- Email: [email protected]
Cerebrospinal fluid purine metabolite and neuron-specific enolase concentrations after febrile seizures
Brain & Development 22 (2000) 427±431
www.elsevier.com/locate/braindev
Original article
Cerebrospinal ¯uid purine metabolite and neuron-speci®c enolase concentrations after febrile seizures Antonio RodrõÂguez-NuÂnÄez a, Elena Cid a, Javier RodrõÂguez-GarcõÂa b, FeÂlix CaminÄa b, Santiago RodrõÂguez-Segade b, Manuel Castro-Gago a,* Department of Pediatrics, Hospital ClIÂnico Universitario, Complejo Hospitalario Universitario de Santiago, Santiago de Compostela, Spain b Department of Biochemistry, School of Pharmacy, University of Santiago, Santiago de Compostela, Spain
a
Received 22 November 1999; received in revised form 2 June 2000; accepted 18 August 2000
Abstract If febrile seizures cause signi®cant compromise of neuronal metabolism (whether permanent or reversible), this should be re¯ected in an increase in the cerebrospinal ¯uid concentrations of neuron-speci®c enolase (NSE) and/or adenosine triphosphate (ATP) breakdown products. In the present study, AMP, IMP, inosine, adenosine, guanosine, adenine, guanine, hypoxanthine, xanthine, uric acid and NSE concentrations were determined in the cerebrospinal ¯uid of 90 children 1 h after febrile seizure (73 simple febrile seizures (SFS); 17 complex febrile seizures (CFS)), and in a control group of 160 children. There was no statistically signi®cant difference between the SFS group and the control group for any of the substances determined, suggesting that SFS neither signi®cantly depletes neuronal ATP concentration, nor signi®cantly increases NSE concentration; thus, SFS do not appear to constitute a threat to neuronal integrity. However, patients with CFS showed signi®cantly lower IMP concentrations and signi®cantly higher adenine concentrations than controls, and signi®cantly higher AMP concentrations than SFS patients; these results suggest that CFS may affect energy metabolism in the brain. However, NSE concentrations were normal in the cerebrospinal ¯uid of both SFS and CFS patients, suggesting that neither type of seizure causes signi®cant neuronal damage, at least early after the seizure. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Febrile seizures; Cerebrospinal ¯uid; Adenosine triphosphate breakdown products; Purine metabolites; Neuron-speci®c enolase
1. Introduction Febrile seizures affect 3±5% of children aged less than 5 years, with good prognosis in the majority of cases [1±5]. Such seizures may be categorized into two groups. Simple febrile seizures (SFS) are single, brief (lasting less than 15 min), bilateral, tonic±clonic or clonic seizures occurring in infants or children with normal development; they are not followed by any transient or permanent neurological sequelae. Complex or complicated febrile seizures (CFS) occur in infants who may have a family history of non-febrile seizures, last longer than 15 min, and/or repeat within 24 h, show partial or unilateral features, and may be followed by transient neurological sequelae or by permanent handicap [1,2,4±8]. CFS are much less common than SFS. Longterm prophylactic treatment of CFS reduces their recurrence, but no evidence is available that such therapy decreases the incidence of non-febrile seizures or mental retardation [1,4]. The effects of febrile seizures on the cerebral metabolism
are poorly understood [9±12], but the available data suggest that probably only prolonged seizures will have lasting consequences [13]. The cerebrospinal ¯uid concentrations of biochemical markers of neuronal damage in children with brief febrile seizures are similar to those of controls [9±11]. Experimental studies have shown that prolonged convulsions are followed by an increase in cerebrospinal ¯uid neuron-speci®c enolase (NSE, a marker of neuronal damage) [14,15], and adenosine triphosphate (ATP) and phosphocreatine depletion with a rise of hypoxanthine concentrations [16]. In children with status epilepticus, cerebrospinal ¯uid concentrations of the ATP metabolites, hypoxanthine and uric acid, are abnormally high [17]. NSE concentrations are likewise high in the serum of adults after convulsive [18±22] and non-convulsive status epilepticus [23,24]. Increased serum NSE concentrations have also been reported in children with SFS and CFS, raising the possibility of transient neuronal damage [25]. In the present study, we investigated whether SFS or CFS are followed by increases in the cerebrospinal ¯uid concentrations of NSE and/or ATP breakdown products. Such
* Corresponding author. Fax: 134-981-531987. 0387-7604/00/$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0387-760 4(00)00172-8
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
428
increases would indicate a transient or permanent compromise of the neuronal metabolism.
boys), aged between 1 month and 13 years, who were attended in the emergency room by paediatricians unaware of the objectives of the present study.
2. Subjects and methods
2.2. Analytical procedures
2.1. Subjects
Cerebrospinal ¯uid samples were sent immediately to our hospital's central laboratory for analysis. Each sample was divided into three subsamples, one (1±2 ml) for conventional cell counts and determination of total protein, the second (0.5 ml) for determination of ATP metabolites by means of high-performance liquid chromatography assay [26], and the third for determination of NSE by enzyme immunoassay (EIA) [27,28]. The second and third subsamples were stored at 208C until analysis. Samples with microor macroscopic evidence of blood contamination were excluded.
