Langerhans Cell Histiocytosis: NACHO Update on Progress, Chaos and Opportunity on the Path to Rational Cures

Abstract

Langerhans cell histiocytosis (LCH) is a myeloid neoplastic disorder characterized by lesions with CD1a+/Langerin (CD207)+ histiocytes and inflammatory infiltrate that can cause local tissue damage and systemic inflammation. Clinical presentations range from single lesions with minimal impact to life-threatening disseminated disease. Therapy for systemic LCH has been established through serial trials empirically testing different chemotherapy agents and duration of therapy. However, fewer than 50% of patients with disseminated disease are cured with the current standard of care vinblastine/prednisone/(mercaptopurine), and treatment failure is associated with long-term morbidity including risk of LCH-associated neurodegeneration. Historically, the nature of LCH, whether a reactive versus neoplastic/malignant condition, was uncertain. Over the past 15 years, seminal discoveries have broadly defined LCH pathogenesis, specifically activating MAPK pathway mutations (most frequently BRAFV600E) in myeloid precursors drive lesion formation. LCH therefore is a clonal neoplastic disorder, though secondary inflammatory features contribute to disease. These paradigm-changing insights offer promise of rationale cures for patients based on individual mutations, clonal “reservoirs” and extent of disease. However, the pace of clinical trial development lags the kinetics of translational discovery. In this review, we discuss the current understanding of LCH biology, clinical characteristics, therapeutic strategies, and opportunities to improve outcomes for every patient through coordinated agent prioritization and clinical trial efforts.

PRECIS

Langerhans cell histiocytosis is an inflammatory myeloid neoplastic disorder driven by activating somatic mutations in MAPK pathway genes. Rapid advances in mechanistic understanding of LCH now offer opportunities to improve outcomes for patients.

Introduction

This manuscript was organized through collaboration between the North American Consortium for Histiocytosis members and partners in the United Kingdom to review the current biological understanding of Langerhans cell histiocytosis (LCH) (progress), data to guide current therapeutic approaches in the context of rapidly-changing biological paradigms and development of novel agents (chaos), and strategies to improve therapy for patients (opportunity).

Background

The epidermal Langerhans cell was first described in 1868 by the Berlin medical student Paul Langerhans as a neuronal cell because of its distinct dendrite-like processes.¹ In a subsequent article, “Corrections”, he withdrew his initial assertion and speculated that this cell might be of hematopoietic origin.² Between 1893 and 1919, Hand, Schüller, and Christian separately published the first clinical descriptions of Langerhans cell histiocytosis (LCH) of patients with exophthalmos, lytic bone lesions and diabetes insipidus.^3–5 In 1924, Letterer reported an infant with extensive skin rash, pancytopenia, hepatosplenomegaly and tachypnea followed nine years later by Siwe describing a toddler with similar symptoms.^6,7 Initially considered distinct entities, isolated bone lesions were called eosinophilic granulomas, multifocal bone disease with diabetes insipidus, Hand-Schüller-Christian Disease, and the most severe multi-system disorder, Letterer-Siwe Disease. In 1941, Farber noted these entities to have similar histologic features. In 1953, Lichtenstein concluded that these were different clinical manifestations of the same disease affecting the reticuloendothelial system and proposed the name “histiocytosis X” to reflect the unknown origin of the lesions cells, thus conceptualizing “histiocytoses”.^8,9 However, electron microscopy revealed Birbeck granules in those cells, leading Nezelof to hypothesize a developmental link with the epidermal Langerhans cell in 1973.¹⁰ This led to the renaming of histiocytosis X to LCH.¹⁰

A long-standing debate as to whether LCH was a malignancy or an inflammatory disease continued even after the clonality of LCH was established in the 1990s^11,12. Subsequently, BRAFV600E (or other MAPK activating mutations) in myeloid precursors were identified as pathogenic drivers, thus placing LCH in the category of myeloid neoplasms and conclusively demonstrating that LCH is an oncologic process.^13–15 This discovery contributed to the current understanding of LCH as both a malignancy AND an inflammatory disorder.¹⁶ While infection or immune dysregulation may not be drivers of disease, inflammation induced by MAPK-activated myeloid cells contributes to pathogenic features.

Diagnosis and Classification

Histology

Once the possibility of LCH is considered, its histopathologic diagnosis is straightforward in most cases. LCH lesions are characterized by CD1a+ Langerin (CD207+) dendritic cells (DC) with reniform nuclei in the setting of a mixed inflammatory infiltrate (Figure 1). Birbeck granules are characteristic of LCH dendritic cells, but electron microscopy is no longer an essential diagnostic tool in the presence of CD1a and CD207 (Langerin) expression which are surrogate diagnostic markers.¹⁷ Peculiarities of staining intensity of Langerin, a component of the Birbeck granule apparatus, can confound the diagnosis in some cases. Surface Langerin is expressed in normal, physiologic epidermal Langerhans cells. Therefore, in addition to Langerin expression, cell features (e.g., dendritic cell (DC) morphology) and pattern of distribution differentiate epidermal LC/DC from LCH cells in skin and draining lymph nodes (Figure 1)¹⁸. Additionally, Langerin expression is induced in precursors and not a fixed feature of LCH cells: Langerin is dim or absent in many cases of liver, bone marrow and CNS disease.^14,19 Increasing use of pan-histiocytosis diagnostic markers is also important in identifying mixed histiocytic lesions (e.g. LCH with juvenile xanthrogranuloma (JXG), Rosai-Dorfman disease (RDD) and/or Erdheim-Chester disease (ECD)). Additionally, identification of molecular alterations (e.g. mutant specific BRAF VE1 stains, BRAFV600E qPCR, targeted next generation sequencing) support histiocytic neoplasm diagnosis and identify potential therapeutic targets.

Figure 1. LCH Histology.

LCH lesions are characterized by “histiocytes” with reniform nuclei and surrounding inflammatory infiltrate. (A) H&E (80x), (B) CD1a immunohistochemistry (80x), (C) Electron microscopy demonstrating Birbeck granule, which is now replaced with Langerin (CD207) stain (not shown). LCH can be challenging to differentiate from reactive Langerhans cells/dendritic cells in skin and lymph nodes (LN). Skin with reactive CD1a+ dendritic cell hyperplasia in the dermis with perivascular histiocytes that are CD1a+ with spindled dendritic shape (D-E) are low to negative Langerin/CD207 staining (F) (10x). By comparison, (G-I) shows a case of skin with superficial dermal infiltrate of LCH with epidermotropism and surface ulceration. The CD1a (H) and Langerin (I) stains highlight the clustering of plump lesional histiocytes in the papillary dermis (10x). Activated DC drain into LN to present antigen to lymphocytes as demonstrated by hyperplasia of reactive dendritic/Langerhans cells in the paracortex ((J) H&E, (K) CD1a, (L) Langerin, 13x digital images). By comparison, nodal LCH is characterized by sinus expansion of LCH with variable spill over into the paracortex ((M) H&E, (N) CD1a, (O) Langerin, 10x). (Images courtesy of Jennifer Picarsic, MD, Cincinnati Children’s Hospital, Cincinnati, OH).

Classification

Histiocytic disorders have historically been classified according to terminal phenotype and similarity to physiologic cells of the mononuclear phagocyte system^20–22 (e.g. “Langerhans cell” histiocytosis because of overlapping features with epidermal LC). However, the past 15 years have seen notable advances in mechanistic understanding of LCH pathogenesis and phylogeny. Based on these developments, a revised reclassification strategy was proposed that includes not only cell phenotype, but also clinical patterns and molecular drivers.²³ LCH is housed in the “L-group” that also includes ECD and disseminated JXG. However, the updated 5^th edition World Health Organization (WHO) classification of mature lymphoid, histiocytic, and dendritic neoplasms separates ECD and LCH into two distinct entities based on integration of clinical, radiological, and histopathological diagnosis.²⁴ (Table 1)

Table 1.

