A retrospective analysis of patients with febrile seizures followed by epilepsy

Seizure 2003; 12: 211 216 doi:10.1016/s1059 1311(02)00226-1 A retrospective analysis of patients with febrile seizures followed by epilepsy SEMA SALTIK, AYDAN ANGAY, ÇIGDEM ÖZKARA, VEYSI DEMİRBİLEK & AYSIN DERVENT Division of Child Neurology, Neurology Department and Neurosurgery Department, Cerrahpasa Medical Faculty, Istanbul University, Istanbul, Turkey Correspondence to: Dr Sema Saltik, Paediatrician, Tibas İs Bankası Evleri, Park Sitesi H Blok Daire: 9, Acıbadem 81010, Uskudar-Istanbul, Turkey. E-mail: saltik@ixir.com This study was performed to investigate some clinical parameters of febrile seizures (FSs) in patients with epilepsy, testing any possible correlation between those parameters and the type of subsequent epilepsy. One hundred and nine patients with epilepsy having a history of FSs were evaluated for age at onset of FSs, interval between first FS and first afebrile seizure, recurrence rate, type of FSs, incidence of febrile status, family history for epilepsy and for FSs and the neurological status of the patient. The epilepsy that developed subsequently, were classified as generalised versus partial and also according to their syndromic subgroups. In temporal lobe epilepsy with mesial temporal sclerosis (TLE MTS), statistical analyses revealed a younger age at onset of FSs, and a high incidence of episodes of febrile status and of complex FSs. Clinical characteristics of FSs followed by partial epilepsies were younger age at onset, presence of focal features and of febrile status, longer interval between the first FS and the first afebrile seizure, and a high incidence of FSs in the family history. In generalised epilepsies, however, a shorter interval between the first FS and the first afebrile seizure, a high incidence of single FS and of a family history of epilepsy were predominating characteristics. Results suggest that certain features of FSs may be predictive of a particular type of subsequent epilepsy. 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved. Key words: febrile seizures; childhood epilepsy; temporal lobe epilepsy with mesial temporal sclerosis; childhood epilepsy with occipital paroxysms. INTRODUCTION Febrile seizures (FSs), the most common convulsive disorder of childhood occur in 2 4% of all children 1 3. According to the Consensus Development Conference in 1980 4, a slightly increased risk of subsequent epilepsy was reported. But the relation between FSs and childhood epilepsies still remains obscure and rather controversial, in spite of many current and past efforts to elucidate the problem 5. Different epidemiological studies 6, 7 demonstrate that the risk of epilepsy in children with FSs is rather low: 2 10% depending on the length of follow up. However, this risk may increase in the presence of certain features. Neuro-developmental abnormalities, family history of epilepsy and complex FSs are reported as the primary predictors of unprovoked seizures in children with FSs 8. The relation between those features of FSs and type of subsequent epilepsy, however, is still a matter of debate. In this study, some clinical parameters of FSs in certain groups of epileptic patients are analysed and any possible correlation between those parameters and the type of following epilepsy is measured. METHOD Among 2600 out-patient files belonging to patients being followed in our Child Neurology Department in Cerrahpasa Medical School, University of Istanbul, within the last 3 years, 109 epileptic patients, under 19 years of age, with prior FSs were identified. The epilepsies and epileptic syndromes were classified according to the classification proposed by the International League Against Epilepsy in 1989 9. The epilepsy was considered as unclassified if information 1059 1311/02/$30.00 2002 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.