We studied 90 patients (35 girls and 56 boys) aged between 4 and 60 months, all of whom had suffered a febrile seizure. In all cases, a lumbar puncture was done within 60 min of seizure onset. The results of routine analysis of the cerebrospinal ¯uid (proteins, glucose, cell counts) were in the corresponding normal range. The SFS group included 73 children (25 girls and 48 boys) aged between 12 and 60 months, and all meeting the clinical criteria for SFS (generalized febrile seizure lasting less than 15 min; aged more than 12 months, but less than 5 years; absence of neurological signs; no personal or family history of non-febrile seizures) [1,6,7]. In 69 patients, the seizure studied was the ®rst, in two patients, it was the second, and in two patients, it was the third or subsequent seizure. Seizure duration ranged between 1 and 10 min. The CFS group included 17 children (ten girls and seven boys) aged between 4 and 60 months, and all meeting the clinical criteria for CFS (aged less than 12 months or more than 5 years; seizures lasting more than 15 min or recurrent episodes in a 24-h interval; focal or unilateral convulsion; transitory neurological signs or permanent neurological handicap; and/or family history of non-febrile seizures) [8]. In all except two patients, the seizure studied was the ®rst. The control group comprised children with fever whose cerebrospinal ¯uid had been analyzed to rule out meningeal infections or other neurological diseases; evidently, ethical reasons exclude cerebrospinal ¯uid analysis in healthy children. In all cases, the results of routine analysis of the cerebrospinal ¯uid (monitored for at least 1 week) were normal, arguing against the existence of underlying neurological disease. This group included 160 patients (71 girls and 89
2.3. Statistics Data normality was assessed using the Kolmogorov± Smirnov test and the Chi-square goodness-of-®t test, following elimination of outliers by Hawkin's method. Data showing a normal distribution are summarized as means ^ SD, while data showing a non-normal distribution are summarized as medians (range). For statistical analysis of variables showing a normal distribution, we used analysis of variance, with the Student's t-test for subsequent pairwise comparisons. For statistical analysis of variables showing a non-normal distribution, we used the Mann±Whitney Utest. The possible relationships between measured compounds and clinical variables were investigated using the Pearson correlation analysis. A P value of less than 0.05 was considered signi®cant. 3. Results The clinical data and results of routine analysis of cerebrospinal ¯uid are listed for each group in Tables 1 and 2.
Table 1 Clinical and analytical ®ndings for children with SFS or CFS, and for children in the control group a
Number of cases Girls/boys Age (months) Temperature (8C) Cerebrospinal ¯uid Cells/mm 3 Proteins (g/l) Duration of seizure (min) Time from onset of seizure to analysis (min) a b c d e
SFS
CFS
Controls
73 25/48 19 (12±60) b 38.9 (37.6±40)
17 10/7 9 (4±30) c 38.8 (37.5±40)
160 71/89 36 (1±156) 38.5 (35.2±40)
1 (1±19) 0.21 ^ 0.13 5 (1±10) 60 (5±300)
1 (1±7) 0.30 ^ 0.23 d 7.5 (1±30) e 60 (2±300)
2 (1±13) 0.21 ^ 0.10 ± ±
Values shown are means ^ SD or medians (range). P , 0:01 compared with controls. P , 0:001 compared with controls; P , 0:001 compared with SFS. P , 0:05 compared with controls; P , 0:05 compared with SFS. P , 0:01 compared with SFS.
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
429
Table 2 Nucleotide, nucleoside, purine base, uric acid and NSE concentrations in cerebrospinal ¯uid in the SFS, CFS groups a SFS AMP IMP Inosine Adenosine Guanosine Adenine Guanine Hypoxanthine Xanthine Urate NSE a b c d
1.33 ^ 0.98 2.31 (0.00±41.40) 0.86 ^ 1.59 1.10 (0.00±46.21) 0.10 (0.00±01.52) 1.75 (0.00±23.02) 0.60 (0.00±12.29) 3.94 ^ 1.56 2.32 ^ 0.85 9.74 ^ 5.58 2.01 ^ 2.02
CFS
Controls b
2.68 ^ 2.75 2.48 (0.00±27.20) c 0.92 ^ 0.96 1.00 (0.00±13.4) 0.10 (0.00±2.38) 3.02 (0.03±10.30) d 0.70 (0.00±01.81) 3.60 ^ 1.60 2.87 ^ 0.72 10.31 ^ 5.31 1.31 ^ 1.29
1.80 ^ 1.11 3.19 (0.00±19.70) 0.64 ^ 0.42 0.90 (0.00±19.86) 0.16 (0.00±01.89) 0.37 (0.00±10.86) 0.58 (0.00±15.71) 3.40 ^ 1.40 2.32 ^ 1.01 10.40 ^ 4.76 1.52 ^ 1.01
Values shown are means ^ SD or medians (range). Results in mmol/l, except NSE (ng/ml). P , 0:01 compared with SFS. P , 0:05 compared with controls. P , 0:001 compared with controls.