Classifications of Histiocytic Disorders.^23,24

WHO 5th Edition Classification of Haematolymphoid Tumors: Dendritic cell and histiocytic neoplasms
Plasmacytoid dendritic cell neoplasms	Mature plasmacytoid dendritic cell proliferation associated with myeloid neoplasm Blastic plasmacytoid dendritic cell neoplasm
Langerhans cell and other dendritic cell neoplasms
Langerhans cell neoplasms	Langerhans cell histiocytosis Langerhans cell sarcoma
Other dendritic cell neoplasms	Indeterminate dendritic cell tumor Interdigitating dendritic cell sarcoma
Histiocytic neoplasms	Juvenile xanthogranuloma Erdheim-Chester disease Rosai-Dorfman disease ALK-positive histiocytosis Histiocytic sarcoma
Histiocyte Society Revised Classification Proposal
L Group	Langerhans cell histiocytosis Erdheim-Chester disease Disseminated juvenile xanthogranuloma ALK-positive histiocytosis Indeterminate cell histiocytosis Mixed LCH/ECD
C Group	Skin limited JXG Other skin-limited xanthogranulomas and non-Langerhans histiocytoses
R Group	Familial Rosai-Dorfman disease (RDD) Classical (nodal) RDD Extranodal RDD Neoplasia-associated RDD Immune-associated RDD
M Group	Primary malignant histiocytosis Secondary malignant histiocytosis (associated with other hematologic malignancy)
H Group	Primary hemophagocytic lymphohistiocytosis (HLH) Secondary HLH HLH of uncertain origin

Evolving Understanding of LCH Biology

Prior to 2010, clues regarding the origins and nature of LCH came primarily from histology. Characteristic but enigmatic features led to recurrent debate regarding LCH as a reactive inflammatory (e.g. immune infiltrate, rare mitoses, no known mutations, rare spontaneous regression) versus neoplastic disorder (e.g. persistence and progression, clonality, destructive tumors, response to chemotherapy, lack of associated infection).¹⁶ Discovery of recurrent BRAFV600E mutations¹³ followed by reports of near universal and mutually exclusive activating MAPK pathway mutations^25–27, and an otherwise quiet genome highlighted the importance of MAPK activation in this disease’s pathophysiology. Physiologic MAPK signaling enables cells to differentiate, proliferate and survive in response to extracellular environmental cues (Figure 2A).²⁸ BRAFV600E mutations were discovered in CD34+ hematopoietic stem cells and in peripheral blood of all patients with high risk (HR) multisystem LCH, less frequently in patients with low risk (LR) single system disease, and very rarely in patients with single lesions. BRAFV600E activates MAPK signaling independent of upstream regulation by receptor tyrosine kinases (RTK) or Ras-family GTPases (Figure 2A).²⁹

Figure 2. LCH Pathogenic Mechanisms.

(A) Diagram of the MAPK signaling pathway. When growth factors/mitogens bind to receptor tyrosine kinases (RTKs), RTKs activate guanine exchange factors to exchange GDP for GTP on Ras, which activates Ras. Normally, GTP bound Ras is required for the downstream activation and homodimerization of Rafs (BRAF or ARAF) and further signaling components of MAPK. The BRAFV600E allows BRAF to function as a monomer independent of Ras activation, which leads to constitutive activation of MAPK signaling pathway. (B) Studies in patients with LCH have shown that acquisition of allelic MAPK-activating allelic variants in myeloid precursors leads to HR LCH due to pathological myeloid cells having access to the circulation, supported by the finding that HR-patients (but never LR-patients) harbor circulating mutation bearing myeloid cells. Circulating CD11a⁺ myeloid cells also contribute the LCH-ND. This risk-stratification can be modeled using genetic knock-in models where expression of BRAFV600E can be enforced in hematopoietic stem cells (Scl^Cre BRAFV600Elox-stop-lox; BRAFV600E^scl or Map17^Cre BRAFV600Elox-stop-lox; BRAFV600E^Map17) or the entire DC lineage (CD11c^Cre BRAFV600Elox-stop-lox; BRAFV600E^CD11c), which lead to lethal LCH-like disease. In comparison, enforced expression of BRAFV600E in tissue limited DCs (CD207^Cre BRAFV600Elox-stop-lox; BRAFV600^CD207) form lesions, but do not die from LCH-related causes. (C) Enforced expression of BRAFV600E in mouse models have found that enhanced myeloid differentiation, so called “myeloid skewing”; activation and exhaustion of T cells, tissue entrenchment by decreased CCR7 expression, and resistance to apoptosis promote the accumulation of pathological DCs in nearly any organ. The latter is part of the BRAFV600E-induced SASP, which is central to LCH pathophysiology.

Enforced expression of BRAFV600E in differentiated CD207+ cells in mice led to development of some lesions, but overall was a mild phenotype. By comparison, expression of BRAFV600E+ in myeloid cells and precursors (CD11c^cre, Scl^cre and MAP17^cre) in mice drove an aggressive HR LCH-like phenotype where mice survived only a few months.^14,30,31 Together, these data from patients and mice support a model in which MAPK activation in myeloid precursors drives disseminated LCH, while less extensive disease is associated with the same MAPK pathway alterations in more differentiated myeloid cells (“the misguided myeloid DC model of LCH pathogenesis”) (Figure 2B).³²

MAPK activation in myeloid precursors enhances differentiation toward the myeloid lineage (e.g. “myeloid skewing”. MAPK-activated myeloid DC lack CCR7 expression, which is required for physiologic migration of DC toward lymph nodes. Further, MAPK activation in lesion DC induces apoptosis resistance via upregulation of Bcl-xL, activation and exhaustion of infiltrating T cells via checkpoint inhibition, and induction of an oncogene-induced senescence program.^30,33 Together, enhanced differentiation, entrenchment, resistance to apoptosis, and the ability to orchestrate tissue immune responses lead to accumulation of pathological DCs and formation, persistence and progression of LCH lesions (Figure 2C).

Beyond somatic mutations, a case-parent trio study using genome-wide association methodology identified inherited risk allele in SMAD6.³⁴ Non-random racial and ethnic distribution (e.g. increased risk in AmerIndian/Hispanic, decreased risk in African ancestry) suggests germline variants may influence LCH risk. LCH does not typically present in families, with the exception of monozygotic twins who share hematopoietic origins. Pediatric LCH arises in approximately 5 children per million, and is estimated to arise in 1 adult per million.^35,36

Clinical Manifestations of LCH in Children

LCH can affect any organ in the body and is typically categorized as single lesion, single-system unifocal, single-system multifocal or multi-system based on extent of disease. (Figure 3). Common sites of disease include skin, bone, lung, liver, and the pituitary gland. Multisystem LCH is further stratified into low- or high-risk groups based on the organs affected and corresponding mortality risks (also referred to as risk organ positive or negative (RO+/RO-); also called “high risk” (HR) and “low risk” (LR)). Specific “risk organ” systems (e.g., liver, spleen, bone marrow) were so classified based on their association with higher mortality risks in the LCH-II trial.³⁷ Within these risk groups, clinicians have recognized classical patterns of presentation. For example, HR disseminated LCH is most common in infants, multifocal disease without risk organ involvement is more common in preschoolers, and more than 50% of patients with an isolated bone lesion are diagnosed after the age of 5 years. Additionally, “CNS risk” lesions, defined as those involving facial bones and the skull base, are associated with development of LCH-associated pituitary dysfunction and neurodegeneration.³⁸ The relatively worse prognosis with liver, spleen and bone marrow disease applies to adults as well as children.³⁹ The diagnosis of LCH can be invoked frequently at residency problem conference “morning report” diagnostic dilemma seminars as the vast spectrum of clinical features can mimic more common conditions (Table 1, Figure 3).

Figure 3. Clinical Features of LCH.

