Vol.2, No.6, 351-357 (2013) Case Reports in Clinical Medicine
Fever of unknown origin as the first manifestation of
Erdheim-Chester disease
Ning Li1*, Mingquan Chen1*, Huaping Sun2, Yun Bao3, Jiming Zhang1#
1Department of Infectious Diseases, Huashan Hospital, Fudan University, Shanghai, China;
#Corresponding Author: jmzhang@fudan.edu.cn
2Department of Radiology, Huashan Hospital, Fudan University, Shanghai, China
3Department of Pathology, Huashan Hospital, Fudan University, Shanghai, China
Received 8 July 2013; revised 1 August 2013; accepted 10 August 2013
Copyright © 2013 Ning Li et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Erdheim-Chester disease (ECD) is an uncom-
mon non-Langerhans cell histiocytosis that af-
fects multiple body systems and can present
clinically in a my riad of ways. W e re port the cas e
of a 29-year-old man presented with fever and
diffuse bone pain. Physical examination showed
bilateral and symmetrical long bone pain, espe-
cially the knees. Radiographs showed multiple
mixed bone lesions with sclerotic and ly tic areas
of the femora, tibiae, and fibula. Biopsy of the
tibial area displayed foamy lipid-laden histio-
cytes, confirming the diagnosis of ECD.
Keywords: Erdheim-Chester Dise ase; Histio cytosi s ;
Erdheim-Chester disease (also known as Erdheim-
Chester syndrome or polyostotic sclerosing histiocytosis,
ECD) is a rare form of non-Langerhans-cell histiocytosis
with particular tropism for connective and adipose tis-
sues. To date, about 450 distinct cases of ECD have been
described in the medical literature [1,2]. The number of
new cases has dramatically increased over the past 10
years because of the better recognition of this condition.
ECD is a systemic disease characterized by proteiform
manifestations, including skeletal involvement with bone
pain, exophthalmos, diabetes insipidus, xanthelasma, in-
terstitial lung disease, bilateral adrenal enlargement, re-
troperitoneal fibrosis with perirenal and/or ureteral ob-
struction, renal impairment, testis infiltration, central ner-
vous system, and/or cardiovascular involvements. The
pathology is characterized by foamy histiocytes or lipid-
laden macrophages or histiocytes, surrounded by fibrosis,
which are those with immunohistochemical findings de-
monstrating the presence of non-Langerhans histiocytes
markers, namely, positivity for anti-CD68 labeling, little
or no protein S100 labeling, and negativity for anti-
CD1a labeling [3,4].
We present the case of a 29-year-old male patient,
who presented initially with decreased vision, diagnosis
of VKHS. To the best of our knowledge, it is the first
description of ECD with VKHS as the initial presenta-
tion. Leveled equations, graphics, and tables are not pre-
scribed, although the various table text styles are provid-
ed. The formatter will need to create these components,
incorporating the applicable criteria that follow.
A 29-year-old man was admitted to our hospital
because of fevers, chills, cough, and pain in the both
lower limbs. He presented a 10-month history of inter-
mittent fevers (about 39˚C), progressive pain in both
lower limbs, and nonproductive cough. There were nu-
merous mildly enlarged bilateral cervical and supracla-
vicular lymph nodes. A doppler B scan system-guided
fine-needle aspiration biopsy of the left cervical lymph
nodes was performed, and pathological examination of
the biopsy specimen revealed reactive lymphoid hyper-
plasia. A radiograph reportedly showed infection-related
lesions in the tibia and fibula. Antibiotics sulperazon and
vancomycin were administered without any improve-
ment. Six days later, twice myeloid culture reportedly
showed Human staphylococcus aureus, and bone scan
reportedly showed lesions with increased activity involv-
ing the large joints. Antibiotics rifampicin, clindamycin,
teicoplanin, or linezolid were sequentially administered,
and the patient was running a mid to high fever, accom-
panied by lower limbs pain and discomfort. Antibiotics
*Ning Li and Mingquan Chen contributed equally to this work.
