Influenza is a common viral infection in children and usually presents as a self-limited respiratory illness. Neurological complications associated with influenza have been increasingly reported over the past several decades. In recent hospitalized cohorts [1], seizures, including febrile, complex febrile, and acute symptomatic seizures, were the most common neurological manifestations, and many affected patients recovered rapidly. A French study similarly reported that most neurological presentations were transient and that febrile seizures or benign encephalopathy often resolved within a few days [2]. However, a subset of patients progressed to more severe phenotypes, including mild encephalitis/encephalopathy with a reversible splenial lesion, brainstem leukoencephalopathy, and acute necrotizing encephalopathy (ANE), often with subsequent neurological sequelae. A Thai cohort further reported that severe encephalopathy was associated with characteristic electroencephalography (EEG) patterns, including diffuse slowing and low-voltage background activity [3]. The most devastating phenotype, influenza-associated ANE, was recently described in a United States (U.S.) multicenter series conducted from 2023 to 2025, which reported a mortality rate of 27%, a median time to death of 3 days, and moderate-to-severe disability in 63% of survivors at 90-day follow-up [4].

Large cohort studies from Korea, the U.S., Thailand, and Europe have reported neurological complications in approximately 8% to 17% of hospitalized pediatric patients with influenza (Table 1) [1-6]. Although most children recover without long-term deficits, mortality rates ranging from 1% to more than 20% have been reported, particularly for severe phenotypes such as ANE [4,7]. Several studies have identified risk factors for influenza-associated neurological complications in children. In a U.S. study, neurological comorbidities increased the risk by nearly five-fold (adjusted odds ratio [aOR], 4.6) [1]. A Thai cohort demonstrated increased risk among children with prior febrile seizures (aOR, 20.3), epilepsy (aOR, 3.6), and other neurological conditions [3]. A Korean multicenter study similarly reported underlying neurological disease as an independent predictor (aOR, 5.41) [5]. Young age has also been associated with an increased risk of neurological complications [1,5]. In addition, lack of seasonal influenza vaccination has been independently associated with neurological complications [1], and vaccination rates were notably low in ANE cases (16%) [4]. Nevertheless, the clinical course of influenza-associated encephalitis/encephalopathy (IAE) remains unpredictable, and some previously healthy children deteriorate rapidly despite mild initial symptoms.

Early antiviral therapy and immunomodulatory treatment are commonly used in moderate-to-severe cases. However, the optimal timing of these interventions and their influence on outcomes remain incompletely defined. In a U.S. pediatric cohort, early antiviral treatment was associated with fewer neurological complications, but definitive conclusions were limited by the study design and sample size [1]. In addition, although immunomodulatory therapies such as corticosteroids and intravenous immunoglobulin have been reported in case series and cohort studies, evidence supporting their efficacy remains limited [2,3].

In this case series, we investigated the clinical spectrum, treatment timing, and risk factors associated with mortality and unfavorable neurological outcomes in children with IAE at a tertiary pediatric center, with particular attention to rapidly progressive fatal cases.

This retrospective study demonstrated marked clinical heterogeneity in pediatric IAE, ranging from complete recovery to hyperacute, fulminant deterioration within hours. Among the 17 children, most recovered fully; however, 17.6% died and 23.5% developed sequelae. Notably, the two fatal cases deteriorated abruptly within hours of presentation, underscoring the potentially fulminant nature of the disease, even in previously healthy children. These findings are consistent with prior cohort studies reporting generally favorable outcomes but occasional catastrophic progression [1-7]. The mortality rate observed in this cohort was comparable to that reported in Asian cohorts with severe disease and in recent U.S. studies focused on ANE [4], highlighting that fatal outcomes remain possible even in modern tertiary centers.

