Cancer in Children: Clinical Management, 5th Edition

Chapter 16. Langerhans cell histiocytosis

Helmut Gadner

Nicole Grois


The term ‘histiocytic syndromes’ or ‘histiocytoses’ describes a group of rare and enigmatic syndromes characterized by accumulation and/or proliferation of cells of the mononuclear phagocytic system. This system comprises two groups of immune cells in the tissues: the macrophages as the cardinal phagocytic cells, and the dendritic cells with low phagocytic capacity but highly effective antigen presentation and a potent capability of initiating primary T-lymphocyte responses. Both macrophages and dendritic cells arise from a common CD34+ myeloid precursor cell in the bone marrow. Dendritic cells are primarily situated in connective and lymphoid tissues, as well as in the epithelium; the blood monocytes, which represent the immediate macrophage precursors, transform into tissue-specific macrophages.1

In 1997 the Histiocyte Society proposed a new classification of the histiocytic disorders separating those of varied behaviour from those of clearly malignant behaviour.2 According to this recommendation, histiocytoses are grouped into three classes (Fig. 16.1).

Class I diseases are of dendritic cell origin, with Langerhans cell histiocytosis (LCH) as the most frequent form. Class II disorders, with ordinary histiocytes as lesional cells, include familial haemophagocytic lymphohistiocytosis, infectionor malignancy-associated haemophagocytic syndrome, and sinus histiocytosis with massive lymphadenopathy. Class III embraces truly malignant histiocytic disorders characterized by the presence of histiocytes with malignant features, i.e. acute monocytic leukaemia (FAB M5), and malignant histiocytosis, which is extremely rare since large-cell anaplastic lymphoma has been recognized as a separate non-histiocyte-related entity.

Langerhans cell histiocytosis

Since the first description by Hand in 1893, LCH has remained a poorly understood disorder. The disease comprises a broad spectrum of clinical presentations ranging from spontaneous regression of a solitary lesion to a widespread life-threatening disorder. In the past several synonyms have been used to describe the disease, such as histiocytosis X, eosinophilic granuloma, Hand–Schueller–Christian disease, Abt–Letterer–Siwe disease, etc. Since 1985, ‘Langerhans cell histiocytosis’ has become the commonly accepted term, acknowledging the central role of the Langerhans cells as the key factor in the various disease forms.3

Incidence and epidemiology

LCH can present at any age ranging from the neonatal period to old age. Predominantly young children aged 1–3 years are affected. The incidence in different age groups is about 0.2–1 per 100 000 children per year (median 0.4 per 100 000) with males outnumbering females (ratio 1.7:1). More than a third of children, especially those <2 years old, are prone to ‘multisystem’ disease with ‘organ dysfunction’ and no gender preference. The age distribution of LCH manifestations in the paediatric population is illustrated in Figure 16.2.

Fig. 16.1 Contemporary classification of histiocytic disorders. Modified from B. E. Favara et al. (1997) Med Pediatr Oncol 29, 157–66.

The disease is essentially sporadic, but has been found associated with congenital anomalies and reported rarely in twins and certain kindreds. In adults, there is some evidence that cigarette smoking plays a key role in the development of pulmonary disease.2 To date, however, very limited and inconclusive data are available regarding other epidemiologic aspects, i.e. preand postnatal risk factors including environmental and genetic risks.4

Aetiology and pathogenesis

LCH can be considered as a ‘clonal’ proliferative neoplasm with highly variable biologic behaviour and clinical severity; however, the question as to what causes the LCH cells to proliferate remains unanswered. Clonality cannot be regarded as a cancer, as there are several examples of inflammatory processes and dermatologic disorders which are clonal but not cancerous.

Fig. 16.2 Age distribution of LCH in paediatric patients.

