Langerhans cell histiocytosis

Meet the patient — the cradle cap that will not go

A four-month-old infant is referred for a scaly, crusted, yellow-brown eruption across the scalp, retroauricular folds, and diaper area that has resisted three weeks of emollients, topical antifungals, and mild steroids. Look closer: scattered petechiae sit within the crusted papules, and there are a few weeping flexural lesions. This is not refractory seborrhoeic dermatitis — it is cutaneous LCH, and the petechiae are the clue that should pull the trigger on a biopsy.[1][4]

Two questions decide this child's prognosis: is more than one organ system involved? and are the risk organs — liver, spleen, or bone marrow — in the disease? A single-system skin or bone lesion is usually curable; risk-organ multisystem disease in an infant is the form that kills. Hold those two questions and the staging, the biopsy, and the treatment all fall into place.[1][5]

It is a myeloid neoplasm — that single fact reframes everything

For decades LCH was debated as reactive versus neoplastic, hidden under the old name "histiocytosis X." That debate ended when BRAF V600E was found in about half of cases, and other MAPK-pathway mutations in nearly all the rest. LCH is a clonal neoplastic proliferation of bone-marrow-derived dendritic-cell precursors that are arrested at an immature myeloid stage — not mature epidermal Langerhans cells gone rogue.[1][3]

The cell of origin is a CD34-positive myeloid progenitor driven toward a Langerhans-cell phenotype but frozen partway. That is why risk-organ multisystem disease carries a higher mutant allele frequency than a single bone lesion — more of the myeloid precursor pool carries the mutation.[3]

The molecular subgroups a candidate must name:[1]

• BRAF V600E — the commonest driver, in about 50 to 60 percent; constitutive RAS-RAF-MEK-ERK activation, linked to higher relapse and chronic-disease risk but not precluding cure.[1][3]
• MAP2K1 (MEK1) — in about a quarter, mutually exclusive with BRAF; responds to MEK inhibition.[1]
• Other MAPK alterations — ARAF, MAP2K2, NRAS, KRAS, and BRAF fusions in the remainder; collectively the MAPK pathway is activated in nearly every case.[1][3]

Why the genetics matters clinically: the same BRAF V600E appears in hairy cell leukaemia, melanoma, papillary thyroid carcinoma, and Erdheim-Chester disease — confirming a true neoplastic basis — and it is the reason a refractory patient now receives vemurafenib or dabrafenib rather than yet another cytotoxic cycle.[8][9]

Classification — single-system, multi-system, and the risk organs

LCH is classified by how many organ systems are involved and whether the risk organs are among them. This single axis does most of the prognostic work.[1][5]

[1] [1]

The skin — where LCH most often first shows itself

Skin involvement is the commonest presenting feature in infants and children, and it is the sign most often misread as a banal dermatitis. The morphology to recognise:[1][4]

• Seborrhoeic dermatitis-like eruption — crusted, scaly, yellow-brown papules and plaques on the scalp, retroauricular area, flexures, and the diaper, inframammary, or perineal regions. The misdiagnoses pile up: seborrhoeic dermatitis, atopic dermatitis, irritant diaper dermatitis.[1]
• Petechiae and purpura within the crusted papules, especially in flexural and perineal skin — the characteristic clue that separates LCH from a plain sebopsoriasis.[4]
• Vesiculopustular variant — may mimic varicella or impetigo in neonates.[1]
• Nodular variant — reddish-brown nodules, sometimes ulcerated.[4]
• Oral mucosa — gingival hypertrophy, mucosal ulceration, and jaw involvement with the classic "floating teeth."[1]
• Congenital self-healing reticulohistiocytosis (Hashimoto-Pritzker) — present at birth with multiple reddish-brown nodules that self-resolve over weeks to months; still a form of cutaneous LCH that demands systemic staging.[3]

The classic trap: every infant with "seborrhoeic dermatitis" that has failed standard treatment, and especially one with petechiae in the flexural papules, earns a skin biopsy with CD1a and Langerin — not a fourth empirical cream.[1][4]

The organs — bone, pituitary, risk organs, lung, and brain

LCH can appear in almost any organ; the patterns below are the ones examiners test.[1][5]

Bone is the commonest organ involved. A solitary lytic lesion — eosinophilic granuloma — is the single most frequent LCH manifestation, and usually the one with the best outlook. The radiological signs are viva gold:[1][4]

• Skull — a punched-out lytic lesion, the "geographic skull" or map-like appearance on X-ray.
• Jaw — alveolar bone loss producing "floating teeth" on X-ray.
• Spine — vertebral body collapse into vertebra plana, the flat vertebra.
• Long bones, pelvis, ribs — painful lytic lesions that may fracture.
• Presentation is localised pain, swelling, and tenderness.[1][4]

