Address correspondence to Leon G. Epstein, MD, The Ann & Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Avenue, Neurology Box 51, Chicago, IL 60611, U.S.A. E-mail: firstname.lastname@example.org
Purpose: In a prospective study, Consequences of Prolonged Febrile Seizures in Childhood (FEBSTAT), we determined the frequency of human herpesvirus (HHV)-6 and HHV-7 infection as a cause of febrile status epilepticus (FSE).
Methods: Children ages 1 month to 5 years presenting with FSE were enrolled within 72 h and received a comprehensive assessment including specimens for HHV-6 and HHV-7. The presence of HHV-6A, HHV-6B, or HHV-7 DNA and RNA (amplified across a spliced junction) determined using quantitative polymerase chain reaction (qPCR) at baseline indicated viremia. Antibody titers to HHV-6 and HHV-7 were used in conjunction with the PCR results to distinguish primary infection from reactivated or prior infection
Key Findings: Of 199 children evaluated, HHV-6 or HHV-7 status could be determined in 169 (84.9%). HHV-6B viremia at baseline was found in 54 children (32.0%), including 38 with primary infection and 16 with reactivated infection. No HHV-6A infections were identified. HHV-7 viremia at baseline was observed in 12 children (7.1%), including eight with primary infection and four with reactivated infection. Two subjects had HHV-6/HHV-7 primary coinfection at baseline. There were no differences in age, characteristics of illness or fever, seizure phenomenology or the proportion of acute EEG or imaging abnormalities in children presenting with FSE with or without HHV infection.
Significance: HHV-6B infection is commonly associated with FSE. HHV-7 infection is less frequently associated with FSE. Together, they account for one third of FSE, a condition associated with an increased risk of both hippocampal injury and subsequent temporal lobe epilepsy.
Febrile seizures (FS) are the single most common seizure type and occur in 2–5% of children younger than age 5 years with a peak incidence in the second year of life (Shinnar, 2003). The majority are brief, generalized convulsions or simple FS, which are thought to be benign. A small proportion of FS are prolonged and 5–8% of cases meet the criteria for status epilepticus (Hesdorffer et al., 2011). Febrile status epilepticus (FSE) accounts for 5% of FS but 25% of all childhood SE and >70% of SE in the second year of life (Shinnar et al., 1997). FSE is associated with a substantially increased risk of epilepsy and, in particular, temporal lobe epilepsy (TLE) (Shinnar, 2003). More recent studies have demonstrated evidence of acute hippocampal injury following FSE (VanLandingham et al., 1998; Lewis et al., 2002; Scott et al., 2002, 2003; Provenzale et al., 2008). How frequently this occurs and its relationship to subsequent hippocampal sclerosis (HS) and TLE are still unknown.
The underlying causes of FSE have not been well established. The cause of the febrile illness may influence not only whether a FS occurs but also its duration and whether associated hippocampal injury occurs (French et al., 1993; Berg et al., 1995; Lewis et al., 2002). Other large epidemiologic studies have established that the outcome of convulsive status epilepticus may depend on the etiology (Chin et al., 2006; Nishiyama et al., 2007; Sadarangani et al., 2008). The role of human herpesvirus (HHV)-6 and HHV-7 in causing FSE, hippocampal injury, and subsequent HS and TLE is of particular interest. HHV-6 and HHV-7 are closely related β-herpesviruses that are universally acquired in early childhood (Hall et al., 1994). The median age for acquisition of HHV-6 is 9 months (Hall et al., 1994) and 26 months for HHV-7 (Caserta et al., 1998), corresponding to the peak incidence of FS and FSE. HHV-6 is the etiologic agent of roseola infantum (Yamanishi et al., 1988). There are two viral subtypes of HHV-6: type A and type B. HHV-6B is a common cause of both febrile illnesses and of FS (Hall et al., 1994; Caserta et al., 1998), whereas HHV-6A has been associated with reactivated infection later in life, predominantly in the central nervous system, often as a result of immunologic suppression (Dewhurst et al., 1993). HHV-7 primary infection is most often asymptomatic (Caserta et al., 1998), but like HHV-6, can present with fever and, in this setting, has an even higher association with FS (Caserta et al., 1998; Hall et al., 2006) Recent studies report evidence of HHV-6B in hippocampal specimens from surgical resections performed on adults with HS and medically refractory TLE, many of whom reported prolonged FS in childhood (Donati et al., 2003; Fotheringham et al., 2007; Provenzale et al., 2008). Together these studies suggest that HHV-6B may be a cause of FSE and in addition may contribute to hippocampal injury and subsequent TLE.