Viral, immunologic, and genetic causes have been considered for LCH, but so far only hypotheses can be presented. Although viruses might initiate a cytokine cascade leading to a Langerhans cell proliferation, no link between viral infections and LCH has been found. It has also been postulated that LCH is caused by an immunologic dysregulation with a cytokine dysequilibrium (‘cytokine storm’) derived from an inborn genetic or acquired immune defect. Also, some of the pathologic sequelae of LCH may be explained by the contribution of cytokines such as interleukin 1α (IL-1α) and tumour necrosis factor-α (TNF-α), which enhance osteoclastic activity. The lesional presence of transforming growth factor-β is considered to be related to the evolution of fibrosis in LCH lesions of bone, liver, and lung.5

An increasing number of studies recently have demonstrated somatic mutations in LCH, suggesting that a certain genetic instability may be at least a component in the aetiology. These include a loss of heterozygosity in different chromosomal regions of LCH cells, possibly acquired in one or more genes that regulate cell growth, survival, or proliferation, and functional defects in the tumour suppressor gene p53. The association between LCH and other malignancies and, although rarely seen, the documentation of familial cases of LCH also suggest a genetic predisposition.4

Another interesting observation in multisystem endstage patients with LCH was the detection of myelodysplastic bone marrow features and the concurrence of LCH with varying degrees of macrophage activation, including fulminant haemophagocytic syndrome. Favara et al6 recently hypothesized that LCH might be a final common pathway for several different primary events.


The lesions in LCH consist of an aggregation of dendritic cells of Langerhans type with a variable admixture of other cells (eosinophils, neutrophils, lymphocytes, fibroblasts, and multinucleated giant cells) forming granulomas with proliferative and locally destructive behaviour. The high cellular content of these granulomas decreases gradually, resulting in a xanthomatous and fibrotic pattern. The typical feature of a Langerhans cell (LC) is a ‘histiocytic’ cell with abundant homogeneous pink cytoplasm in the section stained with haematoxylin and eosin, and a lobulated, ‘coffee-bean’-like nucleus. These cells express a number of phenotypic markers, which are also shared by normal LCs and can be demonstrated by immunohistochemistry and electron microscopy.1

LCH cells produce a wide variety of cytokines including IL-1, IL-2, IL-3, IL-4, IL-6, IL-8, TNF-α and interferon-γ, which probably reflects the activated state of these cells. Class II MHC molecules and the CD1a complex, together with Birbeck granules, are the most specific markers identifying LCs in tissue specimens. Other markers for LCs are surface ATPase and S-100 protein, but neither of these is cell specific. Activated LCs express the CD4 complex, the IL-2 receptor (IL-2R), and placental alkaline phosphatase (PLAP), which represents a very early transient activation marker. Of further interest is the constitutive expression of the co-stimulatory molecules CD86 and CD80 on LCH cells, and the recently described surface molecule langerin and the actin-binding protein fascin.4,5

To avoid difficulties in the distinction of LCH from other histiocytosis syndromes, the histopathologic diagnosis of LCH is based on different confidence levels3 (Fig. 16.3).

Clinical aspects

The clinical manifestations of LCH can range from asymptomatic lesions to significant morbidity depending on the kind and number of organs involved. Almost every organ in the body can be involved, although as yet there are no reports of renal, bladder, gonad, or adrenal involvement. In ‘single-system’ LCH only one organ/system with localized or multifocal lesions is involved, whereas in ‘multisystem’ LCH more than one organ/system is affected (Table 16.1).

Fig. 16.3 Confidence levels of histopathologic diagnosis.

Bone involvement is found in about 80 per cent of patients. The skull is most frequently affected; other common osseous sites include the vertebrae, long bones, and pelvic bones. Symptoms consist of painful swelling with or without functional impairment. Plain radiography typically shows single or multiple irregularly marginated osteolytic lesions which often appear as a ‘punched-out’ defect (Fig. 16.4). Pathologic fractures may occur in the weightbearing bones. In the spine the lytic process can result in compression and collapse of the vertebral body, causing vertebra plana. LCH of the jaws is often associated with hypertrophic gingivitis or ‘floating teeth’, and proptosis is seen when the orbit is involved. Mastoid or petrous bone lesions may cause persistent otitis with aural discharge, and may lead to hearing impairment.7,8 Lesions in the craniofacial region (skull base, temporal, zygomatic, sphenoidal and ethmoidal bones, mastoids, orbits, anterior and middle cranial fossa) or spinal column are considered ‘special site’ diseases as they are often associated with a prominent intracranial or intraspinal soft tissue component which might be difficult to access by local treatment measures.

Soft tissue swelling in the adjacent region of a bone lesion caused by extension of the granuloma or local oedema is a common finding, and is not usually considered as a separate organ involvement. Only rarely is an isolated granuloma found in soft tissue. Differential diagnosis of bony disease includes osteomyelitis, bone cysts, aseptic necrosis, and malignant bone tumours.