The pituitary is the endocrine organ to know. Infiltration of the posterior pituitary or stalk produces central diabetes insipidus — polyuria and polydipsia — the commonest endocrine manifestation, and one that is often permanent even after treatment. Screen with a water deprivation test and pituitary MRI; anterior deficits such as growth hormone deficiency occur less often.[1][5]

The risk organs are what make the disease dangerous. Hepatosplenomegaly with cholestasis or transaminitis, and marrow cytopenias — anaemia, thrombocytopenia, neutropenia — define risk-organ involvement and the worse prognosis, especially in children.[1]

Pulmonary LCH in adults is almost always smoking-related. Over 90 percent of adult pulmonary LCH patients are current or former smokers; the neoplastic Langerhans cells accumulate in the small airways, producing an upper-lobe-predominant cystic and nodular pattern on high-resolution CT, with a 10 to 20 percent risk of spontaneous pneumothorax. Smoking cessation can stabilise it, and isolated pulmonary LCH is usually single-system in adults.[5]

Neurodegenerative LCH — cerebellar ataxia, cognitive and behavioural decline, pyramidal and extrapyramidal signs — may appear years after disease control, is strongly BRAF-associated, and has no effective treatment.[1]

The historical syndromes — one disease, three faces

The eponymous syndromes are now unified under LCH, but a fellowship examiner still expects all three named and ordered by prognosis.[1]

• Eosinophilic granuloma — a solitary bone lesion; the commonest presentation and the best prognosis.[1]
• Hand-Schüller-Christian disease — the classic triad of skull lytic lesions, diabetes insipidus, and exophthalmos from orbital involvement; chronic multisystem, typically in children and young adults.[1]
• Letterer-Siwe disease — acute disseminated multisystem disease in infants under two years: skin plus liver, spleen, and marrow involvement, lymphadenopathy, and fever. The worst prognosis.[1]

[1]

Histopathology — coffee-bean nuclei and the tennis racket

The histological hallmark is the Langerhans cell, and the diagnosis requires it proven by immunohistochemistry — never morphology alone.[1][4]

The three layers of evidence, in the order they confirm the cell:[1][4]

• H and E — sheets and infiltrates of large cells with grooved, reniform "coffee-bean" nuclei and abundant eosinophilic cytoplasm, admixed with eosinophils (sometimes so numerous they name the disease), lymphocytes, neutrophils, and multinucleated giant cells.[1]
• Electron microscopy — Birbeck granules, the pathognomonic cytoplasmic organelle with a rod-shaped or zipper-like body and a dilated terminal end resembling a tennis racket, composed of Langerin.[1]
• Immunohistochemistry — the modern diagnostic standard. CD1a-positive is diagnostic; CD207 or Langerin-positive is the most specific marker, a C-type lectin that forms the Birbeck granule itself; S100-positive is sensitive but not specific. Diagnosis requires CD1a or Langerin positivity.[4]

[1]

The discriminator that ends the workup: a histiocytic infiltrate that is CD1a-negative is not LCH — redirect the diagnosis toward the non-Langerhans family (juvenile xanthogranuloma, Rosai-Dorfman-Destombes, Erdheim-Chester), which are CD68-positive, CD163-positive, and Factor XIIIa-positive.[2][3]

Differential diagnosis — the mimics that delay the biopsy

Cutaneous LCH in infants is frequently misdiagnosed as common inflammatory dermatoses; bone LCH mimics infection and malignancy. A biopsy with CD1a and Langerin is the deciding test in equivocal cases.[1][3][4]

[1]

Investigations and staging

Tissue biopsy is the gateway — skin, bone, or lymph node — with histology and the CD1a, Langerin, and S100 panel. Staging then determines single- versus multi-system disease and plans treatment.[1][5]

• Skeletal survey or whole-body MRI or PET-CT — to detect bone lesions across the skeleton.
• Full blood count with differential — to screen for marrow involvement and cytopenias.
• Liver function tests — to screen for hepatic involvement and cholestasis.
• Water deprivation test plus pituitary MRI — for diabetes insipidus, with urine osmolality.
• Chest high-resolution CT — for pulmonary LCH, looking for cysts and nodules.
• Endocrine assessment — growth hormone and anterior pituitary axes.
• BRAF V600E mutation testing on tissue — guides targeted therapy in refractory disease and is now standard at diagnosis in multisystem or risk-organ disease.[1][3][5]

Management — match the treatment to the extent

Treatment is dictated by extent and risk-organ status — never by a reflex to systemic chemotherapy for a solitary bone lesion that may simply be observed.[1][5]