Positron-emission tomographic (PET) images show a single bone lesion involving the humerus (Panel A, arrow); low-risk lesions involving the orbit, lymph nodes, bone (multifocal lesion), and thymus (Panel B); and high-risk lesions involving the liver, spleen, and bone marrow (Panel C). Other classic presentations include a lytic bone lesion (Panel D, arrow), cystic lung lesions (Panel E), and various skin lesions (Panels F through I). Examples of LCH lesions involving the skull and brain include multifocal skull lesions (Panel J, arrow), an orbital lesion (Panel K, arrow), a pituitary lesion (Panel L, arrow), and LCH-associated neurodegeneration (Panel M, arrow). (Figure from Allen CE et al. N Engl J Med 2018;379:856–868).

Skin, either isolated or with additional affected systems, is a common site of disease in all ages. The diagnosis of skin-limited LCH should not be presumed from a skin biopsy alone as 40% of children with presumptive skin-limited disease were diagnosed with multi-system disease on further investigation.⁴⁰ LCH limited to the skin has a very favorable outcome. In a retrospective study, patients with skin-limited disease had a 3-year progression-free survival (PFS) of 89% compared to 44% for patients with skin and multi-system involvement.⁴⁰ Notably, peripheral blood BRAFV600E+ cells are typically undetectable in patients with truly isolated skin LCH, but are found in patients with skin LCH as part of systemic disease.⁴⁰ This pattern, along with potential for isolated skin LCH to spontaneously resolve, support a concept of isolated skin LCH arising from tissue-restricted erythromyeloid precursors (tissue resident macrophages/epidermal LCs) while systemic disease arises from hematopoietic precursors. (Figure 3) Skin lesions vary in clinical presentation from dermatitis, vesicular eruptions, ulcerative lesions, or petechial rashes. Although any area of skin may be involved, LCH most commonly involves the scalp, axilla, and perineum. In infants, it commonly resembles seborrheic dermatitis and can also be mistaken for persistent cradle cap. A self-limited rash in infants known as Hashimoto-Pritzker is characterized by darker papular lesions. In older children, skin lesions more often develop in skin folds may be misdiagnosed as fungal infection.⁴¹ Nail involvement is rare, presenting as discoloration, hardening of nail beds or loss of nail tissue.⁴² Persistent otorrhea suggests otologic involvement, which can mimic infection. Oral manifestations may include ulcers and gingival hypertrophy with the potential for underlying bone involvement and tooth loss.⁴³

Bone disease may present either as uni- or multifocal disease and can be associated with soft tissue masses. While LCH can occur in any bone, children commonly present with lytic skull lesions manifesting as pain and/or swelling.⁴⁴ Vertebral bone involvement can cause vertebral collapse or vertebra plana.⁴⁵ Bone lesions affecting the skull base, including orbital, mastoid and temporal bones, increase the risk of CNS involvement.³⁸

Lung involvement can occur in any age. In contrast to children, lung LCH in adults can arise in smokers, presumably precipitated by reaction to smoke-related inflammation, and can resolve with smoking cessation alone.⁷ Lung involvement is found in 25% of children with multisystem disease. While pulmonary LCH can cause dramatic lung injury in some cases, lung disease is not considered an independent prognostic factor for moratlity.⁴⁶ Lung involvement may be asymptomatic and diagnosed on imaging as symmetrical areas of nodular fibrosis and bullae formation affecting the upper and middle lobes.⁴⁷ Rarely, patients complain of cough, shortness of breath or develop spontaneous pneumothoraces. In more severe cases, advanced lung damage can lead to hypoxia and restrictive lung disease with irreversible fibrosis in some cases.⁴⁸

Lymph nodes in the cervical chain are the most commonly affected and present as enlarged, matted, soft or hard nodes. Mediastinal involvement with thymic or lymph node infiltration is rare.⁴⁹

Central nervous system involvement can present with lesions in the hypothalamic-pituitary region, dural- based masses, infiltration in the choroid plexus or changes in white mater in the basal ganglia and cerebellum.⁵⁰ Involvement of the hypothalamic-pituitary region can be isolated or a component of multisystem disease. The posterior pituitary is affected most frequently and can lead to diabetes insipidus (DI). Approximately 50–80% of patients with pituitary involvement and DI will develop other manifestations of LCH^51–53 Patients with multisystem disease and craniofacial bone involvement at the time of diagnosis are at increased risk for developing DI (Relative risk 4.6).⁵⁴ On magnetic resonance imaging (MRI), loss of the pituitary bright spot, often with a nodular mass or thickening of the pituitary stalk is noted.⁵⁵ Panhypopituitarism can also develop with growth hormone deficiency occurring in 25% of patients with DI.⁵⁶ Dural-based mass lesions may be noted incidentally or present with symptoms of space-occupying lesions.⁵⁵

Neurodegenerative CNS disease (LCH-ND) develops in approximately 10% of patients with LCH. Patients with pituitary involvement and/or craniofacial bone lesions are at increased risk for LCH-ND, and >90% of patients with LCH-ND have BRAFV600E-driven disease^57–59 This complication typically involves the cerebellum, basal ganglia and pons. “LCH-associated abnormal CNS imaging” (LACI) describes LCH-ND with imaging changes only whereas “LCH-associated abnormal CNS symptoms” (LACS) describes LCH-ND with clinical symptoms. Patients with LACS have a variable clinical course, including dysmetria, tremor, ataxia, dysarthria, behavioral disturbances, cognitive disorders and/or psychosis which can occur many years after the initial diagnosis of systemic LCH. Imaging on MRI shows hyperintensity on T1 weighted images in the dentate nucleus with hyper- or hypointensity on T2 weighted images. MRI changes can precede neurologic symptoms, or in some cases can present in patients who remain asymptomatic with “LCH-associated abnormal CNS imaging” (LACI).³⁸ Some patients develop extension into the white matter of the cerebellum and frequently associated hyperintensity in the basal ganglia on T1 weighted images. Involvement of the pons is associated with particularly severe neurologic impairment.^50,60,61 LCH-ND has historically been considered an immune reaction or paraneoplastic phenomenon in patients with a history of LCH. However, recent studies identified peripheral blood BRAFV600E+ myeloid cells in patients with LCH-ND in the absence of systemic disease, and brain biopsies demonstrated BRAFV600E+ monocyte/macrophage cells enriched at sites of LCH-ND.⁵⁸ Further, in a recent mouse model, senescent BRAFV600E+ CD11A+ monocyte-derived meningeal macrophages were able to break through the blood-brain barrier and drive neurodegeneration in patterns similar to human LCH.³¹ These data together support a hematopoietic origin for clonal myeloid cells with the potential for peripheral precursors to migrate to the brain and promote chronic neurodegeneration in LCH-ND. Notably, BRAFV600E+ T cells are observed in peripheral blood of some patients who develop LCH-ND. It is not known if these T cells are vestigial long-surviving clones or if they might play a role in LCH-ND pathogenesis.⁶²

Bone marrow involvement in LCH presents as cytopenias associated with multisystem disease and carries an inferior prognosis.^63–65 Anemia is the most common cytopenia, followed by thrombocytopenia. Bone marrow findings include hemophagocytosis or increased macrophages, which may be reactive or harboring BRAFV600E or other MAPK gene mutations. Quantitative PCR and/or immunohistochemistry with BRAF VE1 (which stains for BRAF-V600E mutant protein) demonstrate mutated cells in the bone marrow with much higher frequency than conventional morphology or CD1a/Langerin immunohistochemistry.^14,66

Liver involvement can be seen in infants and toddlers, and it is commonly associated with multisystem disease. Transcutaneous biopsy of the liver may be non-diagnostic of CD1a⁺/Langerin⁺ cells given the large duct involvement, but typically will show biliary obstructive features. LCH involvement tends to occur around bile ducts and cause ductal sclerosis. Signs of liver dysfunction include direct hyperbilirubinemia, cholestasis, hypoalbuminemia, coagulopathy and ascites.^67,68 On radiographic imaging, involvement may be nodular or diffusely infiltrative. Sclerosing cholangitis is a rare sequela that may not respond to therapy in cases of long-standing liver damage (similar to entrenched LCH-ND). Liver transplant has been used with varying success.^69,70 Involvement of the spleen usually occurs along with the liver. Hypersplenism can cause cytopenias.