A prospective study, the Consequences of Prolonged Febrile Seizures in Childhood (FEBSTAT), is investigating whether FSE causes hippocampal injury, whether such injury leads to subsequent HS, and the factors associated with such injury. As part of this study, we determined the frequency of primary and reactivated HHV-6A, HHV-6B, and HHV-7 infection as a cause of FSE to test the hypothesis that FSE due to HHV-6A, HHV-6B, or HHV-7 infection is more likely to result in hippocampal injury and TLE. In this article, we report the frequency of HHV-6 and HHV-7 associated with FSE in children.
Study design and study subjects
FEBSTAT is a multicenter, prospective study of children ages 1 month through 5 years who present with FSE, defined as a seizure or a series of seizures without fully regaining consciousness lasting ≥30 min (Commission, 1993; Dodson et al., 1993). FS was defined as a seizure associated with a febrile illness (temperature ≥38.4°C or 101.0°F) without prior history of afebrile seizures and with no evidence of an acute central nervous system (CNS) infection or insult (National Institutes of Health, 1980). FEBSTAT was designed to address the relationship between FSE and subsequent HS and TLE in otherwise normal children; therefore, children with severe neurologic disability were excluded. The five recruiting sites for FEBSTAT are listed in the acknowledgements.
Although hospital-based series of febrile seizures tend to be biased (Ellenberg & Nelson, 1980), convulsive SE by its nature will present to emergency departments at a hospital and therefore hospital-based series are likely to be representative. Epidemiologic studies of SE ascertain status in hospital settings (Hesdorffer et al., 1998; Waterhouse et al., 1999). The FEBSTAT sites, although tertiary care centers, have active emergency departments, serve as primary treatment centers for children with SE, and receive transfers from other hospitals that are not equipped to manage SE. Emergency departments were screened to ascertain SE in FEBSTAT; few children were not admitted and the refusal rate was quite low. Because FEBSTAT is a prospective study, children with SE were not selected from a tertiary care epilepsy program among children with epilepsy who have a past history of FSE or other SE as was the case with older studies (Aicardi & Chevrie, 1970; Fujiwara et al., 1979). Therefore, selection bias is an unlikely problem in FEBSTAT because SE is a medical emergency for which care is obtained in emergency departments, centers from different geographic locations were the enrolling sites, the study is prospective, and there is a high recruitment rate, allowing a fair degree of confidence that the population is representative of children with FSE. HHV-6 and HHV-7 infections are universally acquired, and socioeconomic status has not been found to be a factor in the age of acquiring these infections (Clark et al., 1993; Hall et al., 1994; Cermelli et al., 1996).
The FEBSTAT study, by design, did not have a group of controls. However, the data from a previously published study of HHV-6 infection found no evidence of primary HHV-6 infection in 582 infants and young children with acute nonfebrile illness and in 352 controls without acute illness (Hall et al., 1994). In addition, among 1,653 infants and young children with acute febrile illness, 9.7% had evidence of primary HHV-6 infection (Hall et al., 1994). The data from this study in a group of comparable age establishes that primary HHV-6 infection is very uncommon in the absence of a febrile illness.
Subjects were recruited within 72 h of the episode of FSE. During the initial hospitalization, a detailed survey gathered information about the circumstances of the seizure with an emphasis on duration and focality, prior development, family history and prior illnesses. A complete neurologic and physical examination was conducted. Magnetic resonance imaging (MRI) and electroencephalography (EEG) were performed. Laboratory data including cerebrospinal fluid (CSF) results were gathered, although the decision to perform lumbar puncture was made by the clinical team. Blood specimens for HHV-6 and HHV-7 were obtained at the time of presentation (baseline) and at 1 month. When available, leftover CSF was also analyzed. Details of the overall study methodology have been published previously (Shinnar et al., 2008). The study was approved by institutional review boards at all participating centers, and written informed consent was obtained from all participants.
