Brain Perfusion, Hippocampal Volumetric, and Diffusion-weighted Imaging Findings in Children with Prolonged Febrile Seizures and Focal Febrile Seizures.

BACKGROUND: The current study was conducted to describe the findings of brain perfusion, hippocampal volumetric, and diffusion-weighted imaging (DWI) in children aged six months to seven years with febrile status epilepticus (FSE) and focal febrile seizures (FFS) when compared with age and gender-matched controls.
MATERIALS AND METHODS: This cross-sectional study was conducted among children aged six months to seven years presenting with FSE or FFS within 72 h of the seizure. Cases were subjected to magnetic resonance imaging (MRI) brain and sleep-deprived electroencephalography. Age and gender-matched children who were subjected to MRI brain for nonepileptic indications served as their control. Hippocampal volumes, T2 values, cerebral blood flow, and diffusion characteristics were compared between the cases and controls and also between those with FSE and FFS.
RESULTS: A total of 31 cases (FFS = 20, FSE = 11) and 30 controls were enrolled. There was no significant difference between right and left hippocampal volumes and T2 relaxometry values between cases and controls and also between children with FSE and FFS. Similarly, the cerebral blood flow was also comparable in cases and controls. There was a single case of FSE with hippocampal malrotation; one child showed diffusion restriction in the hippocampus after prolonged (>60 min) FSE.
CONCLUSION: Children with FSE and FFS had comparable hippocampal volume and brain perfusion to healthy children. However, one child with FSE had hippocampal malrotation and another had diffusion restriction. The study findings need to be interpreted in the context of small sample size, and lack of follow-up neuroimaging. Copyright:

INTRODUCTION

Complex febrile seizures include febrile seizures that are prolonged, focal, or have multiple episodes.[1] Children with multiple episodes of febrile seizure within a 24-h time frame are often considered to behave similar to having a simple febrile seizure rather than a complex febrile seizure and have been labeled as simple febrile seizure plus (SFS+).[2] Prolonged febrile convulsion (PFC) or FSE was defined as a seizure in a neurologically healthy child between six months and five years of age lasting at least 30 min, which is associated with fever that is not of CNS origin.[3] The concern of hippocampal injury after prolonged febrile seizure (PFS) arises from the fact that nearly 50% of patients with mesial temporal sclerosis (MTS) have a history of PFC in childhood, suggesting their association.[4]

MRI abnormalities include hippocampal volumetric changes, and T2 signal hyperintensities have been demonstrated in children with complex febrile seizure.[56] Diffusion restriction and perfusion abnormalities are known among children after PFS.[78] An MRI study done in children with PFC within 48 h of seizure showed that hippocampi were large with a prolonged T2 relaxation time suggestive of hippocampal edema.[5] Similarly, hippocampal malrotation has also been associated with complex febrile seizures.[9]

There are limited data on hippocampal abnormalities among children with PFS and FFS. Hence, the current study was designed to describe the findings of brain perfusion, hippocampal volumetric, and DWI in children aged six months to seven years with FSE and FFS when compared with age and gender-matched controls.

MATERIALS AND METHODS

Patient selection

This cross-sectional study was conducted in a tertiary care referral center of North India from January 2018 to April 2019. Children aged six months to seven years who presented to the pediatric emergency with FSE and FFS were enrolled in the study. Age and gender-matched typically developing children without epilepsy undergoing MRI for some other reason served as controls. Children with developmental delay, intellectual disability, autism spectrum disorder, and other pervasive developmental disorders or those with previous severe neurological disabilities or those with progressive neurological disease were excluded from the study.

