Brain & Development 32 (2010) 4250 www.elsevier.

com/locate/braindev

Review article

Treatment of febrile seizures: Historical perspective, current opinions, and potential future directions
Andrew L. Lux *
Department of Paediatric Neurology, Bristol Royal Hospital for Children, UBHT Education Centre, Upper Maudlin Street, Bristol BS2 8AE, UK Received 28 July 2008; received in revised form 9 March 2009; accepted 9 September 2009

Abstract Although most febrile seizures do no harm and two-thirds of initial cases have no witnessed recurrence, the seizures cause much family anxiety, and are sometimes prolonged. In rare cases they are the rst evidence of important epilepsy syndromes or are implicated in the development of epilepsy with mesial temporal sclerosis in later life. There have been trials of prophylactic treatment with antiepileptic drugs including carbamazepine, diazepam, phenobarbital, phenytoin, and sodium valproate. Several strategies have been employed with these drugs, including continuous secondary prophylaxis, intermittent secondary prophylaxis in response to later episodes of fever, and rescue medication early in the course of further seizures. Another treatment strategy has been using one or more antipyretic agents in early response to fever using agents such as acetaminophen and ibuprofen. Over the years, researchers have identied a variety of clinical, genetic, and environmental risk factors for more severe or prolonged febrile seizures and higher risk of recurrence. This review evaluates the rationale for secondary prophylaxis of febrile seizures, the potential eectiveness of such treatment, and whether it can be recommended as a general approach to treating febrile seizures or as an approach to be used in groups identied to be at increased risk. 2009 Elsevier B.V. All rights reserved.
Keywords: Febrile seizures; Treatment; Prophylaxis

1. Historical perspective Febrile seizures have been recognized for centuries. Thomas Willis, in his Of Convulsive Diseases, noted an association with teething but wisely stopped short of attributing teething as a direct cause (Fig. 1) [1]. Lennox, who cited Willis observation on teething, suggested that soothing words such as benign and simple did not present the whole picture with febrile seizures, and that it might be appropriate for clinicians to stratify cases for prognosis and treatment on the basis of such factors as focal features and clustering of seizures (Fig. 2). The immediate treatment of febrile seizures is relatively uncontroversial. For example, there is general
*

Tel.: +44 117 342 0187; fax: +44 117 342 0186. E-mail address: andrew.lux@bristol.ac.uk

acceptance that prolonged febrile seizures should be treated using some form of status epilepticus protocol. It is advisable to check a blood glucose level, particularly if there is a long period of postictal obtundation. And most pediatricians agree that it is essential to rule out CNS infections in the form of meningitis or encephalitis, particularly in younger cases where the signs can be more subtle. The more controversial questions relate to the value of treatments aimed at stopping or reducing the number of recurrent febrile seizures. This can be regarded as an end in itself, since febrile seizures are distressing and frightening for families and carers, but there is also the question of whether recurrent febrile seizures are on a causal pathway that leads to increased risks of later epilepsy. This review evaluates the rationale for secondary prophylaxis of febrile seizures, the potential eectiveness

0387-7604/$ - see front matter 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.braindev.2009.09.016

A.L. Lux / Brain & Development 32 (2010) 4250

43

Fig. 1. Willis comments on the etiology of febrile seizures [Thomas Willis (1684). Of Convulsive Diseases, Ch III, p.24. Cited in [1]].

Fig. 2. Lennoxs comments on labels for febrile seizures [1].

of such treatment, and whether it can be recommended as a general approach to treating febrile seizures or as an approach to be used in groups identied to be at increased risk. It emphasises a few key inuential papers, predominantly randomised-controlled trials, rather than providing a detailed overview. It discusses several treatment guidelines and more detailed reviews. And it suggests potential future directions for treatment strategies. 2. Antipyretic drug treatment Randomised-controlled trials have shown that giving antipyretics with fever, either sporadically or regularly, does not reduce the risk of recurrent febrile seizures. This seems to be true of acetaminophen (paracetamol) and non-steroidal anti-inammatory drugs (NSAIDs) such as aspirin [2]. Schnaiderman et al. showed that administration of paracetamol after an incident febrile seizure had no eect upon the risk of a subsequent febrile seizure within 24 h (paracetamol group 4/53 (8%) versus control group 4/51 (8%)) [3]. A similar lack of eect was described by Uhari et al., with or without the co-administration of rectal or oral diazepam [4]. One study of 70 children reported greater ecacy in terms of temperature reduction with ibuprofen than with paracetamol, with a mean dierence between the two treatment groups of 0.5 C 4 h after treatment [5]. A study of 230 children aged between 1 and 4 years randomly allocated treatments with ibuprofen (5 mg/kg every 6 h) or placebo during periods with rectal temperature above 38.5 C [6]. With a median follow-up of 1 year, there were 67/230 (29%) recurrences of febrile seizure but no signicant dierences between the two treatment groups, both before and after adjustment for non-adherence to the treatment protocol. Although they do not reduce the risk of recurrent febrile seizures, antipyretics are generally considered to make the child feel more comfortable and are usually