Determination of HHV-6 and HHV-7 viremia, primary and reactivated infection HHV-infection
HHV-6A, HHV-6B, and HHV-7 specific real-time quantitative fluorescent probe DNA and RNA polymerase chain reaction (PCR) (Zerr et al., 2000, 2002; Boutolleau et al., 2005) and indirect immunofluorescence for viral antibodies (Hall et al., 1994) was performed on paired blood samples obtained at baseline and at 1 month. The presence of HHV-6A, HHV-6B, or HHV-7 DNA and RNA (amplified across a spliced junction of a late gene transcript) at baseline indicated active viral replication, hereafter referred to as HHV-6 or HHV-7 viremia (Norton et al., 1999; Boutolleau et al., 2006; Hall et al., 2006; Caserta et al., 2010). The absence of HHV-6 or HHV-7 specific antibody at baseline in the presence of viremia indicated primary infection with HHV-6 or HHV-7, respectively. The presence of viral specific antibodies at baseline in the presence of viremia indicated reactivated HHV-6 or HHV-7 infection. The presence of HHV-6 or HHV-7 antibodies at baseline in the absence of viremia indicated prior infection. The absence of viral specific antibodies or viremia indicated no previous infection.
HHV-6 and HHV-7 serologies
Plasma was removed from anticoagulated whole blood and stored at −80°C until used to detect HHV-6 and HHV-7 specific antibodies by indirect fluorescent antibody assay for immunoglobulin G (Advanced Biotechnologies Incorporated, Columbia, MD, U.S.A.). Quantitative assay of HHV antibodies was performed with serial dilutions with a titer ≥7.32 log2 considered positive. The change or rise in antibody titer at 1 month was recorded, but this was not used as a primary endpoint.
The protocol was adapted from previously published literature (Norton et al., 1999; Zerr et al., 2000; Boutolleau et al., 2005, 2006) to detect HHV-6A, HHV-6B, and HHV-7, using quantitative polymerase chain reaction (qPCR) and quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR) with an Applied Biosystems 7500 Real-Time PCR System automatic thermocycler. Primers and probes are shown in Figure 1. The assay was performed using TaqMan (Life Technologies Corporation, Carlsbad, CA, U.S.A.) probes and primers for HHV-6 and HHV-7 that amplify across a spliced junction within the gp105 mRNA transcripts (Norton et al., 1999), a late expressed major structural protein indicative of productive viral replication (Norton et al., 1999). Tenfold serial dilutions of HHV-6 variants A and B (Advanced Biotechnologies Incorporated) produced a standard curve to determine the copy number for our specimens. About 830-nm primers (IDT, Houston, TX, U.S.A.), 100-nm probe (Megabases Inc., Evanston, IL, U.S.A.), 8% glycerol, and a Master Mix (Applied Biosystems, Foster City, CA, U.S.A.) were added to triplicate samples along with negative and positive controls. The cycle threshold (Ct) value was determined by the cycle number required for fluorescence to exceed a selected threshold in the exponential phase of the amplification. The copy number ranged from 1 to 100,000.
Descriptive analysis was conducted through the determination of a percentage, mean, or median. Standard deviations are presented for means and the interquartile range (IQR) for the median. For the comparison of intermittent versus continuous seizures, comparison of means was performed using the t-test, comparison of medians using the Wilcoxon rank-sum test, and comparison of frequencies using the chi-square test or the Fisher’s exact test.
This study is not registered with ClinicalTrials.gov as it is not a clinical trial
Of 199 subjects, 169 (84.9%) had blood samples allowing determination of HHV-6A, HHV-6B, and HHV-7 status at time of FSE. Of these 169, 125 children (74.0%) had a lumbar puncture and CSF was available.
Frequency of HHV-6 and HHV-7 viremia
A total 58 (34.3%) of 169 subjects had viremia with either HHV-6B or HHV-7 at the time of presentation with FSE. No HHV-6A infections were identified. Figure 2 shows the composition of the 58 subjects with viremia according to whether the infection was primary or reactivated at baseline. HHV-6B viremia was present in 54 subjects (32.0%) at baseline, including 38 with primary infection and 16 with reactivated infection. Four of the reactivated infections were associated with primary HHV-7 infection. Twelve subjects (7.1%) had HHV-7 viremia at baseline, including eight with primary infection and four with reactivated infections. The remaining 111 subjects did not have viremia at baseline and were used as the comparison group.
