Prevalence & clinical outcome of autoimmune encephalitis versus viral encephalitis in children with acute encephalitis syndrome: A prospective observational study

Material & Methods

This was a prospective observational multisite hospital-based study carried out in children with acute encephalitic syndrome. These children were between the age group of six months to 14 yr and admitted to five large tertiary care hospitals in two States namely, Odisha and Uttar Pradesh (UP) in India documenting high burden of AES cases. Children with AES as defined by the National Vector born disease control programme (NVBDCP, India; http://www.nvbdcp.gov.in/Doc/AES%20guidelines.pdf) were included in the study. Children with an alternate diagnosis like bacterial or tubercular encephalitis, cerebral malaria, malignancy or metabolic encephalopathy were excluded from the study.

Children with AES who were admitted to the designated tertiary study hospitals in the two States were consecutively included in the study from October 2019 to October 2022. Following admission, comprehensive personal history and clinical examination details were documented and subjected to laboratory analysis in accordance with the hospital’s AES management protocol. Those meeting the specified inclusion and exclusion criteria were enrolled with parental consent. A 2 ml CSF sample and a 2 ml blood sample in a red-top vacutainer were collected at the participating study centres and securely transported to the nodal centres [All India Institute of Medical Sciences (AIIMS), Bhubaneswar/ King George’s Medical University (KGMU), Lucknow] for analysis with appropriate precautions. Comprehensive baseline, clinical, and laboratory assessments were documented. The study was conducted following GCP guidelines with approval from the institute ethics committee (IEC) from both the Institutes.

The testing of viral markers and autoantibodies was done in CSF and serum samples. The investigation of viral markers encompassed viruses, such as HSV-1, HSV-2, VZV, Japanese encephalitis virus, measles, enteroviruses, EBV, cytomegalovirus, parvovirus B19, enterovirus, dengue and HHV-6. Seromolecular investigation included detection of antibodies and PCR assays for DNA/RNA depending on the specific virus type¹¹. We also estimated the level of IgM antibodies for the viruses mentioned above. However, HSV IgG was only done for those children who were positive for anti-NMDAR antibody.

Autoantibodies were screened for NMDA-R, AMPA1, AMPA2, CASPR, LGI-1, and GABAB receptor was conducted on (B1, B2). Testing of autoantibodies was done by indirect immunofluorescence (IIF) assay. These assays transfected cell lines for qualitative determination of human antibodies against glutamate and VGKC receptors in serum/CSF by using Cell Based Assay (CBA) using a commercially available kit (Euroimmune, Germany). The sensitivity and specificity of CBAs using serum or CSF have been well established^12-15. Details of treatment received during hospital stay were recorded in a predesigned proforma. Clinical status at discharge was documented and subsequently, the patients were followed on an outpatient basis at one, three and six months from the date of onset of illness.

Results

Throughout the study period, a total of 867 children with autoimmune and viral aetiology of AES were enrolled at the two designated study centres, with 511 children from Odisha and 356 children from UP. Of these, 37 children (4.2%) received a diagnosis of AIE, and all of them tested positive for anti-NMDA receptor antibodies. While we conducted screenings for other antibodies, including AMPA1, AMPA2, CASPR, LGI-1, and GABAB receptor (GABAB1, B2), none of these antibodies were detected. It’s worth noting that the Odisha site had 32 children who tested positive for anti-NMDA receptor antibodies, whereas the Uttar Pradesh site had only five such cases (Table I).

Within the AES cases, 409 (47.1%) had evidence of a viral infection, of which 254 (29.2%) tested positive for HSV IgM antibodies. In children with AIE, 30 (81%) displayed the presence of anti-NMDA autoantibodies in their CSF, while 6 (16.3%) had them in their serum, and 1 (2.7%) exhibited in both. In instances of non-autoimmune encephalitis, HSV emerged as the most commonly associated virus. HSV IgM antibodies were detected in nearly 254 (29.2%) children in the overall study sample. Among children with AIE, a viral trigger was present in 25 (67.5%) cases. Within this group, eight tested positive for HSV IgM antibodies, 12 for HSV IgG antibodies, and one child had both antibodies identified. Additionally, two children were positive for Japanese encephalitis IgM and parvovirus IgM antibodies within their CSF. We conducted a comparative analysis of the clinical characteristics and outcomes of children with HSV-associated encephalitis and those with AIE (Table II).

Discussion

Autoimmune encephalitis is one of the significant causes of AES in the paediatric age group which has been detected increasingly over the last two decades⁵. Anti-NMDR encephalitis is the most common variant of AIE in the paediatric age group. Post-infectious AIE has been reported in the paediatric age groups in a few series. However, the reports are largely restricted to Western countries^16,17.

In the current study, of the children who had AES, 37 (4.2%) children tested positive for anti-NMDAR antibodies. The estimated prevalence of AIE was 13.7 per million population as reported by Dubey et al¹⁸ in a retrospective study including both children and adults where the prevalence of specific neuronal antibodies against NMDA receptor was reported to be 0.6/100,000¹⁸. A retrospective study from Hong Kong by Peery et al¹⁹ reported an estimated prevalence of 2.2 per million among children. Among seropositive AIE in children, NMDA encephalitis constitutes almost four per cent which is the most common. There was not much data available regarding the exact prevalence of AIE in children and adults from India. In a systematic review of cases of NMDA encephalitis, the majority were of age group below 18 yr (81.5%) as compared to adults. In the current study, we observed a similar prevalence of 4.2 per cent in our cohort²⁰.

