Rosai-Dorfman disease (RDD) is characterized histologically by a dense histiocytic infiltrate with emperipolesis and associated lymphocytes, plasma cells, and neutrophils. Eosinophils are not commonly associated. We report a patient with initial thymus and pituitary gland involvement by RDD, who later developed papules on the groin and axilla. Skin biopsies showed admixed histiocytic infiltrates (lymphocytes, neutrophils, and plasma cells) without emperipolesis. A prominent eosinophilic infiltrate was also observed, a feature not, to our knowledge, previously reported. Immunohistochemistry revealed positivity for CD68 (most cells) and S100 protein (scattered cells) and was negative for anti-CD1a. The diagnosis of RDD was established in the clinical context after comparison with the thymic and pituitary lesions (similar histologic features, albeit with fewer eosinophils, and immunohistochemical profiles). We present the first case, to our knowledge, of multicentric RDD with cutaneous involvement and associated prominent eosinophilic infiltrate. Thus, RDD should be included in the differential diagnosis of mononuclear infiltrates containing eosinophils.

Rosai-Dorfman disease (RDD), also known as sinus histiocytosis with massive lymphadenopathy, is a benign, self-limiting histioproliferative disorder commonly involving the lymph nodes of the head and neck and occasionally extranodal sites, including the skin. Histologically, it is characterized by the presence of a large number of histiocytes, many of which contain engulfed inflammatory cells (emperipolesis), and admixed lymphocytes, plasma cells, and neutrophils in various proportions. We present an unusual case of RDD with an associated prominent eosinophilic infiltrate, a diagnostic feature that has not, to our knowledge, been previously reported. Historically, not only was the presence of eosinophils not considered to be a histologic feature of RDD but also the presence of eosinophils within a histiocytic infiltrate was typically used to help distinguish RDD from Langerhans cell histiocytosis, a histiocytic disorder characterized by heavy eosinophilic infiltrates.

Initial Presentation

A 50-year-old woman presented originally with a thymic mass, resected with sternotomy, and a pituitary mass, removed by transsphenoidal resection. The histologic examination of the mediastinal thymic mass showed a dense infiltrate consisting primarily of histiocytes, lymphocytes, and few plasma cells. Scattered histiocytes had ample cytoplasm and engulfed intracytoplasmic lymphocytes (emperipolesis). Immunohistochemical studies were performed and the histiocytes were labeled by CD68 (Dako, Carpinteria, California) and were focally positive for S100 protein (Biogenex, San Ramon, California), highlighting the emperipolesis, and were negative for CD1a (Immunotech, Brea, California), supporting the diagnosis of Rosai-Dorfman disease. The pituitary mass showed similar histologic and immunohistochemical findings.

Nine months after resection of the thymic and pituitary lesions, the patient presented with multiple (>30) cutaneous papules, measuring up to 0.5 cm, on her bilateral axilla and medial thighs, just below her groin (Figure, A and B). No generalized lymphadenopathy was identified. No topical medications were applied to the lesions. Skin biopsies of axillary and thigh lesions showed a similar histiocytic infiltrate to that found in the pituitary and thymic masses, with the exception of the presence of numerous eosinophils (Figure, C and D). Emperipolesis was also identified (Figure, E). Immunohistochemical studies revealed that many of the lesional cells were labeled by CD68, some large mononuclear cells were highlighted with S100 protein, and anti-CD1a was essentially negative. The diagnosis of RDD was established in the clinical context, after comparison to the thymic and pituitary lesions, which showed similar histologic features (albeit with fewer eosinophils) and immunohistochemical profiles.

Multiple erythematous papular lesions of the axilla (A) and medial thigh (B). C and D, Right axilla skin biopsy with dense, histiocytic, dermal infiltrate admixed with lymphocytes, neutrophils, and numerous eosinophils. E, Emperipolesis (yellow arrow) is also seen (hematoxylin-eosin, original magnifications ×2 [C] and ×40 [D and E]).

Multiple erythematous papular lesions of the axilla (A) and medial thigh (B). C and D, Right axilla skin biopsy with dense, histiocytic, dermal infiltrate admixed with lymphocytes, neutrophils, and numerous eosinophils. E, Emperipolesis (yellow arrow) is also seen (hematoxylin-eosin, original magnifications ×2 [C] and ×40 [D and E]).

Close modal

Although Rosai-Dorfman disease rarely presents a diagnostic challenge, uncharacteristic presentations can cause confusion. Originally described by Juan Rosai and Ronald Dorfman in 1969,1 RDD was observed as a triad of massive cervical lymphadenopathy, expanded lymph node sinuses, and characteristic emperipolesis within histiocytes.1 With a slight male predominance2 and a mean onset of 20 years,3 RDD often presents as prominent bilateral cervical lymphadenopathy accompanied by nonspecific clinical symptoms, such as fever, tenderness, malaise, night sweats, and weight loss.3 

Primarily a histopathologic diagnosis, RDD is characterized by histiocytic proliferation accompanied by abundant admixed infiltrates of lymphocytes, neutrophils, and plasma cells. Yet eosinophils are considered rare and uncharacteristic of the disease.1 In our case, the unusually prominent infiltration of eosinophils in the skin lesions caused us to question a diagnosis of RDD. However, after review of the original pituitary and thymic biopsies, which showed a similar histiocytic infiltrate, albeit with more classic RDD features, including extensive emperipolesis and rare, scattered eosinophils, the diagnosis of RDD involving the skin was established. Although emperipolesis is not unique to RDD, when found in S100-expressing histiocytes, it is considered a diagnostic feature.4 Moreover, in all of our patient's biopsies, the immunohistochemical studies were consistent with RDD, including CD68 and S100 protein labeling of histiocytes and lack of CD1a expression.

