No Neurochemical Evidence of Neuronal Injury or Glial Activation in Children with Paediatric Acute-onset Neuropsychiatric Syndrome

Background Paediatric Acute-onset Neuropsychiatric Syndrome (PANS) is characterized by an acute onset of obsessive compulsive disorder, combined with at least two other neuropsychiatric symptoms with similarly acute onset. Diagnostic criteria also require that no specic medical aetiology is identied. Although there are no veried aetiological biomarkers, PANS is assumed to be a neuroinammatory disorder with a possible autoimmune aetiology. Neurochemical markers such as neurolament light (NfL, a neuronal injury marker) and glial brillary acidic protein (GFAP, an astrocytic activation marker) have not been published for this patient group. Methods Blood samples from 17 children meeting diagnostic criteria for PANS, after assessment at the Child Neuropsychiatry Centre, Sahlgrenska University Hospital, Gothenburg, Sweden, were analysed for serum concentrations of NfL and GFAP. Ten age-matched children without any neurological or psychiatric disorder served as a comparison group. Results No difference was found in mean NfL and mean GFAP serum concentrations between children with PANS and controls. NfL and GFAP concentrations did not correlate with disease duration between PANS onset and time of blood tests.


Introduction
The acronym PANS (Paediatric Acute-onset Neuropsychiatric Syndrome) is clinically de ned by an abrupt, dramatic onset of obsessive-compulsive disorder or severely restricted food intake combined with at least two out of seven other similarly severe symptoms with acute onset; anxiety, emotional lability and/or depression, irritability/aggression and/or severely oppositional behaviours, behavioural (developmental) regression, deterioration in school performance (related to attention-de cit/hyperactivity disorder/ADHD-like symptoms, memory de cits, cognitive changes), sensory or motor abnormalities and somatic signs and symptoms, including sleep disturbances, enuresis, or urinary frequency. In addition, the research diagnostic criteria require that symptoms are not better explained by a known neurological or other medical disorder [1].
The term PANS was a modi cation of a preceding diagnostic term, PANDAS (Paediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcal infections), a similar clinical condition but implicating a temporal association with group A streptococcal infection and an ensuing autoimmune reaction [2]. By the term PANS aetiological associations were eliminated and hitherto no de ned medical causes or reliable biomarkers have been identi ed, although infections and autoimmune reactions have been suggested [3]. Infections, mostly upper respiratory, are commonly reported symptoms preceding the onset of PANS [4]. In addition, autoimmune reactions, known to be associated with some neuropsychiatric disorders, have therefore also been assumed to have a connection with PANS [4]. Gilbert et al [5] emphasized the need to de ne better the clinical manifestations, laboratory and neuroimaging ndings, therapeutic responses, and clinical course in these disorders and that, to date, there is no signi cant evidence supporting any particular aetiologies for PANDAS and PANS.
Neuro lament light chain (NfL) protein, is a component of the neuron cytoskeleton and is released into the cerebrospinal uid (CSF) and blood when neurons are injured, providing a neurodegenerative biomarker of neuronal damage [6]. An important aspect is that the NfL proteins indicate neuronal injury independent of causal pathways [6]. NfL levels in cerebrospinal uid and in serum have been found to correlate signi cantly in paediatric patients with acquired demyelinating syndromes [7].
Glial brillary acidic protein (GFAP) is an astrocytic cytoskeletal protein, the expression and secretion of which is upregulated when astrocytes are activated [8]. In children with symptoms consistent with PANS, speci c causes, such as N-methyl-D-aspartate receptor (NMDAR) encephalitis, the most common form of autoimmune encephalitis, needs to be considered [9]. Several studies have demonstrated that not only neurons, but also astrocytes express functional NMDA receptors and that the pathogenic antibodies against the NMDAR, causing the anti-NMDAR encephalitis, affect not only the neuronal receptors, but also the NMDAR on astrocytes [10].
In the 1990s, GFAP was analysed in cerebrospinal uid (CSF) of children and adolescents with autism, of whom 70% also had intellectual disability (ID) and 21% had epilepsy, and results were contrasted with ndings in similarly aged cases without autism or other neuropsychiatric disorders. GFAP levels in children with autism were found to be at a level almost three times higher than in the group of children without autism. In one aetiological subgroup of children with autism, consisting of three children with a basal ganglia disorder, the highest GFAP levels were found [11]. Also a more recent study, using serum levels of GFAP, demonstrated elevated levels in children with autism compared to controls [12].
To our knowledge it has not been systematically analysed whether children suffering from PANS would have detectable signs of neuronal/axonal or astrocyte cell damage. Since NfL and GFAP can now be analysed in serum, sampling in children is facilitated. Moreover, for both these biomarkers, serum concentrations have been shown to re ect the levels in the CSF [13,14,15]. The aim of the present study was therefore to analyse concentrations of NfL and GFAP in serum of children meeting criteria for PANS.

