Effects of levetiracetam and oxcarbazepine monotherapy on intellectual and cognitive development in children with benign epilepsy with centrotemporal spikes

Children diagnosed with BECTS in the outpatient department of the Affiliated Hospital of Nantong University from October 2018 to February 2020 were selected as the research subjects. The inclusion criteria were (1) a diagnosis in line with the BECTS diagnostic criteria in the new (2017) International Anti-Epileptic Alliance classification of seizures and epilepsy [13]; (2) electroencephalogram (EEG) features that showed that a seizure during an epileptic attack was partial, or was generalized to the whole body, and that the background rhythm was normal; (3) at least two convulsions before enrollment; (4) no abnormality in head magnetic resonance imaging (MRI) or computed tomography examination; and (5) normal liver and kidney function prior to the commencement of medication. The exclusion criteria were (1) the presence of encephalitis, brain injury, cerebral hemorrhage, and other organic diseases of the nervous system; (2) functional insufficiency of the liver, lung, kidney or other important organs; (3) the emergence of mental retardation; (4) the presence of cranial space-occupying lesions; (5) poor medication compliance; and (6) the presence of relevant drug contraindications. Head scans were performed on all children using MRI to differentiate them from other related seizures.

The study was approved by the ethics committee of the Affiliated Hospital of Nantong University, and the parents or guardians of the eligible children signed informed consent forms.

This was a randomized clinical trial. The eligible children were randomized (1:1) via the random number table to the LEV group or the OXC group. Follow-up results were recorded at 1, 3 and 6 months after the commencement of treatment. During follow-up assessments, the frequency and form of seizures, EEG changes, cognitive changes, adverse drug events and drug tolerance data were recorded.

The LEV group was given Keppra (a brand of LEV manufactured by UCB Pharmaceuticals), in the form of tablets containing 250 mg of LEV. The initial dose was set at 10 mg/kg/day. Thereafter, it is advised that this dose is increased once every 7 days, and is maintained at 20–60 mg/kg/day. The OXC group was given OXC (manufactured by Novartis Pharma Schweiz AG), in the form of tablets containing 150 mg of OXC. The initial dose was set at 8–10 mg/kg/day, orally administered twice a day at an interval of 12 h. Thereafter, it is advised to increase the dose to 5–10 mg/kg/day every 5–7 days, depending on the situation, and maintain the dosage at 20–46 mg/kg/day. All children started treatment at a low dose and returned to the clinic for assessment once a week at the beginning of treatment. During the treatment, clinical reactions in each child were closely observed, and the dosage of drug was appropriately adjusted according to the weight of the child and his/her seizure status. If a child exhibited obvious adverse events, the treatment was adjusted. Both groups were followed up for 6 months.

Before the start of the study, the investigators explained the significance of the study to the subjects and explained the importance of taking medication on time to ensure the efficacy. In addition, the investigators called weekly to inquire about medication intake of each subject. Moreover, all the drug packages sent to the patients were returned to the researchers after the trial was over. According to the packages of used and remaining drugs, we could calculate the medication compliance of research subjects.

The monthly frequency of epileptic seizures in the 2 months before treatment was used as the standard, and the monthly frequency of epileptic seizures at 3 and 6 months after treatment was compared between the two groups. A curative outcome was a ≥ 95% decrease in the monthly seizure frequency after treatment. An effective treatment outcome was a 50–74% reduction in the frequency of monthly seizures, while an ineffective treatment outcome was a < 50% decrease or an increase in monthly seizure frequency. The results showed that the monthly frequency of seizures was reduced by 75 to < 95%. The total effective rate of seizure treatment was calculated as [(cured cases + apparently effective cases + effective cases)/total number of cases] × 100.

The amount of interictal epileptiform activity was recorded by EEG before and after treatment in the baseline stationary state, when a child was supine, awake and quiet, and was classified as (1) normal (the epileptic discharge had disappeared completely after treatment); (2) significantly improved: (the epileptic discharge was reduced by 50% compared with that before treatment); (3) improved (the epileptic discharge was reduced by less than 50% compared with that before treatment; (4) not improved (the EEG results showed no improvement compared to those before treatment); or (5) worsened (the EEG results showed that epileptic discharge was increased compared to that before treatment). The total response rate of EEG treatment was calculated as [(normal + significant improvement + improvement)/total number of cases] × 100.

The criteria used for the evaluation of cognitive function efficacy comprised eight indicators, and both groups were assessed according to these criteria before and after treatment. The neuropsychological assessment battery included eight cognitive tests that assessed multiple cognitive abilities: processing speed, spatial skills, calculation, language ability (including semantic comprehension and phonological processing ability), intelligence, visual attention and executive function.

