Abstract
Introduction: Secondary hemophagocytic lymphohistiocytosis (sHLH) is a rare, life-threatening hyperinflammatory syndrome characterized by dysregulated immune cell function in response to a strong antigenic trigger – most commonly malignancies (mHLH), infections, or autoimmune diseases. The underlying pathophysiology involves a failure to downregulate the activation and proliferation of cytotoxic T cells, natural killer (NK) cells, and monocytes/macrophages. This leads to a cytokine storm that often necessitates urgent hospitalization. Without timely and effective intervention, sHLH can result in multiple organ failure and death (Allen 2015; Jordan 2019). There is a lack of comprehensive natural history studies that evaluate sHLH at the patient level using electronic health records (EHRs) with clinician adjudication. Such studies are critical to improving our understanding of prognostic factors, clinical presentation, treatment patterns, healthcare utilization, and survival outcomes. Here, we describe a natural history study designed to address these gaps.
Methods: Potential adult (≥ 18 years) sHLH patients were identified from EHRs from 2006 to 2022 at two United States academic health systems (Site 1: MedStar Health System, Maryland, DC and Virginia; Site 2: Vanderbilt University Medical Center, Nashville, TN) using an algorithm incorporating ICD codes, keywords, and clinical variables with predictive features (i.e., ferritin and sCD25). Patient charts flagged by the algorithm underwent expert clinical review, and sHLH diagnosis was assigned based on comprehensive clinical assessment. Fulfillment of HLH-2004 diagnostic criteria was not required for diagnosis. Patients with primary HLH were excluded. Data collected included demographics, clinical characteristics of sHLH and the underlying trigger, treatment patterns and responses, and survival outcomes. Death dates were recorded if available, otherwise, patient follow up was censored at date of last clinical activity in the EHR. Cox regression models, adjusted for age, sex, race, and Charlson Comorbidity Index, were used to identify factors associated with overall survival (OS).
Results: Clinician review confirmed 57 cases of sHLH at Site 1 and 175 cases at Site 2; median follow-up was 8.0 and 3.6 months, respectively. 50.9% of patients for Site 1 and 62.3% of patients for Site 2 met at least 5 of 8 HLH-2004 diagnostic criteria. At Site 1, 53% of patients were male, with a median age of 37 years at sHLH diagnosis. At Site 2, 47% were male, with a median age of 41 years. At Site 1, 35% of patients were Black, 30% White, 28% Other/Unknown, 11% Hispanic, and 7% Asian; respective proportions at Site 2 were 19%, 66%, 1%, 5%, and 3%. sHLH triggering events varied at Sites 1 and 2 as follows: infection in 26.3% and 40.0%, malignancy in 10.5% and 34.9%, autoimmune disorder in 29.8% and 16.6%, idiopathic in 14.0% and 4.6%, and other in 19.3% and 2.9%, respectively. Notably, Site 1 had a lower percentage of mHLH, potentially due to incomplete identification and adjudication of mHLH cases. Two-month OS for sHLH was 62.1% (95% CI: 48.0–73.4) at Site 1 and 63.5% (95% CI: 56.4–71.4) at Site 2. Inpatient mortality during the initial hospitalization for sHLH was 40.4% at Site 1 and 32.0% at Site 2. At Site 2, which had a larger sample size and where mHLH data were complete, the 2-month OS for mHLH was 54.9% (95% CI: 43.4–69.4) and inpatient mortality was 42.6%. Malignancy as a trigger was associated with significantly higher mortality compared to other triggers, with an adjusted hazard ratio of 2.90 (95% CI: 1.52–5.52).
Conclusions: This study employed novel methods to enhance understanding of sHLH. Approximately 40–50% of adult sHLH cases do not meet the HLH-2004 diagnostic criteria, highlighting a significant risk of underdiagnosis and treatment delays when relying solely on these criteria. HLH experts recommend suspecting HLH when a patient presents with the ‘3 Fs’: fever, elevated ferritin, and falling blood counts (Zoref-Lorenz, 2025). A mortality rate of approximately 40% at two months and during initial hospitalization underscores the critical importance of early recognition and intervention in sHLH. The findings confirm that mHLH carries the poorest prognosis compared to sHLH triggered by other conditions. Importantly, these data highlight the urgent need for more effective therapeutic strategies across all sHLH subtypes.