Table 16.1. Stratification of Langerhans cell hystiocytosis

Single-system disease
Single site

·  Single bone lesion

·  Isolated skin disease

·  Solitary lymph node involvement

·  Solitary lung disease

·  Isolated CNS involvement

Multiple site

·  Multiple bone lesions

·  Multiple lymph node involvement

Multisystem disease
(multiple organ involvement with or without organ dysfunction)
Risk patients   Any age with involvement of at least one risk organ
Low-risk patients   >2 years old with no risk organ involvement
Risk organs   Liver, lungs, haematopoietic system and spleen

Fig. 16.4 ‘Punched-out’ osteolytic LCH lesions in the skull.

Skin involvement occurs in about 25 per cent of LCH patients, either as single-organ involvement or as part of multisystem disease. Any part of the skin can be affected, including the nails.9 The lesions may appear as erythematous scaly seborrhoea-like brown to red papules, especially pronounced on the scalp, on the trunk, and in intertriginous zones (behind the ears and in the axillary, inguinal, and perineal areas). These papules may transform into a vesicular-pustular or crusted appearance. Rarely, LCH manifests as a solitary cutaneous or subcutaneous nodule covered by intact skin. Congenital or perinatal occurrence, leaving residual hyperpigmentation of numerous firm nodules which resolve spontaneously within a few months, is known as Hashimoto–Pritzker syndrome. It is often difficult to differentiate LCH from other skin diseases such as mycosis or seborrhoeic or atopic eczema; nodular lesions have to be distinguished from malignant lymphoma or metastatic solid tumours. Localized or disseminated lesions also manifest on the mucous membranes, usually on the buccal mucosa, palate, and gingiva, but also in the gastrointestinal, urogenital, and vaginal tract, presenting as whitish granulomatous plaques that transform into ulcers with a tendency to bleeding.7,8

Lymph node involvement may be associated with local disease affecting adjacent skin or bone or may be a part of disseminated disease. The cervical nodes are most frequently involved and, rarely, thymic enlargement may be found. In the case of bone marrow involvement the infiltrationwith CD1a-positive cells is typically sparse, but their presence is associated with risk of organ involvement and correlates with an unfavourable outcome (unpublished data, LCH Registry, Vienna, Austria). Severe pancytopenia, frequently observed in infants with multisystem disease, is related to bone marrow dysfunction rather than to infiltration, and is usually associated with gross hepatosplenomegaly and a poor prognosis. Splenomegaly is usually seen in very young children with multisystem LCH and may contribute to pancytopenia. To reduce the need for blood transfusion, the removal of the enlarged spleen may be indicated in some cases.7

Hepatic enlargement is a common finding in patients with multisystem disease. Infiltration with LCH cells, which are CD1a positive but lack Birbeck granules, as well as massive portal lymphadenomegaly can lead to cholestatic hepatomegaly.1 Activation of the cellular immune system by cytokines may result in hyperplasia and hypertrophy of Kupffer cells, presenting the histologic picture of sclerosing cholangitis. Ascites and oedema caused by hypoalbuminaemia, often associated with hypocoagulopathy, are clinical signs of liver dysfunction.

Isolated lung involvement occurs more frequently in adolescents and young adults, with >90 per cent of these patients having a history of smoking. In children, pulmonary LCH is usually part of multisystem disease. Clinical signs include tachypnoea with subcostal recession and persistent cough. Respiratory function tests may show a decreased total lung volume and compliance. The radiologic picture consists of diffuse fine interstitial shadowing due to micronodular granulations, which may appear cystic on high-resolution CT. With advancing disease the cysts increase in number and size and form bullae which appear on radiographs as ‘honeycomb lungs’. Spontaneous pneumothorax resulting from rupture of superficial bullae may occur. Patients with uncontrolled LCH may develop lung fibrosis. Confirmation of diagnosis is mandatory by finding >5 per cent CD1a-positive cells in the bronchoalveolar lavage fluid or by lung biopsy10,11 (Fig. 16.5).

