Neurological Involvement in Adult‐Onset Secondary Hemophagocytic Lymphohistiocytosis: Clinical Features and Prognostic Implications

BACKGROUND: Secondary hemophagocytic lymphohistiocytosis (sHLH) with central nervous system (CNS) involvement poses significant diagnostic and therapeutic challenges. This study aimed to characterize the clinical features, laboratory profiles, and prognostic impact of neurological involvement in adult-onset sHLH. METHODS: We analyzed 130 adult sHLH patients, comparing 28 with CNS involvement to 102 without neurological manifestations. Clinical parameters, neuroimaging, cerebrospinal fluid (CSF) profiles, cytokine levels, treatment responses, and survival outcomes were evaluated. RESULTS: Patients with CNS involvement were older (median age, 54 vs. 46 years; p = 0.013) and had higher disease severity (median HScore, 250 vs. 210; p < 0.001). Malignancy-associated sHLH was more prevalent in the CNS-positive group (42.9% vs. 29.4%; p = 0.038). Neurological manifestations included altered mental status, impaired consciousness, and seizures. Neuroimaging revealed abnormalities in 71.4% of the cases, primarily T2-weighted fluid-attenuated inversion recovery hyperintensities and leptomeningeal enhancement. CNS-positive patients exhibited markedly elevated inflammatory markers, most notably CSF Interleukin-6 (p < 0.001). In multivariable analysis adjusted for malignancy, age, ferritin, and HScore, CNS involvement independently predicted mortality (adjusted HR = 2.0, 95% CI: 1.1-3.7, p = 0.023), with a significantly shorter median overall survival (6.5 vs. 11.5 months, p < 0.0001). Malignancy-associated etiology and HScore ≥ 250 were also independent prognostic factors. The DEP (dexamethasone, etoposide, and polyethylene glycol-asparaginase) regimen achieved a faster median time to initial response than the HLH-94 protocol (9 vs. 14 days, p = 0.02). CONCLUSIONS: CNS involvement defines a severe phenotype of adult-onset sHLH, characterized by malignancy-prone etiology, intense neuroinflammation, and poor prognosis. We establish CNS involvement as an independent predictor of mortality, underscoring the critical need for early recognition and CNS-directed therapies.