administered for that purpose. They are widely administered, but there are small relative risks of hepatotoxicity (with acetaminophen) and metabolic acidosis, renal or respiratory failure, and coma (with ibuprofen) in overdose or where recipients have other risk factors [79]. One potential complication relating to future studies in this area is that benzodiazepine-receptor full agonists such as diazepam and lorazepam have been shown in animal models to have some antipyretic eect [10,11]. Any such eects in humans have yet to be elucidated, but one study has suggested no additive eect of diazepam co-administered with paracetamol [12]. 3. Regular secondary prophylaxis with antiepileptic drugs Several studies have shown that regular secondary prophylaxis with phenobarbital reduces the risk of further febrile seizures. A Canadian randomized doubleblind study of 79 children compared the eects of giving a single daily dose of either phenobarbital or placebo as secondary prophylaxis of simple febrile seizures, with follow-up occurring over a 12-month period [2]. Compliance was monitored using a uorescent tracer linked to riboavin that was combined with the phenobarbital and placebo. Recurrent seizures occurred in 2/39 (5%) in the phenobarbital group compared with 10/40 (25%) in the placebo group (two-sided Fisher exact test, P = 0.025). A study in Manchester, UK enrolled 186 children aged between 6 and 72 months with a febrile seizure in the rst year of life, a complex febrile seizure, or more than one febrile seizure in a 2-year period [13]. They were randomly allocated treatment in three groups: (1) rectal diazepam as rescue treatment of any subsequent prolonged febrile seizure (a form of control group); and regular prophylactic treatment with (2) sodium valproate or (3) phenobarbital. Overall, the risk of recurrence was 30%, and was not signicantly lower in those allocated regular prophylactic AED treatment.

44

A.L. Lux / Brain & Development 32 (2010) 4250

Ngwane and Bower studied 64 children with febrile seizures and randomly allocated them to receive regular sodium valproate or phenobarbital. Twenty-one children were reported as a control group, but they were not formally enrolled in the study or randomly allocated their control status. After 12 months follow-up, the reported recurrence of febrile seizures in those on valproate was 1/18 (6%); for those on phenobarbital, 4/21 (19%); and in the untreated group, 7/21 (33%). They reported a signicant dierence between valproate and untreated groups. This was not an intention-to-treat analysis, and two cases in each of the valproate and phenobarbital treatment groups were withdrawn because of side eects and not included in the analysis of febrile seizure recurrence. It is not clear whether or not those cases had recurrent seizures [14]. There have been relatively few studies of carbamazepine. One study showed no benecial eect of carbamazepine in cases that had recurrent seizures on phenobarbital treatment [15]. A double-blind trial in 40 children compared carbamazepine against phenobarbital as secondary prophylaxis of simple or complex recurrent febrile seizures [16]. Recurrent febrile seizures were reported in 9/19 (47%) children allocated carbamazepine (with documented therapeutic blood levels) compared with 2/21 (10%) in the group treated with phenobarbital (2-sided Fisher exact test, P = 0.012). The authors concluded that these results conrmed previous ndings suggesting that carbamazepine is less eective than phenobarbital as a secondary prophylactic treatment of febrile convulsions. Phenytoin also appears to have no signicant treatment eect. A study of 138 children who had a rst febrile seizure before the age of 2 years investigated the eects of regular secondary prophylaxis with phenobarbital (n = 48), phenytoin (n = 47), or placebo (n = 43) [17]. Monitoring of blood levels showed that phenytoin concentrations were poorly controlled and often outside the ideal therapeutic range. Overall, phenytoin did not signicantly reduce the incidence of further febrile seizures. The authors concluded that it is not an eective agent for prophylaxis of these seizures, noting that many of the recurrent seizures occurred in the context of subtherapeutic phenytoin blood concentrations. Phenobarbital was associated with a reduced risk of recurrent febrile seizures, but this eect seemed to be limited to cases where prophylaxis was started before the age of 14 months. 3.1. Adverse eects of treatment Herranz and colleagues studied 95 children with febrile seizures that were assessed as being complex febrile seizures or having other risk factors for high incidence of recurrence [18]. Comparing the eects of regular secondary prophylaxis with phenobarbital, primidone, or sodium valproate, they reported side eects in three-

fourths of those allocated phenobarbital and over half of those allocated primidone, with no signicant dierences in average phenobarbital blood concentrations between these two treatments (16.4 2.8 mg/ml vs 14.1 3.7 mg/ml) (primidone being a pro-drug of phenobarbital). Sodium valproate was associated with a lower proportion reporting side eects (45%) than phenobarbital. In the Manchester study, adverse eects were reported less often with valproate (24%) than with phenobarbital (61%), but the authors concluded that the benet-risk ratio was insucient to recommend either of these AEDs as secondary prophylaxis [19]. Reported adverse eects of phenobarbital treatment include eects upon mood, behavior and cognition, as well as somatic eects such as hypersensitivity reactions. Behavioral eects have been reported in over one-third of cases in some studies, with high incidences of hyperactivity, irritability, lethargy, and sleep disturbance [7,16,2025]. One particular concern with long-term phenobarbital is the high incidence of depression [21,26,27]. Other studies have reported problems with phenobarbital to include impairment of short-term memory, and impaired attention and concentration [28,29]. One study of 217 children with febrile seizures at ages between 8 and 36 months were randomly allocated treatment with phenobarbital or placebo [20]. After 2 years of treatment, the reported mean IQ in the treatment group was 7 points lower (95% CI 2.5 to 11.5 points; P < 0.01). Six months later, after treatment had been discontinued, the mean IQ in the phenobarbital group was 5.2 points lower (95% CI +0.04 to 10.5; P = 0.052), a dierence of borderline statistical signicance but which might have been subject to a dilutional bias caused by non-compliance and a crossover rate of 30% to treatment from the placebo group. The authors concluded that phenobarbital impaired cognitive performance in children treated for febrile seizures and that this eect may outlast the administration of the drug by several months. Another study investigated interval changes in IQ between 6 months post-treatment initiation and 9 to 12 months later in children treated with phenobarbital (n = 32) or valproic acid (n = 32), comparing these with controls (n = 60) who did not have epilepsy [30]. This study had the disadvantage of not examining pre-treatment baseline IQ, but showed that improvement in IQ on phenobarbital was poorer than in the valproate-treated and control groups. In contrast, Cameld and colleagues found that children with febrile seizures who received regular treatment with phenobarbital for a period of 12 months did not have signicant dierences in IQ (assessed by Binet and Bayley Scales) compared with untreated controls [25]. Phenobarbital was assessed to disturb sleep and to increase fussy behavior. In addition to behavioral problems, analyses suggested that there were some cog-