Because gastrointestinal involvement seldom produces prominent clinical manifestations, it is often underdiagnosed and its actual frequency is unknown. The most common sign is failure to thrive because of malabsorption. Other symptoms include vomiting, diarrhoea, and proteinlosing enteropathy. The diagnosis of gastrointestinal LCH must be supported not only by radiologic evidence of alternating dilated and stenotic segments in the small and large bowel, but also by endoscopic examination and biopsy.7

Diabetes insipidus is the most common of the endocrinopathies associated with LCH occurring before, simultaneous with, or subsequent to the disease manifestation in other organs. Its incidence varies from 15 to 50 per cent in different studies. It can often be found in patients with extensive disease, lesions of the craniofacial bones, and intracranial tumour extension.12 Confirmation of the diagnosis by an appropriate water deprivation test and measurement of urinary arginine vasopressin is essential because partial defects occur and may spontaneously remit.13 A thickening of the hypothalamic–pituitary stalk, absence of the posterior pituitary bright spot, or both can be seen on MRI. Growth failure in children with LCH is commonly reported and appears to have a multifactorial cause. Anterior pituitary function may be compromised, resulting in growth hormone deficiency, hypothyroidism, hypogonadism, and hyperprolactinaemia.14 Finally, panhypopituitarism may occur, and cranial MRI does not necessarily reveal a hypothalamic mass in these patients. However, persistent cytokine release due to chronic disease, occult gut involvement causing malabsorption, vertebral collapse, and prolonged steroid therapy may also contribute to growth retardation in LCH.

Fig. 16.5 Radiograph and high-resolution CT of the lung in a patient with pulmonary involvement.

Apart from the common manifestations in the hypothalamic–pituitary axis, virtually all other parts of the central nervous system (CNS) can be affected by LCH, including the following locations in order of frequency: cerebellum, pons, cerebral hemispheres, choroid plexus, basal ganglia, spinal chord, optic tract, and cranial nerves12 (Fig. 16.6).

Diagnostic procedures

A lesional biopsy is an essential step in obtaining the definitive diagnosis of LCH. In 1987 the Writing Group of the Histiocyte Society outlined an applicable classification of histiocytosis syndromes and standards for histopathologic diagnosis which has been widely accepted.3 According to this proposal, morphologic, immunohistochemical, and clinical criteria are required for the definitive diagnosis (Fig. 16.3).

In a rare disease like LCH with a highly variable clinical presentation, it is crucial to use uniform guidelines for clinical evaluation and assessment of disease extent. Based on an exact history and a meticulous clinical examination mandatory baseline investigations should be performed in every newly diagnosed patient (Table 16.2), and in selected cases diagnostic procedures are required for specific indications15 (Table 16.3).

Permanent consequences and late effects

Permanent consequences are defined as any permanent or irreversible physical and/or neuropsychologic sequelae, attributable to the disease itself rather than treatment, which develop at any time during the disease course. The risk of developing permanent consequences correlates with the duration of disease activity and the frequency of reactivations.

Fig. 16.6 CNS changes in LCH: (a) infundibular thickening apparent on sagittal brain MRI; (b) axial FLAIR weighted MRI (bilateral hyperintensities in cerebellum).

Table 16.2. Diagnostic guidelines

Clinical evaluation

·  Complete history
Fever, pain, irritability, failure to thrive, nutritional status, loss of appetite, diarrhoea, polydipsia, polyuria, activity level, behavioural changes, neurologic changes

·  Complete physical examination
Measurement of temperature, height, weight, and head circumference, pubertal status, skin and scalp rashes, purpura, bleeding, aural discharge, orbital abnormalities, lymphadenopathies, gum and palatal lesions, dentition, soft tissue swelling, dyspnoea, tachypnoea, intercostal retractions, liver and spleen size, ascites, oedema, jaundice, neurologic examination, papilloedema, cranial nerve abnormalities, cerebellar dysfunction


·  Complete blood count

·  Liver enzymes and function tests
Serum glutamic-oxaloacetic transaminase (SGOT), serum glutamate-pyruvate transaminase (SGPT), alkaline phosphatase, bilirubin, total protein, albumin

·  Coagulation studies
Prothrombin time (PT), partial thromboplastin time (PTT), fibrinogen

·  Urine osmolality (measurement after overnight water deprivation)


·  Chest radiograph (posteroanterior and lateral)

·  Skeletal radiograph survey

In single-system LCH permanent consequences occur in about 33 per cent of the patients, and about 80 per cent of these are related to the site of disease. Patients with initial lesions in the long bones or the spinal column often develop orthopaedic problems or growth failure. If the mastoid, maxilla, or mandible are involved, hearing deficits or tooth loss can occur.16 The incidence of permanent consequences in patients with multisystem disease varies between 22 and 26 per cent.17,18 Diabetes insipidus is observed most frequently; others are listed in Figure 16.7.