Age was signi®cantly higher in the SFS group than in the CFS group (P , 0:001), and signi®cantly higher in the control group than in either the SFS group (P , 0:01) or the CFS group (P , 0:001). The cerebrospinal ¯uid total protein concentration was signi®cantly higher in the CFS group than in the control group (P , 0:05) or the SFS group (P , 0:05). Seizures lasted signi®cantly longer in the CFS group than in the SFS group (P , 0:01). Cerebrospinal ¯uid purine metabolite concentrations and NSE concentrations did not differ signi®cantly between the SFS group and the control group. Likewise, cerebrospinal ¯uid concentrations did not differ signi®cantly between the CFS group and the control group, except for: (a), IMP concentration, which was signi®cantly lower in the CFS group than in control group (P , 0:05); and (b), adenine concentration, which was signi®cantly higher in the CFS group than in the control group (P , 0:001). The cerebrospinal ¯uid concentration of AMP was significantly higher in the CFS group than in the SFS group (P , 0:01). No sex differences in metabolite concentrations were found in any of the three groups. There were no signi®cant correlations between purine metabolite or NSE concentrations and any of the clinical variables considered (age, temperature, cell count and protein concentration in cerebrospinal ¯uid, duration of seizure, time elapsed from the start of seizure to the moment of lumbar puncture, or clinical diagnosis). 4. Discussion Febrile seizures are common in infancy and early childhood, and usually have a good prognosis [1±5]. The hypothetical in¯uence of these seizures on the neuronal metabolism has been investigated previously by several authors, whose results are comparable with those obtained
in the present study; immediately after an SFS, no signi®cant alterations in cerebrospinal ¯uid ATP metabolite concentrations could be detected [9±11,17,29,30]. In addition, we measured cerebrospinal ¯uid NSE concentrations and found no signi®cant difference between the SFS and control groups. These results suggest that SFS do not significantly alter energy metabolism in the brain, and that they should not be considered as a risk factor for neuronal damage. Our data do not support the results of Hizli et al., who observed an increase in serum NSE concentrations in children with SFS [25]. One possibility is that these results are due to production of NSE in the serum by cells outside the brain, such as lymphocytes, thrombocytes and endocrine cells [31]. Previous experimental studies have demonstrated that long-lasting seizures may cause neuronal injury, as indicated by a decrease in ATP and phosphocreatine concentrations in cerebrospinal ¯uid [32,33], and an increase in the NSE concentration in cerebrospinal ¯uid [15,34] and serum [18±22]. In the present study, we found that adenine concentration was signi®cantly higher and IMP concentration signi®cantly lower in the CFS group than in the control group. These alterations may be related to the breakdown of AMP to adenosine and adenine, and to the breakdown of IMP to inosine and its ®nal products (oxypurines and uric acid), and/or to the consumption of IMP in the re-synthesis of AMP. When the CFS and SFS groups were compared, the only signi®cant difference detected was in AMP concentration, which was signi®cantly higher in the CFS group. The physiopathological implications of this difference are dif®cult to explain, since there were no signi®cant differences in the concentrations of other related metabolites. Cerebrospinal ¯uid concentrations of NSE in the SFS and CFS groups did not differ signi®cantly from those in the control group. These results may indicate that febrile seizures do not greatly disturb the neuronal metabolic status,
430
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
and do not cause signi®cant neuronal damage. Alternative explanations are that the seizure durations were not long enough to cause detectable disturbances, or that the cerebrospinal ¯uid was sampled too soon (in the ®rst hour after the seizure). In this connection, Niebroj-Dobosz et al., in a study of adult patients with brain infarction, found that cerebrospinal ¯uid NSE concentrations were signi®cantly increased in the cerebrospinal ¯uid 48 h after the ischemic event [35]. However, there would be no clinical justi®cation (and thus, no ethical justi®cation) for performing delayed or sequential cerebrospinal ¯uid analyses in children with febrile seizures. To our knowledge, only one previous study has detected increased cerebrospinal ¯uid NSE concentrations in adult patients with long-lasting seizures [34]. In a previous study of children, serum NSE was found to increase 2 h after a complex febrile seizure [25]. Surprisingly, and as far as we know, there have been no other studies involving the measurement of cerebrospinal ¯uid NSE concentrations in children with febrile seizures, although it seems clear that this variable is an excellent indicator of any effects of such seizures. Our results suggest that SFS do not constitute, at least by themselves, a threat to the integrity of the central nervous system. In contrast, CFS may alter the energy metabolism in the brain, as re¯ected by the increase in cerebrospinal ¯uid AMP concentrations; however, this change may be readily compensated for by homeostasis, since we found normal concentrations of the other ATP metabolites in our CFS patients. The normal cerebrospinal ¯uid concentrations of NSE observed in both SFS and CFS patients suggest that neither type of seizure causes neuronal damage, although we cannot rule out the possibility that NSE concentrations rose after we took our samples.
[8] [9] [10]
[11] [12] [13] [14]
[15] [16] [17] [18]
[19] [20]
Acknowledgements This work was supported by the Fondo de Investigaciones Sanitarias (Spain), FIS 00/0297.
References [1] Berg AT, Shinnar S, Hauser WA, Alemany M, Shapiro ED, Salomon ME, et al. A prospective study of recurrent febrile seizures. New Engl J Med 1992;327:1122±1127. [2] Hirtz DG. Generalized tonic±clonic and febrile seizures. Pediatr Clin North Am 1989;36:365±382. [3] Nelson KB, Ellenberg JH. Prognosis in children with febrile seizures. Pediatrics 1978;61:720±727. [4] Wolf SM, Forsythe A. Epilepsy and mental retardation following febrile seizures in childhood. Acta Paediatr Scand 1989;78:291±295. [5] Aicardi J. Epilepsy in children, 2nd ed. New York: Raven Press, 1994. pp. 253±275. [6] Holmes GI. Febrile seizures. In: Holmes GI, editor. Diagnosis and management of seizures in children. Philadelphia, PA: W.B. Saunders, 1986. pp. 226±236. [7] Herranz FernaÂndez JL, Armijo SimoÂn JA, Arteaga ManjoÂn-Cabeza R.