Intestinal involvement presents with bloody diarrhea, failure to thrive and malabsorption. Histological diagnosis can be difficult due to patchy involvement. However, demonstration of mutant histiocytic cell involvement, either bys ensitive molecular PCR or BRAF VE1 staining in known BRAFV600E LCH cases, may be used to infer intestinal disease.^71–73

Other organs less commonly involved in LCH include the thyroid, which can present with thyromegaly and/or hypothyroidism⁷⁴ and eyes, which can result in visual impairment.⁷⁵

LCH in Adults

Adults can develop LCH de novo or carry unresolved LCH from childhood. Clinical presentation and diagnostic approaches are similar between adults and children with some caveats.³⁹ In general, LCH arises in children due to accidents of hematopoietic development and in adults due to hematopoietic decline. Adults typically have more somatic mutations than children, though still with a relatively low burden.⁷⁶ LCH can arise in some cases along with evolving hematopoietic stem cell mutations during clonal hematopoiesis.⁷⁷ LCH arising as part of a “mixed histiocytic” disorder (e.g. ECD and LCH features in the same lesion or ECD and LCH histology in distinct lesions) occurs in approximately 5% of cases.⁷⁸ Isolated pulmonary LCH is a unique adult feature, almost exclusively in smokers. While data is lacking to describe quality of life and patient-perceived symptoms for adults with LCH, a patient-reported outcome study identified significant pain and fatigue in the majority of patients with ECD.⁷⁹ Patients with long-standing systemic disease and/or concerning neurologic symptoms and/or peripheral blood BRAFV600E+ cells may warrant MRI evaluation for LCH-ND.

Staging

A comprehensive evaluation for each patient is important to determine the extent of disease and organ systems involved. This is done through history, physical examination, laboratory studies and imaging. PET-CT is highly sensitive for systemic lesions.⁸⁰ PET-MR is also feasible if available and if the patient is able to tolerate the increased acquisition time.⁸¹ Skeletal survey and abdominal ultrasound may be used when PET-CT is not available and in infants for whom PET-CT may not be practical. MRI is preferred for evaluation of brain and spinal cord lesions, and T2-weighted brain MRI is optimal for LCH-ND.³⁸ Risks and benefits of any imaging study needs to take into account risks (e.g. radiation exposure, need for prolonged anesthesia) along with diagnostic sensitivity. Bone marrow aspirate and biopsy may be informative for patients with cytopenias at presentation and may be considered for any patient under 2 years of age with multisystem disease. Bone marrow aspirate and peripheral blood BRAFV600E qPCR or ddPCR can identify LCH precursors in patients with BRAFV600E+ lesions that may become useful for risk stratification and evaluation of minimal residual disease which is discussed below. For adults, bone marrow studies can also identify potential co-incident development of clonal hematopoiesis/myelodysplasia.^14,82–84

Current Therapeutic Strategies

Identifying optimal therapy for LCH has been confounded, until recently, by limited mechanistic understanding of the disease. Uncertainty of LCH as a reactive immune disorder versus neoplastic disorder (e.g., cancer, or cancer family) has led to questions regarding the optimal therapeutic approach (e.g., immune suppression vs chemotherapy), treatment goal (e.g., management of flare vs cure), and response nomenclature (e.g. “reactivation” vs “relapse”).

Front-Line – Limited Disease

Single lesions.

There have been few clinical trials to guide therapy for single LCH lesions, so practice is based largely on consensus and expert opinion.⁸⁵ Biopsy/curettage with or without local steroid injection may be sufficient therapy for single bone lesions.⁸⁶ Due to association of lesions in skull base and facial bones with risk of LCH-ND, systemic chemotherapy is recommended, even for single bone lesions in those locations. Similarly, systemic therapy is typically recommended for LCH involving the pituitary to decrease risks of developing DI and LCH-ND.³⁸

Bone lesions.

Patients with single site and multi-focal bone LCH should undergo a full staging work-up. A recent report recommended observation only in presumed isolated non-CNS risk skull lesions of the calvarium with the assumption that these will spontaneously resolve.⁸⁷ However, biopsy can diagnose other conditions (e.g. infection, vascular malformation, other malignancy) that may require urgent alternative therapy, and comprehensive imaging is important to confirm that a lesion is in fact “isolated”. The risks of biopsy are minimal, and the procedure may offer therapeutic benefit in true isolated LCH lesions. On the other side of the surgical spectrum, gross total resections are also not warranted because most single lesions with intact margins typically resolve and remodel following curettage/local steroid injection or systemic chemotherapy. Low-dose radiation therapy, which can also be effective in emergency situations for resolution of individual lesions, is mostly restricted to adult patients. Multifocal bone disease is a common presentation of LCH, and we typically consider single-system, multi-focal bone LCH to represent systemic disease.

Skin lesions.

As discussed above, skin LCH can represent the tip of a multi-system LCH “iceberg” or truly be skin-limited.⁴⁰ Most cases of skin-limited disease arise in infants and resolve spontaneously. Topical steroids, methotrexate and/or hydroxyurea may improve symptoms. Conventional systemic LCH therapy can be used in rare refractory cases (although these cases likely represent disseminated disease).

Front-Line – Systemic Disease

The Histiocyte Society Trials (LCH I-III) built on the DAL-HX studies (collaborations between European institutions) to test risk-adapted approaches through randomized trials (Table 3)^88,89. We consider the current standard of care for front-line therapy for pediatric patients with multifocal disease or single lesion in risk organ or CNS-risk site to be 1 year of vinblastine/prednisone (with 6-mercaptopurine (6MP) for HR LCH). This practice is based on the Histiocyte Society LCH-III trial.⁹⁰ A major conclusion from the LCH-III trial is that extended therapy (e.g. 12 vs 6 months) was associated with decreased risk of relapse in patients with multisystem LR LCH. We extrapolate this result to suggest that increased duration of therapy may be beneficial in patients with systemic LCH in general (e.g. multifocal single system and CNS-risk in addition to multisystem LCH). Verification of this hypothesis is being tested in LCH-IV.

Table 3. Front-line pediatric LCH trials.

PFS-progression-free survival; OS-overall survival; EFS-event-free survival; HR-high risk (risk-organ positive); LR-low risk (risk-organ negative); MS-multisystem; SS MF-single system multifocal bone; 6MP-6-mercaptopurine; vin-vinblastine; P-prednisone/prednisolone; Etop-etoposide; mo-month