Table 16.3. Evaluations upon specific indication



Bone marrow aspirate and trephine biopsy

Anaemia, leucopenia, or thrombocytopenia

Pulmonary function tests

Abnormal chest radiograph, tachypnoea, intercostal retractions

Lung biopsy, preceded by bronchoalveolar
lavage, when available; when diagnostic
obviates lung biopsy

Patients with abnormal chest radiograph in whom
chemotherapy is being considered to exclude
opportunistic infection

Small bowel series and biopsy

Unexplained chronic diarrhoea or failure to thrive,
evidence of malabsorption

Liver biopsy

Liver dysfunction, including hypoproteinaemia not due to protein-losing enteropathy, to differentiate active
LCH of the liver from cirrhosis

MRI of brain/hypothalamic–pituitary axis
with i.v. [Gd]DTPA

Hormonal, visual, or neurologic abnormalities

Panoramic dental radiography of
mandible and maxilla, oral surgery

Oral involvement

Endocrine evaluation

Short stature, growth failure, diabetes insipidus,
hypothalamic syndromes, galactorrhoea, precocious
or delayed puberty; CT or MRI abnormality of

Otolaryngology consultation and

Aural discharge, deafness

Endocrinopathies include hypothyroidism, hypogonadism and growth hormone deficiency. Neurologic problems usually associated with neurodegenerative CNS changes consist of hypoor hyperreflexia, tremor, ataxia, and dysarthria. Convulsions, palsy, and cranial nerve deficits may be seen in patients with intracranial tumorous LCH lesions. Psychosocial and intellectual impairment has also been observed.12

Various malignancies have been reported in 5 per cent of long-term survivors of LCH. In a retrospective analysis a high association of malignancy and LCH even without any treatment and, not infrequently, preceding the diagnosis of LCH was recognized.19 Recently, there has been concern regarding an association of secondary acute myeloid leukemia (sAML) with etoposide administration. However, it has been shown that occurrence of sAML may be linked with high cumulative doses, short intervals, or combination with other topoisomerase-IIinhibiting cytotoxic drugs.20,21

Current management and prognosis


The course of LCH is highly variable and unpredictable in the individual patient. Spontaneous regression is observed as well as chronic reactivations and, rarely, fulminant lethal courses. The underlying factors responsible for the varying biologic behaviour of LCH are unknown.18 The lack of knowledge of the pathogenesis of LCH and the failure to establish generally accepted diagnostic criteria have inhibited the development of a rational treatment policy. Therefore the treatment of patients with LCH has varied over the past century according to what was believed to be the cause of the disease. Only with the introduction of new concepts of staging and diagnostic criteria has it become possible to collect sufficiently large numbers of patients to carry out prospective clinical trials.21 Empirically it has been shown that the treatment of LCH should depend on the extent of the disease, and patients should be stratified as having ‘singlesystem’ or ‘multisystem’ disease (see Table 16.1).

Fig. 16.7 Permanent consequences of LCH: percentage of most frequent sequelae related to the totality of permanent consequences (unpublished data from the LCH Registry, Vienna, Austria).

As the disease may follow variable courses, a new definition and assessment of response to a given treatment had to be established. The following criteria were defined by the Histiocyte Society: complete resolution of the disease [no active disease (NAD)], disease regression [active disease (AD-better)], intermediate response with regression of some and reappearance of other lesions (AD-intermediate, mixed) or unchanged disease (AD-intermediate, stable), and progression of the disease (AD-worse).8,21

Single-system disease

A solitary bony lesion does not usually require treatment other than curettage at the time of obtaining diagnostic biopsy. Lesions that are painful may respond to intralesional instillation of steroids (crystalline methylprednisolone 75–150 mg), but polyostotic disease may require a short course of systemic therapy. Various systemic regimens have been used including steroids (40 mg/m2/day), vinblastine (6 mg/m2/week), etoposide, and methotrexate. Indications for systemic treatment (e.g. prednisone, vinblastine) also include risky locations with imminent spontaneous fracture, spinal cord compression, huge non-resectable tumour mass, and involvement of ‘special sites’ (craniofacial or vertebral lesions) with impending functional impairment. In bony disease indomethacin (2 mg/m2/day) may be an alternative to steroids both for its analgesic effect and as an antiprostaglandin, and recently bisphosphonates have also been recommended. Primary surgical excision of lesions in the craniofacial region or spinal column (‘special sites’), beyond a biopsy, is not recommended because of the accompanying prominent intracranial or intraspinal soft tissue component. Although used in only a restricted way during the last decades (because of the risk of late sequelae and secondary malignancy), emergency lowdose radiation (6–12 Gy) should be reserved for critical circumstances (e.g. optic nerve or spinal cord).8 Patients with solitary bone lesions show a survival rate of 100 per cent; reactivations mostly occur during the first 2 years after diagnosis and are usually restricted to the skeleton.16