[21] [22]
[23] [24]
[25]
[26] [27] [28]
Factores que in¯uyen en la e®cacia del tratamiento pro®laÂctico de las convulsiones febriles. An Esp Pediatr 1989;31:116±121. O'Donohoe NV. Febrile convulsions. In: Roger M, Dravet C, Dreifuss FE, Perret A, Wolf P, editors. Epileptic syndromes in infancy, childhood and adolescence, 2nd ed. London: Libbey, 1992. pp. 45±52. Livingston JH, Brown JK, Harkness RA, McCreannor GM, O'Hare AE. Cerebrospinal ¯uid nucleotide metabolites following short febrile convulsions. Dev Med Child Neurol 1989;31:161±167. RodrõÂguez-NuÂnÄez A, CaminÄa F, Lojo S, RodrõÂguez-Segade S, CastroGago M. Purine metabolites and pyrimidine bases in cerebrospinal ¯uid of children with simple febrile seizures. Dev Med Child Neurol 1991;33:908±911. Castro-Gago M, Cid E, Trabazo S, PavoÂn P, CaminÄa F, RodrõÂguezSegade S, et al. Cerebrospinal ¯uid purine metabolites and pyrimidine bases after brief febrile convulsions. Epilepsia 1995;36:471±474. Castro-Gago M, RodrõÂguez-Segade S, CaminÄa F, Cid E, EirõÂs J, PavoÂn P, et al. Cerebrospinal ¯uid purine metabolites after complex febrile convulsions. Childs Nerv Syst 1996;12:315±317. Uthman BM, Wilder BJ. Emergency management of seizures: an overview. Epilepsia 1989;30:S33±S37. Steinberg R, Gueniau C, Scarna H, Keller A, Worcel M, Pujol JF. Experimental brain ischemia: neuron-speci®c enolase level in cerebrospinal ¯uid as an index of neuronal damage. J Neurochem 1984;43:19±24. Rabinowicz AL, Correale JD, Couldwell WT, DeGiorgio CM. CSF neuron-speci®c enolase after methohexital activation during electrocorticography. Neurology 1994;44:1167±1169. Ingvar M, SiesjoÈ BK. Pathophysiology of epileptic brain damage. In: Wasterlain CG, Vert P, editors. Neonatal seizures. New York: Raven Press, 1990. pp. 113±122. Manzke H, Staemmler W. Oxypurine concentration in the CSF in children with different diseases of the nervous system. Neuropediatrics 1981;12:209±214. DeGiorgio CM, Gott PS, Rabinowicz AL, Heck CN, Smith TD, Correale JD. Neuron-speci®c enolase, a marker of acute neuronal injury, is increased in complex partial epilepticus. Epilepsia 1996;37:606±609. DeGiorgio CM, Correale JD, Gott PS, Ginsburg DL, Bracht KA, Smith T, et al. Serum neuron-speci®c enolase in human status epilepticus. Neurology 1995;45:1134±1137. DeGiorgio CM, Correale J, Ginsberg D, Rabinowizc AL. Neuronspeci®c enolase in human status epilepticus: report of an ongoing study Abstract. Epilepsia 1993;34(Suppl 2):72. DeGiorgio CM, Correale J, Ginsberg D. Serum neuron speci®c enolase in status epilepticus. Neurology 1994;4(Suppl 2):205. Rabinowizc AL, Correale JD, Boutros RB, Couldwell WT, Henderse CW, DeGiorgio CM. Neuron-speci®c enolase is increased after single seizures during inpatient video/EEG monitoring. Epilepsia 1996;37:122±125. Rabinowizc AL, Correale JD, Bracht KA, Smith TD, DeGiorgio CM. Neuron-speci®c enolase is increased after non-convulsive status epilepticus. Epilepsia 1995;36:474±479. Hizli T, GuÈlen H, Akan P, Fadiloglu M, Dirik E. Serum neuronspeci®c enolase changes in children with non-convulsive status epilepticus. Abstract for the 2nd Congress of the European Paediatric Neurology Society (EPNS), Maastricht, 1997. p. A74. Hizli T, GuÈlen H, Akan P, Fadiloglu M, Dirick E, Serum neuronspeci®c enolase after febrile seizures in children. Abstract for the 2nd Congress of the European Paediatric Neurology Society (EPNS), Maastricht, 1997. p. A73. Kuracka L, Kalnovicova T, Liska B, Turcani P. HPLC method for measurement of purine nucleotide degradation products in cerebrospinal ¯uid. Clin Chem 1996;42:756±760. Orlino EN, Olmstead CE, Lazareff JA, Peacock WJ, Fisher RS, Fluharty AL. An enzyme immunoassay for neuron-speci®c enolase in cerebrospinal ¯uid. Biol Mol Med 1997;61:41±46. RodrõÂguez-NuÂnÄez A, Cid E, EirõÂs J, RodrõÂguez-GarcõÂa J, CaminÄa F,
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
[29] [30] [31] [32]
RodrIÂguez-Segade S, et al. Neuron-speci®c enolase levels in the cerebrospinal ¯uid of neurologically healthy children. Brain Dev 1999;21:16±19. Meberg A, Saugstad OD. Hypoxanthine in cerebrospinal ¯uid in children. Scand J Clin Lab Invest 1978;38:437±440. Manzke H, Staemmler W, Dorner K. Increased nucleotide catabolism after cerebral convulsions. Neuropediatrics 1981;12:119±131. Kaiser E, Kuzmits R, Pregant P, Burghuber O, Worofka W. Clinical biochemistry of neuron speci®c enolase. Clin Chim Acta 1989;183:13±32. Volpe J. Neonatal seizures. In: Volpe J, editor. Neurology of the newborn, 2nd ed. Philadelphia, PA: W.B. Saunders, 1987. pp. 172± 207.