There was no CSF pleocytosis in children with FSE including those with HHV-6 and HHV-7 infection. In the 108 children with nontraumatic lumbar punctures (<1,000 RBCs) and with information available on white blood cell (WBC) counts, the median number of WBCs was 1.0 (interquartile range [IQR] 0.0–2.0) in the 76 children with no or prior HHV-6 or HHV-7 infection and was 1.0 (IQR 0.0–2.0) in the 32 children with primary HHV-6 or HHV-7 infection or reactivation at baseline (p = 0.93). Neither HHV-6 nor HHV-7 DNA or RNA was detected in any of the 75 subjects with CSF available at baseline, including the 23 subjects with HHV-6B or HHV-7 viremia (11 subjects with HHV-6B primary infection, 7 subjects with reactivated HHV-6 infection, 4 subjects with HHV-7 primary infection and 1 with HHV-7 reactivated infection determined from blood samples). There were also no clear differences in the mean peripheral WBC count between groups (mean WBC 13,1K in those with primary or reactivated HHV-6B or HHV-7 infection versus 14,8K in those without primary or reactivated HHV-6B or HHV-7 infection, p = 0.18).
A summary of the clinical characteristics of children with HHV-6B or HHV-7 viremia (primary or reactivated infection) and of those with prior or no HHV-6B or HHV-7 infections is shown in Table 1. There were no significant differences between the two groups with respect to age, gender, development, history of prior FS or family history of FS or epilepsy. In addition, there were no significant differences in the characteristics of the acute illness, including duration of illness, peak temperature, and temperature in the emergency department. It is important to note that the characteristics of the episode of FSE itself, including median seizure duration, intermittent versus continuous FSE, and focal onset FSE was not associated with the presence or absence of HHV-6B or HHV-7 viremia.
Table 1. Descriptive information for children with FSE with HHV-6 or HHV-7 infections (primary or reactivated) at baseline compared to children with prior or no HHV-6 and HHV-7 infections
Total (N = 169)
HHV-6+ or HHV-7+ or reactivation (N = 58)
HHV-6− and HHV-7− or prior infections (N = 111)
FSE, febrile status epilepticus; FS, febrile seizure.
aOne with missing information.
Median (interquartile range)
Age ≥18 months
Prior febrile seizures
Sick behavior before FSEa
Duration of recognized fever before FSE
Peak temperature during illness
Mean peak temperature °F (SD)
Mean ED temperature °F (SD)a
Duration of FSE (min)
Median (interquartile range)
Continuous versus intermittent FSE
Intermittent without full recovery in between
Intermittent with full recovery in between
Yes (definitely or probably)
No (maybe or no)
Family history of FS or FSEa
Evidence of acute neurologic injury
There was no significant difference in the proportion of children with acute EEG abnormalities or with acute hippocampal MRI abnormalities between children with HHV-6B or HHV-7 viremia and those without. For MRI this incudes proportion with acute hippocampal T2 signal as well as of developmental abnormalities. For EEG this includes presence of focal slowing or attenuation and of epileptiform activity.
This prospective study demonstrates that HHV-6B and HHV-7 infection are commonly associated with FSE with HHV-6B predominating. HHV-6 and HHV-7 establish persistent infection with the potential for reactivation (Caserta et al., 2004; Hall et al., 2006). Primary HHV-6 infection is a known cause of simple and complex FS (Hall et al., 1994). In the 38 subjects with primary HHV-6B infection we conclude that this is the likely cause of FSE. It is not known if reactivated HHV-6B infection can cause seizures. Sixteen subjects in this study had FSE associated with reactivated HHV-6B infection; however, in four of these subjects HHV-6B reactivation occurred during primary HHV-7 infection, the more likely cause of FSE. The cause of the reactivated infections in the other 10 subjects is not known and the role of reactivated HHV-6B or dual-reactivated infection as a cause of FSE will require additional study. Similarly HHV-7 primary infection has been previously shown to cause simple and prolonged FS (Caserta et al., 1998). Hall and colleagues using a similar strategy to distinguish primary from reactivated HHV-7 infection identified 30 children (of 2,806) with HHV-7 viremia (Hall et al., 2006). These authors found that 12 of these children with HHV-7 viremia (11 primary, 1 reactivated) presented with FS, of whom 2 had FSE (Hall et al., 2006).