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7
were discontinued, and antipyretic-analgesic was admi-
nistrated. He had no prior medical history, and no expo-
sure to occupational, dust, or animal hazards. He had no
known allergies to medications. Medications included
antipyretic-analgesic as needed for pain and fever, pred-
nisone, potassium chloride sustained-release tablets, and
pantoprazole. Laboratory examination results revealed
erythrocyte sedimentation rate (ESR) of 90 mm/h and C-
reactive protein (CRP) 96.5 mg/l (reference range, <8).
Approximately 8 weeks before admission, increased
severe persistent pain in his right frontal, temporal, and
occipital lobs were suffer from. Specimens of cerebro-
spinal fluid were stained for acid-fast bacilli, and cultur-
ed for bacteria, mycobacteria, and fungi and remained
sterile. At the time of admission to our ward he appeared
unwell, still fever, increasing fatigue, malaise, weakness,
sweats, myalgias, and leg pain developed. The patient
reported that the pain was located primarily in the region
of both leg, occurred at rest, and awoke him at night. He
described it as a constant aching and throbbing and gave
it a score of 5 (on a scale of 1 to 10, where 10 is the most
severe pain). The pain lessened with rest and narcotic
analgesia and increased when the patient walked and
moved his leg.
His temperature was 39.5˚C, blood pressure 125/65,
pulse 106 per minute, respiratory rate 22 per minute. He
had peripheral lymphadenopathy. B-ultrasound of the
superficial lymph node showed multiple nodules through-
out both bilateral cervical (up to 1.6 cm in diameter),
supraclavicular lymph nodes (up to 2.1 cm in diameter),
axillae (up to 1.4 cm in diameter) and the abdomen (up
to 1.6 cm in diameter), and a thyroid nodule (1.3 cm). On
abdominal examination, liver and spleen tip was not
palpable. In our laboratory findings, results were normal
for measurements of red and white blood cell count,
platelets, glucose, plasma electrolytes, total protein, albu-
min, globulin, bilirubin, lactate dehydrogenase, immu-
noglobulins, and urine Bence Jones protein; studies of
serum protein electrophoresis and tests of renal and liver
function were also normal, as were serum levels of an-
giotensinconverting enzyme, antinuclear antibodies, rheu-
matoid factor, antineutrophil cytoplasmic antibodies
(ANCA). Repeated culture of blood, sputum, urine, and
stool were negative for possible pathogens.
To evaluate the extent of the disease, positron emis-
sion tomography/computed tomography (PET/CT) was
performed 1 h after administration of 370 MBq 18F-fluo-
rodeoxyglucose (FDG), which showed asystematic lytic
bone lesion in the long bone and the pelvis with in-
tensively FDG uptake (standardize uptake value, SUV
approximately 15.5 to 17.1) (Figures 1(A)-(C)). There
was mild FDG uptake (SUV, 2.4) with hyperdense
nodules in bilateral posterior chamber of eyeball around
optic nerves (Figure 1(D)). Malignant tumor was con-
Figure 1. FDG PET/CT scan showed involvement of various organ systems in patients with Erdheim-Chester
disease. (A)-(C) Typical bilateral and symmetric uptake in the diaphyseal and metaphyseal regions of the long
bones was seen; (D) Neurologic involvement, with multiple nodules in bilateral posterior chamber of eyeball
around optic nerves, is demonstrated; (E) A Tc-99m bone scintigraphy displayed a symmetrically increased up-
take in the epiphyses of long bones, sparing the middiaphysis and the axial skeleton.
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7 353
sidered via PET/CT scan. However, anterior and post-
erior whole-body images of Tc-99m bone scintigraphy
displayed the characteristic findings of ECD with sym-
metric increased radionuclide uptake in the epiphyses of
long bones (Figure 1(E)). Findings of MRI showed re-
placement of the normal fatty bone marrow of the dia-
physeal and metaphyseal bone segments by a markedly
low signal intensity (hypointense compared with muscle
and heterogeneous signal intensity on T1-weighted ima-
ges) and heterogeneous intermediate or high signal in-
tensity (isointense or hyperintense compared with muscle)
on fatsuppressed T2-weighted images. Areas of high ir-
regular signal intensity on T1-weighted images after con-
trast (Figure 2).