431
[33] Delivora-Papadopoulos M, Younkin DP, Chance B. Cerebral metabolic studies during neonatal seizures with 31-P NMR spectroscopy. In: Wasterlain CG, Vert P, editors. Neonatal seizures. New York: Raven Press, 1990. pp. 181±189. [34] Correale J, Rabinowicz AL, Heck CN, Smith TD, Loskota WJ, DeGiorgio CM. Status epilepticus increases CSF levels of neuronspeci®c enolase and alters the blood±brain barrier. Neurology 1998;50:1388±1391. [35] Niebroj-Dobosz I, Rafalowska J, Lukasiuk M, Pfeffer A, Mossakowki MJ. Immunochemical analysis of some proteins in cerebrospinal ¯uid and serum of patients with ischemic strokes. Folia Neuropathol 1994;32:129±137.
www.elsevier.com/locate/braindev
Original article
Cerebrospinal ¯uid purine metabolite and neuron-speci®c enolase concentrations after febrile seizures Antonio RodrõÂguez-NuÂnÄez a, Elena Cid a, Javier RodrõÂguez-GarcõÂa b, FeÂlix CaminÄa b, Santiago RodrõÂguez-Segade b, Manuel Castro-Gago a,* Department of Pediatrics, Hospital ClIÂnico Universitario, Complejo Hospitalario Universitario de Santiago, Santiago de Compostela, Spain b Department of Biochemistry, School of Pharmacy, University of Santiago, Santiago de Compostela, Spain
a
Received 22 November 1999; received in revised form 2 June 2000; accepted 18 August 2000
Abstract If febrile seizures cause signi®cant compromise of neuronal metabolism (whether permanent or reversible), this should be re¯ected in an increase in the cerebrospinal ¯uid concentrations of neuron-speci®c enolase (NSE) and/or adenosine triphosphate (ATP) breakdown products. In the present study, AMP, IMP, inosine, adenosine, guanosine, adenine, guanine, hypoxanthine, xanthine, uric acid and NSE concentrations were determined in the cerebrospinal ¯uid of 90 children 1 h after febrile seizure (73 simple febrile seizures (SFS); 17 complex febrile seizures (CFS)), and in a control group of 160 children. There was no statistically signi®cant difference between the SFS group and the control group for any of the substances determined, suggesting that SFS neither signi®cantly depletes neuronal ATP concentration, nor signi®cantly increases NSE concentration; thus, SFS do not appear to constitute a threat to neuronal integrity. However, patients with CFS showed signi®cantly lower IMP concentrations and signi®cantly higher adenine concentrations than controls, and signi®cantly higher AMP concentrations than SFS patients; these results suggest that CFS may affect energy metabolism in the brain. However, NSE concentrations were normal in the cerebrospinal ¯uid of both SFS and CFS patients, suggesting that neither type of seizure causes signi®cant neuronal damage, at least early after the seizure. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Febrile seizures; Cerebrospinal ¯uid; Adenosine triphosphate breakdown products; Purine metabolites; Neuron-speci®c enolase
1. Introduction Febrile seizures affect 3±5% of children aged less than 5 years, with good prognosis in the majority of cases [1±5]. Such seizures may be categorized into two groups. Simple febrile seizures (SFS) are single, brief (lasting less than 15 min), bilateral, tonic±clonic or clonic seizures occurring in infants or children with normal development; they are not followed by any transient or permanent neurological sequelae. Complex or complicated febrile seizures (CFS) occur in infants who may have a family history of non-febrile seizures, last longer than 15 min, and/or repeat within 24 h, show partial or unilateral features, and may be followed by transient neurological sequelae or by permanent handicap [1,2,4±8]. CFS are much less common than SFS. Longterm prophylactic treatment of CFS reduces their recurrence, but no evidence is available that such therapy decreases the incidence of non-febrile seizures or mental retardation [1,4]. The effects of febrile seizures on the cerebral metabolism
are poorly understood [9±12], but the available data suggest that probably only prolonged seizures will have lasting consequences [13]. The cerebrospinal ¯uid concentrations of biochemical markers of neuronal damage in children with brief febrile seizures are similar to those of controls [9±11]. Experimental studies have shown that prolonged convulsions are followed by an increase in cerebrospinal ¯uid neuron-speci®c enolase (NSE, a marker of neuronal damage) [14,15], and adenosine triphosphate (ATP) and phosphocreatine depletion with a rise of hypoxanthine concentrations [16]. In children with status epilepticus, cerebrospinal ¯uid concentrations of the ATP metabolites, hypoxanthine and uric acid, are abnormally high [17]. NSE concentrations are likewise high in the serum of adults after convulsive [18±22] and non-convulsive status epilepticus [23,24]. Increased serum NSE concentrations have also been reported in children with SFS and CFS, raising the possibility of transient neuronal damage [25]. In the present study, we investigated whether SFS or CFS are followed by increases in the cerebrospinal ¯uid concentrations of NSE and/or ATP breakdown products. Such
* Corresponding author. Fax: 134-981-531987. 0387-7604/00/$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0387-760 4(00)00172-8
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
428
increases would indicate a transient or permanent compromise of the neuronal metabolism.
boys), aged between 1 month and 13 years, who were attended in the emergency room by paediatricians unaware of the objectives of the present study.
2. Subjects and methods
2.2. Analytical procedures
2.1. Subjects
Cerebrospinal ¯uid samples were sent immediately to our hospital's central laboratory for analysis. Each sample was divided into three subsamples, one (1±2 ml) for conventional cell counts and determination of total protein, the second (0.5 ml) for determination of ATP metabolites by means of high-performance liquid chromatography assay [26], and the third for determination of NSE by enzyme immunoassay (EIA) [27,28]. The second and third subsamples were stored at 208C until analysis. Samples with microor macroscopic evidence of blood contamination were excluded.