The demonstration of primary infection with either of these viruses at the time of presentation with FSE strongly suggests that the virus is the proximate cause of the fever, which in turn is associated with FSE. However we cannot distinguish whether the FSE is, at least in part, due to a direct effect of the virus on the brain or secondary to the inflammatory process and high fever caused by the viral infection. Studies of febrile seizures and FSE in animal models have shown that, in addition to fever, inflammatory cytokines in particular IL-1B may contribute to epileptogenesis, reviewed in (McClelland et al., 2011). Cytokines have been implicated in the acute encephalopathy that occasionally follows prolonged febrile seizures (Ichiyama et al., 2008) and specifically in the rare encephalopathy in children associated with HHV-6 infection (Ichiyama et al., 2009).
Ward and colleagues performed a 3-year prospective study of 205 children 2–35 months old hospitalized with suspected encephalitis and/or severe illness with convulsions reported via the British Paediatric Surveillance Unit Network (Ward et al., 2005). These authors found that 17% (26 of 156) of these children had primary infection with either HHV-6 or HHV-7, all 26 had fever, 25 had convulsions, and 18 had status epilepticus (Ward et al., 2005). Although Ward and colleagues found primary HHV-6 and HHV-7 infection to be approximately equal in causing encephalitis in their study, the subjects were not selected on the basis of presenting with febrile status epilepticus; hence the studies are not directly comparable. The finding of febrile status epilepticus in 18 of 26 children <2 years of age with primary HHV-6 or HHV-7 infection is consistent with the results reported in this article.
HHV-6 DNA has been detected in the CSF of children presenting with FS, suggesting neuroinvasion during primary infection (Hall et al., 1994). We did not detect HHV-6B or HHV-7 DNA in the CSF of the 23 subjects who presented in FSE with documented HHV-6B or HHV-7 viremia. Hall et al. reported detection of HHV-6B DNA in two of seven samples from children with seizures during primary HHV-6B infection (Hall et al., 1994). We would have anticipated detecting HHV-6B DNA in the CSF of a few of the subjects in the FEBSTAT cohort. This may be due to the multicenter design of FEBSTAT, which results in several days’ delay in CSF reaching the virology laboratory and may cause further loss of viable cells in the CSF. Other data from a pediatric study of primary HHV-6 infection suggest that the presence of viral DNA in the CSF is transient (Mannonen et al., 2007). No definitive conclusions can be drawn about the presence or absence of neuroinvasion by HHV-6B or HHV-7 based on our data.
The lack of clinical or laboratory features that distinguish children with FSE associated with HHV-6B or HHV-7 viremia from those without infection is particularly interesting. Prior studies of HHV-6 and HHV-7 have focused on children with fever and have found differences in age, height of fever, and other clinical features that distinguished between those with HHV-6 or HHV-7 infection as the cause of the fever from those with fever due to other causes (Hall et al., 1994; Caserta et al., 1998). In FEBSTAT, where all children had both fever and FSE, there were no clinical or laboratory differences that distinguished the two groups.
One of the most controversial issues in epilepsy is whether or not prolonged FS causes HS and associated TLE (Shinnar, 2003). MRI evidence of acute postictal hippocampal abnormalities followed by the development of HS after FSE provides strong evidence for a causal connection (VanLandingham et al., 1998; Lewis et al., 2002; Scott et al., 2002, 2003). An aim of the FEBSTAT study is to determine whether FSE due to HHV-6 or HHV-7 infection is more likely to result in hippocampal injury, HS, and TLE than FSE without HHV-6 or HHV-7 infection. The findings from this study indicate that HHV-6B is a commonly associated with FSE. The causative link between FSE and HS remains to be determined and is the next step in the FEBSTAT study.