Study	Time Period	Groups (number)	Therapies	Responses	OS	Conclusions
DAL-HX83 DAL-HX90 88	1983–1991	LR (Multifocal bone), n=28 LR (Bone + soft tissue), n=57 HR, n=21	Induction – vinblastine, prednisone, etoposide (all) Continuation vinblastine, prednisone, 6MP vinblastine, prednisone, etoposide vinblastine, prednisone, 6MP, etoposide, methotrexate 1 year	Disease resolution on study Total-86% A-89% B-91% C-67%	8 years Total-81% A-100% B-96% C-62%	Extent of disease at presentation is an important prognostic factor. Response to induction is an important prognostic factor
LCH-I 89	1991–1995	LR+HR, n=143	Single high dose methylprednisolone pulse for both arms Randomization: Vinblastine Etoposide 24 weeks	Response at last eval (median 4 year 11 month) Vinblastine – 58% Etoposide – 65% 3 year “reactivation after NAD” Vinblastine – 61% Etoposide – 55%	3 years Vinblastine- 76% Etoposide- 83%	No statistical difference in survival probability between groups. Worse 6 week response (50 vs 80%) and more relapse (50 vs 30%) than more intense and longer DAL-HX83. Lack of response by 6 weeks recognized as a poor risk factor.
LCH-II 37	1996–2001	LR+HR, n=193	Randomization Vinblastine, prednisone, 6MP Vinblastine, prednisone, 6MP, etoposide 24 weeks	1 year “NAD” Vin/P/6MP – 49% Vin/P/6MP/Etop– 62% 3 year “reactivation after NAD” Vin/P/6MP – 46% Vin/P/6MP/Etop– 46%	5 years Vin/P/6MP – 74% HR-64% LR-100% Vin/P/6MP/Etop – 41% HR-73% LR-100%	Therapy intensification associated with improved response rate and decreased mortality relative to LCH-I 6 months of therapy appeared inferior to historic 12 month DAL-HX studies Supported “high risk/RO+” and “low risk/RO-” designations
JLSG-96 91	1996–2001	SS MF and MS, n = 91	Prospective, non-randomized Cytarabine, Vincristine, Prednisone, methotrexate Intensification Regimen B if poor response 24 weeks	5 year EFS SS MF – 68.5% MS – 28.8%	5 years SS MF – 100% MS – 94.4%	Response rates and OS similar to contemporary front-line regimen Higher relapse rates may be due to short duration of treatment
JLSG-02 122,123	2002–2009	SS MF and MS, n = 91	Prospective, non-randomized Cytarabine, Vincristine, Prednisone, methotrexate Intensification regimen B if poor response 48 weeks	5 year EFS SS MF – 66.7% MS – 46.2%	5 years SS MF – 100% MS – 91.7 %	Longer duration of therapy associated with improved EFS in MS, but not in SS MF pediatric LCH patients
LCH-III 90	2001–2008	HR, n=235 MS LR, n=187	HR Randomization Vinblastine, prednisone, 6MP, Vinblastine, prednisone, 6MP, methotrexate 1 year LR Randomization Vinblastine, prednisone × 6 months Vinblastine, prednisone × 1 year	1 year “NAD” HR-no MTX – 48% HR-MTX – 48% 5 year reactivation after “NAD” LR-6 mo arm – 54% at 5 yr LR-12 mo arm – 37% at 5 yr	5 years HR-no MTX – 88% HR-MTX – 82% LR-6 mo – 100% LR-12 mo – 99%	Addition of methotrexate did not significantly impact outcomes for HR LCH Therapy prolongation decreased reactivation rates.

The majority of patients with multi-system LCH treated with one year of vinblastine/prednisone/6MP (according to LCH-III) will require salvage therapy.⁹⁰ This has prompted the question of whether vinblastine/prednisone/6MP should be considered the standard of care. However, prospective clinical trial data supporting alternative regimens is currently lacking. Additionally, likely due to its roots as a potential “inflammatory disorder”, initial failure has been tolerated for LCH more than is typical for front-line cancer therapy. The LCH-III-based therapy has proven to be relatively well-tolerated in children with low risk of long-term sequelae.⁹⁰

The Histiocyte Society LCH-IV trial tests hypotheses generated from prior European and Histiocyte Society trials, studies conducted through groups such as the Japan Langerhans Cell Histiocytosis Study Group (JLSG),⁹¹ as well as observations from retrospective series.^92,93 The LCH-IV front-line stratum is now evaluating the impact of 24 vs 12 months of vinblastine/prednisone and inclusion or exclusion of 6MP in pediatric LCH with prospective double-randomization (NCT02205762). Another multi-site study, LCH-REASON (NCT02670707) is evaluating vinblastine/prednisone/6MP versus cytarabine monotherapy. The recent shift in practice at some institutions to treat patients exclusively with MAPK inhibition is discussed below.

Chemotherapy Salvage/Hematopoietic Stem Cell Transplant

Data to guide treatment after front-line failure is limited (Table 4). Case series and early-phase studies demonstrate high response rates to nucleoside analogs (e.g., cladribine, cytarabine, clofarabine), but durability of response is related to dose, duration and intensity. For example, only 4% of patients treated with low-dose cladribine (5 mg/m2/day × 5 days/cycle) on the LCH-S-98 trial were cured after 6 months.⁹⁴ By comparison, in the LCH-S-2005 trial, the 5-year overall survival and progression-free survival were 85% using high-dose cladribine (9 mg/m2/day) with cytarabine (1 g/m2/day). However, this regimen was associated with significant toxicity (8% treatment-related mortality).⁹⁵ Case series support potential therapeutic potential of intermediate dose of single agent cytarabine⁹⁶ or clofarabine^97–99 that can typically be tolerated in outpatient settings. Oral hydroxyurea, a standard of care for myeloproliferative neoplastic disorders in the pre-tyrosine kinase inhibitor era, may also be effective for patients with LCH.¹⁰⁰ The relative importance and risks of intensity versus duration in LCH therapies is a fundamental question that remains unanswered. Prospective trials are needed to determine the “just right” agent, dose and schedule for specific risk groups.

Table 4. Salvage chemotherapy trials for pediatric LCH.

HR-high risk (risk-organ-positive); LR-low risk (risk-organ-negative); NAD-no active disease; ORR-overall response rate ((NAD + active disease-better); or (CR + PR)); OS-overall survival; PFS-progression-free survival; CR-complete response; PR-partial response; SD-stable disease; PD-progressive disease; TRM-treatment-related mortality

Study	Design	Subjects	Therapies	Response Timepoint	Outcomes	Conclusions
LCH-S-98 94	Phase 2	Total-83 46 HR 37 LR	Cladribine (5 mg/m2/day × 5 days) 21 day cycles 2–6 cycles	6 month 2 year	NAD – 4% LR – 10% HR – 0% ORR – 40% LR- 62% HR-23% OS – 68% LR-97% HR-48%	Lower dose cladribine achieved responses in many patients with relapsed/refractory LCH, but achieved durable complete responses in very few. An adult LCH series reported 67% ORR for cladribine salvage.124
LCH-S-2005 95	Phase 2	Total – 27 HR	Cladribine (9 mg/m2/day) + cytarabine (1 g/m2/day) × 5 days 28 day cycles	128 day (median) 2 month 5 year	NAD – 85% ORR – 92% OS – 85%	Very high dose nucleoside analog combination therapy yielded relatively high response rates, and also significant rates of treatment-related mortality (7.4%). Most patients successfully completing therapy were “cured” of LCH.
Cytarabine monotherapy 96	Retrospective	Total – 22 6 HR 16 LR	Cytarabine (100–170 mg/m2/day) × 5 days every 4 weeks for 6–12 months.	3 month 3 year 3 year	ORR – 59% PFS-41% OS-100%	Intermediate (typically outpatient) dose cytarabine monotherapy can elicit durable responses in patients with relapsed and refractory LCH.
Clofarabine monotherapy 99	Retrospective	Total – 11 3 HR 8 LR	Clofarabine (25 mg/m2/day) × 5 days every 4 weeks for 6–12 months.	2 month 1 year 1 year	ORR – 73% PFS-76% 91%	Intermediate (typically outpatient) dose clofarabine monotherapy can elicit durable responses in patients with relapsed and refractory LCH. Other series with similar outcomes.97,98
Hydroxyurea oral therapy 100	Retrospective	Total – 15 Adult – 12 Pediatric – 3 HR-2 LR-13	Hydroxyurea (500 mg twice daily or 20 mg/kg) 5 also treated with oral methotrexate	1–24 months	ORR – 80% CR-8 PR-4 SD-2 PD-1	Hydroxyurea was safe and elicited durable responses in some patients who had persistent or progressive disease following multiple previous therapies.100
Hematopoietic Cell Transplant 101	Retrospective	Total – 87 MAC – 20 RIC - 67	HSCT with MAC or RIC	3 year	OS MAC – 77% RIC – 71% TRM MAC-28% RIC-8%	HSCT offers potential for cure for patients with relapsed and refractory LCH. Myeloablative conditioning is associated with higher treatment-related mortality; reduced intensity conditioning is associated with higher rates of relapse.