We studied 90 patients (35 girls and 56 boys) aged between 4 and 60 months, all of whom had suffered a febrile seizure. In all cases, a lumbar puncture was done within 60 min of seizure onset. The results of routine analysis of the cerebrospinal ¯uid (proteins, glucose, cell counts) were in the corresponding normal range. The SFS group included 73 children (25 girls and 48 boys) aged between 12 and 60 months, and all meeting the clinical criteria for SFS (generalized febrile seizure lasting less than 15 min; aged more than 12 months, but less than 5 years; absence of neurological signs; no personal or family history of non-febrile seizures) [1,6,7]. In 69 patients, the seizure studied was the ®rst, in two patients, it was the second, and in two patients, it was the third or subsequent seizure. Seizure duration ranged between 1 and 10 min. The CFS group included 17 children (ten girls and seven boys) aged between 4 and 60 months, and all meeting the clinical criteria for CFS (aged less than 12 months or more than 5 years; seizures lasting more than 15 min or recurrent episodes in a 24-h interval; focal or unilateral convulsion; transitory neurological signs or permanent neurological handicap; and/or family history of non-febrile seizures) [8]. In all except two patients, the seizure studied was the ®rst. The control group comprised children with fever whose cerebrospinal ¯uid had been analyzed to rule out meningeal infections or other neurological diseases; evidently, ethical reasons exclude cerebrospinal ¯uid analysis in healthy children. In all cases, the results of routine analysis of the cerebrospinal ¯uid (monitored for at least 1 week) were normal, arguing against the existence of underlying neurological disease. This group included 160 patients (71 girls and 89
2.3. Statistics Data normality was assessed using the Kolmogorov± Smirnov test and the Chi-square goodness-of-®t test, following elimination of outliers by Hawkin's method. Data showing a normal distribution are summarized as means ^ SD, while data showing a non-normal distribution are summarized as medians (range). For statistical analysis of variables showing a normal distribution, we used analysis of variance, with the Student's t-test for subsequent pairwise comparisons. For statistical analysis of variables showing a non-normal distribution, we used the Mann±Whitney Utest. The possible relationships between measured compounds and clinical variables were investigated using the Pearson correlation analysis. A P value of less than 0.05 was considered signi®cant. 3. Results The clinical data and results of routine analysis of cerebrospinal ¯uid are listed for each group in Tables 1 and 2.
Table 1 Clinical and analytical ®ndings for children with SFS or CFS, and for children in the control group a
Number of cases Girls/boys Age (months) Temperature (8C) Cerebrospinal ¯uid Cells/mm 3 Proteins (g/l) Duration of seizure (min) Time from onset of seizure to analysis (min) a b c d e
SFS
CFS
Controls
73 25/48 19 (12±60) b 38.9 (37.6±40)
17 10/7 9 (4±30) c 38.8 (37.5±40)
160 71/89 36 (1±156) 38.5 (35.2±40)
1 (1±19) 0.21 ^ 0.13 5 (1±10) 60 (5±300)
1 (1±7) 0.30 ^ 0.23 d 7.5 (1±30) e 60 (2±300)
2 (1±13) 0.21 ^ 0.10 ± ±
Values shown are means ^ SD or medians (range). P , 0:01 compared with controls. P , 0:001 compared with controls; P , 0:001 compared with SFS. P , 0:05 compared with controls; P , 0:05 compared with SFS. P , 0:01 compared with SFS.
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
429
Table 2 Nucleotide, nucleoside, purine base, uric acid and NSE concentrations in cerebrospinal ¯uid in the SFS, CFS groups a SFS AMP IMP Inosine Adenosine Guanosine Adenine Guanine Hypoxanthine Xanthine Urate NSE a b c d
1.33 ^ 0.98 2.31 (0.00±41.40) 0.86 ^ 1.59 1.10 (0.00±46.21) 0.10 (0.00±01.52) 1.75 (0.00±23.02) 0.60 (0.00±12.29) 3.94 ^ 1.56 2.32 ^ 0.85 9.74 ^ 5.58 2.01 ^ 2.02
CFS
Controls b
2.68 ^ 2.75 2.48 (0.00±27.20) c 0.92 ^ 0.96 1.00 (0.00±13.4) 0.10 (0.00±2.38) 3.02 (0.03±10.30) d 0.70 (0.00±01.81) 3.60 ^ 1.60 2.87 ^ 0.72 10.31 ^ 5.31 1.31 ^ 1.29
1.80 ^ 1.11 3.19 (0.00±19.70) 0.64 ^ 0.42 0.90 (0.00±19.86) 0.16 (0.00±01.89) 0.37 (0.00±10.86) 0.58 (0.00±15.71) 3.40 ^ 1.40 2.32 ^ 1.01 10.40 ^ 4.76 1.52 ^ 1.01
Values shown are means ^ SD or medians (range). Results in mmol/l, except NSE (ng/ml). P , 0:01 compared with SFS. P , 0:05 compared with controls. P , 0:001 compared with controls.