The febrile seizure was considered as provoked seizure in which the sole acute provocation was fever (temperature >38.4°C, 101.0°F) with no evidence of a severe CNS infection or insult. The FFS was defined as focal seizure semiology in cases, otherwise fulfilling the criteria of febrile seizure.[1] FSE was defined as a single seizure or a series of seizures without full recovery, lasting for more than 30 min that otherwise fulfills the definition of febrile seizure.[1] Considering 22 of 226 (9.7%) children demonstrating T2 hippocampal hyperintensity after FSE in the FEBSTAT study, a sample size of 28 (rounded to 30) was computed while considering a 95% confidence level and a 10% margin of error.[6]

Clinical information

Written informed consent was obtained from the caregivers, and an institutional ethical clearance was sought before the commencement of the study. All eligible children were enrolled consecutively and were subjected to detailed clinical history, including the medical history, developmental history, and family history. A complete neurological evaluation was done among all enrolled children. Children were subjected to an MRI of the brain within 72 h of seizures or hospital admission once the child was stable.

MRI protocol

All MRI investigations were performed on a 3.0T MR scanner (Discovery MR 750w; GE, Milwaukee, USA) by using a protocol, including T1W SE (TR 600 ms, TE 10 ms, echoes 1/1, flip angle 68o, reconstruction matrix 320 × 224, NSA-2, FOV (22 × 22) cm2, slice thickness 5 mm); T2W (TR 4805 ms, TE 114 ms, echo 1/1, flip angle 142o, reconstruction matrix 512 × 512, NSA-2, FOV (22 × 22)cm2, slice thickness 5 mm); T2W/FLAIR (TR 10,000 ms, TE 120 ms, echos 1/1, flip angle 160o, reconstruction matrix 320 × 228, NSA 1, FOV (22 × 22)cm2, slice thickness 5 mm); T2 Map (TR 800 ms, TE 5.7–45.8 ms, number of TEs 8, phase FOV 1, frequency 256, phase 192, NSA 2, bandwidth 62.50, slice thickness 5 mm, spacing 0.6 mm); DW (time of repetition 5061 ms, time to echoes 83 ms, echo 1/1, b value 1000, reconstruction matrix 192 × 192, NSA 2, field of view (22 × 22) cm2); ASL (FOV = 24, slice thickness = 4.0 mm, TR = 4852 ms, TE = 10.7 ms, NSA = 3, bandwidth = 62.5, post-label delay = 2025 ms). Hippocampal volumetry was performed by using the images obtained from FSPGR-BRAVO (TR 7.2 ms, TE 1 ms, echoes 1/1, flip angle 12o, acquisition reconstruction voxel size 0.9 mm, reconstruction matrix 256 × 256, NSA 1). Regions of interest (ROIs) were drawn manually on successive coronal slices, using axial and sagittal views for further refinement, to encompass the entire hippocampus.[1011]

Outcome

The following parameters on the MRI brain were compared between the cases and controls: Volumetric analysis of right and left hippocampus, T2 relaxometric analysis in bilateral hippocampii, perfusion imaging in bilateral hippocampii using ASL imaging, and DWI of hippocampii. All MRI scans were interpreted by the principal investigator who was blinded to the clinical details, minimizing the interobserver variance.

Electroencephalography

Electroencephalography (EEG) was done at the earliest feasible appointment in the neurophysiology laboratory. All EEGs were performed according to the standards of the American Clinical Neurophysiological Society (ACNS) for the recording of EEGs.[10] Electrodes were applied according to the universal 10–20 system by using standard measurements. Subjects were sedated by using Triclofos (0.2 mL/kg) for the purpose of EEG as per standard procedure when required.

Analysis

All data was entered in Microsoft Excel SPSS software version 15 for statistical analysis. Categorical variables were expressed as numbers (percentage), and continuous variables were expressed as mean (SD) or median (IQR). Categorical variables such as prevalence of mesial temporal sclerosis and changes in hippocampi were compared between two groups by using χ2 or Fisher’s exact test. Continuous variables, including hippocampal volumes and percentage of hippocampal volume, were compared between cases and controls by using Student’s t-test or Wilcoxon rank-sum test.