When RDD is suspected, it is important to exclude other histiocytic disorders, especially Langerhans cell histiocytosis (LCH), a potentially fatal disease.5 The histiocytes in LCH are characterized by irregular nuclear contours (nuclear grooves and “coffee-bean”–shaped nuclei) that label diffusely for CD1a and contain Birbeck granules on electron microscopy. The associated inflammatory infiltrate is dominated by eosinophils and lymphocytes, with emperipolesis characteristically lacking. During examination of our patient's skin biopsy, the many infiltrating eosinophils put LCH at the top of our differential diagnosis list. However, the lack of prominent nuclear grooves and the presence of emperipolesis made LCH unlikely. Furthermore, after the review of the immunohistochemical studies and comparison with our patient's previous pituitary and thymic lesions, we excluded the diagnosis of LCH. Thus, we consider that any histiocytic lesion with admixed prominent eosinophilic infiltrate requires immunohistochemical studies to rule out LCH.

Another differential diagnosis considered was the xanthogranulomatous family of diseases. This collection of cutaneous histiocytic disorders can display a wide variety of morphologic characteristics, including mononuclear cells that may be vacuolated, xanthomatous, spindle-shaped (all characteristically found in juvenile xanthogranulomas), and display oncocytic changes (more commonly seen in adult xanthogranulomas).6 Although multinucleated giant cells are often found and help classify the disorder (Touton giant cells in juvenile xanthogranuloma and Langhans giant cells in adult xanthogranuloma), they are, by no means, specific. Multinucleated giant cells, as well as foamy histiocytes, can be seen in RDD; therefore, the presence or lack of emperipolesis is probably the single most-important histologic feature that can help distinguish RDD and xanthogranulomatous diseases. Immunohistochemistry can also help because xanthogranulomatous diseases are usually positive for factor XIIIa, CD68, CD163, fascin, and CD14 and negative for CD1a and S100.7 Notable exceptions are solitary and multicentric reticulohistiocytosis, which are usually negative for factor XIIIa.8 Within the appropriate clinical context, the diagnosis is often apparent.

Malignant histiocytic disorders must also be ruled out, including histiocytic sarcoma, a rare lymphohematopoietic malignant tumor of cells with histiocytic differentiation,9 frequent extranodal presentation, and generally aggressive clinical course (median survival, 4–5 months).10 Multinucleated giant cells, foamy cells, and spindle-shaped cells may be encountered and, depending on sampling issues, might be confused with RDD or the aforementioned xanthogranulomatous diseases. Many of the histiocytic cells display a high degree of cellular atypia. The diagnosis must be supported by CD68 positivity, which is also accompanied by positive lysozyme staining in 80% of cases.11 Because lysozyme has been reported positive in RDD, there is little value in using that histochemical feature to distinguish the 2 diseases.12 Few cases of histiocytic sarcoma arising in the background of RDD have been reported, but that is a possible, exceedingly rare, occurrence.13 The key features for distinction of malignant histiocytic entities from benign histiocytic proliferations, such as RDD, are significant cytologic atypia, pleomorphism, and evidence of tissue invasion or destruction.14 Hornick et al11 collected 14 cases of extranodal histiocytic sarcoma and noticed that mitotic figures ranged from 1 to 64 per 10 high-power fields (median 11 per 10 high-power fields), necrosis was present in 8 cases (57%), and nearly all tumors showed a striking inflammatory infiltrate, most often neutrophils or lymphocytes, important features not seen in RDD.

Rosai-Dorfman disease may be confused with infectious diseases, such as tuberculosis,15 but clinicopathologic correlation, along with special stains, can help to differentiate the 2 conditions. Fite stain can help rule out lepromatous leprosy, where foamy histiocytes predominate but are separated from the epidermis by a distinct grenz zone. Giemsa staining can also help rule out Leishmania parasites, where amastigote organisms are usually present within histiocytes. Benign inflammatory conditions can also be confused with RDD and include granuloma annulare (palisading granulomas with necrobiosis and mucin accumulation) and inflammatory pseudotumor (vimentin and smooth muscle actin positive, S100 negative). Special attention is needed to rule out malignant neoplasms that can commonly mimic RDD, including melanoma (HMB45 and MART-1 positive) and epithelioid sarcoma (epithelial membrane antigen, cytokeratin, vimentin, CD34, and actin positive; most often, S100 negative). None of these cases have the characteristic large histiocytes with emperipolesis and S100 protein expression.