A.L. Lux / Brain & Development 32 (2010) 4250

45

nitive problems. For example, there were reported problems with verbal language processing that were thought to be associated with the duration of treatment. And there were reported memory decits that were reported to be associated with blood levels of phenobarbital. These adverse eects were reported to be more tolerable after 12 months of treatment than earlier. 4. Intermittent secondary prophylaxis with antiepileptic drugs Adverse eects of regularly administered phenobarbital might be overcome by intermittent administration at times of fever. However, this has been shown to be ineective. A Californian study showed that four-fths of recurrent febrile seizures occurred within 24 h of the identied fever. Phenobarbital administered orally at the time of fever was no more eective than no treatment at all, in contrast with the reduction reported with regularly administered phenobarbital [31]. This lack of eect is probably due to the long half-life of phenobarbital and consequent long lead-time to eective blood and brain concentrations. Even though its half-life is shorter, valproate is similarly disadvantaged when administered orally in response to fever. A Danish study of 169 children compared valproic acid suppositories with rectal diazepam solution given in response to temperatures above 38.5 C [32]. By intention-to-treat analysis at 12 months, there were no signicant dierences in proportions with recurrent febrile seizures 16/80 (20%) versus 24/89 (27%) (Pearson v2 1.13, P = 0.29). A signicantly lower proportion had reported side eects with valproate (37/80 (46%) versus 69/89 (78%), v2 17.6, P < 0.001). However, there were more recurrent febrile seizures classied as complex in the valproate group, and only a very small proportion in both groups had received treatments according to protocol. Knudsen and colleagues studied 195 children with rst febrile seizures occurring between 6 and 30 months of age, investigating the relative eects of intermittent or regular secondary prophylaxis [23]. One group was allocated rectal diazepam (5 mg suppository every 8 h) in response to a rectal temperature above 38.5 C, and the other group was allocated oral phenobarbital (2.5 to 4.5 mg/kg/day). A full intention-to-treat analysis was not performed. Thirty-nine children did not complete the study, and analysis was performed on the remaining 156 children (83 allocated diazepam and 73 allocated phenobarbital treatment). Febrile seizure recurrence rates were similar in both groups: 11% and 9% in diazepam and phenobarbital groups, respectively, at 6 months, and between 15% and 16% in each group at 12 months. There were no signicant dierences in seizure severity or duration. The authors concluded that regular long-term prophylaxis with phenobarbital had

no benets over intermittent treatment with rectal diazepam. Intermittent secondary prophylaxis with benzodiazepines has been more extensively investigated. Rosman and colleagues reported a randomized, double-blind, placebo-controlled trial in 406 children with one or more febrile seizures and mean age of 24 months [33]. Oral diazepam (0.33 mg/kg every 8 h) was administered in response to any further febrile illness. With a mean follow-up of just under 2 years, they estimated a 44% reduction in relative risk for further recurrent febrile seizures (RR 0.56, 95% CI 0.38 to 0.81; P = 0.002). Almost 40% of the 153 children who received diazepam had reported side eects of moderate seriousness, such as ataxia, irritability and lethargy; but none were reported to have severe side eects, and these eects tended to settle with dose reduction. The authors concluded that oral diazepam given as intermittent secondary prophylaxis was eective and safe. An eight-centre French study found less impressive results with an oral diazepam regimen of 0.5 mg/kg followed by 0.2 mg/kg every 12 h [34]. This randomised, placebo-controlled, double-blind trial enrolled 185 children aged between 8 months and 3 years with a rst febrile seizure and normal neurologic development. Treatment was given in response to any rectal temperature above 38 C. One thousand days of prophylactic treatment resulted from 462 episodes of fever. After 12 months, there was no signicant dierence in the proportions progressing to a further febrile seizure (diazepam group 16% versus placebo 19.5%) but adherence to the treatment protocol was found to be poor. Only 1/15 children with recurrent febrile seizures in the diazepam group received treatment according to the trial protocol, and 7/18 children in the placebo group. In seven cases in each group, the seizure occurred before any fever was identied by the parents or carers, providing a further example of how challenging it is for families and carers to eectively implement a strategy of secondary prophylaxis. Another Danish study investigated the eectiveness of short-term diazepam prophylaxis in febrile convulsions in a prospective, controlled study of 289 children who were enrolled after their rst febrile seizure [35,36]. Rectal diazepam was allocated to 137 children with instructions that it be administered in response to any body temperature above 38.5 C, and 152 children did not receive diazepam in response to fever but were permitted to receive diazepam as rescue medication in the event of seizure recurrence. The active treatment group received a mean number of ve doses of rectal diazepam per year. The proportions with recurrent seizures after 18 months were 12% versus 39% in treatment and control groups, respectively. Within 2 years, 3% of children in both treatment groups had developed nonfebrile epileptic seizures. None of these cases developed