One notable finding was the difference in influenza vaccination status across outcome groups. In particular, none of the patients in the mortality group had received influenza vaccination for at least 2 years before admission, whereas vaccination rates were higher in the sequelae and recovery groups. However, given the small-sample size and limited number of events, these findings should be interpreted as exploratory and do not establish a causal relationship. In the multivariable analyses, although statistical significance was not achieved, the direction of the estimates was consistent with a potential association between prior vaccination and more favorable neurological outcomes. Previous studies have reported that lack of influenza vaccination is associated with an increased risk of neurological complications in children. Vaccination may attenuate viral load and modulate systemic inflammatory responses, potentially reducing the risk of excessive neuroinflammatory cascades that contribute to encephalitic injury. This interpretation is supported by previous epidemiological studies. In a large U.S. pediatric cohort, lack of seasonal influenza vaccination was independently associated with neurological complications (aOR, 1.6) [1]. Similarly, in a recent U.S. multicenter series of influenza-associated ANE, only 16% of patients had received age-appropriate seasonal influenza vaccination, highlighting low vaccine uptake among the most severe cases [4]. National surveillance data from the U.S. have also demonstrated that a substantial proportion of children with IAE or fatal influenza are unvaccinated, reinforcing the potential protective role of vaccination against severe neurological disease [4,9,10]. Moreover, vaccine-effectiveness studies have consistently shown that influenza vaccination reduces influenza-related hospitalizations and intensive care unit admissions in children, providing biological plausibility that vaccination may indirectly decrease the risk of catastrophic neurological outcomes by mitigating overall disease severity [11,12]. Japanese ecological analyses have suggested that discontinuation of influenza vaccination programs in schoolchildren was temporally associated with increased influenza-related mortality and severe IAE in children, supporting the broader public health impact of vaccination on the most severe pediatric outcomes [13]. Taken together, these findings suggest that influenza vaccination may be associated with differences in clinical severity, although further studies with larger sample sizes are needed to clarify this relationship and determine whether a protective effect exists.

Treatment timing is clinically important in pediatric IAE. The first two fatal cases died before antiviral or immunomodulatory therapy could be initiated. After these early catastrophic cases, institutional practice shifted toward prompt treatment initiation in children with suspected IAE, even while diagnostic evaluations were ongoing. In subsequent patients, early therapy was associated with a protective trend in regression analyses. These findings support the possibility that IAE may involve a narrow therapeutic window during which early intervention could mitigate secondary immune-mediated brain injury. The findings remained consistent in a sensitivity analysis that excluded patients with hyperacute fatal progression before treatment initiation, supporting the robustness of the observed trends despite the limited sample size. Because antiviral and immunomodulatory therapies were initiated concurrently in this cohort, the independent effect of immunomodulatory therapy could not be determined.

The pathophysiology of IAE involves not only direct viral neuroinvasion but also an exaggerated systemic inflammatory response. Evidence suggests that dysregulated cytokine release, often described as a ‘cytokine storm,’ plays a central role in severe influenza-associated neurological injury [14-16]. Elevated levels of pro-inflammatory cytokines, including interleukin-6 and tumor necrosis factor-α, have been reported in both serum and CSF from affected patients and are thought to disrupt the blood–brain barrier, promote cerebral edema, and trigger secondary neuronal injury [15,16]. Recent reports have suggested that tocilizumab, an interleukin-6 receptor antagonist, may be beneficial in severe cases such as ANE; however, none of the patients in this cohort received this treatment, and its clinical impact could not be evaluated in this study. In fulminant phenotypes such as ANE, symmetric deep gray matter lesions, particularly those involving the thalamus, are thought to reflect cytokine-mediated vascular and metabolic dysfunction rather than direct viral invasion [14,17]. Therefore, the hyperacute deterioration observed in some fatal cases may represent an overwhelming neuroinflammatory cascade occurring within a narrow therapeutic window. In the two fatal cases in this cohort, laboratory findings suggestive of multi-organ involvement, including elevated lactate and cardiac enzyme levels, were observed and may reflect rapid progression to systemic failure in fulminant IAE. This mechanism further supports the rationale for early antiviral therapy combined with timely immunomodulatory treatment in patients with suspected severe IAE [18].

Brainstem involvement on MRI and ventilator support was associated with increased odds of mortality, consistent with prior literature identifying deep gray matter and brainstem lesions as markers of severe disease. Although these associations did not reach statistical significance, likely because of the limited number of events, the direction of the findings is consistent with previous reports suggesting an association between brainstem involvement and disease severity.

This study had several limitations. First, the sample size was small, with only three death events, resulting in wide CIs and limited statistical power. To address potential small-sample bias and separation issues, Firth-penalized logistic regression was applied; however, the estimates remained imprecise and should be interpreted as exploratory. Second, this was a single-center retrospective study, and institutional practice patterns, particularly the shift toward earlier treatment after the initial fatal cases, may have influenced outcome comparisons. Third, treatment timing was not randomized, and residual confounding could not be excluded. Patients who received early therapy may have differed systematically from those who did not. Despite these limitations, the consistent direction of the observed associations, particularly for vaccination status and early treatment initiation, provides clinically meaningful signals that warrant further investigation in larger multicenter cohorts.

In conclusion, this single-center cohort study highlights that pediatric IAE, although many patients recover fully, can progress unpredictably and requires immediate clinical vigilance. Prompt recognition, early initiation of antiviral and immunomodulatory therapies, and reinforcement of influenza vaccination may help reduce severe neurological complications. Future multicenter studies with larger sample sizes are needed to validate these observations and better define optimal risk stratification and treatment timing in pediatric IAE.