLE than controls (P<0.05 and 0.01, respectively), but P-Tau 181 showed no statistical difference. P-Tau 181/T-Tau ratio was increased in CSF of patients with TLE than controls (P<0.01). Patients with TLE have severer WMIs than controls (P<0.01). Linear by linear association analysis showed a correlation between WMI and TLE (P<0.01), Spearman's correlation analysis showed a correlation between T-Tau, P-Tau 181/T-Tau ratio and scores of WMIs (P<0.01). Receiver operating curves indicated apparent diagnostic value of T-Tau and P-Tau 181/T-Tau ratio in differentiating patients with TLE from controls (area under curve=1 and 0.936, respectively). ZAG also showed a lower but favorable diagnostic value (area under curve=0.771). ZAG, T-Tau, and P-Tau 181/T-Tau ratio in CSF may be potential biomarkers of WMI of TLE if further researched, and these findings may also be involved in potential mechanism of brain damage in TLE.

Cerebrospinal fluid microRNAs are potential biomarkers of temporal lobe epilepsy and status epilepticus.

There is a need for diagnostic biomarkers of epilepsy and status epilepticus to support clinical examination, electroencephalography and neuroimaging. Extracellular microRNAs may be potentially ideal biomarkers since some are expressed uniquely within specific brain regions and cell types. Cerebrospinal fluid offers a source of microRNA biomarkers with the advantage of being in close contact with the target tissue and sites of pathology. Here we profiled microRNA levels in cerebrospinal fluid from patients with temporal lobe epilepsy or status epilepticus, and compared findings to matched controls. Differential expression of 20 microRNAs was detected between patient groups and controls. A validation phase included an expanded cohort and samples from patients with other neurological diseases. This identified lower levels of miR-19b in temporal lobe epilepsy compared to controls, status epilepticus and other neurological diseases. Levels of miR-451a were higher in status epilepticus compared to other groups whereas miR-21-5p differed in status epilepticus compared to temporal lobe epilepsy but not to other neurological diseases. Targets of these microRNAs include proteins regulating neuronal death, tissue remodelling, gliosis and inflammation. The present study indicates cerebrospinal fluid contains microRNAs that can support differential diagnosis of temporal lobe epilepsy and status epilepticus from other neurological and non-neurological diseases.

Treatment of immune-mediated temporal lobe epilepsy with GAD antibodies.

PURPOSE: Temporal lobe epilepsy with antibodies (abs) against the glutamic acid decarboxylase 65 isoform (GAD-TLE) is known as an immune-mediated neurological syndrome. Here we evaluate the therapy response to various immunotherapies and epilepsy surgery in this syndrome. METHOD: All patients with GAD-TLE and follow-up data and stored serum and CSF samples, identified and treated at the Bonn centre from 2002 to 2010, were studied retrospectively. Seizure freedom for ≥1 year and reduction of ≥50%, i.e. therapy response, were assessed. GAD-ab titres and neuropsychological performances were documented prior and after individual interventions. RESULTS: Thirteen patients with GAD-TLE were identified with the following seizure responses: corticosteroids (5 responders out of 11 treated patients); i.v. immunoglobulins (1/5), apheresis therapy (1/8); and natalizumab (1/1), selective amygdala-hippocampectomy (2/3). None of the patients achieved sustained seizure freedom apart from one patient. This patient was on antiepileptic drug treatment after discontinuation of immunotherapy. CONCLUSION: The seizure response to immunotherapies in patients with GAD-TLE was poor. Corticosteroids were the most effective regarding seizure response. Especially the poor effects of apheresis therapies support the idea that GAD-abs are not directly pathogenic. None of three patients was seizure-free after temporal lobe surgery suggesting that GAD-TLE patients respond worse than others to this type of intervention. Our results reflect the chronic course of the disease with low likelihood for patients with GAD-TLE to attain long-term seizure freedom.

Altered expression of CX3CL1 in patients with epilepsy and in a rat model.

