Flu, Autoimmunity and Salt Imbalance All Converge on One Reversible Brain Lesion in Children

When children with a mysterious brain lesion arrive at the hospital, the imaging looks strikingly similar no matter what caused it: a small, bright spot on diffusion-weighted MRI at the splenium, the rear portion of the corpus callosum, the thick band of fibers connecting the brain’s two hemispheres. A new case series from Tianjin Medical University General Hospital, published in Frontiers in Pediatrics, follows eight children with this pattern—known as reversible splenial lesion syndrome, or RESLES—and shows how radically different triggers, from influenza A to autoimmune inflammation to profound sodium depletion, can funnel into the same transient imaging signature.

The retrospective single-center study reviewed children admitted between March 2022 and December 2025 who met strict criteria: a localized or splenium-predominant lesion on brain MRI, documented radiological resolution on follow-up imaging, no prior epilepsy, and at least six months of clinical follow-up. Children whose lesions persisted, or whose imaging pointed instead to acute disseminated encephalomyelitis, multiple sclerosis spectrum disease, ischemia, tumor, trauma or another primary callosal disorder, were excluded. Eight children made the cut—five girls and three boys, aged 1.5 to 168 months, with a median age of 82 months at onset.

The etiologic spread was wide. Four children had confirmed influenza A infection, one had respiratory syncytial virus combined with adenovirus, and one had fever, sore throat, headache and meningeal irritation diagnosed clinically as herpangina, with enterovirus unable to be excluded despite negative pathogen testing. Definite pathogen positivity was documented in five of the eight cases, or 62.5 percent, while infection-related or highly suspected infection accounted for six cases, or 75 percent. Beyond the infectious group, one child had autoimmune glial fibrillary acidic protein astrocytopathy, or GFAP-A, and one afebrile infant developed lesions in both the genu and splenium of the corpus callosum during a severe encephalopathy marked by respiratory failure, seizures, dehydration and multisystem abnormalities.

On MRI, six children had type 1 RESLES—an isolated splenial lesion—while two had type 2, in which the splenium is affected alongside additional lesions elsewhere in the corpus callosum, centrum semiovale, periventricular white matter, brainstem or spinal cord. The typical appearance was T2 or FLAIR hyperintensity, bright signal on diffusion-weighted imaging and reduced apparent diffusion coefficient values, the hallmark of restricted diffusion consistent with cytotoxic injury or edema. Follow-up MRI confirmed that every lesion resolved, with intervals ranging from less than three full days to approximately three months. In one influenza case, the splenial signal had vanished on repeat imaging in under five days.

Laboratory findings wove the disparate cases together. Hyponatremia struck four children, with nadir serum sodium values of 124.8, 127.5, 117 and 110.2 mmol/L—the last and most severe occurring in the GFAP-A patient, who also showed syndrome of inappropriate antidiuretic hormone secretion, evidenced by low plasma osmolality, low serum uric acid, elevated urine sodium and concentrated urine. Interleukin-6, a proinflammatory cytokine, was elevated in six patients, and three of those also had raised interleukin-10, an anti-inflammatory and immunoregulatory cytokine. Cerebrospinal fluid lactate dehydrogenase, measured in six patients, exceeded the reference upper limit in every case tested—a marker the authors note can indicate tissue injury, cellular breakdown or inflammation within the central nervous system.

Clinical presentations were as varied as the causes. Seven children had fever or an infection-related prodrome, with peak temperatures generally at or above 39.3 °C and reaching 40.4 °C in one patient. Neurological manifestations ranged from fatigue, dizziness and somnolence to incoherent speech, altered consciousness, personality change, behavioral abnormality, depressed mood, tearfulness and impaired memory or calculation. Two children had convulsions or afebrile seizures and two had neck stiffness. The GFAP-A case stood out with diplopia, difficulty urinating and defecating, urinary retention and brainstem or spinal cord involvement—a picture of meningoencephalomyelitis backed by markedly elevated cerebrospinal fluid pressure, mononuclear cell-predominant pleocytosis, raised protein and GFAP antibodies positive in both serum and cerebrospinal fluid.