Connatal LCH confined to skin tends to regress spontaneously within a few months in most of the cases. However, progression to multisystem involvement with fatal outcome may occur, and close observation for a longer follow-up period is mandatory. Erythematous lesions usually respond to topical steroids. In severe persisting or progressing skin involvement a topical 20 per cent solution of nitrogen mustard or psoralen–ultraviolet A (PUVA) photochemotherapy may be useful, but can be recommended only for short-term therapy because of concern about possible carcinogenicity. In severe resistant skin disease mild systemic steroid therapy with or without vinca alkaloids is needed. This therapy should also be used if multiple enlarged or bulky lymph nodes are present.8,21 Surgical excision is the treatment of choice for skin nodules and isolated lymph node involvement, with a cure rate of nearly 100 per cent.16

In young children, isolated lung involvement should be treated with chemotherapy according to protocols for multisystem patients. As in adults, it is essential that adolescents refrain from smoking, and mild systemic therapy with steroids with or without vinblastine can be effective.11 There are anecdotal reports of a positive effect of ciclosporin A and bisphosphonates in progressing disease.22

Active LCH mass lesions in the CNS have been shown to respond to conventional LCH chemotherapy. The choice of therapy depends on the individual case and possible pretreatment. A beneficial effect of combined chemotherapy [prednisone, vinblastine, etoposide, and 2-chlorodeoxyadenosine (2-CdA)] has been reported.23 Radiotherapy or systemic therapy are usually unable to restore pituitary function, and hormone replacement therapy is required.13,24 No specific therapy can yet be recommended for neurodegenerative CNS disease.

Multisystem disease

Two major approaches to the treatment of multisystem disease have existed for the last 20 years: a conservative approach with treatment used only during disease exacerbations (single-center study),18 and an intensive chemotherapy induction followed by treatment (two large cooperative clinical trials, the Italian AIEOP-CNR-HX 83 study25 and the German–Austrian DAL-HX 83/90 study17). Surprisingly, the overall mortality was about 20 per cent in both the conservative and the aggressive treatment approaches. In contrast, the incidence of disease-related late sequelae was 67 per cent in the conservative treatment and only 33 per cent in the DAL-HX studies. The low incidence of disease reactivations in the DAL-HX studies (overall 23 per cent) provided evidence that effective treatment may beneficially influence the natural course of the disease.

The aim of the first international randomized chemotherapy trial (LCH I), initiated by the Histiocyte Society in 1991, was to compare the efficacy of monotherapy with vinblastine and etoposide with respect to the course of the disease and outcome.26 No significant difference was found between the two treatment arms. In the next international trial (LCH II) the effect of continuous oral prednisone combined with vinblastine with or without the addition of etoposide was compared in a randomized way in severely affected multisystem disease patients. A new stratification system was adopted, distinguishing between ‘risk’ patients with involvement of ‘risk organs’ like liver, spleen, lungs, or haematopoietic system, or age <2 years, and ‘low-risk’ patients without involvement of such organs and age >2 years.8 Risk patients were eligible for randomization between the two-drug and the three-drug arm, low-risk patients received initial treatment according to the two-drug arm only, and all patients went on continuation therapy with 6-mercaptopurine and prednisone–vinblastine pulses. Treatment duration was limited to 24 weeks as in LCH I.