Age was signi®cantly higher in the SFS group than in the CFS group (P , 0:001), and signi®cantly higher in the control group than in either the SFS group (P , 0:01) or the CFS group (P , 0:001). The cerebrospinal ¯uid total protein concentration was signi®cantly higher in the CFS group than in the control group (P , 0:05) or the SFS group (P , 0:05). Seizures lasted signi®cantly longer in the CFS group than in the SFS group (P , 0:01). Cerebrospinal ¯uid purine metabolite concentrations and NSE concentrations did not differ signi®cantly between the SFS group and the control group. Likewise, cerebrospinal ¯uid concentrations did not differ signi®cantly between the CFS group and the control group, except for: (a), IMP concentration, which was signi®cantly lower in the CFS group than in control group (P , 0:05); and (b), adenine concentration, which was signi®cantly higher in the CFS group than in the control group (P , 0:001). The cerebrospinal ¯uid concentration of AMP was significantly higher in the CFS group than in the SFS group (P , 0:01). No sex differences in metabolite concentrations were found in any of the three groups. There were no signi®cant correlations between purine metabolite or NSE concentrations and any of the clinical variables considered (age, temperature, cell count and protein concentration in cerebrospinal ¯uid, duration of seizure, time elapsed from the start of seizure to the moment of lumbar puncture, or clinical diagnosis). 4. Discussion Febrile seizures are common in infancy and early childhood, and usually have a good prognosis [1±5]. The hypothetical in¯uence of these seizures on the neuronal metabolism has been investigated previously by several authors, whose results are comparable with those obtained
in the present study; immediately after an SFS, no signi®cant alterations in cerebrospinal ¯uid ATP metabolite concentrations could be detected [9±11,17,29,30]. In addition, we measured cerebrospinal ¯uid NSE concentrations and found no signi®cant difference between the SFS and control groups. These results suggest that SFS do not significantly alter energy metabolism in the brain, and that they should not be considered as a risk factor for neuronal damage. Our data do not support the results of Hizli et al., who observed an increase in serum NSE concentrations in children with SFS [25]. One possibility is that these results are due to production of NSE in the serum by cells outside the brain, such as lymphocytes, thrombocytes and endocrine cells [31]. Previous experimental studies have demonstrated that long-lasting seizures may cause neuronal injury, as indicated by a decrease in ATP and phosphocreatine concentrations in cerebrospinal ¯uid [32,33], and an increase in the NSE concentration in cerebrospinal ¯uid [15,34] and serum [18±22]. In the present study, we found that adenine concentration was signi®cantly higher and IMP concentration signi®cantly lower in the CFS group than in the control group. These alterations may be related to the breakdown of AMP to adenosine and adenine, and to the breakdown of IMP to inosine and its ®nal products (oxypurines and uric acid), and/or to the consumption of IMP in the re-synthesis of AMP. When the CFS and SFS groups were compared, the only signi®cant difference detected was in AMP concentration, which was signi®cantly higher in the CFS group. The physiopathological implications of this difference are dif®cult to explain, since there were no signi®cant differences in the concentrations of other related metabolites. Cerebrospinal ¯uid concentrations of NSE in the SFS and CFS groups did not differ signi®cantly from those in the control group. These results may indicate that febrile seizures do not greatly disturb the neuronal metabolic status,
430
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
and do not cause signi®cant neuronal damage. Alternative explanations are that the seizure durations were not long enough to cause detectable disturbances, or that the cerebrospinal ¯uid was sampled too soon (in the ®rst hour after the seizure). In this connection, Niebroj-Dobosz et al., in a study of adult patients with brain infarction, found that cerebrospinal ¯uid NSE concentrations were signi®cantly increased in the cerebrospinal ¯uid 48 h after the ischemic event [35]. However, there would be no clinical justi®cation (and thus, no ethical justi®cation) for performing delayed or sequential cerebrospinal ¯uid analyses in children with febrile seizures. To our knowledge, only one previous study has detected increased cerebrospinal ¯uid NSE concentrations in adult patients with long-lasting seizures [34]. In a previous study of children, serum NSE was found to increase 2 h after a complex febrile seizure [25]. Surprisingly, and as far as we know, there have been no other studies involving the measurement of cerebrospinal ¯uid NSE concentrations in children with febrile seizures, although it seems clear that this variable is an excellent indicator of any effects of such seizures. Our results suggest that SFS do not constitute, at least by themselves, a threat to the integrity of the central nervous system. In contrast, CFS may alter the energy metabolism in the brain, as re¯ected by the increase in cerebrospinal ¯uid AMP concentrations; however, this change may be readily compensated for by homeostasis, since we found normal concentrations of the other ATP metabolites in our CFS patients. The normal cerebrospinal ¯uid concentrations of NSE observed in both SFS and CFS patients suggest that neither type of seizure causes neuronal damage, although we cannot rule out the possibility that NSE concentrations rose after we took our samples.
[8] [9] [10]
[11] [12] [13] [14]
[15] [16] [17] [18]
[19] [20]
Acknowledgements This work was supported by the Fondo de Investigaciones Sanitarias (Spain), FIS 00/0297.
References [1] Berg AT, Shinnar S, Hauser WA, Alemany M, Shapiro ED, Salomon ME, et al. A prospective study of recurrent febrile seizures. New Engl J Med 1992;327:1122±1127. [2] Hirtz DG. Generalized tonic±clonic and febrile seizures. Pediatr Clin North Am 1989;36:365±382. [3] Nelson KB, Ellenberg JH. Prognosis in children with febrile seizures. Pediatrics 1978;61:720±727. [4] Wolf SM, Forsythe A. Epilepsy and mental retardation following febrile seizures in childhood. Acta Paediatr Scand 1989;78:291±295. [5] Aicardi J. Epilepsy in children, 2nd ed. New York: Raven Press, 1994. pp. 253±275. [6] Holmes GI. Febrile seizures. In: Holmes GI, editor. Diagnosis and management of seizures in children. Philadelphia, PA: W.B. Saunders, 1986. pp. 226±236. [7] Herranz FernaÂndez JL, Armijo SimoÂn JA, Arteaga ManjoÂn-Cabeza R.