Electroencephalography was normal in five patients. The three abnormal recordings included paroxysmal medium- to high-amplitude theta activity over the frontal regions during sleep in one child, scattered medium-amplitude slow waves during wakefulness with a borderline amplitude-integrated EEG in another, and paroxysmal high-amplitude broad sharp waves across multiple leads in the infant, whose ambulatory EEG captured events lasting up to roughly seven seconds, right-side predominant, with additional focal sharp waves favoring the central-parietal-occipital regions on the left. EEG abnormalities clustered in the children with impaired consciousness, seizure-like episodes or severe encephalopathy.

Treatment was tailored to the suspected trigger rather than the lesion itself. All four influenza-associated patients received oseltamivir; the child with the adenovirus and respiratory syncytial virus coinfection received empiric off-label ganciclovir given the severity of illness, alongside corticosteroids and supportive care, though the specific contribution of ganciclovir could not be determined. Every patient received corticosteroids during the acute phase, and three with more severe disease or stronger immune-inflammatory features received intravenous immunoglobulin. The GFAP-A patient underwent pulse methylprednisolone with gradual tapering plus intravenous immunoglobulin, while the critically ill infant required mechanical ventilation and antiseizure medications on top of steroids and immunoglobulin. Supportive care included careful correction of electrolyte disturbances and maintenance of fluid and internal milieu stability.

During follow-up ranging from six months to 3.5 years, no definite neurological sequelae were documented in any of the eight children. The authors caution, however, that radiological reversibility should not be mistaken for a benign acute illness: seizures, abnormal EEG findings, profound hyponatremia, autoimmune inflammation and type 2 involvement all flagged greater clinical complexity. The two type 2 cases were indeed the sickest—the autoimmune patient with multifocal central nervous system involvement and severe salt disturbance, and the infant with cerebral herniation, type II respiratory failure, hepatic dysfunction, anemia, hypoproteinemia, ascites and pericardial effusion. The authors frame the link between extrasplenial lesions and greater severity as hypothesis-generating only, requiring larger cohorts to test.

The series supports a model in which pediatric RESLES is a shared imaging phenotype rather than a single disease: diverse insults—infection, autoimmunity, hyponatremia, dehydration, metabolic acidosis, hypoxia and critical illness—converge on homeostatic disturbance and focal cytotoxic edema in the splenium, possibly through cytokine release, neurovascular unit dysfunction, blood-brain barrier alteration, astrocyte injury, osmotic imbalance, energy stress and abnormal water transport. The densely packed glial cells and callosal axons of the central splenium may be uniquely vulnerable, with proposed mechanisms including cytotoxic swelling of astrocytes and neurons and intramyelinic fluid shifts, though these remain incompletely established and restricted diffusion alone does not demonstrate demyelination. The authors stress that management should target the trigger—systematic infectious workup, dynamic sodium monitoring, autoimmune antibody testing when meningeal signs, spinal or brainstem involvement or disproportionate cerebrospinal fluid inflammation appear, and cautious correction of hyponatremia to avoid osmotic demyelination—while acknowledging that the retrospective design, small sample, unstandardized testing and near-universal steroid use prevent any causal claim about whether immunotherapy accelerated recovery. Routine immunotherapy for mild, infection-associated type 1 RESLES remains uncertain, and multicenter prospective studies with standardized protocols and long-term developmental follow-up are needed.

Subject of Research: Etiologic heterogeneity and shared imaging features in pediatric reversible splenial lesion syndrome

Article Title: Etiologic heterogeneity and a common imaging endpoint in reversible splenial lesion syndrome

Article References: Sun, Y., Mao, Z., Dong, S., Kan, X., Jiang, L., & Mu, Q. (2026). Etiologic heterogeneity and a common imaging endpoint in reversible splenial lesion syndrome. Frontiers in Pediatrics, 14, 1931158. https://doi.org/10.3389/fped.2026.1931158

Image Credits: AI Generated

DOI: 10.3389/fped.2026.1931158

Keywords: RESLES, corpus callosum, splenium, pediatric neurology, influenza A, GFAP astrocytopathy, hyponatremia, cytokines, diffusion-weighted MRI, encephalopathy, cytotoxic edema, case series