The patient had not been to rural areas or been ex-
posed to animals recently. He reported no highrisk sexual
activity, occupational exposures. Tests for histoplasma
antigen were negative.
Five days later, at a follow-up visit, the patient re-
ported increased leg pain. During the next 3 days, He
was readmitted to this hospital. The vital signs were
normal. There was mild percussion pain on both lower
extremities, the remainder of the examination was un-
changed. An electrocardiogram was normal. The white-
cell count was 9200 per cubic millimeter, with a normal
differential count. Specimens of blood and sputum were
cultured for bacteria, mycobacteria, and fungi and re-
mained sterile. Tests for respiratory viruses and anti-
bodies to HIV were negative, as were multiple specimens
of sputum stained for acid-fast bacilli. Two days later, an
open excisional biopsy and curettage of the lesion of the
left tibia were performed at this hospital; cefazolin was
administered before and after the surgery; pathological
examination of the biopsy specimen revealed infiltration
of foamy histiocytes with irregular spindle cell prolife-
ration and infiltration of mononuclear inflammatory cells
Figure 2. MRI findings. (A) Diffuse low signal intensity on axial T1-weighted images; and (B)
intermediate enhancement on T1-weighted images after contrast are visible in diaphysis and
metaphysis of distal femora; (C) Diffuse low signal intensity on axial T1-weighted images;
and (D) with heterogeneous enhancement on T1-weighted images enhanced after gadolinium
injection are visible in diaphysis and metaphysis of bilateral fibulas and tibias; (E) Diffuse low
signal intensity on coronal T1-weighted images; and (F) irregular enhancement on T1-
weighted images after gadolinium injection are visible in diaphysis and metaphysis of distal
femora and proximal tibiae; (G) Fat-suppressed T1-weighted images after contrast show he-
terogeneous signal intensity, irregular stippled fat-suppressed signal (arrow).
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7
and eosinophils (Figure 3). Immunohistochemical stain-
ing showed infiltrating cells to be positive for CD68,
lysozyme (LYS), but negative for S-100, CD1a. Based on
the imaging and pathologic data, the diagnosis of ECD
was made.template measures proportionately more than
is customary. This measurement and others are deliberate,
using specifications that anticipate your paper as one part
of the entire journals, and not as an independent docu-
ment. Please do not revise any of the current designa-
The first case of ECD was reported by W. Chester and
Jakob Erdheim in 1930 [5]. ECD is a rare systemic non-
Langerhans cell histiocytosis (LCH) with distinguishing
clinicopathologic and radiographic features of unknown
etiology. It is associated with many clinical presentations
with involvement of 1 or more organ systems. In parti-
cular, bone is affected, but skin, soft tissue, central ner-
vous system, heart, lungs, or abdominal viscera can also
be involved.
While the exact ontogeny is incompletely understood,
histiocytes are believed to arise from a common CD-34
positive progenitor stem cell within the bone marrow.
Two major types of histiocytes are described. Dendritic
cells (e.g., Langerhans cells in skin and bronchial epi-
thelium; and interdigitating and dendritic reticulum cells
in spleen and lymph nodes) are the main antigen-present-
ing cells and are characterized by immunoreactivity to
S-100 protein and CD1a [6]. The second main histologic
cell type includes cells of the monocyte/macrophage line-
age (e.g., circulating monocytes, alveolar macrophages, he-
patic Kupffer cells, osteoclasts, and microglia) [7]. These
antigen-phagocytosing cells are characterized by immu-
noreactivity to CD68. The causative cell of Langerhans
cell histiocytosis (LCH) is derived from the Langerhans
Figure 3. Histological photomicrographs. (A), (B) Intraoperative smear
preparation showing foamy histiocytes (arrows) and small numbers of
lymphocytes. Note presence of multinucleate giant cells (arrows) and
emperipolesis. Paraffin-embedded, HE stain, 100× original magnifica-
tion; (C) CD68 immunohistochemitry. Many of the foamy cells are im-
munoreactive. 100× original magnification; (D) CD1a immunohisto-
chemistry. There is scarce immunoreactivity to CD1a, a marker for cer-
tain dendritic cells. 100× original magnification; (E) S-100 immunohis-
tochemistry. Only rare isolated cells are S-100 immunoreactive. 100×
original magnification; (F) LYS immunohistochemistry. The cells are
immunoreactive to LYS, 100× original magnification.