Study participants
Our initial cohort of children with PANS consisted of 23 children, referred between 2015 and 2017 and assessed at the Child Neuropsychiatry Centre (CNC)/Gillberg Neuropsychiatry Centre in Gothenburg.
During the study period, CNC was the referral center for children with suspected PANS in Gothenburg and the southwestern region of Sweden and these 23 children were deemed to meet clinical criteria for PANS. Data from the clinical assessments of these children have been reported previously [16].
For the present study 11 of the 23 children (7 boys), mean age 13 years, gave consent to have blood test taken for analysis of NfL and GFAP. In addition, 6 children (3 boys), mean age 9.6 years, later included in the cohort, were able to participate in blood sampling. Thus, 17 children could be analysed for serum NfL and GFAP.
The time from onset of PANS symptoms to the time point of blood sampling varied between 1 month and 5 years. At the time of assessment, however, all children had moderate to severe impairments that markedly in uenced their school attendance and parental and family situation.
Four of the 17 patients had autism, no child in the study group had diagnosed intellectual disability. In addition to OCD and at least two other neuropsychiatric symptoms, four children had psychotic symptoms (delusions, hallucinations).
The comparison group included 10 children (5 boys, 5 girls), mean age 9.4 years, who did not have any neurological or psychiatric disorders. Their blood samples were taken in connection with investigation for non-neurological/non-psychiatric disorders, mainly allergic diseases, con rmed in their paediatric records, at a paediatric outpatient clinic in Gothenburg.

Serum analyses
NfL and GFAP were measured using ultrasensitive Single molecule array (Simoa) technology, using commercially available kits, according to instructions from the manufacturer (Quanterix, Billerica, MA).
The analyses were performed in one round of experiments, using one batch of reagents by board-certi ed laboratory technicians who were blinded to clinical data. For a quality control (QC) sample with an NfL concentration of 15.0 pg/mL, the repeatability was 7.2%. For a QC sample with an NfL concentration of 47.3 pg/mL, the repeatability was 1.7%. For a QC sample with a GFAP concentration of 131 pg/mL, the repeatability was 6.1%. For a QC sample with a GFAP concentration of 365 pg/mL, the repeatability was 4.3%.

Statistical analysis
We conducted two-tailed t-tests in order to test mean differences between the PANS group and the comparison group. Since sample sizes were small, we cross-checked the t-test results with the nonparametric Mann-Whitney U test, which yielded very similar p-values. We calculated Spearman correlations between the time point of PANS onset and the time point of blood sampling.

Discussion
This study was based on 17 children referred to a child and adolescent neuropsychiatric clinic, at the time of the study the main referral center for children with suspected PANS in Gothenburg and the southwestern part of Sweden. All included children met the research criteria for PANS 1 and had had an abrupt onset of OCD and at least two associated, severe neuropsychiatric symptoms. Co-occurring psychotic symptoms (delusions, hallucinations) were reported in 23.5% of the children in the PANS group, which is in line with other reports [3,17]. A basal ganglia involvement is presumed in PANDAS and there are studies implicating immune-mediated basal ganglia and striatal involvement also in PANS aetiology and in the pathogenesis of psychotic disorders [18]. However, as discussed by Gilbert et al [5] studies of PANDAS and PANS are inconclusive regarding pathophysiology, diagnostic markers and treatment. The need for cross-institutional collaboration and well-conducted investigations of underlying biological mechanisms in children with PANDAS/PANS has also been emphasized, given the severe nature of the symptoms and considerable effects on family and child functioning [19].
In the present study, we could not demonstrate increased levels of NfL and GFAP in serum of children with PANS, compared to a control group of children without symptoms of any neurological or psychiatric disorder. Levels of NfL and GFAP did not correlate with the elapsed time between onset of PANS and blood tests. In summary, our study did not reveal any signs of neuronal/axonal or astrocyte damage in children with PANS.
The four children with autism in our PANS group had GFAP concentrations that did not differ signi cantly from GFAP levels in the children with PANS without autism. However, the small study groups with only four children diagnosed with autism do not allow any conclusion to be drawn. These results differ from previous studies, in which elevated levels of GFAP were found in CSF [11] and serum [12] in children with autism compared to controls. However, none of the children with autism and PANS in our study had intellectual disability, while this was present in 70% of the children with autism reported in the 1990s [11].
Clear limitations of this study are the small study groups and the large differences in time periods between onset of PANS symptoms and blood sampling. However, all children had moderate to severe symptoms and impairment at blood sampling. A strength of the study is that all patients were clinically well de ned and assessed by the same paediatricians, and that it was possible to use analyses of serum, since previous research has demonstrated correlation between these two biomarkers in cerebrospinal uid and serum.

Conclusions
No indications of neuronal injury and astrocyte activation in PANS were found in this study. The study group was small, and even if ndings may be reassuring for parents and patients, they should be interpreted with caution and veri ed in larger cohort and possibly with other markers in both serum and CSF.

Declarations
Ethics approval and consent to participate The study was approved by the Gothenburg ethical review board. All parents and participants provided informed consent after receiving oral and written information of the study.
Funding Swedish Brain Foundation. The funder of the study had no role in study design, data collection, data analysis, data interpretation, writing of the report, or the decision to submit for publication.

Authors' contributions
KB and HZ carried out the laboratory data analyses. All authors contributed to study design, data interpretation, drafting and revision of the manuscript. The corresponding author had full access to all the data in the study and had nal responsibility for the decision to submit for publication.

Consent for publication
All authors gave nal approval of the version to be submitted for publication.