Chemokine C-X3-C motif ligand 1 (CX3CL1, alias fractalkine), is highly expressed in the central nervous system and participates in inflammatory responses. Recent studies indicated that inflammatory processes within the brain constitute a common and crucial mechanism in the pathophysiological characteristics of epilepsy. This study investigated the expression pattern of CX3CL1 in epilepsy and its relationship with neuronal loss. Double immunolabeling, IHC, and immunoblotting results showed that CX3CL1 expression was up-regulated in the temporal neocortex of patients with temporal lobe epilepsy. In a rat model of epilepsy, CX3CL1 up-regulation began 6 hours after epilepsy, with relatively high expression for 60 days. In addition, ELISA revealed that the concentrations of CX3CL1 in cerebrospinal fluid and serum were higher in epileptic patients than in patients with neurosis but lower than in patients with inflammatory neurological diseases. Moreover, H&E staining demonstrated significant neuronal loss in the brains of epileptic patients and in the rat model. Finally, the expression of tumor necrosis factor-related apoptosis-inducing ligand was significantly increased in both patients and the animal model, suggesting that tumor necrosis factor-related apoptosis-inducing ligand may play a role in CX3CL1-induced cell death. Thus, our results indicate that CX3CL1 may serve as a possible biomarker of brain inflammation in epileptic patients.

Gelsolin in cerebrospinal fluid as a potential biomarker of epilepsy.

Gelsolin is an actin regulatory protein that generally distributed in a wide variety of body tissues, especially the brain tissues and cerebrospinal fluid. In this study we found that lumbar CSF-gelsolin concentrations markedly decreased in epileptic patients by enzyme linked immunosorbent assay. In order to help judge the result, we determined gelsolin expression in temporal lobe tissues of patients with temporal lobe epilepsy using double-label immunofluorescence to location and using western blot to quantitation. Then we observed that gelsolin was co-expressed with microtubule-associated protein-2 in axons and cytoplasms of neurons and gelsolin protein level was also down-regulated in temporal lobe tissues of epileptic patients. Our findings suggested that CSF-gelsolin level might reflect the alteration of gelsolin in brain tissue of epileptic patients and CSF-gelsolin seems to be a potential biomarker for epilepsy.

Cerebrospinal fluid levels of the endocannabinoid anandamide are reduced in patients with untreated newly diagnosed temporal lobe epilepsy.

PURPOSE: The endocannabinoid system is involved in excitatory/inhibitory balance mechanisms within the central nervous system (CNS). Growing evidence shows that its perturbation leads to development of epileptic seizures in experimental models, thus indicating that endocannabinoids play an intrinsic protective role in suppressing pathologic neuronal excitability. Experimental data also demonstrate that the endocannabinoid anandamide (AEA) can antagonize epileptic discharges in hippocampal tissue. The objective of our study was to measure endocannabinoids levels in the cerebrospinal fluid (CSF) of drug-naive patients affected by temporal lobe epilepsy (TLE). METHODS: We measured the levels of both AEA and the other endocannabinoid, 2-arachidonoylglycerol (2-AG), in the CSF of drug-naive patients with TLE. RESULTS: A significant reduction of AEA was found in the CSF of patients with compared with healthy controls (epileptic patients = 2.55 +/- 1.78 pmol/ml; healthy controls = 11.65 +/- 7.53 pmol/ml; n = 9 for both groups, p < 0.01). 2-AG levels, however, were not affected (epileptic patients = 209.5 +/- 146.56; healthy controls = 159.6 +/- 110.2) (n = 6 for both groups, p = 0.48). DISCUSSION: Our findings seem to be consistent with experimental evidence demonstrating a significant prevention of epileptic seizures induced by endocannabinoids in models of epilepsy. Furthermore, they support the hypothesis that AEA may be involved in its pathogenesis, suggesting a hypothetical primary impairment of the endocannabinoid system in untreated TLE. The actual role of this in vivo dysregulation still remains unclear.

Proteomic analysis of cerebrospinal fluid from patients with idiopathic temporal lobe epilepsy.

Proteomic analysis of cerebrospinal fluid (CSF) from patients with temporal lobe epilepsy (TLE) and controls was carried out using two-dimensional gel electrophoresis followed by liquid chromatography electrospray ionization tandem mass spectrometry. Five protein spots showed significant differential expression (p<0.05): vitamin D-binding protein (DBP) was elevated in the CSF of TLE patients whereas cathepsin D, apolipoprotein J, Fam3c, and superoxide dismutase 1 (SOD1) were decreased in the CSF of TLE patients. Additional six protein spots presented only in the CSF of epilepsy patients were identified as tetranectin (TN), talin-2, apolipoprotein E, immunoglobulin lambda light chain (IGL@), immunoglobulin kappa variable light chain 1-5 (IGKV1-5), and procollagen C-endopeptidase enhancer 1 (PCOLCE). Expression of DBP, SOD1 and talin-2 was validated by western blot. Our results may provide better understanding of the pathophysiologic mechanisms underlying epileptogenesis and possible epilepsy biomarkers.

Cisternal S100 protein and neuron-specific enolase are elevated and site-specific markers in intractable temporal lobe epilepsy.