Single-system skin LCH is treated topically — potent corticosteroids, topical nitrogen mustard (mechlorethamine), topical calcineurin inhibitors, or phototherapy (narrowband UVB or PUVA). The prognosis is good, but the patient is monitored for progression to multisystem disease.[1]

Single-system bone LCH (eosinophilic granuloma) often needs little: many solitary lesions regress spontaneously. Symptomatic or structurally important lesions are managed with curettage and bone grafting, intralesional corticosteroid injection (triamcinolone), bisphosphonates for pain, or rarely localised radiotherapy for inaccessible or refractory lesions.[5]

Multi-system LCH requires systemic chemotherapy. Cytarabine or cladribine (2-CdA) as a single agent is first-line in most modern protocols for its efficacy and tolerability; the historical vinblastine plus prednisolone backbone still appears in some paediatric protocols. Refractory multisystem disease with risk-organ involvement may need haematopoietic stem cell transplant.[1][5]

The trials and the agents — LCH-III and beyond

Multisystem-LCH management rests on the international Histiocyte Society studies. A candidate should hold the two headline messages.[1][6]

LCH-III (Gadner and colleagues, Blood 2013) is the current evidence base for first-line multisystem therapy. It showed that twelve months of continuation therapy reduced five-year reactivation in non-risk-organ multisystem disease, and that risk-organ multisystem five-year survival improved to about 84 percent on a vinblastine-plus-prednisone backbone; adding methotrexate brought no survival benefit but added toxicity, so the backbone alone remained standard.[6]

When frontline therapy fails — no disease resolution by week 12 of induction, or reactivation within six months of stopping — purine analogues are the established second line. The Donadieu international phase 2 study of cladribine plus cytarabine in refractory risk-organ multisystem disease reported an overall response rate around 92 percent and five-year survival around 85 percent, at the cost of substantial haematologic toxicity.[7]

[1] [7]

Targeted therapy — the BRAF and MEK inhibitors

For refractory or relapsed disease with the right mutation, targeted therapy has transformed the outlook. Testing BRAF V600E on tissue at diagnosis is now standard for multisystem, risk-organ, or relapsed disease because it determines this eligibility.[3][5]

• BRAF inhibitors — vemurafenib and dabrafenib — are highly effective for BRAF V600E-mutated refractory LCH. The VE-BASKET phase 2 study and the early dramatic-response reports established their place; the Whitlock paediatric study confirmed dabrafenib activity in children. Responses appear within weeks, but disease recurs within weeks to months off drug, so therapy is suppressive rather than curative.[8][9][10]
• MEK inhibitors — trametinib and cobimetinib — are the rational choice for MAP2K1-mutated disease or when a BRAF inhibitor fails, and work regardless of BRAF status because they act downstream.[3][5]

[8] [9] [10] [11]

A confession worth carrying into the viva: in adults, long-term BRAF-inhibitor use carries a real keratoacanthoma and cutaneous squamous-cell carcinoma risk, which is why serial skin monitoring is mandatory and why a MEK inhibitor may be preferred in a BRAF-positive adult with a heavy ultraviolet-exposure history.[3][11]

Diabetes insipidus is managed symptomatically with desmopressin (DDAVP) for polyuria and polydipsia; because the pituitary damage is often irreversible, it frequently persists despite disease control.[1]

Prognosis — the risk organs decide it

Prognosis tracks the extent and the risk organs far more than the tumour bulk.[1]

• Single-system LCH — excellent, over 90 percent survival, with many bone lesions regressing spontaneously.[1]
• Multi-system without risk organs — good, over 80 percent survival.[1]
• Multi-system with risk organs — worse; mortality around 10 to 20 percent in children, with age under two years an additional risk factor.[1]
• Neurodegenerative LCH — progressive, with no effective treatment and a poor quality-of-life outcome.[1]
• Long-term sequelae — often-permanent diabetes insipidus, growth hormone deficiency, skeletal deformity, hearing loss, neurodegeneration, and secondary malignancies after chemotherapy or radiotherapy.[5]

The wider histiocytosis family — know the boundaries

LCH sits within a family of histiocytic disorders that share MAPK-driven myeloid neoplasia but have distinct phenotypes. Recognising the boundary matters for prognosis and treatment.[3][5]

[1] [2]

The mantra, and the memory device

[1]

The mantra: biopsy the rash, stage the organs, fear the risk organs, target the mutation.[1]

Etymology for viva gold: Langerhans cells are named for Paul Langerhans, the German medical student who described them in 1868 (and who, in a happy accident, mistook them for cutaneous nerve cells). Birbeck granules take their name from Susanne Birbeck, who captured them under the electron microscope in 1961. Both names survive the science that renamed the disease a myeloid neoplasm.[1]