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7 355
dendritic cell line and that of ECD is from the mono-
cyte-macrophage lineage, however, the clinical presenta-
tions can be similar, with bony involvement most fre-
quent in both disease states. There are several key fea-
tures, particularly immunohistochemical profile, that al-
low LCH and ECD to be distinguished.
The diagnosis of ECD is based on the combination of
radiographic features, nuclear medicine features and
nearly pathognomonic immunohistochemical profile as
follow: evidence of bilateral and symmetrical cortical
osteosclerosis of the diaphyseal and metaphyseal regions
of the long bones and/or the symmetrical and strong
uptake on the distal ends of the long bones of the legs
and in some cases of the arms, on technetium-99 skeletal
scintigraphy; and typical histological findings: xantho-
granulomatous infiltration with foamy histiocytes sur-
rounded by fibrosis; CD68-positive and CD1a-negative
immunohistochemical labelling.
Patients with ECD typically develop characteristic
multifocal osteosclerotic lesions of long bones [8] and
may also develop extraskeletal histiocytic infiltrates [9].
Extra-osseous manifestations are numerous: central ner-
vous system involvement may lead to diabetes insipidus
[10], cerebellar symptoms, venous sinus thrombosis [11],
and dural based masses, which may mimic multiple
meningiomas [12,13]. Exopthalmos, which may be bila-
teral, occurs secondary to orbital infiltration. Pulmonary
involvement is manifest as worsening dyspnoea, with
interstitial infiltrates seen on chest radiographs [14]. Re-
troperitoneal fibrosis is a well-recognized sequela, and
can cause bilateral adrenal infiltration [15], and also can
lead to ureteric obstruction and progressive renal failure,
which is a common cause of death in these patients.
When our case presented with only blurred and de-
creased vision involvement diagnosis of VKHS, after
that persistent fever, increasing fatigue, malaise, weak-
ness, and leg pain developed. The diagnosis of ECD was
made until this case biopsy and curettage, according to
the results of immunophenotypic analysis, which mat-
ches the ECD description. We think that immune factors
may serve as an important stimulus for VKHS. Mean-
while, ECD lesions express a wide array of cyto-che-
mokines, which may orchestrate histiocyte recruitment
and activation [16,17], and which is also related with
immune factors. One of the unique points in our case is
the progression from VKHS to ECD. The three most
frequent clinical features of ECD are bone pain, diabetes
insipidus and exophthalmos, these features are called the
classic ECD triad. This patient only present with bone
pain companied by fever, and twice myeloid culture
reportedly showed Human staphylococcus aureus, which
might have contributed to the delayed diagnosis; ECD is
a very rare disease and difficult to diagnose, even with
classic symptoms. However, the diagnosis of ECD can
be made with near certainty because of the characteristic
radiographic findings. The bilateral diffuse or patchy
symmetric sclerosis of the medullary cavity of the major
long bones, with relative epiphyseal sparing, is almost
pathognomonic for ECD and can be confirmed by PET-
CT. The distinct histological findings of ECD are also
useful in making the diagnosis. Biopsy samples from the
organs involved will reveal diffuse infiltrations of foamy
histiocytes, Touton giant cells, lymphocytes and plasma
cells, associated with dense fibrosis and reactive sclero-
sis of the cortical and cancellous bone. Immunohisto-
chemistry of the biopsy samples confirms the monocyte/
macrophage lineage of the histiocytes by their expression
of CD68, LYS, and S100 protein (variable), and by their
lack of the Langerhans cell marker CD1a. This patient’s
biopsy material exhibited diffuse infiltration of CD68-
and LYS-expressing foamy histiocytes, and were nega-
tive for CD1a and S100. These findings were associated
with dense fibrosis, which matches the ECD description.
In this patient, additional immunohistochemical staining
with antibodies against CD68 and CD1a supplemented
the original histological findings. The present data, which
are, to the best of our knowledge, it is the first de-
scription of ECD with VKHS as the initial presentation.
It was difficult to assess the exact relationship between
these 2 conditions.