In the brain, S100 protein and neuron-specific enolase (NSE) are mainly found in glial cells and neurons, respectively. We investigated concentrations of S100 protein and NSE in cisternal cerebrospinal fluid obtained during implantation of foramen ovale electrodes in eight patients with temporal lobe epilepsy (TLE). In addition, the meningeal markers cystatin-C and beta-trace as well as total protein were measured. Patients with trigeminal neuralgia (TN) undergoing glycerol rhizotomy served as controls. S100 protein and NSE levels ipsilateral to the site of seizure onset were significantly higher than in TN. Contralateral TLE values were also markedly but not significantly elevated. The meningeal markers cystatin-C and beta-trace protein as well as total protein did not differ in TLE and TN. We conclude that interictal temporal lobe dysfunction corresponds with neuronal and glial marker elevations in the extracellular space and that site-specific elevations may predict the site of seizure origin biochemically.

Interictal behavioral alterations and cerebrospinal fluid amino acid changes in a chronic seizure model of temporal lobe epilepsy.

This study extends our previous work in which we described the presence of an interictal behavioral disturbance in a chronic animal model of temporal lobe epilepsy (TLE). In this study, we investigated the cerebrospinal fluid (CSF) neurotransmitter changes underlying the development of chronic recurrent seizures of temporal lobe origin and interictal behavioral disturbance in cats made epileptic after intrahippocampal injection of kainic acid (KA). Using high-performance liquid chromatography, we measured 22 putative neurotransmitter amino acids. After intrahippocampal KA injection, cats developed an initial acute period of intense seizure activity. Cisternal CSF amino acids, which were repeatedly sampled during the acute period through a permanent indwelling cannula, were unchanged apart from a mild elevation in CSF alanine. The high-level seizure activity gradually decreased, and cats entered a chronic epileptic period characterized by recurrent yet intermittent temporal lobe seizures. CSF GABA levels during the chronic epileptic period were significantly decreased. In contrast, CSF levels of other amino acids--alanine, tyrosine, taurine, aspartic acid, and glutamic acid--did not change significantly. Behavioral testing also showed a heightened interictal defensive reactivity during the chronic epileptic period. To the extent that CSF GABA concentration reflects brain GABA concentration, this study suggests that a decrease in brain GABA may contribute both to the epilepsy and interictal emotional lability of animals with a chronic seizure disorder of temporal lobe origin.

Prognostic significance of acute epilepsia partialis continua.

We present 3 patients in whom epilepsia partialis continua was the presenting sign of an acute, rapidly evolving and catastrophic neurologic illness. Initial seizures were partial simple (i.e., eye deviation in one, finger twitching in one) which progressed to multifocal partial seizures. The course of the epilepsia partialis continua was 36-41 days. Prognosis was uniformly poor (i.e., death in 2, vegetative state in 1); therefore, epilepsia partialis continua in the context of an acute neurologic illness may herald a grim outcome.

A study on opioid peptides in CSF of patients with epilepsy.

The contents of L-EK, M-EK and beta-EP in CSF of 32 epileptics and 24 controls were determined by RIA. The mean L-EK content of epileptics was significantly higher than that of the controls (P less than 0.01). There were no obvious changes with respect to mean M-EK and beta-EP contents. No significant differences were seen in L-EK contents between generalized and partial seizures, treated and untreated with antiepileptic drugs, normal and abnormal CT manifestation patient groups. These data indicated that endogenous L-EK content was related to human epilepsy, and changes in opioid peptides were selective changes shared by different types of seizures. The increase of L-EK content was not caused by taking antiepileptic drugs, nor due to structural pathological changes of the brain that might be found on CT scanning, but a manifestation of neurochemical disorders of the brain that resulted in epilepsy.

Quinolinic acid concentrations in brain and cerebrospinal fluid of patients with intractable complex partial seizures.

Quinolinic acid (QUIN) is a neurotoxin and convulsant when injected directly into the brains of experimental animals and as such has been implicated in the etiology of human seizure disorders. In the present study, we quantified QUIN in cerebrospinal fluid (CSF) and in spiking (focus) and nonspiking (nonfocus) regions of surgically resected human temporal neocortex. L-tryptophan (L-TRP), the putative precursor of QUIN, was also measured in brain, along with CSF concentrations of L-TRP, 5-hydroxyindoleacetic acid (5-HIAA), and homovanillic acid (HVA). In brain tissue, no differences were found in the concentrations of QUIN and L-TRP between focus and nonfocus regions in 15 pairs of samples. No differences were found in CSF, L-TRP, 5-HIAA, or HVA concentrations between 11 neurologically normal controls and 15 interictal (no seizures for greater than 24 h) and 20 postictal (within 50 min of seizure) samples from epileptic patients. However, CSF QUIN concentrations were significantly lower (32%) in the epileptic patients as compared with controls, which may indicate a generalized disturbance in brain QUIN metabolism or perhaps a response to antiepileptic drugs.

