Hemophagocytic Lymphohistiocytosis (HLH) in Langerhans Cell Histiocytosis (LCH): A Multicenter Retrospective Descriptional Study

Introduction: Langerhans cell histiocytosis (LCH) is an inflammatory myeloid neoplasia characterized by the accumulation of CD1a+ CD207+ histiocytes. Hemophagocytic lymphohistiocytosis (HLH), a non-malignant histiocytic disorder, is typified by the accumulation and activation of CD8+ T cells and macrophages, which secrete high levels of pro-inflammatory cytokines. The co-existence of LCH and HLH has been reported, albeit rarely, and is believed to be associated with a poorer outcome. To better understand the relationship between these two conditions, in this study we sought to describe the incidence, risk factors for development, and outcome of HLH when it develops in children and young adults with multisystem-LCH (MS-LCH).

Methods: We conducted a retrospective study involving 14 centers and collected data on 384 MS-LCH patients aged less than 30 years and who were diagnosed between year 2000 and 2015. Data collected on the eligible patients included clinical information at the time of LCH diagnosis, clinical and laboratory parameters at HLH diagnosis (for those who developed HLH), treatment and disease outcome. Patients who developed HLH were classified as having "true-HLH", which was defined as disease fulfilling 5 of 8 HLH-2004 diagnostic criteria or as "HLH-like" disorder, which was defined as fulfilling <5 of 8 HLH diagnostic criteria but whose disease status was suggestive of HLH and treated with HLH- and/or LCH-directed therapy.

Results: Of 384 MS-LCH patients, 44 (11%) were identified with HLH (29 with true HLH and 15 with an HLH-like disorder), ranging in age from 15 days to 20.6 years (median, 1.12 years). The majority of MS-LCH patients who also had HLH were females (n=27) and had accompanying risk organ (liver, spleen and/or hematopoietic system) involvement (RO+) (n=40), as opposed to non-HLH MS-LCH patients. Among nine HLH patients tested for BRAF V600E mutation status, eight were found to be positive. Twenty (45%) patients developed HLH (true or HLH-like) concurrent (±7 days) with LCH diagnosis, while 24 (55%) developed HLH >7 days before or after LCH diagnosis. The 3-year cumulative incidence of HLH (true or HLH-like) in MS-LCH was 16.8%. The 5-year overall survival of LCH patients without HLH was 98 ± 9%, while survival for those with an HLH-like disorder or true-HLH was 75 ± 12% and 70 ± 14%, respectively (P<0.0001). Age <2 years, female gender, RO+ and lack of bone involvement at LCH diagnosis were each independently associated with increased risk for HLH. Among 20 HLH patients with available data, the median soluble interleukin-2 receptor level (sIL-2R) was 16,220 U/mL (range, 1,149 to 60,420 U/mL) (normal reference <2,400 U/mL), ferritin was 505 ng/mL (range, 28 to 26,660 ng/mL) (normal reference <500 ng/mL), and sIL-2R/ferritin ratio was 42.

Conclusion: The development of HLH in patients with MS-LCH was not uncommon and associated with a poorer prognosis. Young females with RO+ MS-LCH who lack bone lesions at LCH diagnosis were at increased risk of developing HLH. Ferritin levels appear to be lower in comparison to patients who develop HLH in other contexts. There are overlapping features between MS-LCH and HLH that make the clinical distinction between these disorders difficult. Accordingly, improved biomarkers are needed to facilitate the identification of HLH in patients with MS-LCH. It is anticipated that early identification of HLH and prompt intervention may improve the outcome for affected individuals. Future prospective studies are needed to better understand the underlying mechanisms and identify more effective therapies.