For highly refractory high-risk cases, hematopoietic stem cell transplant (HSCT) which replaces the neoplastic clone may be curative. Myeloablative conditioning (MAC) and reduced Intensity (RIC) approaches have similar historic overall survival, though MAC was associated with increased treatment-related mortality and RIC was associated with increased risk of relapse.¹⁰¹ Utilization of HSCT for LCH has decreased significantly as more refractory HR patients are now treated with MAPK inhibition, on- and off-study.

Treatment of Adults with LCH

Adults are frequently intolerant to the pediatric front-line vinblastine/prednisone therapy. Expert opinion therefore recommends methotrexate, cytarabine or cladribine monotherapy for front-line treatment of adults with LCH who require systemic therapy.^102,103 Vemurafenib is FDA-approved for treatment of BRAFV600E+ ECD, and cobimetinib is now approved for adults with histiocytic neoplasms. In patients where a quick response is desirable, vemurafenib may be used for BRAFV600E mutated disease. Empiric MEK inhibition is also a reasonable treatment if mutation status is unknown. For adult smokers with isolated pulmonary LCH, smoking cessation may be sufficient. Bisphosphonates or gabapentin may be helpful for LCH-associated pain. A recent study demonstrated an 80% ORR in adults with bone LCH treated with the RANKL inhibitor denosumab.¹⁰⁴

LCH-ND

LCH-ND is a highly significant remaining clinical challenge for LCH patients and their physicians due to unpredictable onset and clinical consequences of progressive neurodegeneration. For patients who are cured of systemic LCH, uncertainty of risk for LCH-ND can be a sword of Damocles. As discussed above, recent biological insights demonstrate LCH-ND as a manifestation of clonal LCH.⁵⁸ Management has historically been aimed at supportive measures including physical therapy and IVIG.³⁸ Clinical series report potential efficacy of LCH-directed therapies including cytarabine¹⁰⁵ and vemurafenib;⁵⁸ however, responses to MAPK inhibition are not as reliable or complete as for systemic LCH. Further, potential for significant clinical improvement may be limited in patients with established brain injury. If LCH-ND arises from hematopoietic precursors, then systematic approaches to prevention (e.g., clinical risk-stratification, peripheral blood BRAFV600E testing, serial MRI imaging) may become important to avoid development of entrenched pathogenic BRAFV600E+ CNS macrophages.^31,58

MAPK Pathway Inhibition

Discoveries over the past 13 years have demonstrated that MAPK hyperactivation in myeloid precursors plays a central role in driving LCH lesion formation and pathogenesis. Blocking MAPK hyperactivation therefore offers a rationale “hammer” for this prominent, newly exposed “nail” (Table 5). Adults with LCH and ECD were included in the Memorial Sloan Kettering Phase 2 basket trial for the BRAFV600E inhibitor vemurafenib. Responses were good when measured by RECIST criteria (ORR 61%) and extraordinary when measured by PET-CT (ORR 100%).^106,107 Similar experiences were reported in a French series of adults with ECD and mixed ECD/LCH.¹⁰⁸ It should be noted that vemurafenib, dabrafenib and other first-generation BRAF inhibitors specifically block BRAFV600E, with paradoxical activation of MAPK signaling in normal cells. This limits the potential patient population and contributes to toxicities.¹⁰⁹ However, no such activation accompanies MEK inhibition, where toxicities are related to MAPK attenuation in normal cells. Diamond and colleagues developed a trial to treat “all-comers” with adult histiocytosis with MEK inhibitors based on the rationale that almost all activating mutations associated with histiocytic disorders arise in genes encoding MEK or proximal MAPK pathway genes. This trial demonstrated a metabolic ORR 91% regardless of mutational status in adults and led to FDA approval of cobimetinib for histiocytic disorders in adults. The first child treated with a MAPK inhibitor was reported by Heritier and colleagues in 2015 with dramatic response.¹¹⁰ Subsequent pediatric series supported reproducible high response rates to MAPK inhibitors, even in patients with aggressive disease.^83,84,111 Some patients with LCH-ND also responded, though potential for recovery in patients with long-standing disease is perhaps limited by drug penetration of the blood-brain barrier and pre-existing CNS injury.⁵⁸

Table 5. MAPK inhibition studies for LCH and related conditions.

HR – high risk (risk-organ-positive); LR – low risk (risk-organ negative; LCH-ND – neurodegenerative LCH; ORR – overall response rate; PFS – progression free survival; CR – complete response; PR – partial response.

Study	Design	Subjects	Therapies	Response Criteria	Response Timepoint	Outcomes	Conclusions
VE-Basket 106,107	Phase 2 Clinical trial	26 Total - Adults 22 ECD 4 LCH	Vemurafenib	RECIST PERCIST	2–40 months 2–40 months	ORR – 61.5% ORR – 100%	Consistent responses to vemurafenib in adults with BRAF-V600E+ LCH and ECD Frequent but generally manageable toxicities (skin, GI); rare ocular, secondary squamous cell carcinoma. No evidence of acquired resistance to therapy. PET-CT-based responses more sensitive than RECIST.
Targeted Therapy for ECD 108	Retrospective	8 Total - Adults 4 ECD 4 mixed ECD/LCH	Vemurafenib	PET response	6 months 10.5 months (median)	ORR-100% PFS 100%	Consistent responses in vemurafenib adults with BRAF-V600E+ LCH and ECD
LOVE Study 112	Retrospective series	54 Total, all ECD Adults 20 in cessation cohort	Various MAPK inhibitors	PERCIST	6 months after cessation of MAPK inhibition therapy	PFS 25%	Rapid relapse in majority of patients with ECD with cessation of MAPK inhibition.
Cobimetinib for Histiocytoses 125	Phase 2 clinical trial	18 total - Adults 12 ECD 2 LCH 2 RDD 2 mixed	Cobimetinib	PERCIST	1.6–15.9 months	ORR – 88%	Cobimetinib (MEK inhibitor) is effective in the majority of adults with histiocytic disorders with a spectrum of driver mutations. This study supported FDA approval for cobimetinib for histiocytic disorders in adults.
Pediatric Case Study 110	Case study	1 HR LCH case - Pediatric	Vemurafenib	RECIST	Day 60	CR	1st pediatric case of severe HR LCH with positive response to vemurafenib.
NACHO-LIBRE 84	Retrospective series	21 LCH - Pediatric 13 LCH-ND without systemic lesions 8 Systemic LCH	MAPK Inhibitors	Modified PERCIST	0.6–48 months	ORR-86%	High response rates in pediatric cohort with LCH-ND and systemic LCH to MAPK inhibition. Persistent peripheral blood mononuclear cell BRAFV600E were detected in patients on MAPK inhibitor therapy, even in patients with positive clinical responses.
European Vemurafenib Study 83	Observational study	54 LCH - Pediatric 44 HR 10 LR	Vemurafenib	Disease Activity Score	8 weeks 0.1 – 7.3 months after cessation of vemurafenib	ORR – 100% 38 CR 16 PR PFS 20%	High response rate in pediatric cohort with systemic LCH. Rapid relapse with cessation of therapy
MAPK Inhibitors for Histiocytic Disorders 111	Retrospective series	30 LCH – Pediatric (n=27) and Adult (n=3) 16 front-line 14 salvage	Dabrafenib (10) Trametinib (17) Trametinib/Dabrafenib (3)	Histiocyte Society Criteria (including PET-CT)	0.3 month-6.4 years	ORR-100% 25/30 (83%) LCH patients achieved NAD without persistent symptoms of disease	High response rate to front-line or salvage MAPK inhibition in largely pediatric cohort with LCH.
Dabrafenib and Trametinib Pediatric Study 113	Phase 2	25 LCH – Pediatric 13 Dabrafenib monotherapy 12 Dabrafenib + Trametinib	Dabrafenib Dabrafenib + Trametinib	Histiocyte Society Criteria	7–65 months 1.8–35.9 months	ORR Dabrafenib – 76.9% Dual therapy – 58.3%	Responses likely lower than comparable pediatric cohorts due to HS Response Criteria versus PET-based approaches. Dual therapy did not appear to add benefit in this study.