46

A.L. Lux / Brain & Development 32 (2010) 4250

Table 1 Summary of conclusions from several key studies of treatment for febrile seizures. Extensive tabulated data from relevant pre-1999 studies can be found Baumanns technical report for the American Academy of Pediatrics [53]. Treatment approach Antipyretic drugs Acetaminophen (paracetamol) Conclusions Key studies

Ibuprofen

Diazepam as an antipyretic drug Regular AED secondary prophylaxis Phenobarbital

No eect, when given 4-hourly in response to temperature of 37.9 C or above, upon risk of febrile seizure occurring within 24 h [3]. No eect (with or without rectal diazepam co-administration) when given in response to fever of 38.5 C or higher [4] Temperature reduction greater (mean dierence 0.5 C) than with acetaminophen; no signicant dierence in incidence of subsequent febrile seizures [5] Evidence of antipyretic eect from animal studies only [10,11]

Schnaiderman et al. [3] Uhari et al. [4]

Van Esch et al. [5]

Jackson and Nutt [10,11]

Sodium valproate (or valproic acid) Carbamazepine

Phenytoin

Reduced risk of subsequent febrile seizure (5% vs 25%) when given at 5 mg/ kg/day and with adherence monitored [2]. When given at 2.54.5 mg/kg/ day, recurrence risk similar to intermittent prophylactic rectal diazepam (9% vs 11% at 6 months) [23]. No signicant reduced risk of FS; 61% incidence of reported side eects, 20% of which led to early treatment withdrawal [13]. Reduced risk of recurrent FS in cases where prophylaxis started before age 14 months [17]. Side eects reported in 77% of cases [18]. Treatment discontinued in 32% because of side eects [31] Small study with eect estimate for recurrence of 6% (vs 19% for phenobarbital and 33% for untreated group) [14]. No signicant reduced risk of FS; side eects reported in 24%, with 6% having treatment withdrawn [13]. Side eects reported in 45% [18] No signicant benet in cases with recurrent FS where earlier recurrence occurred on regular phenobarbital prophylaxis [15]. Signicantly higher incidence of FS recurrence (47% vs 10%) when given as initial secondary prophylaxis and compared with phenobarbital prophylaxis [16] Blood concentrations dicult to control; no signicant reduction in incidence of FS [17]

Cameld et al. [2] McKinlay and Newton [13] Bacon et al. [17] Herranz et al. [18] Knudsen and Vestermark[23] Wolf et al. [31]

Ngwane and Bower [14] McKinlay and Newton [13] Herranz et al. [18] Cameld et al. [15] Antony and Hawke [16]

Bacon et al. [17]

Intermittent AED secondary prophylaxis Diazepam Intermittent rectal administration as 5 mg suppository every 8 h in response to rectal temperature of 38.5 C or above had similar FS recurrence rate as regular oral phenobarbital (11% vs 9% at 6 months; both groups 1516% at 12 months) [23]. Given orally as 0.33 mg/kg every 8 h in response to fever, FS recurrence relative risk reduced by 44% (RR 0.56; 95% CI 0.38 to 0.81); moderate side eects in 40% [33]. Oral regimen of 0.5 mg/kg followed by 0.2 mg/kg associated with poor treatment adherence and no signicant benet over placebo (FS recurrence in 16% vs 19.5%) [34]. Given rectally at 5 mg (7.5 mg if age over 3 years) every 12 h until temperature below 38.5 C, FS recurrence at 18 months signicantly lower than control (12% vs 39%) [35]. Risk of respiratory depression suggested to be low from registry study [37] Clobazam Non-randomized prospective studies with febrile episodes, rather than enrolled cases, as unit of statistical analysis (prone to trial-ination and other biases) [44,45] Nitrazepam Non-randomized study with no ITT analysis; high incidence of side eects probably contributed to poor treatment adherence [43] Phenobarbital No signicant reduction in recurrent FS [31] Sodium Given rectally, not signicantly more eective than rectal diazepam (20% vs valproate 27%); signicantly fewer side eects than rectal diazepam (46% vs 78%) [32] FS = febrile seizure(s); ITT = intention-to-treat.

Knudsen and Vestermark [23] Rosman et al. [33] Autret et al. [34] Knudsen [35] Knudsen [37]

Manreza et al. [44] Rose et al. [45] Vanasse et al. [43] Wolf et al. [31] Daugbjerg et al. [32]

from the 230 children with simple febrile seizures. Risk factors for later epilepsy were initial complex febrile seizures and being assessed as having severe interictal EEG abnormalities.

4.1. Safety of intermittent benzodiazepine administration The incidence of signicant respiratory depression associated with benzodiazepine administration in the