212 S. Saltik et al. was inadequate. Groups were defined according to the type of epilepsies as partial and generalised; and, also, according to their syndromic characteristics. All patients were evaluated for the following parameters: age at onset of FSs, interval between the first febrile and the first afebrile seizures, recurrence rate, type of FSs (simple vs. complex, partial vs. generalised), incidence of febrile status, family history for epilepsy and FSs and neurological status of the patient. Statistics Chi-square and Fisher s exact tests were used for evaluation of difference between the clinical parameters pertaining to FSs in patients with partial epilepsies versus those with generalised epilepsies and Student s t-test, for the age of onset. Patient groups with childhood epilepsy with occipital paroxysms (CEOP) and with temporal lobe epilepsy with mesial temporal sclerosis (TLE MTS) were statistically evaluated by similar procedures. All statistical analyses were done in the Biostatistics Department of Cerrahpasa Medical School. Definitions To be precise for the inclusion criteria, definitions were restricted as follows: Febrile seizure Clinical seizure with tonic clonic, clonic, tonic or atonic manifestations accompanying body temperature of 38 C or above, occurring between the ages of 3 months and 6 years without any evidence of intracranial infection. Simple FSs Isolated generalised seizures, lasting less than 15 minutes, and leaving no post-ictal neurological deficit. Complex FSs FSs that were either focal, prolonged (lasting for 15 30 minutes), recurring within 24 hours, or associated with post-ictal neurological symptoms or signs. Febrile status An FS lasting longer than 30 minutes or recurrent seizures lasting in total more than 30 minutes without the child fully regaining consciousness. RESULTS When the total group of 109 patients was classified according to localisation, partial epilepsies outnumbered generalised epilepsies by a ratio of 4.4:1 (Table 1). The male:female ratio was 1:1.6. Epileptic syndromes, which constituted 37% of our patients, were CEOP, benign childhood epilepsy with centro-temporal spikes, childhood absence epilepsy, benign infantile myoclonic epilepsy, grand mal on awakening, reflex epilepsies and TLE MTS (Table 2). The relatively high incidence of patients with CEOP and TLE MTS does probably reflect a selected population from our previous studies, rather than a random accumulation. The clinical features of FSs in the antecedent of the epileptics were as follows: Age at onset of FSs was between 3 to 72 months with the majority around 12 months. FSs apparently started earlier in partial epilepsies i.e. within the first 6 months of life as compared to generalised epilepsies. All patients with TLE MTS had their first FS before 2 years of age, which was a finding of statistical significance (Table 3). Table 1: Classification of the patient population according to the type of epilepsies. Type of epilepsy Number of patients (%) Partial epilepsy 75 (68.8%) Idiopathic 21 Cryptogenic 39 Symptomatic 15 Generalised epilepsy 17 (15.5%) Idiopathic 6 Cryptogenic 10 Symptomatic 1 Undetermined 4 (3.6%) Unclassified 13 (11.9%) Total 109 Table 2: Distribution of the patients according to the epileptic syndromes. Epileptic syndromes Number (%) of all patients CEOP 17 (15.5) BCECT 4 (3.6) CAE 2 (1.8) BIME 1 (0.9) GMA 2 (1.8) RfE 1 (0.9) TLE MTS 13 (9.1) Total 40 (36.6) CEOP: childhood epilepsy with occipital paroxysms; BCECT: benign partial epilepsy with centro-temporal spikes; CAE: childhood absence epilepsy; BIME: benign infantile myoclonic epilepsy; GMA: grand mal during awaking; RfE: reflex epilepsies; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis.