In this study, 81 per cent of CSI samples tested positive for anti-NMDAR antibodies whereas in serum it was 16.3 per cent. Similar results of higher sensitivity for CSF neuronal antibodies ranging from 98.5-100 per cent have been reported in multiple studies^14,21. Chowdhury et al²⁰ reported a CSF positivity of 73 per cent in children in a systematic review from India²⁰.

HSV infection is known to trigger anti-NMDAR encephalitis both in children and adults. Twenty one (56.7%) children with anti-NMDAR encephalitis had either HSV IgM or IgG positive whereas two had antibodies positive for JE and Parvo B19. HSV IgM antibodies were detected in 29.2 per cent of our children with AES which may be considered a possible herpes infection as we could not detect any HSV PCR positivity in CSF which is confirmatory. This may be due to delayed presentation to tertiary care centres, late sample collection and processing, and prior antiviral therapy. JE is the predominant aetiology for AES in different parts of India. In this study, JE was positive in 27 (3.1%) children with AES. HSV encephalitis is one of the common viral aetiologies for AES which is reported in almost 11-14 per cent of children with AES in various studies across the globe. Almost 30 per cent of the children with HSV encephalitis had anti-NMDAR encephalitis as reported in a retrospective study by Pruss et al²². Previous epidemiological studies mostly focused on confirmatory cases of HSV, hence using PCR as a modality of investigation resulting in a lower rate of positivity. The current study not only focused on the confirmatory cases but also looked at exposure to HSV which is considered as one of the triggering factors for AIE in literature, mostly in retrospective studies. As we had screened for antibodies (both IgM and IgG) against HSV, our results showed a higher prevalence of positivity suggesting possible infection. This may be due to the endemicity of the virus attributed to the trigger. This finding further substantiates the proposed hypothesis of HSV-triggered AIE which may need further confirmation in future studies. Other than HSV, parvovirus, JE virus, HSV II, and HHV7 were also positive in our study sample which is similar to previous reports indicating association with anti-NMDAR encephalitis²³.

There was a significant delay in diagnosis of AIE from the onset of illness in this study as compared to viral aetiologies. This may be due to late suspicion or delayed presentation to tertiary care centres. Most of these children were managed initially as AES with possible viral or bacterial origin and referred when there was poor response to treatment or recurrence of symptoms. Armangue et al²⁴ reported the median interval between HSV encephalitis and AIE to be 26 (24-32) days which is similar to the findings in this study²⁴.

Children with anti-NMDAR encephalitis presented with seizure as one of the common clinical features which is observed in almost 80 per cent of the children in this study. Similar results were reported by Chowdhury et al²⁰ in the retrospective analysis of cases from India among the paediatric age group²⁰. Apart from seizure and altered sensorium, movement disorders, and behavioural problems were the other features usually associated with AIE. In the current study, movement disorder was only observed in 27 per cent of children which is lower than what is reported in a previous study²⁰. This may be due to early initiation of treatment in most cases with AIE. There was not much difference in clinical features of anti-NMDAR encephalitis and HSV encephalitis except for behavioural problems and movement disorder which were more common in earlier group²⁵. There was not much difference in the CSF examination except for higher protein in HSV encephalitis. CSF pleocytosis with or without elevated protein has been reported in few cases of anti-NMDAR encephalitis but it has not been observed uniformly as in ours to contribute to diagnosis²⁶.

Though there are no standard recommendations for treatment for AIE, immunosuppressants remain the cornerstone of management including anti-NMDAR encephalitis. Steroids and IVIg are considered to be the first line of therapy. In our cohort almost 31 (84%) children received IVMP followed by oral steroid. A similar treatment protocol has been reported by Titulaer et al²⁷ where 87 per cent of the children had received steroids.

Following treatment, 32 (86.4%) of the children in this study were discharged from the hospital. Eleven (29.7%) had complete recovery at discharge while 21 (56.7%) had sequelae in the form of poor sensorium or abnormal movement. However, at six-month follow up, 22 (59.4%) children had complete recovery without recurrence whereas 10 (27%) had neurological sequelae (altered sensorium or behavioural abnormality). Chowdhury et al²⁰ reported similar outcomes in their retrospective analysis of paediatric cases where 56 per cent had improved whereas 40 per cent had sequelae. However, these are the retrospective data where the duration of follow up and recurrence were the limitation and have not been recorded. In a prospective cohort study by Titulaer et al²⁷, the treatment response was 53, 80 and 90 per cent at four, 12 and 24 months follow up, respectively. However, this study includes both children as well as adults. Separate treatment response rates have not been mentioned in the study. Complete recovery in children with possible HSV encephalitis was 64.6 per cent at discharge whereas neurological sequelae were there in 19.5 per cent of children. A meta-analysis of nine studies reported a prevalence of neurological sequelae up to 50.7 (39.2-62.2%) per cent. However, all these studies included children with confirmed HSV encephalitis by PCR evaluation only²⁸.

The major strength of the present study lies in the fact that it is a prospective study where children with anti-NMDAR encephalitis were evaluated for associated CNS viral infection which may trigger AIE. Children were followed up for six months to look for treatment outcomes and relapse. The limitation of the study is that late presentation of cases and investigation thereof would have resulted in missing some early warning signs if so to suspect AIE early as well as picking up viral detection in CSF by PCR assay.

Overall, anti-NMDAR encephalitis is one of the major causes of AES in children. Post-viral triggered autoimmune encephalitis is more common than what is reported and usually, these children respond to immunosuppressants if treated early. Though clinically it is difficult to differentiate from common mimickers like viral encephalitis, children presenting with behavioural problems and movement disorders should be evaluated for AIE. Early diagnosis and treatment with immunosuppressants result in better neurological outcomes. Children with viral encephalitis should be treated adequately and followed up regularly for early diagnosis of autoimmune encephalitis.