ECD is a rare non-Langerhans histiocytosis of un-
known etiology. The outcome of patients with ECD is
worse than that for Langerhans-cell histiocytosis with
59% of patients in the former group dead after a mean
follow-up of 32 months [3]. Although our study was not
designed to assess the efficacy of treatments in ECD
complicated by VKHS, these data suggest that corti-
costeroids which have been recognized as effective in
other literature of VKHS, and have only a minor impact
on the course of morbid involvement in ECD patients.
Treatment strategies for ECD have been challenged by
rare occurrence and poor understanding of etiopathology.
Therapy with interferon-α (IFN-α) effective in ECD
appear to have minimal benefit [18-20]. A rationale for
this approach is that IFN-α induces terminal differen-
tiation of some histiocytes and dendritic cells. Target-
ing interleukin-1 (IL-1), a proinflammatory cytokine ex-
pressed by ECD histiocytes, with recombinant inter-
leukin-1 receptor (IL-1R) has also recently been reported
as a promising therapeutic strategy [1]. Arnaud and col-
leagues quantitatively analyze 23 cytokines in serum
samples from 37 patients with ECD, and studied the
impact of treatment on cytokine production [21]. They
found no significant difference between cytokine levels
measured in untreated and IFN-α-treated patients. This
suggests that IFN-α has limited impact on the systemic
immunologic perturbations that are characteristic of ECD.
They also found many significant differences in cyto-
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7
kine/chemokine levels in ECD patients compared with
controls, with principal component analysis identifying
an ECD-specific signature based on increased expression
of IFN-α, IL-1/IL-1RA, IL-6, IL-12, monocyte chemo-
tactic protein-1, and decreased expression of IL-4 and
IL-7. The authors conclude that the systemic immune
Th-1-oriented perturbation associated with this condition,
and provides clues for the choice of more focused
therapeutic agents.
ECD is a rare form of non-Langerhans’ cell histiocy-
tosis. ECD appears to be the pathognomonic clinical,
radiographic and histological features. Diagnosis of ECD
is based on the identification in tissue biopsy of histio-
cytes, which are typically foamy and immunostain for
CD68+ CD1a. Central nervous system involvement is a
major prognostic factor in ECD. Because the incidence
of ECD is low, the data from literature remains limited.
More case data for this histiocytosis need therefore be
obtained for a better understanding of its characteristics
and an exploration of appropriate therapy.
[1] Aouba, A., Georgin-Lavialle, S., Pagnoux, C., Silva,
N.M., Renand, A., Galateau-Salle, F., et al. (2010) Ra-
tionale and efficacy of interleukin-1 targeting in Erdheim-
Chester disease. Blood, 116, 4070-4076.
[2] Haroche, J., Arnaud, L. and Amoura, Z. (2012) Erdheim-
Chester disease. Current Opinion in Rheumatology, 24,
53-59. doi:10.1097/BOR.0b013e32834d861d
[3] Veyssier Belot, C., Cacoub, P., Caparros Lefebvre, D.,
Wechsler, J., Brun, B., Remy, M., et al. (1996) Erdheim-
Chester disease—Clinical and radiologic characteristics
of 59 cases. Medicine, 75, 157-169.
[4] Wimpissinger, T.F., Schernthaner, G., Feichtinger, H. and
Stackl, W. (2005) Compression of kidneys in Erdheim-
Chester disease of retroperitoneum: Open surgical ap-
proach. Urology, 65, E29-E31.
[5] Chester, W. (1930) The lipogranulomatosis. Virchows
Archiv für Pathologische Anatomie und Physiologie und
für Klinische Medizin, 279, 561-602.
[6] Kairouz, S., Hashash, J., Kabbara, W., McHayleh, W. and
Tabbara, I.A. (2007) Dendritic cell neoplasms: An over-
view. American Journal of Hematology, 82, 924-928.
[7] Zelger, B.W.H., Sidoroff, A., Orchard, G. and Cerio, R.