Data from other studies highlight the potential importance of HHV-6B as a potential cause of HS. In samples of resected temporal lobe from patients with intractable TLE, HHV-6B was identified in hippocampal astrocytes, but was not detected in surgical tissue from patients with neocortical epilepsy (Donati et al., 2003). HHV-6B tropism for astrocytes and inhibition of glutamate reuptake, provides a potential pathogenic mechanism for excitotoxic hippocampal injury (Donati et al., 2005). These data are consistent with HHV-6B localization in the temporal lobe during primary infection with later reactivation and suggest that children with FSE due to HHV-6B infection may be at increased risk for subsequent TLE similar to the mechanism postulated for HHV 6B–associated limbic seizures in immunosuppressed patients (Wainwright et al., 2001).
Given the long latency period (8–11 years) between the episode of FSE and subsequent TLE (Commission, 1993; French et al., 1993; Mathern et al., 1995) it will be some time before the role of HHV-6B in the development of TLE will be fully understood. FEBSTAT will follow the 199 children with FSE of which 30–40% are expected to ultimately develop epilepsy including TLE (Shinnar, 2003). The current data showing that that 32.0% of all FSE is due to HHV-6B infection lays the foundation to determine whether these subjects are more likely to develop hippocampal injury and TLE. If an association between HHV-6B infection and TLE is found in the FEBSTAT study, it will inform the design of future trials aimed at prevention of intractable TLE following FSE.
Supported by NINDS grant NS43209 (P.I. S. Shinnar, M.D., Ph.D.).
FEBSTAT Study Team: Montefiore Medical Center and Jacobi Medical Center, Albert Einstein College of Medicine, Bronx, NY: Shlomo Shinnar M.D., Ph.D. (PI), Jennifer Ayala B.A., Jacqueline Bello M.D., Ronda L. Facchini Ph.D., James Hannigan RT, Sharyn Katz R-EEGT, Ann Mancini M.A., David Masur Ph.D., Solomon L. Moshé M.D., Ruth C. Shinnar RN MSN, Maryana Sigalova M.A., Rachel Steinman B.A., Erica Weiss M.A.
Children’s Memorial Hospital, Northwestern University Feinberg School of Medicine, Chicago, IL: Douglas Nordli M.D. (PI), John Curran M.D., Sarah Ahlm M.A., LCSW, Leon G. Epstein M.D., Aaliyah Hamidullah M.S., Andrew Kim M.D., Julie Renaldi Ph.D., Diana Umanzor.
Columbia University School of Medicine, New York, NY: Dale C. Hesdorffer Ph.D. (PI), Emilia Bagiella Ph.D., Emma K. T. Benn MPH, Stephen Chan M.D., Veronica J. Hinton Ph.D., Christine Roman MPH.
Duke University Medical Center, Durham, NC: Darrell Lewis M.D. (PI), Melanie Bonner Ph.D., William Gallentine DO, James MacFall Ph.D., James Provenzale M.D., Elizabeth Rende RN MSN CPNP, Allen Song Ph.D., James Voyvodic Ph.D., Yuan Xu BS.
Eastern Virginia Medical School, Norfolk, VA: L. Matthew Frank M.D. (PI), Joanne Andy RT, Terrie Conklin RN, Susan Grasso M.D., Diane James R-EEG T, David Kushner M.D., Susan Landers RT, Virginia Van de Water Ph.D.
Virginia Commonwealth University, Richmond, VA: John M. Pellock M.D. (PI), Tanya Bazemore REEG-T, James Culbert Ph.D., Kathryn O’Hara RN, Syndi Seinfeld M.D., Jean Snow RT-R.
International Epilepsy Consortium at Virginia Commonwealth University, Richmond, VA: Shumei Sun PhD, Brian J Bush MSMIT, Sreedevi Chandrasekaran, Lori L. Davis, John M. Pellock M.D., Christiane Rogers, Cynthia Shier Sabo MS, Helen Wang.
Collaborators: Joan Conry M.D., Childrens National Medical Center, Washington DC – Safety Monitor. Tracy Glauser M.D., Cincinnati Children’s Hospital, Cincinnati OH – Genomics substudy. Jeffrey L. Noebels M.D., Ph.D., Baylor College of Medicine, Houston, TX – Genetics substudy.
None of the authors has any conflict of interest to disclose.
The authors confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with the guidelines of the Committee on Publication Ethics (COPE) guidelines for ethical publication (http://publicationethics.org/).