Over 4 decades, French children teach about LCH

In this issue of Blood, Thalhammer and colleagues¹ describe the characteristics and outcomes of 331 children with Langerhans cell histiocytosis (LCH), out of a total of 2313 enrolled in the French National Histiocytosis Registry, who had features of hematologic involvement (HI). The study, which includes patients enrolled over 4 decades between 1983 and 2023, reflects the power of forethought, patience, and meticulous record keeping.

Prior to 2010, the fundamental nature of LCH was a mystery. It had long been recognized that although the tumors of patients with LCH share histopathologic features, patients’ clinical courses can vary dramatically. A subset of patients, usually babies and very young children, experience an aggressive multisystem disease. This entity, first described in the early 20th century and dubbed Letterer-Siwe disease, typically features pancytopenia and hepatosplenomegaly.^2,3 In 1975, Lahey identified HI as an important adverse feature in childhood LCH and defined it as hemoglobin <10 g/dL (<9 g/dL in infants) not due to other causes, leukocytes <4000/μL, and platelets <100 G/L.⁴ The identification of prognostic factors, including involvement of the hematopoietic system, liver, and/or spleen (so-called risk organs), created the opportunity to risk-stratify treatment. Beginning in the LCH-II clinical trial (1996-2001),⁵ the Histiocyte Society adopted a risk-stratification algorithm and showed that, in comparison with LCH-I, intensified treatment improved outcomes in risk organ–positive patients. But since 1975, the definition of HI has not been refined and initial treatment of patients with HI who are at highest risk of treatment failure also has not changed. Nor have additional risk factors been established.

LCH was firmly placed in the category of myeloid neoplasia in 2010 after the demonstration that about half of cases harbor the BRAF^V600E mutation.⁶ Multiple investigators confirmed and expanded on this finding, demonstrating that activation of the MAPK signaling pathway is a critical driver of LCH.⁷ Whether the presence of BRAF^V600E or other genomic lesions is of prognostic significance remains unsettled.

The French National Histiocytosis Registry has been enrolling children with LCH for decades. This unique database contains 2313 patients seen between 1983 and 2023. For this study, the authors sought to identify patients with LCH at highest risk of an adverse or fatal outcome. Of the 331 registry participants with HI, they identified a subset of 154 (46.5%) who had severe HI (SHI), which they defined as hemoglobin <7 g/dL or platelets <20 G/L (ie, transfusion eligible). Of the patients with SHI, median age at diagnosis was 0.9 years, 80% had spleen involvement, and 79% had liver involvement. Except for 2 newborns who died before treatment initiation, all SHI cases received systemic therapy, which varied over the study period. In comparison with cases with mild HI (MHI), patients with SHI were more likely to be refractory to first-line treatment (35% vs 6.7%), more likely to require second-line therapy (76% vs 47%), and more likely to die early in treatment (of the 46 cohort patients who died, 43 had SHI). Examination of bone marrow aspirates or biopsies did not add prognostic value over assessment of peripheral blood counts.

Molecular testing was completed in 133 cases in the study cohort, including 58 with SHI, among whom 55 (98%) were positive for BRAF^V600E; of 75 patients with MHI, 55 (73%) carried BRAF^V600. Among 110 cohort patients with HI and BRAF^V600E, the level of the mutation detected on cell-free DNA analysis correlated with HI severity; levels >1% were only present in SHI cases. Over the study period, outcomes for SHI cases improved first with the application of intensive second-line chemotherapy beginning in 1998⁸ then improved further in the era of MAPK inhibitors (MAPKi) beginning in 2014. Early in the study period, nearly half of patients with LCH with HI died. In the contemporary era, deaths became far less frequent and almost completely restricted to those with SHI. The longitudinal nature of the study provides intriguing data about the risk of LCH-related neurodegeneration (ND), a poorly understand and potentially devastating condition that may arise years after initial LCH diagnosis. Overall, a concerningly high proportion of survivors in the cohort went on to experience ND: 11% developed clinical symptoms, and 33.6% developed radiographic findings consistent with ND.

What did we learn from this study? How could the findings influence next steps in clinical research in LCH? First, it may be time to update the Lahey criteria. The peripheral blood count criteria for SHI defined in the article, although they are rigid thresholds applied to continuous variables, are supported by the data presented and are reasonable. Further, bone marrow examination was shown to be unnecessary in this large cohort of patients with LCH with cytopenia(s). Therefore, this procedure should be performed for the purpose of excluding alternative diagnoses or for research. The study also shows the relevance of molecular data in LCH cases. This is the only way to understand the biology of the disease and advance therapy for all children with LCH, including, but not limited to, those at risk of dying of the disease. The study supports the view⁹ but does not prove (due to study design) that BRAF^V600E is a prognostic factor in childhood LCH since almost all study subjects with SHI were positive for BRAF^V600E. Also, the study suggests that the presence of BRAF^V600E is a treatment factor in childhood LCH; by looking at treatment outcome over time, the beneficial impact of MAPKi is strongly implied. Prospective trials will be needed to determine when and how best to deploy these drugs. In our opinion, these trials are urgently needed and long overdue.

Conflict-of-interest disclosure: The authors declare no competing financial interests.