Analysis of patients with febrile seizures 213 Table 3: Age at onset of FSs as compared to the type of epilepsies. Age (month) Incidence, n = 109 (%) PE, n = 75 (%) GE, n = 17 (%) TLE MTS, n = 10 (%) CEOP, n = 17 (%) 6 and 25 (22.9) 20 (26.6) 3 (17.6) 6 (46) 2 (12) 12 and 56 (51.3) 39 (52) 9 (52) 11 (85) 8 (47) 24 and 84 (77) 58 (77) 14 (82) 13 (100) a 14 (82) 36 and 98 (89) 69 (92) 15 (88) 16 (94) PE: partial epilepsy; GE: generalised epilepsy; CEOP: childhood epilepsy with occipital paroxysms; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis. a Significant early-onset of FSs in TLE MTS group as compared to the all other total patient population (P <0.001). Table 4: Interval between first FS and first afebrile seizure as compared to the type of epilepsies. Year Incidence, n (%) PE, n (%) GE, n (%) TLE MTS, n (%) CEOP, n (%) 1 30 (27.5) 15 (20) 8 (47) 1 (7.6) 2 (11.7) 2 43 (39) 21 (28) 10 (58.8) 1 (7.6) 3 (17.5) 5 70 (64.4) 43 (57.3) 13 (76.4) 6 (46) 10 (58.8) 10 98 (89.9) 66 (88) 16 (94) 10 (77) 17 (100) No statistical significance between parameters compared (PE/GE, TLE MTS/the other epilepsies, CEOP/the other epilepsies). PE: partial epilepsy; GE: generalised epilepsy; CEOP: childhood epilepsy with occipital paroxysms; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis. Results regarding the interval between the initial FS and the initial afebrile seizure are summarised in Table 4. Patients with generalised epilepsies displayed, in average, a shorter interval as compared to partial epilepsies; and, in nearly half of those patients, epilepsy developed within a single year. Among patients with an interval longer than 10 years, cases with TLE MTS occupied the majority, whereas no patient with CEOP was in this group. Recurrence of FSs occurred in 66% of the patient population; in Table 5: Recurrence rate as compared to the type of epilepsies. Number of FSs 1 2 5 >5 All patients 48 45 16 PE 32 32 11 GE 10 5 2 TLE MTS 7 4 2 CEOP 7 9 1 No statistical significance between parameters compared (PE/GE, TLE MTS/the other epilepsies, CEOP/the other epilepsies). PE: partial epilepsy; GE: generalised epilepsy; CEOP: childhood epilepsy with occipital paroxysms; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis. other words, 44% of our patients had only a single FS (Table 5). Recurrence rate was highest in cases with TLE MTS and higher in patients with generalised epilepsies as compared to partial epilepsies. An exceptionally low incidence of frequent FSs (i.e. more than five) was a prominent finding for patients with CEOP. Analysis of FS types revealed a 59% incidence of simple FSs which was slightly higher in the CEOP group. In contrast, complex FSs were significantly more common in the group of patients with TLE MTS. The ratio of generalised versus partial FSs was 5.2:1 in our patient population. This ratio was higher in cases with CEOP and with generalised epilepsies when compared to partial epilepsies. It was, however, approximately 1:1 in the TLE MTS group (Table 6). A finding worth of special mention was that febrile status, encountered in 15 patients, was followed by partial epilepsy in 14 (93.3%) of them. Cranial magnetic resonance imaging (MRI) supported the diagnosis of mesial temporal sclerosis in five of those patients. Also, a relatively high incidence of previous febrile status in patients with TLE MTS was statistically significant (33.3%). Table 6: Type of FSs as compared to the type of epilepsies. Simple FS, n (%) Complex FS, n (%) Partial FS, n (%) Generalised FS, n (%) All patients 64 (58.8%) 45 (41.2) 17 (16.2) 88 (83.8) PE 39 (52.7) 35 (47.3) 13 (17.3) 58 (77.3) GE 10 (58.8) 7 (41.2) 1 (5.8) 14 (82.3) TLE MTS 2 (15) 11 (85) a 2 (28) 5 (71) CEOP 11 (65) 6 (35) 1 (5.8) 16 (94) PE: partial epilepsy; GE: generalised epilepsy; CEOP: childhood epilepsy with occipital paroxysms; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis. a Statistical significance high incidence of complex FSs in TLE MTS (P <0.005).