These early experiences with MAPK inhibitors have been game-changing for some patients with LCH. Death in infants with HR LCH is becoming increasingly rare. However, the LOVE study, in which adults with ECD were followed after achieving complete responses with MAPK inhibitors, signaled a cautionary note with >75% of patients relapsing within 6 months of stopping therapy.¹¹² Similarly the European vemurafenib study also reported rapid relapse in the majority of patients following treatment cessation.⁸³ Unlike other conditions with more complex mutation landscape, resistance to MAPK inhibition does not typically develop in LCH, and patients respond to retreatment with the same MAPK inhibitor following relapse or progression off-therapy. Where combining MAPK inhibitors (e.g. BRAFV600E and MEK inhibitors) have demonstrated superior outcomes in more genomically complex disorders such as melanoma, dabrafenib alone and dabrafenib/trametinib yielded similar outcomes in pediatric LCH patients.¹¹³

Chaos and Opportunity

LCH Treatment Approaches: 2024

Prior to 2010, when we had few insights into mechanisms of LCH pathogenesis, there was near global consensus regarding optimal therapeutic approaches for systemic LCH, based on the best results of Histiocyte Society trials. Over the past decade, the pace of discovery has outrun our clinical trial capacity. There are numerous and increasingly available MAPK inhibiting agents that are being used in various settings (e.g. front-line, second-line, beyond), in different types of patients (e.g. LCH, LCH-ND, ECD, RDD, etc.), and in different combinations (single agent, combined targeted agents, combined with cytotoxic chemotherapy, etc.). At the same time, clinical trials to rigorously define responses and toxicities associated with MAPK inhibitor therapy in LCH are limited, especially in children. Therapy recommendations and level of evidence for the range of LCH scenarios are outlined in Table 6. We strongly support enrollment of patients on clinical trials when possible. In the meantime, we favor the conventional approach of using treatment strategies with the best evidence: For front-line therapy, this remains the LCH-III-supported approach of vinblastine/prednisone. However, we acknowledge that in cases where patients are critically ill, such as infants with rapidly progressive multi-system disease and bone marrow and/or liver involvement requiring immediate responses, off-label and off-study use of MAPK inhibition for some period may be reasonable. Standardized treatment recommendations become more difficult for salvage therapy, where data are retrospective and anecdotal.

Table 6.

LCH Treatment Strategies

Front-line	Recommendation	Evidence
Single bone	Curettage +/− local steroid injection (Do NOT aim for gross total resection)	Expert opinion
Single bone CNS risk Not surgically accessible/morbidity risk	Vinblastine/prednisone × 12 months (according to LCH-III)	Extrapolated from randomized control trial (LCH-III)
Multifocal bone	Vinblastine/prednisone × 12 months (according to LCH-III)	Extrapolated from randomized control trial (LCH-III)
Multisystem (Low-risk)	Vinblastine/prednisone × 12 months (according to LCH-III)	Randomized control trial
Multisystem (High-risk)	Vinblastine/prednisone/6MP × 12 months (according to LCH-III)	Randomized control trial
Multisystem (High-risk) Meets HLH criteria Critically ill	Consider MAPK inhibition to manage acute severe symptoms (ideally on clinical trial)	Expert opinion
1st Salvage	Recommendation	Evidence
Recurrent single bone	Vinblastine/prednisone × 12 months – or- Cytarabine monotherapy (100 mg/m2/day × 5 days/cycle) × 12 months if previous exposure to vinblastine/prednisone. (If single bone progresses or relapses, we consider this evolution of systemic disease)	Extrapolated from randomized control trial (LCH-III) Retrospective series
Multifocal bone	Cytarabine monotherapy (100 mg/m2/day × 5 days/cycle) × 12 MAPK inhibitor if on clinical trial	Retrospective series Expert opinion
Multisystem (Low-risk)	Cytarabine monotherapy (100 mg/m2/day × 5 days/cycle) × 12 MAPK inhibitor if on clinical trial	Retrospective series Observational cohort
Multisystem (High-risk)	Cytarabine monotherapy (100 mg/m2/day × 5 days/cycle) × 12 -or- Clofarabine monotherapy (25 mg/m2/day × 5 days/cycle) × 6; if CR at 6 months, then 2 days/cycle × 6 more cycles; if PR at 6 months, then continue 5 day cycles × 6 more cycles. – or - MAPK inhibitor if on clinical trial; or MAPK inhibitor off-study if disease is rapidly progressive or if patient is critically ill.	Retrospective series Retrospective series Observational cohort
2nd Salvage and Beyond	Recommendation	Evidence
Systemic relapse/progression	Clofarabine monotherapy (25 mg/m2/day × 5 days/cycle) × 6; if CR at 6 months, then 2 days/cycle × 6 more cycles; if PR at 6 months, then continue 5 day cycles × 6 more cycles. – or – MAPK inhibitor, or other therapeutic strategy on clinical trial. – or - MAPK inhibitor off-study	Expert opinion Observational cohort Observational cohort
Special Cases	Recommendation	Evidence
LCH-ND: LACI	Observe, confirm progression of abnormal MRI findings with 2 or more worsening MRI over 3 months Cytarabine monotherapy (150 mg/m2/day × 5 days/cycle) × 12 -or- MAPK inhibitor, or other therapeutic strategy on clinical trial. – or - MAPK inhibitor off-study	Expert opinion Retrospective series Observational cohort Observational cohort
LCH-ND: LACS	Cytarabine monotherapy (150 mg/m2/day × 5 days/cycle) × 12 -or- MAPK inhibitor, or other therapeutic strategy on clinical trial. – or - MAPK inhibitor off-study	Retrospective series Observational cohort Observational cohort
Adult LCH	Same as above with exceptions: Cytarabine or cladribine monotherapy as front-line for systemic disease MAPK inhibition as 2nd line and greater for systemic disease	Observational cohort Phase 1–2 clinical trials
Skin-limited LCH	Observe If symptomatic, consider topical steroids and/or oral methotrexate and/or oral hydroxyurea as needed If persistent/progressive, be sure to image to rule out multi-system LCH	Expert opinion
Systemic disease in lower resource settings	Consider oral hydroxyurea or subcutaneous cytarabine	Retrospective series

Even with limited data, the efficacy of MAPK inhibitors in most LCH patients seems predictable. However, associated toxicities are unknown. Thus, balancing the costs and benefits for patients with diverse clinical manifestations and across risk groups presents a significant challenge. While acute toxicities are typically manageable, many patients taking MAPK inhibitors experience side effects. Skin rash and GI symptoms are common and severe toxicities including the development of second malignancies have been reported.¹¹⁴ Additionally, identification of other potential side effects of chronic inhibitor therapy, including impact on growth and development in children are not known. Prospective clinical trials with extended follow-up are therefore essential to understand the impact of extended use of MAPK inhibitors in children and adults.¹¹⁵ Traditional chemotherapy also has predictable toxicities, but responses for most patients with LCH do not appear as reliable or rapid compared to MAPK inhibition. However, less than 50% of patients with disseminated LCH are cured with 1 year of vinblastine/prednisone/6MP treatments (48% with “no active disease” status at 1 year in LCHIII).⁹⁰ Decisions regarding therapy with traditional chemotherapy with the aim for cure versus indefinite management with MAPK inhibition are being made with limited data describing the safety and efficacy of these novel targeted agents. Data on the patient experience with therapy alternatives will be critically important to guide treatment decisions and to optimize quality of life during and after therapy.