The utility of cerebrospinal fluid examination in patients with partial epilepsy.

The initial evaluation of patients with seizure disorders frequently includes cerebrospinal fluid (CSF) examination in order to identify an underlying cerebral lesion. With increasing use of computed tomography (CT) scanning to detect cerebral neoplasms, the value of CSF examination has become less certain. The significance of mild CSF abnormalities in patients with a normal CT scan remains unknown. We reviewed the records of 95 patients with adult onset partial epilepsy whose initial evaluation included CSF examination and CT scan. A CSF abnormality not temporally related to convulsive seizure was seen in 24 patients (25%). The CSF study confirmed a clinically suspect subarachnoid hemorrhage in 4 patients. Isolated mild (49-106 mg/dl) increases in CSF protein were seen in 19 patients. Of these 19 patients, 8 had a structural lesion on CT scan. Clinical follow-up of the other 11 patients (mean 5 years) has revealed no evidence of a focal lesion or increasing seizure frequency. This suggests that in an adult population with partial epilepsy routine CSF examination may not be necessary and should be reserved for situations in which there is particular clinical indication.

Somatostatin-like immunoreactivity in cerebrospinal fluid of patients with complex partial epilepsy.

To investigate the role of somatostatin in human epilepsy, we measured somatostatin-like immunoreactivity (SLI) by radioimmunoassay of the cerebrospinal fluid (CSF) of 60 patients with complex partial seizures (CPS), 5 patients with other neurological diseases (OND), and 23 controls. The SLI levels were measured in groups of epileptic patients that differed in their history of disease, electroencephalogram (EEG), computerized tomography (CT) finding, psychological test scores, or anticonvulsant medication. SLI was lower in the epilepsy group (p less than 0.05) than in the controls. Patients with carbamazepine-clonazepam therapy had lower SLI than did other epileptics (p less than 0.02) or controls (p less than 0.005). Patients with central atrophy (p less than 0.01) in CT and infection (p less than 0.01) as an etiologic cause of epilepsy also seemed to have lower levels of SLI in the CSF than did other epileptics. No correlation was found between psychological memory scores and SLI levels in the CSF of patients with CPS. The present study shows that somatostatin levels are lowered in the CSF of epileptic patients, possibly owing to the lowered somatostatin content or the decreased number of somatostatinergic nerve cells in the epileptic human brain. However, studies in unmedicated patients with different types of seizures are needed to further clarify the possible role of somatostatin in human epilepsy.

Beta-endorphin, somatostatin, and prolactin levels in cerebrospinal fluid of epileptic patients after generalised convulsion.

The possible role of different peptidergic systems in the postictal stage of human epilepsy was studied by measuring beta-endorphin, somatostatin, and prolactin levels by radioimmunoassay of cerebrospinal fluid (CSF) from nine epileptic patients. The first sample was taken within 2 hours after generalised tonic-clonic convulsion, and the second sample was obtained interictally after 1-4 days without any kind of clinically observable seizures. beta-endorphin was elevated postictally (p = 0.044) compared with interictal levels. SLI and PROL were similar in both samples. The present study suggests that in humans beta-endorphin is released into CSF during generalised seizures. This may indicate that neurons containing beta-endorphin are activated during a seizure.

Carbamazepine lowering effect on CSF somatostatin-like immunoreactivity in temporal lobe epileptics.

The effect of carbamazepine treatment on CSF-somatostatin-like immunoreactivity (SLI) in patients suffering from temporal lobe epilepsy was investigated. A baseline lumbar puncture was performed on 12 patients and 10 normal volunteers. A second tap was repeated only in patients when they were on peak of carbamazepine concentration for 10 days. Levels of CSF-SLI were measured by RIA. No significant differences were found in CSF-SLI basal concentrations between epileptics and controls, whereas a significant decrease (p less than .0002 Duncan's multiple range test) of CSF peptide levels occurred in 9 of 12 patients under medication. Although the neural mechanism through which carbamazepine lowers CSF-SLI is still unknown, the results of the present study suggest that the reported effect might be part of the apparatus by which carbamazepine exerts its anticonvulsant action.