[21] [22]
[23] [24]
[25]
[26] [27] [28]
Factores que in¯uyen en la e®cacia del tratamiento pro®laÂctico de las convulsiones febriles. An Esp Pediatr 1989;31:116±121. O'Donohoe NV. Febrile convulsions. In: Roger M, Dravet C, Dreifuss FE, Perret A, Wolf P, editors. Epileptic syndromes in infancy, childhood and adolescence, 2nd ed. London: Libbey, 1992. pp. 45±52. Livingston JH, Brown JK, Harkness RA, McCreannor GM, O'Hare AE. Cerebrospinal ¯uid nucleotide metabolites following short febrile convulsions. Dev Med Child Neurol 1989;31:161±167. RodrõÂguez-NuÂnÄez A, CaminÄa F, Lojo S, RodrõÂguez-Segade S, CastroGago M. Purine metabolites and pyrimidine bases in cerebrospinal ¯uid of children with simple febrile seizures. Dev Med Child Neurol 1991;33:908±911. Castro-Gago M, Cid E, Trabazo S, PavoÂn P, CaminÄa F, RodrõÂguezSegade S, et al. Cerebrospinal ¯uid purine metabolites and pyrimidine bases after brief febrile convulsions. Epilepsia 1995;36:471±474. Castro-Gago M, RodrõÂguez-Segade S, CaminÄa F, Cid E, EirõÂs J, PavoÂn P, et al. Cerebrospinal ¯uid purine metabolites after complex febrile convulsions. Childs Nerv Syst 1996;12:315±317. Uthman BM, Wilder BJ. Emergency management of seizures: an overview. Epilepsia 1989;30:S33±S37. Steinberg R, Gueniau C, Scarna H, Keller A, Worcel M, Pujol JF. Experimental brain ischemia: neuron-speci®c enolase level in cerebrospinal ¯uid as an index of neuronal damage. J Neurochem 1984;43:19±24. Rabinowicz AL, Correale JD, Couldwell WT, DeGiorgio CM. CSF neuron-speci®c enolase after methohexital activation during electrocorticography. Neurology 1994;44:1167±1169. Ingvar M, SiesjoÈ BK. Pathophysiology of epileptic brain damage. In: Wasterlain CG, Vert P, editors. Neonatal seizures. New York: Raven Press, 1990. pp. 113±122. Manzke H, Staemmler W. Oxypurine concentration in the CSF in children with different diseases of the nervous system. Neuropediatrics 1981;12:209±214. DeGiorgio CM, Gott PS, Rabinowicz AL, Heck CN, Smith TD, Correale JD. Neuron-speci®c enolase, a marker of acute neuronal injury, is increased in complex partial epilepticus. Epilepsia 1996;37:606±609. DeGiorgio CM, Correale JD, Gott PS, Ginsburg DL, Bracht KA, Smith T, et al. Serum neuron-speci®c enolase in human status epilepticus. Neurology 1995;45:1134±1137. DeGiorgio CM, Correale J, Ginsberg D, Rabinowizc AL. Neuronspeci®c enolase in human status epilepticus: report of an ongoing study Abstract. Epilepsia 1993;34(Suppl 2):72. DeGiorgio CM, Correale J, Ginsberg D. Serum neuron speci®c enolase in status epilepticus. Neurology 1994;4(Suppl 2):205. Rabinowizc AL, Correale JD, Boutros RB, Couldwell WT, Henderse CW, DeGiorgio CM. Neuron-speci®c enolase is increased after single seizures during inpatient video/EEG monitoring. Epilepsia 1996;37:122±125. Rabinowizc AL, Correale JD, Bracht KA, Smith TD, DeGiorgio CM. Neuron-speci®c enolase is increased after non-convulsive status epilepticus. Epilepsia 1995;36:474±479. Hizli T, GuÈlen H, Akan P, Fadiloglu M, Dirik E. Serum neuronspeci®c enolase changes in children with non-convulsive status epilepticus. Abstract for the 2nd Congress of the European Paediatric Neurology Society (EPNS), Maastricht, 1997. p. A74. Hizli T, GuÈlen H, Akan P, Fadiloglu M, Dirick E, Serum neuronspeci®c enolase after febrile seizures in children. Abstract for the 2nd Congress of the European Paediatric Neurology Society (EPNS), Maastricht, 1997. p. A73. Kuracka L, Kalnovicova T, Liska B, Turcani P. HPLC method for measurement of purine nucleotide degradation products in cerebrospinal ¯uid. Clin Chem 1996;42:756±760. Orlino EN, Olmstead CE, Lazareff JA, Peacock WJ, Fisher RS, Fluharty AL. An enzyme immunoassay for neuron-speci®c enolase in cerebrospinal ¯uid. Biol Mol Med 1997;61:41±46. RodrõÂguez-NuÂnÄez A, Cid E, EirõÂs J, RodrõÂguez-GarcõÂa J, CaminÄa F,
A. RodrIÂguez-NuÂnÄez et al. / Brain & Development 22 (2000) 427±431
[29] [30] [31] [32]
RodrIÂguez-Segade S, et al. Neuron-speci®c enolase levels in the cerebrospinal ¯uid of neurologically healthy children. Brain Dev 1999;21:16±19. Meberg A, Saugstad OD. Hypoxanthine in cerebrospinal ¯uid in children. Scand J Clin Lab Invest 1978;38:437±440. Manzke H, Staemmler W, Dorner K. Increased nucleotide catabolism after cerebral convulsions. Neuropediatrics 1981;12:119±131. Kaiser E, Kuzmits R, Pregant P, Burghuber O, Worofka W. Clinical biochemistry of neuron speci®c enolase. Clin Chim Acta 1989;183:13±32. Volpe J. Neonatal seizures. In: Volpe J, editor. Neurology of the newborn, 2nd ed. Philadelphia, PA: W.B. Saunders, 1987. pp. 172± 207.
431
[33] Delivora-Papadopoulos M, Younkin DP, Chance B. Cerebral metabolic studies during neonatal seizures with 31-P NMR spectroscopy. In: Wasterlain CG, Vert P, editors. Neonatal seizures. New York: Raven Press, 1990. pp. 181±189. [34] Correale J, Rabinowicz AL, Heck CN, Smith TD, Loskota WJ, DeGiorgio CM. Status epilepticus increases CSF levels of neuronspeci®c enolase and alters the blood±brain barrier. Neurology 1998;50:1388±1391. [35] Niebroj-Dobosz I, Rafalowska J, Lukasiuk M, Pfeffer A, Mossakowki MJ. Immunochemical analysis of some proteins in cerebrospinal ¯uid and serum of patients with ischemic strokes. Folia Neuropathol 1994;32:129±137.