Determination of neurobiomarkers in children with COVID-19

Abstract

The coronavirus infection (COVID-19) pandemic has become a real challenge for medical professionals and researchers, given the frequency of deaths and complications. Damage to the nervous system is one of the most common. The frequency of neurological complications due to COVID-19, according to the literature, is up to 82% and is characterized in the acute period by delirium and seizures (34%), fatigue (32%), myalgia (20%), impaired smell or taste, and headache (13 %). Guillain-Barre syndrome is also registered in 10% of cases and stroke in 2% of cases [1]. Complications can occur with a frequency of more than 33% within six months after COVID-19. According to a published study, in patients treated in the intensive care unit, the frequency of damage to the nervous system was 46% and was characterized by ischemic stroke, dementia, intracranial hemorrhages, parkinsonism, psychotic and anxiety disorders [2]. For the purpose of in-depth diagnosis of neurological manifestations, neuron-specific proteins are widely used as markers of damage to the nervous system. Disruption of the blood-brain barrier (BB) in response to brain injury can increase the concentration of specific molecules in the circulation, and the assessment of the concentration of these molecules in the serum can contribute to the diagnosis of the lesion. Astrocyte-specific proteins such as S100β are found at high levels in the brain, and their presence in the blood may indicate a loss of blood-brain barrier function or the presence of trauma. Several neuron-specific proteins, including neuron-specific enolase (NSE), are released from damaged neurons and enter the bloodstream if the integrity of the blood-brain barrier is damaged. The presence of serum albumin, which is detected at high levels in the blood, in the cerebrospinal fluid also indicates damage to the blood-brain barrier [3]. In a study by Antonio Aceti et al. (2020), serum S100b concentration correlated with disease severity as indicated by clinical and laboratory parameters. Researchers have shown correlations of S100b with indicators of distress, including non-neuronal indicators such as ALT, D-dimer, and platelets [4]. In other studies, scientists also suggest the marker S100b and NSE as reliable predictors of clinical severity. The serum S100b level showed significantly higher mean values in the cohorts with severe COVID-19 than in the group with mild symptoms. In a similar pilot study, researchers found an increased level of NSE in blood serum in a group of patients with a severe course than in a group without complications (p = 0.034) [5,6]. The purpose of the study was to determine the level of S100b and NSE proteins in children with COVID-19 and to investigate the correlation of indicators with the severity of COVID-19. Research materials and methods. We conducted a retrospective cohort, observational study. We examined 88 children aged 1 month to 17 years with laboratory-confirmed COVID-19 who underwent inpatient treatment at the Kyiv City Children's Clinical Infectious Disease Hospital in 2021-2022. Children were divided according to the course of the disease into two groups - the control group, which had a complicated course of COVID-19, and the main one without complications. We also made a division by age groups - from birth to 12 months, from 1 to 6 years, from 6 to 10 years and from 10 to 17 years. The main laboratory indicators, data of anamnesis and objective examination were taken into account. During the complex routine examination of the patients during the first day of their stay in the hospital, blood serum of the patients was collected for the purpose of its further examination for the level of neurobiomarkers s100b and NSE by enzyme immunoassay. Fujirebio's "CanAg S100 EIA kit" and "CanAg NSE EIA kit" with a working measurement range of 1--3500 ng/L for s100 and 1-150 μg/L for NSE were used. The study was approved by the bioethical committee of the hospital and informed consent was obtained from the patients. For statistical processing of the results, we used the biostatistical package Statistical software EZR v. 1.54 and performed interval estimation of the distribution, multiple comparisons, and calculated the Pearson and Dunn correlation coefficient.