RESULTS

A total of 46 children with complex febrile seizure were admitted during the study period, of whom 31 were eligible for enrollment. Thirty-one cases (FFS = 20, FSE = 11) and 30 controls were enrolled in the study. There were 20 boys among the cases and 16 among the controls. Both right and left hippocampal volumes [Figure 1] and T2 relaxometry values were comparable between children with FSE/ FFS and age-matched controls [Table 1] as well as between children with FSE and FFS [Table 2]. Similarly, the cerebral blood flow as determined by ASL was comparable in cases (FSE+FFS) and controls [Table 1]. There was a single case of FSE whose MRI revealed hippocampal malrotation [Figure 2]. Another child showed diffusion restriction in hippocampus after prolonged (>60 min) FSE [Figure 3]. Only one child with FSE had generalized epileptiform abnormality on EEG, with the remaining 30 children having normal EEG. Despite lack of statistical significance, there was negative correlation between the hippocampal volumes and the duration and frequency of seizure. Also, there was no significant correlation between the hippocampal volume and temperature at the time of seizure [Table 3].

Figure 1

(A and B) Images showing B/L hippocampal volumes; R = 1.965 cm,[3] L = 2.478 cm[3]

Table 2

Table comparing the MRI findings in children with FFS and FSE

Parameter	Children with FSE (n = 11)	Children with FFS (n = 20)	P value
Hippocampal volume right (cm3)	2.3 ± 0.59	2.47 ± 0.45	0.20
Hippocampal volume left (cm3)	2.01 ± 0.59	2.14 ± 0.61	0.83
Mean T2 relaxometry value	102.62	98.82	0.42

Table 1

Table comparing the MRI findings among children with FSE/FFS and age-matched healthy controls

Parameter	Children with FSE/FFS	Control	P value
Hippocampal volume right (cm3)
0–24 months	2.12	2.04	0.23
24–48 months	2.50	2.21	0.22
49–72 months	2.48	3.09	0.12
Hippocampal volume left (cm3)
0–24 months	2.01	2.00	0.08
24–48 months	2.51	2.13	0.10
49–72 months	2.19	2.89	0.22
T2 relaxometry values in Hippocampus right
0–24 months	100.6	98.02	0.14
25–48 months	100.3	97.02	0.15
49–72 months	101.6	98.12	0.17
T2 relaxometry values in Hippocampus left
0–24 months	91.20	94.60	0.23
25–48 months	90.20	93.60	0.33
49–72 months	92.40	93.58	0.35
Cerebral blood flow right (mL/100 g/min)
0–24 months	3.50	3.67	0.63
25–48 months	2.73	4.06	0.87
49–72 months	3.39	4.64	0.59
Cerebral blood flow left (mL/100 g/min)
0–24 months	3.05	3.39	0.98
25–48 months	2.41	4.27	0.13
49–72 months	2.98	4.25	0.54

Figure 2

Coronal T2W images showing normal horizontally oriented right hippocampus and vertically oriented left hippocampus and collateral sulcus suggesting left hippocampal malrotation

Figure 3

(A) DW and (B) ADC map showing diffusion restriction in right hippocampus

Table 3

Table comparing hippocampus volumes with duration, frequency of seizure, and temperature at seizure (n = 31)

	Mean	Correlation coefficient with duration of seizure	Correlation coefficient with frequency of seizure	Correlation coefficient with temperature at seizure
Right hippocampal volume	2.26	−0.392	-0.31	0.18
Left hippocampal volume	2.17	−0.36	-0.15	0.22

DISCUSSION

The current study reveals lack of any significant hippocampal and perfusion abnormalities immediately after FFS and FSE. This is in contrast to findings of hippocampal T2 hyperintensities after PFS that evolve to hippocampal sclerosis over a period of one year.[11] In the FEBSTAT study, MRI abnormalities in the form of T2 hyperintensities were reported in 22 of 199 children within 72 h after FSE.[6]

In a study by Natsume et al.,[5] 12 children with PFS who had seizures beyond a 60-min duration demonstrated larger hippocampal volume. In contrast, another study on five children with complex febrile seizure demonstrated that total hippocampal volume tends to be smaller than in controls.[12] The current study, in contrast, did not demonstrate any significant hippocampal volumetric changes among children with FSE. Despite lack of statistical significance, we found a negative correlation of hippocampal volumes with duration and frequency of seizures.

