High-dose methylprednisolone and tocilizumab improve survival of patients with high-risk pediatric acute necrotizing encephalopathy

1. Study population

This retrospective cohort study included pediatric patients diagnosed with ANE who were admitted to the pediatric intensive care unit (PICU) of Beijing Children’s Hospital, Capital Medical University, between January 2016 and December 2024. Diagnosis was established according to the modified Mizuguchi criteria [1], which include: (1) acute encephalopathy following a febrile illness, with rapid deterioration in consciousness and seizures; (2) cranial imaging using computed tomography or magnetic resonance imaging demonstrating multifocal brain lesions, with required bilateral symmetric involvement of the thalamic; (3) cerebrospinal fluid (CSF) analysis typically revealing elevated protein without pleocytosis; (4) elevated serum aminotransferases levels in the absence of hyperammonemia. All cases were independently validated by 2 attending pediatricians through review of clinical and neuroimaging records.

Eligible participants were aged between 29 days and 18 years diagnosed with virus-associated ANE. Patients with positive respiratory detection were included even if incomplete CSF data were available. Patients with alternative diagnoses, such as infectious encephalopathy, inherited metabolic encephalopathy, hypoxic-ischemic encephalopathy, autoimmune encephalitis, or other infection-triggered encephalopathy syndromes, were excluded.

2. Data collection

Clinical data were extracted from the electronic medical records and included: (1) demographic and baseline clinical characteristics; (2) clinical manifestations at admission; (3) neuroimaging findings and laboratory examination results, including serum cytokine and CSF routine before immunotherapy; (4) treatment regimens, including immunomodulatory therapies with MP, IVIG, PLEX, and tocilizumab, as well as supportive treatments (antiviral agents, anticonvulsants, and mechanical ventilation); (5) clinical outcomes.

3. Treatment

All patients initiated immunomodulatory therapy within 24 hours of admission. High-dose MP was defined as an initial dosage of ≥20 mg/kg/day; MP pulse therapy was administered for 3 days, followed by tapering over 7–10 days. IVIG is administered at 1g/kg/day for 2 days. Tocilizumab (<30 kg: 12 mg/kg; ≥30 kg: 8 mg/kg) was infused in 100-mL saline over ≥1 hour, with a maximum dose of 800 mg. The administration of tocilizumab and PLEX was determined based on clinical deterioration or cytokine marker levels.

4. Outcome measures

The primary outcome was 28-day postdischarge mortality, chosen over in-hospital mortality to account for potential underestimation due to treatment abandonment. Secondary outcomes included the presence of neurological sequelae during a minimum follow-up period of 6 months. Survivors underwent structured neurological evaluation either in person at outpatient clinics or via standardized telephone interviews, using the Pediatric Overall Performance Category (POPC) scale [13]. Outcomes were dichotomized into: good outcomes (POPC scores 1–2, indicating normal to mild disability) and poor outcomes (POPC scores 3–6, representing moderate disability to brain death).

5. Severity stratification: ANE-SS

The ANE severity score (ANE-SS) was used to stratify illness severity, with total scores ranging from 0 to 9 points based on the following parameters: shock at onset (3 points), brainstem involvement (2 points), age >48 months (2 points), platelet count <100,000/μL (1 point), and CSF protein >60 mg/dL (1 point) [14]. Patients were categorized into 3 risk groups: low risk (0–1 point), intermediate risk (2–4 points), and high risk (5–9 points).

6. Ethics

All patients received with high-dose MP, PLEX, and tocilizumab, written informed consent was routinely obtained from their legal guardians as part of standard clinical practice. This study was reviewed by the Ethics Committee of Beijing Children’s Hospital and classified as exempt.

7. Statistical analysis

Continuous variables were presented as mean±standard deviation for parametric data or as median with interquartile range (IQR) for nonparametric data. Categorical variables were expressed as counts and percentages. Student t test was used for parametric continuous variables, and the Mann-Whitney U test was employed for nonparametric continuous variables. Categorical variables were analyzed using the chi-square test or Fisher exact test. Statistical significance was set at P<0.05. All the statistical analyses were conducted using IBM SPSS Statistics ver. 27.0 (IBM Co., USA).

1. Patient characteristics and disease severity

The study cohort included 82 pediatric patients diagnosed with ANE. Seventeen patients, who died within 24 hours of admission with a Glasgow Coma Scale (GCS) of 2T, were excluded from subsequent analysis of immunotherapy efficacy. The remaining 65 patients were included in the efficacy analysis (Fig. 1). The median age of the cohort was 4.8 years (IQR, 2.8–7.7 years), with a nearly equal sex distribution (male-to-female ratio, 1:1.02). Influenza A and SARS-CoV-2 were identified as the predominant infectious pathogens.

Fig. 1.

Flow diagram of screening and enrollment of patients with pediatric acute necrotizing encephalopathy (ANE). A total of 105 children with suspected ANE were admitted to our hospital during the study period. Of them, 21 did not undergo imaging due to critical status or the absence of typical imaging findings, while 2 cases of suspected genetic metabolic encephalopathy were excluded. Finally, 82 children presented with clinically and radiologically confirmed ANE. Among them, 17 had a Glasgow Coma Score of 2T upon admission, exhibited dilated and fixed pupils, and died early during treatment or within 24-hour posttreatment. Consequently, 65 patients were included in the immunotherapy efficacy analysis. PICU, pediatric intensive care unit; ANESS, ANE severity score.