The results in the low-risk group were satisfying, with 89 per cent responders at week 6 and no fatalities. No statistical difference with respect to initial response, survival, and reactivationfree survival was found between the two treatment arms in risk patients. However, it was shown that patients with involvement of risk organs, who did not show disease regression by weeks 6 or 12 of therapy, had a high risk of poor outcome and mortality. Notably, all the patients who died in the LCH I and LCH II studies had involvement of at least one risk organ, irrespective of age. Therefore it seems justified to regard risk organ involvement and response to initial treatment as the most important prognostic factors, whereas age <2 years did not prove to be of independent prognostic importance. Interestingly, the overall probability of survival of the multisystem patients did not differ significantly between the DAL-HX, LCH I, and LCH II studies, and was 80 per cent. This indicates that there is a therapy-resistant ‘high-risk’ population of 20 per cent of the multisystem patients who cannot be rescued by standard treatment. When the results were compared with those of the DAL-HX studies, there was a clear superiority of therapy given for 1 year, with respect to the rate of reactivations, over treatment for 6 months.27

The third international randomized trial (LCH-III) was started in April 2001 ( (Fig. 16.8). Patients are stratified into three groups: multisystem risk patients (patients with involvement of one or more risk organs), multisystem low-risk patients (patients with multiple organs involved but without involvement of risk organs), and patients with single-system ‘multifocal bone disease’ or localized ‘special site’ involvement (craniofacial or vertebral lesions). Risk patients are randomized between two different treatment arms including prednisone, vinblastine, and 6-mercaptopurine with or without methotrexate. Initial treatment consists of one or two courses depending on response and is followed by continuation therapy (treatment duration, 12 months). Treatment for low-risk patients includes prednisone and vinblastine; the overall duration of therapy for this patient group is 6 or 12 months as randomly assigned. Patients with multifocal bone disease or special site involvement are treated with prednisone and vinblastine for 24 weeks.

Resistant disease

Patients who do not respond to initial treatment are considered to have a high risk of mortality (~75 per cent). Ciclosporin A has been suggested as an alternative treatment approach. However, convincing data are lacking, especially for patients with advanced chemotherapyresistant multisystem disease. Regarding the role of bone marrow transplantation only scanty and inconsistent data are available. Regimen-related mortality in these studies is high;22 only recently the application of intensity reduced conditioning showed promising results.29 Other studies report on the successful treatment of LCH with interferon-α, anti-CD1a and anti-TNF-α antibodies, but further investigation is necessary.8,21,28). The use of 2-CdA and 2′-deoxycoformycin has recently appeared to be successful in refractory LCH.23According to salvage treatment protocol of the Histiocyte Society, 2-CdA is given as monotherapy (5 mg/m2 2-CdA daily for 5 days at intervals of 3–4 weeks; two, four or six courses) to non-responding multisystem patients or patients with recurrent disease was not convincing.23 A new salvage approach including 2-CdA combined with cytarabinoside is in progress.30

Fig. 16.8 LCH-III protocol: (a) initial treatment plan for risk patients; (b) protocol continuation treatment plan for risk patients. PDN, prednisone; MTX, methotrexate; VBL, vinblastine; 6-MP, 6-mercaptopurine.


1. Schmitz L, Favara BE (1998). Nosology and pathology of Langerhans cell histiocytosis. Hematol Oncol Clin North Am 12, 221–46.

2. Favara BE, Feller AC, Pauli M, et al. (1997). Contemporary classification of histiocytic disorders. The WHO Committee on Histiocytic/Reticulum Cell Proliferations. Reclassification Working Group of the Histiocyte Society. Med Pediatr Oncol 29, 157–66.

3. Writing Group of the Histiocyte Society (1987). Histiocytosis syndromes in children. Lancet i, 208–9.

4. Arceci RJ, Longley BJ, Emanuel PD (2002). Atypical cellular disorders. Hematology (Am Soc Hematol Educ Program), 297–314.

5. Willman CL, McClain KL (1998). An update on clonality, cytokines, and viral etiology in Langerhans cell histiocytosis. Hematol Oncol Clin North Am 12, 407–16.

6. Favara BE, Jaffe R, Egeler RM (2002). Macrophage activation and hemophagocytic syndrome in Langerhans cell histiocytosis: report of 30 cases. Pediatr Dev Pathol 5, 130–40.

7. Egeler RM, D'Angio GJ (1995). Langerhans cell histiocytosis. J Pediatr 127, 1–11.

8. Gadner H, Grois N (2004). Langerhans cell histiocytosis. Pinkerton R, Plowman RN, Pieters R. Pediatric Oncology 3rd edition, 469–483.

9. Munn S, Chu AC (1998). Langerhans cell histiocytosis of the skin. Hematol Oncol Clin North Am 12, 269–86.

10. McClain KL, Gonzalez JM, Jonkers R, et al. (2002). Need for a cooperative study: Pulmonary Langerhans cell histiocytosis and its management in adults. Med Pediatr Oncol 39, 35–9.