All tests were programmed using web-based applications in the Online Experimental Psychological System (www.​dweipsy.​com/​lattice) [14]. In the examination room, patients were given a series of tests over two 45-min sessions. For each test, the interviewer had to give the instructions and then practice. Participants’ responses were automatically recorded and sent via the Internet to a server. The tests were conducted by trained interviewers. For processing speed, a white dot appeared to either the left or the right of a fixed cross on a black screen. Participants were asked to press Q if the dot appeared on the left, or P if the dot appeared on the right. A total of 30 trials were conducted sequentially and the time used was summed up to calculate the choice reaction time. For spatial skills, participants were asked to select which 3D image to match the top image from the bottom of the screen by pressing the Q or P key; Matching images can only be identified by mental rotation. Simple subtraction was conducted to assess calculation ability with 92 simple subtraction problems. Participants were asked to press the Q key to select the answer on the left and the P key to select the answer on the right. For each test of word semantics, a sentence was displayed in the middle of the computer screen with one word missing. Participants were asked to choose one of the two candidate words by pressing either the Q key or the P key. Raven’s Progressive Matrices test was used to evaluate figure reasoning ability. Participants must determine that the missing parts of the character will be completed. Two candidate answers are displayed side by side below each question. If the missing part is on the left, instruct the participant to press Q; If on the right, the P key is indicated. For visual tracing, several curves in a square intersect each other, starting at the left side of the square and ending at the right. Participants were asked to follow the target line from start to finish with only their eyes, and then mark the correct end. WCST is one of the most widely used performing functional tests. The task consists of a stimulus card and a response card; According to one of three principles (i.e., color, form, or quantity), the participant was instructed to flip the card so that the response card matched one of the stimulus cards.

Inter-group comparisons were performed for age, frequency of attack, and intelligence test score data, such as means and standard deviations (SDs), using independent sample t tests. Intra-group comparisons were performed using paired t tests, and enumeration data, such as sex, seizure type, follow-up epileptic seizure data and EEG results, were expressed with respect to the total number of cases (percent) [n (%)]. Comparisons were made using the chi-square test or chi-square correction formula, or Fisher’s exact test. Statistical Package for Social Sciences (SPSS) 20.0 statistical software was used for data analysis, with p < 0.05 regarded as an indicator of a statistically significant difference.

Seventy children were randomized to the LEV (N = 35) group or the OXC (N = 35) group, and 32 children in each group were ultimately followed (Fig. 1).

The LEV group comprised 21 boys and 11 girls, who had an average age of 8.47 ± 2.13 years and an average age at onset of 6.98 ± 1.82 years. The OXC group comprised 19 boys and 13 girls, who had an average age of 8.62 ± 2.21 years and an average age at onset of 7.13 ± 1.75 years. There was no significant difference in sex, age, intelligence quotient, incidence frequency and type, EEG status, and family history between the two groups before formal enrollment (p > 0.05) (Table 1).

At the 3-month follow-up, the effective treatment rate was 50% in the OXC group and 37.5% in the LEV group, but these clinical differences were not statistically significant (p = 0.313). After 6 months of follow-up, the effective treatment rate in the OXC group was 78.12%, while that in the LEV group was 53.12%, and these clinical differences were statistically significant (p = 0.035) (Table 2).

After 3 months of follow-up, the effective rate of EEG improvement was 40.63% in the OXC group and 31.25% in the LEV group. After 6 months of follow-up, the effective rate of EEG improvement was 59.38% in the OXC group and 43.75% in the LEV group. There was no significant difference between these rates of EEG improvement (p > 0.05) (Table 2).

We used eight indicators to compare the children’s intelligence and cognitive function development (Table 3). Compared with pre-intervention, word semantics, Raven’s Progressive Matrices, visual tracing, and Wisconsin Card Sorting Test were significant improved after intervention for patients in both LEV and OXC group, and significant improvement were further observed for choice reaction time and mental rotation for patients in OXC group (all p < 0.05). In terms of processing speed (choice reaction time) (p = 0.038), spatial skills (mental rotation) (p = 0.034) and executive function (WCST) (p = 0.046), the OXC group improved more than the LEV group.

The adverse events that were observed included behavioral changes (e.g., drowsiness, mood instability, irritability, fatigue and dizziness), neuropsychiatric symptoms (e.g., headache), gastrointestinal reactions (e.g., loss of appetite, nausea and diarrhea), and rashes, but all of these adverse events were of mild intensity. The incidence of adverse events was 18.75% in the LEV group and 21.88% in the OXC group, with no statistically significant inter-group difference (p = 0.200) (Table 4).