At admission, 46.2% (30 of 65) were classified as low- to medium-risk (ANE-SS<5), while 53.8% (35 of 65) were categorized as high-risk (ANE-SS≥5). Patients in the high-risk group were significantly older (median age: 5.0 years vs. 3.0 years, P=0.020) and exhibited more severe clinical manifestations compared to those in low-to-medium-risk group (ANE-SS<5). Specifically, high-risk patients presented with deeper coma (GCS≤5: 76.9% vs. 33.3%, P<0.001), a higher incidence of shock (69.2% vs. 6.7%, P<0.001), and more frequent brainstem involvement (84.6% vs. 26.7%, P<0.001) (Table 1). Additionally, levels of systemic inflammatory biomarkers, including procalcitonin, IL-6, IL-8, IL-10, as well as organ injury markers such as aspartate aminotransferase, lactate dehydrogenase, D-dimer were significantly elevated in high-risk group (all P<0.05) (Supplementary Table 1).

Table 1.

Demographic and clinical characteristics by ANE-SS status

2. Trends in immunotherapy utilization and mortality (2016–2024)

An analysis of treatment trends from 2016 to 2024 revealed significant changes in immunotherapy strategies and associated mortality outcomes (Table 2). The use of first-line immunomodulatory therapy, defined as high-dose MP (≥20 mg/kg/day) combined with IVIG, increased steadily from 33.3% during 2016–2018 to 100% by 2021. Notably, the use of 30-mg/kg/day MP regimen rose sharply, from 15.4% in 2022 to 62.5% by 2024.

Table 2.

Composition of immunotherapy for ANE in pediatric intensive care unit: 2016–2024

The application of PLEX gradually increased over the study period, ranging from 20.0% to 52.4%, with the exception of 2021, when it was not utilized. Tocilizumab, first introduced in 2023, was rapidly incorporated into clinical practice, with its use rising from 38.1% to 62.5% within 1 year.

Mortality rates range from 40% and 75%, peaking in 2019 when the use of immunotherapy was lowest. In subsequent years, mortality gradually declined with the increasing use of intensive immunotherapy regimens, particularly those combining 30-mg/kg/day MP, PLEX, and tocilizumab, reaching 38.9% in 2022, 36.8% in 2023, and 16.7% in 2024. Notably, indicators of disease severity, including ANE-SS≥5, brainstem involvement, and mechanical ventilation, remained relatively consistent over the study period.

3. Clinical outcomes of immunomodulation therapy by disease severity

All enrolled patients (n=65) in this study received first-line therapy with MP plus IVIG within 24 hours of admission. Based on this standardized combination, we further investigated the effect of different MP dosages and the additional use of second-line therapies, including tocilizumab and PLEX, across different disease severity.

1) Impact of MP dosage

In the low-risk group (ANE-SS<5), high-dose MP regimens demonstrated superior survival benefits. Treatment with 30-mg/kg/day MP achieved 100% survival rate, while 20 mg/kg/day resulted in 90.9% survival, compared to only 40% survival with lower doses (10–20 mg/kg/day) (P=0.029) (Table 3). In the high-risk patients (ANE-SS≥5), survival was clearly dose-dependent: 75% with 30 mg/kg/day, 25% with 20 mg/kg/day, and only 14.3% with 10–20 mg/kg/day (P=0.024). These associations remained statistically significant in univariate Cox regression analysis (P<0.001), but lost significance in multivariate model (P=0.321), likely due to limited sample sizes in some subgroups (Table 4).

Table 3.

Clinical outcomes of children with ANE at 28 days postdischarge by clinical phenotype and immunotherapy regimen

Table 4.

Univariate and stepwise multivariate Cox hazard analysis of mortality by immunotherapy regimen at 28 days postdischarge

4. Comparative efficacy of tocilizumab in high-risk patients with ANE

Among high-risk patients (ANE-SS≥5), those who received combination therapy with tocilizumab (n=11), had a better overall outcome compared to those who did not receive tocilizumab (n=8, Table 5).

Table 5.

Clinical outcomes of tocilizumab plus corticosteroid immunomodulatory therapy in patients with high-risk ANE (ANE-SS ≥5): 28-day mortality and 6-month sequelae

Survival analysis revealed that in the tocilizumab group, survival rates were 80% for patients receiving 20-mg/kg/day MP and 83.3% for those receiving 30 mg/kg/day. In contrast, survival rates in the non-tocilizumab group were significantly lower, at 0% for the 20-mg/kg/day subgroup and 50% for the 30-mg/kg/day subgroup.

We further investigated neurological recovery at the 6-month follow-up after discharge. Among patients receiving combined tocilizumab with 30 mg/kg/day of MP, 60.0 % had a good outcome (POPC score≤2), compared to 25% in the 20-mg/kg/day subgroup. In the nontocilizumab group, only one child survived in the 30-mg/kg/day subgroup. Although this observed difference was not statistically significant, likely due to the small sample size (Table 5). These results suggest that tocilizumab combined with high-dose MP (30 mg/kg/day) may enhance both survival and possibly long-term neurological recovery in high-risk patients with ANE.

5. Treatment-related adverse events

Three ANE patients treated with high-dose MP (1 at 20 mg/kg/day and 2 at 30 mg/kg/day) developed stress hyperglycemia. Two ANE patients with concomitant Mycoplasma pneumoniae infection receiving combined therapy (20-mg/kg/day MP, PLEX, and tocilizumab) developed cerebral arterial thrombosis, likely due to infection-induced hypercoagulability rather than procedural complications.