(1996) Non-Langerhans cell histiocytoses—A new uni-
fying concept. American Journal of Dermatopathology,
18, 490-504. doi:10.1097/00000372-199610000-00008
[8] Dion, E., Graef, C., Miquel, A., Haroche, J., Wechsler, B.,
Amoura, Z., et al. (2006) Bone involvement in Erdheirn-
Chester disease: Imaging findings including periostitis
and partial epiphyseal involvement. Radiology, 238, 632-
639. doi:10.1148/radiol.2382041525
[9] Allen, C.E. and McClain, K.L. (2011) Erdheim-Chester:
Beyond the lesion. Blood, 11 7, 2745-2746.
[10] Wright, R.A., Hermann, R.C. and Parisi, J.E. (1999) Neu-
rological manifestations of Erdheim-Chester disease. Jour-
nal of Neurology, Neurosurgery & Psychiatry, 66, 72-75.
[11] Perras, B., Petersen, D., Lorch, H. and Fehm, H.L. (2002)
Psychoneuroendocrine disturbances in a patient with a
rare granulomatous disease. Experimental and Clinical
Endocrinology & Diabetes, 110, 248-252.
[12] Naqi, R., Azeemuddin, M., Idrees, R. and Wasay, M.
(2010) Meningioma-like lesions in Erdheim Chester dis-
ease. Acta Neurochirurgica, 152, 1619-1621.
[13] Donaldson, G., Bullock, P. and Monson, J.P. (2010) Erd-
heim-Chester disease mimicking multiple meningiomas.
British Journal of Neurosurgery, 24, 296-297.
[14] Arnaud, L., Pierre, I., Beigelman-Aubry, C., Capron, F.,
Brun, A.L., Rigolet, A., et al. (2010) Pulmonary in-
volvement in Erdheim-Chester disease: A single-center
study of thirty-four patients and a review of the literature.
Arthritis & Rheumatism, 62, 3504-3512.
[15] Haroche, J., Amoura, Z., Touraine, P., Seilhean, D., Graef,
C., Birmele, B., et al. (2007) Bilateral adrenal infiltration
in Erdheim-Chester disease. Report of seven cases and
literature review. Journal of Clinical Endocrinology &
Metabolism, 92, 2007-2012. doi:10.1210/jc.2006-2018
[16] Stoppacciaro, A., Ferrarini, M., Salmaggi, C., Colarossi,
C., Praderio, L., Tresoldi, M., et al. (2006) Immunohis-
tochemical evidence of a cytokine and chemokine net-
work in three patients with Erdheim-Chester disease—
Implications for pathogenesis. Arthritis and Rheumatism,
54, 4018-4022. doi:10.1002/art.22280
[17] Dagna, L., Girlanda, S., Langheim, S., et al. (2010) Erd-
heim-Chester disease: Report on a case and new insights
on its immunopathogenesis. Rheumatology, 49, 1203-
1206. doi:10.1093/rheumatology/kep461
[18] Braiteh, F., Boxrud, C., Esmaeli, B. and Kurzrock, R.
(2005) Successful treatment of Erdheim-Chester disease,
a non-Langerhans-cell histiocytosis, with interferon-alpha.
Blood, 106, 2992-2994. doi:10.1182/blood-2005-06-2238
[19] Jeon, I.-S., Lee, S.S. and Lee, M.K. (2010) Chemother-
apy and Interferon-alpha treatment of Erdheim-Chester
disease. Pediatric Blood & Cancer, 55, 745-747.
[20] Suzuki, H.I., Hosoya, N., Miyagawa, K., Ota, S., Naka-
shima, H., Makita, N., et al. (2010) Erdheim-Chester dis-
ease: Multisystem involvement and management with in-
terferon-alpha. Leukemia Research, 34, E21-E24.
Copyright © 2013 SciRes. Openly accessible at http://www.sc irp.or g/journal/crcm/
N. Li et al. / Case Reports in Clinical Medicine 2 (2013) 35 1-35 7
Copyright © 2013 SciRes. http://www.scirp.org/jour nal/crcm/ Openly accessible at
[21] Arnaud, L., Gorochov, G., Charlotte, F., Lvovschi, V.,
Parizot, C., Larsen, M., et al. (2011) Systemic perturba-
tion of cytokine and chemokine networks in Erdheim-
Chester disease: A single-center series of 37 patients.
Blood, 117, 2783-2790.