Patients
We retrospectively recruited the neonates who had seizures within the first month of life in Children’s Hospital of Fudan University in China from Jan 2015 to Mar 2020. Of all the patients with convulsions, we selected those have KCNQ2 gene variants identified by molecular genetic testing and agreed to have VEEG examination during the neonatal period in our study. They were followed up to at least 12 months of age.
Genetic testing: Genomic DNA were extracted from peripheral blood. Whole exome sequencing (WES) or clinical exome sequencing (2742 targed genes) was performed to identify mutations. We used public databases (the dbSNP137 reported in the UCSC Genome Browser, the exome aggregation consortium, and the 1000 Genome Project) and our local database to filter the variants. Sanger sequencing was performed to confirm the candidate variants. In this cohort, patients enrolled were met one of the following four criteria: 1) patients with a previously established heterozygous pathogenic variant in KCNQ2 gene, which from both public database and internal database; 2) same amino acid change as a previously established pathogenic variant, however different nucleotide change; 3) patients with novel (both public database and internal database) heterozygous null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon) in KCNQ2 gene; 4) patients with novel and de novo heterozygous variant in the negative family history; or inherited from the affected parents. All the genetic results were interpreted by two experienced molecular genetic clinicians simultaneously.
Previous studies showed that pathogenic variants in EIEE7 cluster in four functionally important protein domains: voltage sensor, the pore, C-terminus proximal region (important for modulation by second messengers), and calmodulin-binding B helix region[11]. Common location of KCNQ2- encephalopathy missense variants reported by two recent studies cluster at voltage sensor (S4), the pore domain (from S5 to S6, including S5, S5-H5 linker, H5, H5-S6 linker, S6)[20–21]. We selected common location as key spots.
We categorized the variant location of KCNQ2: 0, non-key spots; 1, key spots.
VEEG
Scalp EEGs were recorded at 8 electrodes (F3/F4, C3/C4, T3/T4, P3/P4,) according to the international 10–20 system using the Nicolet VEEG monitor (Nicoletone, Middleton, Wisconsin, United States). Each neonate had at least one VEEG recording. Each recording lasted more than 4 hours including at least one full cycle of wakefulness and sleep.
Most patients have been treated with anti-epileptic drugs(AEDs) in other hospitals. In order to reduce the deviation caused by AEDs, for the patients who have not used anticonvulsant drugs before, we use the VEEG data of more than 3 days after taking AEDs in our hospital as early interictal VEEG, while used the first VEEG data in our hospital for the patient who have been treated with AEDs in other hospital as early interictal VEEG.
Each VEEG data was read by two experienced neurophysiologists individually. If the results from the two neurophysiologists were consistent, they would directly apply it. If not, they would discuss together to give the final results. If they can't reach an agreement in the end, the data would be discussed by the VEEG team which including more than three more experienced neurophysiologists.
According to VEEG background classification, we divide VEEG into normal, mild abnormal, moderate abnormal, and severe abnormal[22]. Moderately abnormal VEEG was divided into two groups according to whether the sleep cycle can be distinguished.
We assigned points according to the following conditions:
0 , normal
1, Mildly abnormal: mild multifocal sharp waves (Fig. 1.A);
2, Moderately abnormal: multifocal sharp waves,with sleep cycle (Fig.
1.B);
3, Moderately abnormal: multifocal sharp waves,without sleep cycle ( Fig. 1.C);
4, Markedly abnormal: burst-suppression pattern (Fig. 1.D).
Delivery type
0, Vaginal delivery; 1, Cesarean section.
Other clinical data
Clinical manifestations, including gender, birth weight, age of the onset, and gestational age, of each proband were ascertained comprehensively through physician by the review of the medical records.
outcome:
Prognosis referred to the prognosis of at least 12 months of follow-up. All patients were older than 12 months old at time of follow-up. The Bayley Scales of Infant and Toddler Development-Third Edition (BSID-III) was used to assess the developmental level of patients between 12 and 42 months of age[23]. A composite score below 70 (2SD below the mean) in the cognitive indicates moderate of profound developmental delay; a score in the cognitive below 85 (1SD below the mean) indicates mild impairment; a score in the cognitive of 100 or more indicates normal. The Wechsler preschool and primary scale of intelligence-fourth edition (WPPSI-Ⅳ) was used to assess the development level of children more than 42 months of age (normal (FSIQ (full scale IQ score)≥70); mild (FSIQ 50-69); moderate or profound (FSIQ <50)).
Prognosis was graded into four categories according to different evaluation scores:
0: normal group (any evaluation score was normal)
1: mild impairment group (any evaluation score was mild)
2: encephalopathy group (any evaluation score was moderate or profound, Including patients with frequent convulsions and underdevelopment at the time of death)
Statistics: Normally distributed data are presented as means± standard deviation (SD), skewed data are presented as medians (interquartile range).
Prognosis was regressed on predictors with ordinal logistic model. Considered variables include gender, gestational age, birth weight, age of the initial seizures, early interictal VEEG, variant location, delivery type. We modelled gestational age, birth weight, and age of the initial seizures as continuous variables, early interictal VEEG as ordinal variables, and variant location, and delivery type as categorical variables. Predictors significant at α level 0.05 in univariate analyses were entered into a multivariate model and retained at the significance level of 0.05. Statistical analysis was performed by SPSS statistics version 21.0.