A.L. Lux / Brain & Development 32 (2010) 4250

47

context of febrile seizures is likely to be low. Knudsen reported that there were only two reports of diazepaminduced respiratory problems in Denmark between 1975 and 2000, a very low number considering that over two decades more than 30,000 have been exposed to rectal diazepam in Denmark alone [37]. This nding is congruent with other studies of diazepam safety, though not specically in the context of febrile seizures [3842]. The use of benzodiazepines other than diazepam is more restricted by the high incidence of side eects. Nitrazepam, for example, has been studied as secondary prophylaxis of febrile seizures with high reported rates of drowsiness, agitation and ataxia [43]. Brazilian and Indian studies of clobazam reported relatively high incidences of side eects that included ataxia, hyperactivity, somnolence and recurrent vomiting [44,45]. The magnitude of treatment eects in these studies is dicult to compare with other studies because they used the number of later febrile episodes as the denominator for the reported treatment eect, rather than the number of enrolled cases with one or more subsequent febrile seizure. 5. Physician preferences and treatment guidelines 5.1. Descriptive studies of physician preferences In the mid-1980s, a broad survey of management of febrile seizures was mailed to 10,000 child neurologists, neurologists, pediatricians, and family and general practitioners, with an overall response rate of approximately 50%. One-third or less of physicians prescribed anticonvulsive therapy only at the time of febrile illness, although this practice was much less common among recent graduates. Most cases with prolonged or focal seizures were prescribed regular AEDs or referred to a pediatric neurologist for further assessment [46]. In the early 1990s, Millichap performed a survey of pediatricians in Illinois and found that over half of the respondents reported using EEGs to inform the decision whether to give regular phenobarbital prophylaxis, with 90% of respondents stating that they would treat a complex febrile seizure with phenobarbital given as regular prophylaxis [47]. The reported mean duration of such treatment was 2 years. A broader questionnaire survey of members of the Child Neurology Society in North America showed that the preferred treatments for prolonged febrile seizures were intravenous phenobarbital, lorazepam and diazepam [48]. Long-term phenobarbital was prescribed by 89% for prevention of complex febrile seizures and by 43% for simple febrile seizures, with two-thirds of respondents stating that these choices were modied by the degree of perceived parental anxiety. A survey published in 1998 reported the practice of 500 doctors in 14 Mediterranean countries [49]. It

reported that antiepileptic drugs were generally not prescribed as regular secondary prophylaxis of febrile seizures, though this practice was common in Turkey and with a minority of doctor in France, Syria and Tunisia. A large number of doctors prescribed intermittent prophylaxis with a benzodiazepine. These surveys are relatively old and it is likely that regular secondary prophylaxis of febrile seizures, including complex febrile seizures, is now administered much less commonly. However, it is also likely that there are substantial international and regional variations in practice. Guidelines have been published, though many of the recommendations pertain to simple febrile seizures rather than addressing the more challenging question of managing complex febrile seizures. 5.2. Guidelines Guidelines relating to the neurodiagnostic evaluation of a child with a rst simple febrile seizure have been published by the American Academy of Pediatrics Provisional Committee on Quality Improvement, Subcommittee on Febrile Seizures [50]. The guidelines recommend that lumbar puncture should be considered in any case presenting under the age of 18 months, and that it be strongly considered in children with the rst simple febrile seizure presenting before the age of 12 months. On the basis of consensus and evidence, it recommended that EEG is not routinely performed in a neurologically well child with a rst simple febrile seizure. It recommended that serum electrolytes, calcium, phosphorus, magnesium, complete blood count, and blood glucose not be routinely performed, but that blood glucose should be measured if there is a period of prolonged obtundation postictally. However, this recommendation does not preclude testing for the source of fever. It recommended that routine neuroimaging not be performed. A clinical practice guideline relating to the long-term treatment of children with simple febrile seizures was published by the AAP in 1999 and has just been updated with similar conclusions (Box 1) [7,51]. These reports focus specically on simple febrile seizures, which accounts for approximately 90% of incident rst febrile seizures [52]. The technical report associated with the 1999 recommendation has a comprehensive review of evidence to that date [53]. Regular secondary prophylaxis with phenobarbital or valproic acid (or sodium valproate) was regarded as eective in reducing the risk of recurrent febrile seizures, but the benet-risk ratio for these medications was regarded as insucient to justify recommending such treatment. No medication has been shown to reduce the risk of later epilepsy. The use of intermittent secondary prophylaxis with diazepam was considered likely to reduce the risk of further febrile seizures, but was con-

48

A.L. Lux / Brain & Development 32 (2010) 4250

Box 1. Conclusion of 2008 American Academy of Pediatrics Practice Parameter on simple febrile seizures [7].

sidered to have a relatively high risk of side eects and by families to have relatively low benets. Antipyretic therapy was considered to have some benets, but did not reduce the risks of further simple febrile seizures. In eect, the recommendation was that long-term treatment of simple febrile seizures is unjustied [54]. Recommendations from Japan were published in 1996 [55]. These suggested that most cases of febrile seizures can be followed with a laissez-faire approach, with consideration of treatment with intermittent diazepam, or regular phenobarbital or sodium valproate, being given in selected cases considered at higher risk. The authors identied the following risk factors, though explained that they thought the connotations of the term risk to be rather too strong in this context where recurrence of febrile seizures does not necessarily lead to longer-term harm. The suggested factors relating to recurrence were: (1) febrile seizures starting before 6 months of age; and (2) history of febrile seizures in one or both parents. Warning factors for progression to epilepsy were: (1) neurological symptoms or signs, or developmental delay, before the onset of febrile seizures; (2) focal, prolonged, or clustered seizures; and (3) a history of epilepsy in rst-degree relatives. 6. Potential future directions Febrile seizures are common and subsequent serious adverse outcomes are rare. When there are serious outcomes, such as the later development of epilepsy, it is doubtful whether earlier treatment interventions, such as the secondary prophylactic antiepileptic treatments reviewed here, signicantly modify the risks of such outcomes. However, there remains doubt and future studies are required to answer these questions eectively. In particular, the risks of febrile seizure recurrence are increased by such factors as earlier age at rst febrile seizure, the rst febrile seizure having features that make it complex in nature, and a family history of epilepsy or of febrile seizures in a rstdegree relative. Knudsen found that these factors, along with a history of day care nursery attendance,