The North American Consortium for Histiocytosis (NACHO) consortium developed the NACHO-COBI trial (NCT04079179), to test the safety and efficacy of cobimetinib monotherapy in 1)children and young adults with relapse/refractory LCH; 2)children and adults with LCH-ND; 3)children and young adults with non-LCH histiocytoses (e.g. Rosai-Dorfman disease, juvenile xanthogranuloma, histiocytic sarcoma); and 4)adults with histiocytic disorders (LCH and non-LCH). The first LCH trial in the Children’s Oncology Group (COG) recently opened (ANHL2121, NCT05828069), evaluating the safety and efficacy of a second-generation pan-RAF inhibitor (tovorafenib) in children and young adults with relapsed and refractory LCH. A prospective single center phase 2 trial is testing the MEK inhibitor mirdametinib in newly diagnosed and relapsed and refractory LCH (NCT06153173). While off-label access to MAPK inhibitors in pediatric patients with histiocytic disorders is increasing in the United States, prospective clinical trials are essential to rigorously assess efficacy and toxicity and establish standards of care. Importantly, these studies will accelerate our understanding of biological and clinical correlates of disease response, resistance and outcomes.

Opportunities

After decades of incremental advances, mechanistic insights now offer footholds for opportunities to optimize therapies (and cures) for every patient with LCH.

Risk stratification.

Histiocyte Society clinical trials identified patients at highest risk of death from LCH based on presenting sites of disease.^37,90 Molecular data may now inform refined risk stratification strategies. For example, somatic BRAFV600E mutations in LCH lesions have been associated with increased risks of systemic disease and LCH-ND.^14,58,59 Further, detection of BRAFV600E in blood via cell-free DNA or in PBMCs is also associated with systemic disease, with BRAFV600E allele frequency tracking disease burden in patients treated with chemotherapy.^14,82 Custom ddPCR primers can be used to characterize similar patterns in patients with alternative MAPK pathway mutations.¹¹⁶ The limited correlation between clinical responses and BRAFV600E allele frequency in blood in patients treated with MAPK inhibition is consistent with rapid relapse with therapy cessation.^83,84 MAPK inhibitors therefore appear to be set to “stun” rather than “kill”, suppressing the distal pathway of LCH development rather than eliminating the precursor pool. Inclusion of “minimal detectable disease” and “minimal residual disease” strategies on prospective clinical trials will help develop valuable risk stratification tools to guide therapy. Further, the ability to predict risk of LCH-ND may guide surveillance strategies.

New Agents and Combinations.

After decades of vinblastine and prednisone in pediatric disease, our current challenge is prioritizing the clinical testing of dozens of novel agents with potential efficacy in LCH. For most drugs currently used in LCH salvage, the best available data are limited to case series. The general signal of MAPK inhibitor activity in LCH is revolutionary, but we do not know how distinct pharmacodynamic features or distribution (e.g., CNS penetration) impact efficacy, toxicity and durability of response. Beyond chemotherapy and MAPK inhibitors, emerging biology supports the potential for alternative therapeutic strategies including targeting the survival advantage induced by MAPK activation (e.g., apoptosis dis-inhibition with BH3 mimetics)³⁰, other senolytic agents to inhibit the senescence-associated secretory phenotype (e.g. sirolimus)³⁰, differentiation blockade (e.g., HDAC inhibitors)¹⁴, antigen targets (e.g. CSF1R, CD207, SIRPα)^117,118, and immune checkpoint inhibitors (e.g. PD-1)¹¹⁹. Chemotherapy for LCH is not always effective but provides the chance of cure, whereas MAPK inhibition is highly active but does not seem to cure. Therefore, identifying how to most effectively combine agents to achieve safe, reliable and durable cure is a key priority. Combining intermediate-dose cytarabine/cladribine with vemurafenib was feasible in a prospective cohort (n=19) with 2-year “reactivation/progression-free survival” (RFS) 76.9% for risk-organ+ (RO+) and 83.3% for RO-; one patient developed myelodysplastic syndrome with uncertain relationship to therapy¹²⁰. Pre-clinical models (primary cell culture and mice) may be able to contribute to drug and drug combination prioritization.^30,33

Quality of Life/ Survivorship.

There are now many paths to prevent death in patients with LCH. Current front-line practices range from a year of chemotherapy to a life of MAPK inhibition. Which is better? Current response evaluations are focused on ORR and PFS. If potential for death has been largely eliminated, clinical trials may not answer the most important questions for patients regarding their experiences on therapy and after. Building patient-reported quality of life data into clinical trials will be essential to understand which approaches are the “best”. Similarly, systematic collection of epidemiologic data and long-term outcomes is required to understand the life-long impacts of this disease and different treatments.

Access to Therapy.

Most patients in the world who develop histiocytic disorders do not have access to centers or drugs required for chemotherapy infusions or targeted oral therapies. Oral hydroxyurea offers an opportunity to treat patients with “smoldering” disease – and may offer insights into the relative benefit of duration of cytotoxic therapy against myeloid precursors versus intensity.¹⁰⁰ As another option for rural or lower-resource settings, cytarabine may be given subcutaneously, avoiding need for frequent IV access or port placement.¹²¹

Organizational Collaboration and Standardized Endpoints.

Prospective clinical trials are essential to identify optimal therapies for LCH. NACHO was created to catalyze improvements in outcomes for patients with histiocytic disorders. We grew from 11 original members in 2014 to 65 current member institutions. This growth was not initially planned or expected, reflecting the great interest in participating in LCH research and previously unmet need. Initiation of the European Consortium for Histiocytosis (ECHO) and other regional/national groups, and inclusion of LCH in COG are tremendous opportunities to advance research. Collaboration and coordination are essential to amplify the impact of research organizations where multiple complementary studies can be conducted in parallel. Harmonized endpoints and common data dictionaries will be important to facilitate comparisons between studies.

Conclusions

After more than 100 years of developing therapeutic strategies for patients with LCH in the “dark”, we now have the challenge of testing, prioritizing and implementing strategies based on mechanistic insights that are becoming rapidly illuminated. In comparison to the frenetic pace of biological discoveries, the clinical trial and regulatory environments have limited acceleration capacity. We therefore have the challenge to strategically prioritize and test novel concepts as efficiently and effectively as possible. In the meantime, as patients continue to arrive in our offices, we rely on conventional approaches to evaluate existing clinical data and contribute patient experiences to cohorts and trials whenever possible. Given that many new therapies now effectively avoid death and produce unprecedented response rates, the voices of our patients and families are increasingly important to inform the best experiences during and after therapy. Our goal over the next decade is to identify and implement the best possible therapy for each patient with LCH to ensure long-term survival with optimal quality of life.

Table 2.

Differential diagnoses and LCH.

Area of Involvement	Clinical Manifestation	Possible Differential Diagnoses
Skin	Dermatitis	Seborrheic dermatitis
	Vesicles	Varicella, herpes simplex, erythema toxicum
	Petechiae	Immune thrombocytopenia, leukemia
	Ulcerative lesions	Fungal infection
	Nodules	Juvenile xanthogranuloma, infant leukemia, mastocytosis, neuroblastoma, subcutaneous panniculitis-like T cell lymphoma
Bone	Lytic lesions/vertebra plana	Acute osteomyelitis, Chronic relapsing multifocal osteomyelitis, Atypical mycobacterial infection, Bone angiomatosis (Gorham disease), Aneurysmal bone cyst, Juvenile xanthogranuloma, Malignancy
Lung	Cavitary nodules	Mycobacterial or other infection, sarcoidosis Pneumocystis jirovecii
Liver	Jaundice/hypoalbuminemia	Hepatitis, metabolic disease, malignancy, chronic active EBV, NK/T cell lymphoma, toxic injury
Pituitary	Diabetes insipidus	Central nervous system germ cell tumor, hypophysitis