Ward-round test — three stems, thirty seconds each

Stem 1 — the cradle cap that will not clear (answer)

A four-month-old has a scaly, crusted scalp and diaper eruption refractory to antifungals and emollients. There are petechiae within the flexural papules. What is the diagnosis to exclude, and what is the next step? Model: This is cutaneous Langerhans cell histiocytosis until proven otherwise — the seborrhoeic-like eruption with petechiae in the flexures is the classic infantile presentation. The next step is a skin biopsy with CD1a and Langerin immunohistochemistry, followed by systemic staging (skeletal survey, full blood count, liver function tests, urine osmolality) to distinguish single-system skin disease from multisystem or risk-organ involvement. Empirical creams have had their chance; the biopsy is overdue.[1][4]

Stem 2 — skull lesion, thirst, and a bulging eye (answer)

A five-year-old presents with a painless skull lump, new-onset polyuria and polydipsia, and a protruding right eye. Name the triad, the disease, and the first imaging test for the pituitary component. Model: This is the Hand-Schüller-Christian triad of skull lytic lesions, diabetes insipidus, and exophthalmos — chronic multisystem Langerhans cell histiocytosis. The diabetes insipidus is central, from posterior pituitary or stalk infiltration, and is often permanent. Confirm with a water deprivation test and a pituitary MRI, biopsy a lesion for CD1a and Langerin, and stage the skeleton and risk organs. Treatment is systemic chemotherapy guided by extent and risk-organ status.[1][5]

Stem 3 — refractory multisystem LCH with BRAF V600E (answer)

A child with risk-organ multisystem LCH has not responded to twelve weeks of vinblastine and prednisone. Tissue is BRAF V600E-positive. What changes in the treatment plan, and what is the caveat? Model: BRAF V600E positivity opens the targeted-therapy pathway — vemurafenib or dabrafenib, with rapid clinical and metabolic responses documented in the VE-BASKET study and the Whitlock paediatric series. The caveat is that these drugs are suppressive, not curative: disease recurs within weeks to months of stopping, so long-duration or continuous therapy is the rule, and in adults the keratoacanthoma and cutaneous squamous-cell carcinoma risk demands serial skin monitoring. A MEK inhibitor such as trametinib is the alternative if a BRAF inhibitor is unsuitable.[8][9][10]

References

1. [1]Rodriguez-Galindo C, Allen CE. Langerhans cell histiocytosis Blood, 2020.PMID 32106306
2. [2]Moore PF. Histiocytic Diseases Vet Clin North Am Small Anim Pract, 2023.PMID 36270835
3. [3]McClain KL, Bigenwald C, Collin M, et al. Histiocytic disorders Nat Rev Dis Primers, 2021.PMID 34620874
4. [4]McKinney RA, Wang G. Langerhans Cell Histiocytosis and Other Histiocytic Lesions Head Neck Pathol, 2025.PMID 39998733
5. [5]Goyal G, Tazi A, Go RS, et al. International expert consensus recommendations for the diagnosis and treatment of Langerhans cell histiocytosis in adults Blood, 2022.PMID 35271698
6. [6]Gadner H, Minkov M, Grois N, et al. Therapy prolongation improves outcome in multisystem Langerhans cell histiocytosis Blood, 2013.PMID 23589673
7. [7]Donadieu J, Bernard F, van Noesel M, et al. Cladribine and cytarabine in refractory multisystem Langerhans cell histiocytosis: results of an international phase 2 study Blood, 2015.PMID 26194764
8. [8]Diamond EL, Subbiah V, Lockhart AC, et al. Vemurafenib for BRAF V600-Mutant Erdheim-Chester Disease and Langerhans Cell Histiocytosis: Analysis of Data From the Histology-Independent, Phase 2, Open-label VE-BASKET Study JAMA Oncol, 2018.PMID 29188284
9. [9]Haroche J, Cohen-Aubart F, Emile JF, et al. Dramatic efficacy of vemurafenib in both multisystemic and refractory Erdheim-Chester disease and Langerhans cell histiocytosis harboring the BRAF V600E mutation Blood, 2013.PMID 23258922
10. [10]Whitlock JA, Geoerger B, Dunkel IJ, et al. Dabrafenib, alone or in combination with trametinib, in BRAF V600-mutated pediatric Langerhans cell histiocytosis Blood Adv, 2023.PMID 36884302
11. [11]Haroche J, Cohen-Aubart F, Emile JF, et al. Reproducible and sustained efficacy of targeted therapy with vemurafenib in patients with BRAF(V600E)-mutated Erdheim-Chester disease J Clin Oncol, 2015.PMID 25422482