increased the risks of recurrent febrile seizures substantially, with 18-monthy recurrence rates of between 80 and 100% if there were three to ve risk factors in the individual; 50% if there were two risk factors; 25% where there was only one factor; and 12% if there were none [36]. If such risk factors are reliable, future study designs might reserve secondary prophylaxis for cases with higher risks of seizure recurrence, possibly modifying risks of later epilepsy that have been undetected in studies with greater dilutional bias. The associations between febrile status epilepticus, acute hippocampal injury, mesial temporal sclerosis, and later temporal lobe epilepsy remain controversial [56]. Future studies will help to elucidate these relationships and might suggest eective strategies for preventing progression to epilepsy. However, there are many challenges. For example, results from the North American multicenter prospective FEBSTAT study show that most cases of febrile status epilepticus occur as the rst febrile seizure episode [57]. Another potential future development is screening for genetic risk factors such as channelopathies. Combining knowledge of such genetic predisposition with or without an associated family history of febrile seizures or epilepsy with other risk factors might lead to further targeting of cases and increases in the benet-to-cost ratio for prophylactic treatment interventions. Potential interventions might relate to immunizations or reduction of exposures to viruses that might modify risks of longer-term epilepsy. Human herpesvirus 6 (HHV6), for example, is known to be associated with febrile seizures and might be a target for immunization or other interventions [58]. The interleukin-1 system has been shown, in animal models, to have an eect that might be unrelated to the endogenous pyrogen eect of these proteins, and it has been shown that the ratio of IL-1b to IL-1 receptor antagonists has an eect upon hippocampal neuronal excitability [59]. At present, it appears that few of these factors are likely to lead to innovative treatment interventions, but a combination of basic science and epidemiologic knowledge might lead to a signicant breakthrough that does reduce the risk of progression to later epilepsy.

A.L. Lux / Brain & Development 32 (2010) 4250

49

7. Summary and conclusions The case for secondary prophylaxis of febrile seizures is controversial and not compelling [37]. In most cases, febrile seizures are benign, with low risks for progression to epilepsy and no convincing evidence that secondarily preventing recurrent febrile seizures modies any such risk. However, they are very distressing to families and carers, and the overall recurrence rate is high, at around one-third for an incident case of rst febrile seizure. Secondary prophylaxis is generally inappropriate in the case of simple febrile seizures, but for complex febrile seizures, and especially cases in which there are prolonged febrile seizures with focal features, secondary prophylaxis is considered more seriously because of the increased risks of later epilepsy (Table 1). Antipyretic treatments are not proven to reduce the recurrence risk for febrile seizures, but they do help control fever and will usually make the child more comfortable during febrile illnesses. In most cases there is no indication for regular prophylaxis with antiepileptic drugs. Phenytoin and carbamazepine have been found to have no benet as secondary prophylaxis. There is good evidence that phenobarbital reduces the incidence of subsequent febrile seizures when given as regular secondary prophylaxis, and some evidence from smaller studies that valproate is also eective. However, it is generally felt that their side eects outweigh likely benets. There is no evidence that phenobarbital or valproate reduce the risk subsequent non-febrile seizures, or of the longer-term outcome of epilepsy. Intermittent, rather than regular, prophylaxis with phenobarbital or valproate is not proven to be eective at reducing the incidence of subsequent febrile seizures. Intermittent use of oral or rectal diazepam, and potentially of other benzodiazepines, is an option that can be oered to families where there is very high anxiety about further febrile seizures, although studies have not consistently reported benets. This approach is not likely to lead to signicant harm or longer-term problems with behavior or cognition, but there is a relatively high incidence of clinically signicant short-term side eects that is likely to lead to poor adherence with such interventions. Also, the strategy is often undermined by the tendency for febrile seizures to recur before the fever is identied. For families where there are high anxiety levels about further febrile seizures, however, intermittent secondary prophylaxis can be oered as a practical therapeutic option. Acknowledgements I thank Professors Yukio Fukuyama, Yoshihiro Takeuchi, and Hitoshi Yamamoto, and other members of the Infantile Seizures Society for their support and