214 S. Saltik et al. Table 7: Positive family history of FS and epilepsy as compared to the type of epilepsies. Type of epilepsy Positive family Positive family history history of FS (%) of epilepsy (%) All patients 29.4 24.8 PE 29.3 26.6 GE 23.5 35 TLE MTS 50 10 CEOP 23.5 29.4 No statistical significance between parameters compared (PE/GE, TLE MTS/the other epilepsies, CEOP/the other epilepsies). PE: partial epilepsy; GE: generalised epilepsy; CEOP: childhood epilepsy with occipital paroxysms; TLE MTS: temporal lobe epilepsy with mesial temporal sclerosis. Incidence of FSs was slightly more common than the incidence of epilepsy among the family members of our patients with a preponderance in the TLE MTS group. In contrast, a family history of epilepsy predominated in cases with CEOP (Table 7). Without any significant difference in distribution regarding the type of epilepsy, abnormality in neurological status was encountered in only 18 patients (16.5%). DISCUSSION The relationship between FSs and epilepsy has been a matter of debate for clinicians for some time, and seizure susceptibility triggered by fever at a certain age of development has been a subject of investigation in various animal studies 10, 11. Although a wide range of hypotheses has been developed in the field, from considering FSs as a mild form of epilepsy to suggesting those two entities are completely different phenomena on the other end. The majority of studies have suggested evidence of either a common pathogenetic mechanism, or a cause and effect relationship 12 15. Recent clinical and molecular genetic studies provide further support for a genetic relationship between FSs and epilepsies, in favour of the argument that there might be some specificity to the types of epilepsy that follow FSs, rather than FSs being a non-specific marker of a lowered seizure threshold 15. Clinical studies up to present, however, do not produce conclusive evidence for a specific relationship of FSs to the type of later epilepsy or, special characteristics of FSs which might be early indicators of a developing epilepsy except such features as unilaterality, long-duration and high-recurrence rate. The important clinical problem, at this point, appears to be the identification of which children with FSs may develop later epilepsy, and if possible, which type of epilepsy may develop. In such cases, then, the usual optimistic prognosis of FSs would be tempered and perhaps earlier diagnosis and treatment would follow. Generalised epilepsy with febrile seizures is a recently identified genetic disorder, for example, drawing attention to the concurrence of various generalised epilepsies with FSs, or FSs lasting beyond the age of 6 years 16. In a longitudinal study on the time course of EEG changes of children with FSs, genetic epileptogenic patterns and photosensitivity were reported with 49 and 42% incidences, respectively 17. On the other hand, significant risk factors, on clinical grounds, were attributed to differentiate children at high risk of developing epilepsy, or afebrile seizures, from those at low risk. The high risk group, for which one study showed a 13% incidence of epilepsy, is characterised by the presence of at least two of the following risk factors: a family history of non-febrile seizures, abnormal neurologic and developmental status prior to FSs and the presence of complex FSs. Only 2 3% of children who had none or only one of the earlier discussed risk factors were reported to subsequently develop non-febrile seizures 4. The risk of afebrile seizures developing was 2.4% after simple FSs, 6 8% after FSs with a single complex feature, 17 22% with two complex features and 49% with all three complex features in another study 7. When the matter is questioned from a different stand-point, our results show that only 2.7% of our patients had all three risk factors, 17.4% of patients had two risk factors but a significant proportion (79.8%) of our patients had none or only one of the earlier discussed risk factors. These results may be considered as suggestive of a possibility of epilepsy developing even after apparently benign FSs. It has been reported that the younger the child at the onset of FSs, the more likely epilepsy will develop, particularly if FSs occur in the first year of life 18. According to our findings, the onset of FSs was mainly around 1 year of age, of which, 51.3% occurred within the first year (Table 3). Similar results with 40 70% occurrence within the first year, 67 77% within 3 years, 85% within 4 years and 8% after 10 years of age were reported by others 5, 7, 19, 20. A clear trend between the number of FSs and the risk of unprovoked seizures was proposed by some authors 21 ; whereas, such a correlation was not confirmed in the NCPP 6 study on children thought to have been normal prior to FSs. A high recurrence rate (more than five) for FSs occurred in only 14.7% of our patients, and the incidence of more than one seizure was 66%. Comparable studies have shown a wide variation (36.5 70%) 3, 22. A family history of epilepsy, but not of FSs, was reported to be another parameter strongly associated with subsequent epilepsy in children with FSs 23. Results of our study revealed a similar prevalence (24.8% for epilepsy and 29.4% for FSs) within the families of our patient population. Complexity of FSs, which is considered as another risk factor relating to epilepsy,