In an interesting study by Yokoi et al.,[13] among 22 children with FSE, six patients (27%) demonstrated unilateral hippocampal hyperintensity and three of them demonstrated unilateral thalamic hyperintensity. These hyperintensities have also been demonstrated in the cortex, apart from thalamus and hippocampus.[13] Natsume et al.[9] found that in patients with PFS for 60 min or longer, DWI showed hyperintensity in the unilateral-hippocampus in three patients with intractable seizure, ipsilateral thalamus in two, and cingulate in one. The EEG showed abnormalities in temporal areas ipsilateral to DWI abnormalities in these patients. They concluded that large hippocampal volumes and hippocampal hyperintensities on DWI were seen in patients with refractory PFS and also suggested that refractory PFS lasting for 60 min or longer may cause structural changes in limbic structures.

A four-year-old boy with prolonged FSE had demonstrated transient reduced diffusion in the cortex, thalamus, and hippocampus on DWI.[11] In our cohort, only one patient with FSE with seizures lasting more than 60 min demonstrated similar hyperintensity on DWI images in the right hippocampus, supporting the hypothesis of cytotoxic edema after FSE.[1415] This patient developed T2 hyperintensity in the right hippocampus (T2 value in the right hippocampus is 107 ± 8.7 whereas that in the left hippocampus is 86 ± 3.0) with slightly hippocampal volume loss (right hippocampal volume of 2.151 cm[3] vs left hippocampal volume of 2.516 cm[3]) on follow-up MRI done after six months. These observations support Natsume’s conclusion and suggest that only status epilepticus lasting more than 60 min showed diffusion abnormalities, that too extending far and wide in the limbic circuit and not restricted to only hippocampus (though signal abnormality was seen only in hippocampus in our case). Interestingly, none of the children with FFS demonstrated any focal abnormalities on imaging.

Developmental abnormalities of the hippocampus were common in the FSE group than controls, with hippocampal malrotation (HIMAL) being the most common.[6] Developmental malformation, including hippocampal malrotation, has been demonstrated in 20 of the 226 (8.8%) children with FSE.[10] It was observed that boys and those with seizure lasting for more than 60 min were significant risk factors for HIMAL among children with FSE. Most of these patients with HIMAL demonstrated smaller hippocampal volumes. Our cohort demonstrated this abnormality in one of 11(9.1%) children with FSE and none of those with FFS. All these studies, including ours, showed less than 10% cases of hippocampal malrotation in patients with FSE.

Asymmetrical cerebral blood flow distribution has been demonstrated in the post-ictal phase of FSE using ASL sequences.[7] In this preliminary report on three pediatric patients with FSE, it was observed that the epileptic zone demonstrated higher regional blood flow when compared with the contralateral side. In contrast, regional cerebral flow was comparable between cases and controls in our cohort. The ASL sequences are gaining popularity in pediatric neurology, considering it to be a noninvasive investigation.

The strengths of the study were blinding of the radiologist to clinical details, all readings taken by a single radiologist, and experience of the radiologist. Limitations include small sample size, and lack of healthy children as controls considering ethical concerns of subjecting healthy children to neuroimaging. In addition, the results would have been more meaningful when 2:1 or higher ratio of controls are enrolled in the study but for the limitation of feasibility. Moreover, follow-up neuroimaging was not performed in the current study, thus limiting the long-term implications of the study findings. Hence, the results of this current study need to be interpreted in the context of these limitations.

CONCLUSION

The current cross-sectional study demonstrated comparable hippocampal volumetric, T2 signal changes and brain perfusion parameters among children with prolonged and FFS with their controls. One child with PFS demonstrated diffusion restriction and hippocampal malrotation each. Further longitudinal studies with larger sample size and long-term follow-up with repeat neuroimaging would probably help our understanding of this association of hippocampal atrophy or hippocampal sclerosis to prolonged and FFS.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.