kind hospitality at the Societys 11th Annual Meeting on Febrile Seizures and Related Condition, Otsu-city, Shiga, Japan in April 2008. References
[1] Lennox WG. Febrile convulsions (hyperthermic-precipitation epilepsy). In: Lennox WG, editor. Epilepsy and related conditions. Boston: Little, Brown & Company; 1960. p. 393413. [2] Cameld PR, Cameld CS, Shapiro SH, Cummings C. The rst febrile seizure antipyretic instruction plus either phenobarbital or placebo to prevent recurrence. J Pediatr 1980;97:1621. [3] Schnaiderman D, Lahat E, Sheefer T, Aladjem M. Antipyretic eectiveness of acetaminophen in febrile seizures: ongoing prophylaxis versus sporadic usage. Eur J Pediatr 1993;152:7479. [4] Uhari M, Rantala H, Vainionpaa L, Kurttila R. Eect of acetaminophen and of low intermittent doses of diazepam on prevention of recurrences of febrile seizures. J Pediatr 1995;126:9915. [5] Van Esch A, Van Steensel-Moll HA, Steyerberg EW, Oringa M, Habbema JD, Derksen-Lubsen G. Antipyretic ecacy of ibuprofen and acetaminophen in children with febrile seizures. Arch Pediatr Adolesc Med 1995;149:6327. [6] van Stuijvenberg M, Derksen-Lubsen G, Steyerberg EW, Habbema JD, Moll HA. Randomized, controlled trial of ibuprofen syrup administered during febrile illnesses to prevent febrile seizure recurrences. Pediatrics 1998;102:E51. [7] Steering Committee on Quality Improvement and Management, Subcommittee on Febrile Seizures American Academy of Pediatrics. Febrile seizures: clinical practice guideline for the long-term management of the child with simple febrile seizures. Pediatrics 2008;121:12816. [8] Easley RB, Altemeier 3rd WA. Central nervous system manifestations of an ibuprofen overdose reversed by naloxone. Pediatr Emerg Care 2000;16(1):3941. [9] AAP. Acetaminophen toxicity in children. Pediatrics 2001;108(4):10204. [10] Jackson HC, Nutt DJ. Body temperature discriminates between full and partial benzodiazepine receptor agonists. Eur J Pharmacol 1990;185(23):2436. [11] Jackson HC, Nutt DJ. Comparison of the eects of benzodiazepine and beta-carboline inverse agonists on body temperature in mice. Eur J Pharmacol 1991;205(2):2136. [12] Keinanen-Kiukaanniemi S, Simila S, Luoma P, Kangas L, Saukkonen AL. Antipyretic eect and plasma concentrations of rectal acetaminophen and diazepam in children. Epilepsia 1979;20(6):60712. [13] McKinlay I, Newton R. Intention to treat febrile convulsions with rectal diazepam, valproate or phenobarbitone. Dev Med Child Neurol 1989;31(5):61725. [14] Ngwane E, Bower B. Continuous sodium valproate or phenobarbitone in the prevention of simple febrile convulsions. Comparison by a double-blind trial. Arch Dis Child 1980;55(3):1714. [15] Cameld PR, Cameld CS, Tibbles JA. Carbamazepine does not prevent febrile seizures in phenobarbital failures. Neurology 1982;32(3):2889. [16] Antony JH, Hawke SH. Phenobarbital compared with carbamazepine in prevention of recurrent febrile convulsions. A doubleblind study. Am J Dis Child 1983;137(9):8925. [17] Bacon CJ, Hierons AM, Mucklow JC, Webb JK, Rawlins MD, Weightman D. Placebo-controlled study of phenobarbitone and phenytoin in the prophylaxis of febrile convulsions. Lancet 1981;2(8247):6004. [18] Herranz JL, Armijo JA, Arteaga R. Eectiveness and toxicity of phenobarbital, primidone, and sodium valproate in the preven-