Analysis of patients with febrile seizures 215 was present in 41.2% of our patients. Although this is apparently larger than expected in a general population of FSs, some studies have demonstrated that complex FSs in patients with subsequent epilepsy can be more common than simple FSs 5. Prolonged FSs were reported to lead to brain damage or neurological deficits 24, 25, but fortunately, not always 26. The number of patients with a seizure duration of more than 30 minutes was slightly more than expected (13.7%) in our patients with epilepsy, as compared to the incidence in the general FS population 6, 27. Results of the present data do not contribute much to the known characteristics of FSs, when followed by epilepsy, except some unique features as a slightly higher incidence of febrile status and complex FSs as well a higher incidence of a family history of FSs among our epileptic population, when compared to previous studies. Very limited information, on the other hand, concerning a possible relationship between clinical features of FSs and the type of developing epilepsy exists in the literature, and that is solely restricted to TLE MTS. Although such a presumptive relationship was a specific question throughout the evaluation procedure in this study, available data did not reveal a significant correlation between the type of epilepsy and epileptic syndrome and the clinical features of FSs, except in the TLE MTS group. Statistically significant features in patients with TLE MTS were younger age at onset of FSs, high incidence of febrile status complex FSs. Based on retrospective studies indicating that patients with TLE MTS have relatively high incidence of prolonged FSs in infancy, it has been suggested that FSs, particularly those that last longer than 15 minutes, lead to subsequent hippocampal sclerosis and chronic recurrent complex partial seizures 28. In our patient population, there were 13 patients with TLE MTS as the subsequent epilepsy and 5 individuals were identified as having febrile status in infancy and 11 had complex FSs. Other features compatible with TLE MTS were a longer interval between first febrile and afebrile seizure, a high incidence of single FS and of a family history of FSs. Due to the uniformity of our patient population, CEOP is the other epileptic syndrome in the present study that deserves mentioning in comparison to TLE MTS, since the former is a good example of an age-related idiopathic epilepsy and the latter, of a medically intractable cryptogenic (or symptomatic) condition. The patients with CEOP in this study had the lowest recurrence rate of FSs and the highest incidence of generalised and simple FSs within the total population of patients. Those data were suggestive of benign features for FSs antedating CEOP, which might suggest those seizures might have been the initial seizure of the epilepsy itself, triggered by fever at the age of highest susceptibility. Differences in the risk factors associated with generalised and partial epilepsies in patients with prior FSs have been reported previously. A positive correlation between a family history of seizures and a high frequency (three or more) of simple FSs, were noted in association with generalised epilepsies, whereas, partial epilepsies were preceded by complex FSs only 7. FSs with early-onset, with partial features and with post-ictal deficits were found to be related to partial epilepsies in another study, in contrast to FSs with late onset and with prolonged duration as well as the presence of a shorter interval between the onset of epilepsy correlating with generalised epilepsies 29. Significant correlation with early-onset of FSs and partial epilepsies has also been reported earlier 30. In this study, clinical features of FSs followed by partial epilepsies were younger age at onset, partial or unilateral features, longer interval between the first febrile and afebrile seizures, presence of febrile status and a high incidence of FSs in the family history. In generalised epilepsies, however, a shorter interval between the first febrile and afebrile seizure, a high incidence of single FSs and a family history of epilepsy were predominating characteristics. Although being far from conclusive, the present results seem to suggest that some characteristics of the FSs may reflect the type of epilepsy that subsequently develop; and, especially if they suggest a partial epilepsy and TLE MTS specifically, may provide evidence for closer follow up for possible future consequences of the condition. REFERENCES 1. Hauser, W. A. and Kurland, L. T. The epidemiology of epilepsy in Rochester Minnesota, 1935 through 1967. Epilepsia 1975; 16: 1 66. 2. Verity, C. M., Butler, N. R. and Golding, J. Febrile convulsions in a national cohort followed up from birth. Prevalence and recurrence in the first five years of life. British Medical Journal 1985; 290: 1307 1310. 3. Offringa, M., Hazebroeck-Kampschreurt, A. A. J. M. and Derek-Lubsen, G. Prevalence of febrile seizures in Dutch school children. Paediatric and Perinatal Epidemiology 1991; 5: 181 188. 4. Freeman, J. M. Febrile seizures: a consensus of their significance, evaluation and treatment. Pediatrics 1980; 66 (6): 1009 1013. 5. Sofijanov, N., Sadikario, A., Dukovski, M. et al. Febrile convulsions and later development of epilepsy. American Journal of Disease in Childhood 1983; 137: 123 126. 6. Nelson, K. B. and Ellenberg, J. H. Predictors of epilepsy in children who have experienced febrile seizures. New England Journal of Medicine 1976; 255: 1029 1033. 7. Annegers, J. F., Hauser, W. A., Shirts, S. B. et al. Factors prognostic of unprovoked seizures after febrile convulsions. New England Journal of Medicine 1987; 316: 494 498.