11. Vassallo R, Ryu JH, Colby TV, et al. (2000). Pulmonary Langerhans'-cell histiocytosis. N Engl J Med 342, 1969–78.

12. Prayer D, Grois N, Prosch H, et al. (2004). MR imaging presentation of interacranial disease associated with Langerhans cell histiocytosis. Am J Neuroradiol 25, 880–1.

13. Broadbent V, Pritchard J (1997). Diabetes insipidus associated with Langerhans cell histiocytosis: is it reversible? Med Pediatr Oncol 28, 289–93.

14. Nanduri VR, Bareille P, Pritchard J, et al. (2000). Growth and endocrine disorders in multisystem Langerhans' cell histiocytosis. Clin Endocrinol (Oxf) 53, 509–15.

15. Writing Group of the Histiocyte Society (1989). Histiocytosis syndromes in children: II. Approach to the clinical and laboratory evaluation of children with Langerhans cell histiocytosis. Med Pediatr Oncol 17, 492–5.

16. Titgemeyer C, Grois N, Minkov M, et al. (2001). Pattern and course of single-system disease in Langerhans cell histiocytosis data from the DAL-HX 83and 90-study. Med Pediatr Oncol 37, 108–14.

17. Gadner H, Heitger A, Grois N, et al. (1994). Treatment strategy for disseminated Langerhans cell histiocytosis. DAL HX-83 Study Group. Med Pediatr Oncol 23, 72–80.

18. McLelland J, Broadbent V, Yeomans E, et al. (1990). Langerhans cell histiocytosis: the case for conservative treatment. Arch Dis Child 65, 301–3.

19. Egeler RM, Neglia JP, Arico M, et al. (1998). The relation of Langerhans cell histiocytosis to acute leukemia, lymphomas, and other solid tumors. The LCH-Malignancy Study Group of the Histiocyte Society. Hematol Oncol Clin North Am 12, 369–78.

20. Haupt R, Fears TR, Heise A, et al. (1997). Risk of secondary leukemia after treatment with etoposide (VP-16) for Langerhans' cell histiocytosis in Italian and Austrian–German populations. Int J Cancer 71, 9–13.

21. Ladisch S, Gadner H (1994). Treatment of Langerhans cell histiocytosis—evolution and current approaches. Br J Cancer 23 (Suppl), S41–6.

22. Minkov M, Grois N, Broadbent V, et al. (1999). Cyclosporine A therapy for multisystem Langerhans cell histiocytosis. Med Pediatr Oncol 33, 482–5.

23. Weitzman S, Wayne AS, Arceci R, et al. (1999). Nucleoside analogues in the therapy of Langerhans cell histiocytosis: a survey of members of the Histiocyte Society and review of the literature. Med Pediatr Oncol 33, 476–81.

24. Howell SJ, Wilton P, Shalet SM (1998). Growth hormone replacement in patients with Langerhan's cell histiocytosis. Arch Dis Child 78, 469–73.

25. Ceci A, De Terlizzi M, Colella R, et al. (1993). Langerhans cell histiocytosis in childhood: results from the Italian Cooperative AIEOP-CNR-H.X ‘83 study. Med Pediatr Oncol 21, 259–64.

26. Gadner H, Grois N, Arico M, et al. (2001). A randomized trial of treatment for multisystem Langerhans' cell histiocytosis. J Pediatr 138, 728–34.

27. Minkov M, Grois N, Heitger A, et al. (2002). Response to initial treatment: an important prognostic predictor in multisystem Langerhans cell histiocytosis. Med Pediatr Oncol 39, 581–5.

28. Henter JI, Karlen J, Calming U, et al. (2001). Successful treatment of Langerhans-cell histiocytosis with etanercept. N Engl J Med 345, 1577–8.

29. Steiner M, Matthes-Martin S, Attarbaschi A, et al.(2005). Improved outcome of treatment-resistant high-risk Langerhans cell histiocytosis after allogeneic stem cell transplantation with reduced-intensity conditioning. Bone Marrow Transplant 6, (in press)

30. Bernhard F, Thomas C, Bertrand Y, et al.(2005). Report of multicenter pilot study of 2-CdA and Ara-C combined chemotherapy in refractory Langerhans cell histiocytosis with haematological dysfunctions. Eur J Cancer (in press) (accepted for publication Nov. 2004).

If you find an error or have any questions, please email us at Thank you!