50

A.L. Lux / Brain & Development 32 (2010) 4250 tion of febrile convulsions, controlled by plasma levels. Epilepsia 1984;25(1):8995. Newton RW, McKinlay I. Subsequent management of children with febrile convulsions. Dev Med Child Neurol 1988;30(3):4026. Farwell JR, Lee YJ, Hirtz DG, Sulzbacher SI, Ellenberg JH, Nelson KB. Phenobarbital for febrile seizures eects on intelligence and on seizure recurrence. N Engl J Med 1990;322(6):3649. Barabas G, Matthews WS. Barbiturate anticonvulsants as a cause of severe depression. Pediatrics 1988;82(2):2845. AAP. Behavioral and cognitive eects of anticonvulsant therapy. American Academy of Pediatrics Committee on Drugs. Pediatrics 1995;96(3 Pt 1):53840. Knudsen FU, Vestermark S. Prophylactic diazepam or phenobarbitone in febrile convulsions: a prospective, controlled study. Arch Dis Child 1978;53(8):6603. Lee K, Melchior JC. Sodium valproate versus phenobarbital in the prophylactic treatment of febrile convulsions in childhood. Eur J Pediatr 1981;137(2):1513. Cameld CS, Chaplin S, Doyle AB, Shapiro SH, Cummings C, Cameld PR. Side eects of phenobarbital in toddlers; behavioral and cognitive aspects. J Pediatr 1979;95(3):3615. Brent DA, Crumrine PK, Varma R, Brown RV, Allan MJ. Phenobarbital treatment and major depressive disorder in children with epilepsy: a naturalistic follow-up. Pediatrics 1990;85(6):108691. Brent DA, Crumrine PK, Varma RR, Allan M, Allman C. Phenobarbital treatment and major depressive disorder in children with epilepsy. Pediatrics 1987;80(6):90917. Vining EP, Mellitis ED, Dorsen MM, Cataldo MF, Quaskey SA, Spielberg SP, et al. Psychologic and behavioral eects of antiepileptic drugs in children: a double-blind comparison between phenobarbital and valproic acid. Pediatrics 1987;80(2):16574. Riva D, Devoti M. Discontinuation of phenobarbital in children: eects on neurocognitive behavior. Pediatr Neurol 1996;14(1):3640. Calandre EP, Dominguez-Granados R, Gomez-Rubio M, Molina-Font JA. Cognitive eects of long-term treatment with phenobarbital and valproic acid in school children. Acta Neurol Scand 1990;81(6):5046. Wolf SM, Carr A, Davis DC, Davidson S, Dale EP, Forsythe A, et al. The value of phenobarbital in the child who has had a single febrile seizure: a controlled prospective study. Pediatrics 1977;59(3):37885. Daugbjerg P, Brems M, Mai J, Ankerhus J, Knudsen FU. Intermittent prophylaxis in febrile convulsions: diazepam or valproic acid? Acta Neurol Scand 1990;82(1):1720. Rosman NP, Colton T, Labazzo J, Gilbert PL, Gardella NB, Kaye EM, et al. A controlled trial of diazepam administered during febrile illnesses to prevent recurrence of febrile seizures. N Engl J Med 1993;329(2):7984. Autret E, Billard C, Bertrand P, Motte J, Pouplard F, Jonville AP. Double-blind, randomized trial of diazepam versus placebo for prevention of recurrence of febrile seizures. J Pediatr 1990;117(3):4904. Knudsen FU. Eective short-term diazepam prophylaxis in febrile convulsions. J Pediatr 1985;106(3):48790. Knudsen FU. Recurrence risk after rst febrile seizure and eect of short term diazepam prophylaxis. Arch Dis Child 1985;60(11):10459. Knudsen FU. Practical management approaches to simple and complex febrile seizures. In: Baram TZ, Shinnar S, editors. Febrile seizures. San Diego: Academic Press; 2002. p. 273304. Kriel RL, Cloyd JC, Pellock JM, Mitchell WG, Cereghino JJ, Rosman NP. Rectal diazepam gel for treatment of acute repetitive seizures. The North American Diastat Study Group. Pediatr Neurol 1999;20(4):2828. [39] Mitchell WG, Conry JA, Crumrine PK, Kriel RL, Cereghino JJ, Groves L, et al. An open-label study of repeated use of diazepam rectal gel (Diastat) for episodes of acute breakthrough seizures and clusters: safety, ecacy, and tolerance. North American Diastat Group. Epilepsia 1999;40(11):16107. [40] Kriel RL, Cloyd JC, Pellock JM. Respiratory depression in children receiving diazepam for acute seizures: a prospective study. Dev Med Child Neurol 2000;42(6):42930. [41] Cereghino JJ, Mitchell WG, Murphy J, Kriel RL, Rosenfeld WE, Trevathan E. Treating repetitive seizures with a rectal diazepam formulation: a randomized study. The North American Diastat Study Group. Neurology 1998;51(5):127482. [42] Dreifuss FE, Rosman NP, Cloyd JC, Pellock JM, Kuzniecky RI, Lo WD, et al. A comparison of rectal diazepam gel and placebo for acute repetitive seizures. N Engl J Med 1998;338(26):186975. [43] Vanasse M, Masson P, Georoy G, Larbrisseau A, David PC. Intermittent treatment of febrile convulsions with nitrazepam. Can J Neurol Sci 1984;11(3):3779. [44] Manreza ML, Gherpelli JL, Machado-Haertel LR, Pedreira CC, Heise CO, Diament A. Treatment of febrile seizures with intermittent clobazam. Arq Neuropsiquiatr 1997;55(4):75761. [45] Rose W, Kirubakaran C, Scott JX. Intermittent clobazam therapy in febrile seizures. Indian J Pediatr 2005;72(1):313. [46] Hirtz DG, Lee YJ, Ellenberg JH, Nelson KB. Survey on the management of febrile seizures. Am J Dis Child 1986;140(9):90914. [47] Millichap JG. Management of febrile seizures: current concepts and recommendations for phenobarbital and the electroencephalogram. Clin Electroencephalogr 1991;22(1):512. [48] Millichap JG, Colliver JA. Management of febrile seizures: survey of current practice and phenobarbital usage. Pediatr Neurol 1991;7(4):2438. [49] Baldy-Moulinier M, Covanis A, DUrso S, Eskazan E, Fattore C, Gatti G, et al. Therapeutic strategies against epilepsy in Mediterranean countries: a report from an international collaborative survey. Seizure 1998;7(6):51320. [50] AAP. Practice parameter: the neurodiagnostic evaluation of the child with a rst simple febrile seizure. American Academy of Pediatrics. Provisional committee on quality improvement, subcommittee on febrile seizures. Pediatrics 1996;97(5):76972 [discussion 735]. [51] AAP. Practice parameter: long-term treatment of the child with simple febrile seizures. American Academy of Pediatrics. Committee on quality improvement, subcommittee on febrile seizures. Pediatrics 1999;103(6 Pt 1):13079. [52] Verity CM, Golding J. Risk of epilepsy after febrile convulsions: a national cohort study. BMJ 1991;303(6814):13736. [53] Baumann RJ. Technical report: treatment of the child with simple febrile seizures. Pediatrics 1999;103(6):e86. [54] Baumann RJ, Duner PK. Treatment of children with simple febrile seizures: the AAP practice parameter. Am Acad Pediatr Neurol 2000;23(1):117. [55] Fukuyama Y, Seki T, Ohtsuka C, Miura H, Hara M. Practical guidelines for physicians in the management of febrile seizures. Brain Dev 1996;18(6):47984. [56] Shinnar S, Glauser TA. Febrile Seizures. In: Pellock JM, Bourgeois BFD, Dodson WE, editors. Pediatric epilepsy: diagnosis and therapy. New York: Demos; 2008. p. 293301. [57] Shinnar S, Hesdorer DC, Nordli Jr DR, Pellock JM, ODell C, Lewis DV, et al. Phenomenology of prolonged febrile seizures: results of the FEBSTAT study. Neurology 2008;71(3):1706. [58] Suga S, Suzuki K, Ihira M, Yoshikawa T, Kajita Y, Ozaki T, et al. Clinical characteristics of febrile convulsions during primary HHV-6 infection. Arch Dis Child 2000;82(1):626. [59] Gatti S, Vezzani A, Bartfai T. Mechanisms of fever and febrile seizures: putative role of the interleukin-1 system. In: Baram TZ, Shinnar S, editors. Febrile seizures. San Diego: Academic Press; 2002. p. 16988.

[19] [20]

[21] [22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35] [36]

[37]

[38]