216 S. Saltik et al. 8. Tsai, M. L. and Hung, K. L. Risk factors for subsequent epilepsy after febrile convulsions. Journal of the Formosan Medical Association 1995; 94 (6): 327 331. 9. Commission on the Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989; 30: 389 399. 10. Morimoto, T., Nagao, H., Sano, N. et al. Electroencephalographic study of rat hyperthermic seizures. Epilepsia 1991; 32: 289 293. 11. Baram, T. Z., Gerth, A. and Schultz, L. Febrile seizures: an appropriate-aged model suitable for long-term studies. Brain Research: Developmental Brain Research 1997; 98: 265 270. 12. Gibbs, F. A. and Low, N. L. Electroencephalography in children. Pediatric Clinics of North America 1955; 2: 291 303. 13. Breg, W. and Yannet, H. The child in a convulsion. Pediatric Clinics of North America 1962; 9: 101 114. 14. Bird, T. D. Genetic considerations in childhood epilepsy. Epilepsia 1987; 28 (Suppl. 1): 71 81. 15. Berkovic, S. F. and Scheffer, I. E. Febrile seizures: genetics and relationship to other epilepsy syndromes. Current Opinion in Neurology 1998; 11: 129 134. 16. Singh, R., Scheffer, I. E., Crossland, K. et al. Generalized epilepsy with febrile seizures plus: a common childhood-onset genetic epilepsy. Annals of Neurology 1999; 45 (1): 75 81. 17. Doose, H., Ritter, K. and Völzke, E. EEG longitudinal studies in febrile convulsions. Genetic aspects. Neuropediatrics 1983; 14 (2): 81 87. 18. Wallace, S. J. The Child with Febrile Seizures. London, John Wright, 1988: pp. 109 126. 19. Lennox, W. G. Significance of febrile seizures. Pediatrics 1953; 11: 341 357. 20. Schmidt, D., Tsai, J. J. and Janz, D. Febrile seizures in patients with complex partial seizures. Acta Neurologica Scandinavia 1985; 72: 68 71. 21. Berg, A. T. and Shinnar, S. Unprovoked seizures in children with febrile seizures: short term outcome. Neurology 1996; 47: 562 568. 22. Wolf, S. and Forsythe, A. Epilepsy and mental retardation following febrile seizures in childhood. Acta Paediatrica Scandinavia 1989; 78: 291 295. 23. Nelson, K. B. and Ellenberg, J. H. Prognosis in children with febrile seizures. Pediatrics 1978; 61: 720 727. 24. Freeman, J. M. Febrile seizures: an end to confusion. Pediatrics 1978; 61: 806 808. 25. Knudsen, F. U. Effective short-term diazepam prophylaxis in febrile convulsions. Journal of Pediatrics 1985; 106: 487 490. 26. Aicardi, J. and Chevrie, J. J. Consequences of status epilepticus in infants and children. Advances in Neurology 1983; 34: 115 128. 27. Wallace, S. J. Spontaneous fits after convulsions with fever. Archives of Disease in Childhood 1977; 52: 192 196. 28. Darrell, V. L. Febrile convulsions and mesial temporal sclerosis. Current Opinion in Neurology 1999; 12: 197 201. 29. Konishi, T., Naganuma, Y., Hongo, K. et al. Clinical evaluation of the patients with epilepsy following febrile convulsions. No To Hattatsu 1990; 22 (3): 230 234. 30. Aicardi, J. and Chevrie, J. J. Febrile convulsions: neurological sequel and mental retardation. In: Brain Disfunction in Infantile Febrile Convulsions (Eds M. A. B. Brazier and F. Coceani). New York, Raven Press, 1976: pp. 247 257.