We present an unusual case of multicentric RDD with cutaneous involvement and associated prominent eosinophilic infiltrate. The correct diagnosis of RDD required correlating immunohistochemical studies with the overall clinical picture, including thymic and pituitary lesions, which showed a more classic histologic picture of the disease (ie, with minimal eosinophils). When considering RDD, it is important to exclude other common mimickers, including LCH, xanthogranulomatous disease, histiocytic sarcoma, infections, and melanoma. This case brings to light the importance of including RDD in the differential diagnosis of any mononuclear histiocytic infiltrate with extensive infiltration of eosinophils.

1.
Rosai
J
,
Dorfman
RF
.
Sinus histiocytosis with massive lymphadenopathy: a newly recognized benign clinicopathological entity
.
Arch Pathol
.
1969
;
87
(
1
):
63
70
.
2.
Foucar
E
,
Rosai
J
,
Dorfman
RF
.
Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): review of the entity
.
Semin Diagn Pathol
.
1990
;
7
(
1
):
19
73
.
3.
Gaitonde
S
.
Multifocal, extranodal sinus histiocytosis with massive lymphadenopathy: an overview [review]
.
Arch Pathol Lab Med
.
2007
;
131
(
7
):
1117
1121
.
4.
Juskevicius
R
,
Finley
JL
.
Rosai-Dorfman disease of the parotid gland: cytologic and histopathologic findings with immunohistochemical correlation
.
Arch Pathol Lab Med
.
2001
;
125
(
10
):
1348
1350
.
5.
Willman
CL
,
Busque
L
,
Griffith
BB
, et al.
Langerhans'-cell histiocytosis (histiocytosis X)—a clonal proliferative disease
.
N Engl J Med
.
1994
;
331
(
3
):
154
160
.
6.
Zelger
BW
,
Sidoroff
A
,
Orchard
G
,
Cerio
R
.
Non-Langerhans cell histiocytoses: a new unifying concept [review]
.
Am J Dermatopathol
.
1996
;
18
(
5
):
490
504
.
7.
Weitzman
S
,
Jaffe
R
.
Uncommon histiocytic disorders: the non-Langerhans cell histiocytoses
.
Pediatr Blood Cancer
.
2005
;
45
(
3
):
256
264
.
8.
Luz
FB
,
Gaspar
AP
,
Ramos-e-Silva
M
, et al.
Immunohistochemical profile of multicentric reticulohistiocytosis [review]
.
Skinmed
.
2005
;
4
(
2
):
71
77
.
9.
Huang
SC
,
Chang
CL
,
Huang
CH
,
Chang
CE
.
Histiocytic sarcoma: a case with evenly distributed multinucleated giant cells
.
Pathol Res Pract
.
2007
;
203
(
9
):
683
689
.
10.
Weiss
LM
,
Grogan
TM
,
Muller-Hermeilink
HK
,
Stein
H
.
Histiocytic sarcoma: pathology and genetics of tumours of haematopoietic and lymphoid tissues
.
In
:
Jaffe
ES
,
Harris
NL
,
Stein
H
,
Vardiman
J
,
eds
.
Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues
.
Lyon, France
:
IARC Press
;
2001
:
278
279
.
World Health Organization Classification of Tumours; vol 3.
11.
Hornick
JL
,
Jaffe
ES
,
Fletcher
CD
.
Extranodal histiocytic sarcoma: clinicopathologic analysis of 14 cases of a rare epithelioid malignancy
.
Am J Surg Pathol
.
2004
;
28
(
9
):
1133
1144
.
12.
Montgomery
EA
,
Meis
JM
,
Frizzera
G
.
Rosai-Dorfman disease of soft tissue
.
Am J Surg Pathol
.
1992
;
16
(
2
):
122
129
.
13.
Venkataraman
G
,
McClain
KL
,
Pittaluga
S
,
Rao
VK
,
Jaffe
ES
.
Development of disseminated histiocytic sarcoma in a patient with autoimmune lymphoproliferative syndrome and associated Rosai-Dorfman disease
.
Am J Surg Pathol
.
2010
;
34
(
4
):
589
594
.
14.
Vos
JA
,
Abbondanzo
SL
,
Barekman
CL
,
Andriko
JW
,
Miettinen
M
,
Aguilera
NS
.
Histiocytic sarcoma: a study of five cases including the histiocyte marker CD163
.
Mod Pathol
.
2005
;
18
(
5
):
693
704
.
15.
Safdar
A
,
Gillenwater
AM
,
Jones
DM
,
Jorgensen
JL
,
Romaguera
JE
.
Rosai-Dorfman disease misdiagnosed as active tuberculosis
.
Leuk Lymphoma
.
2006
;
47
(
7
):
1441
1442
.

Author notes

From the Departments of Pathology (Drs Cangelosi, Prieto, and Ivan) and Dermatology (Drs Prieto and Ivan), The University of Texas M. D. Anderson Cancer Center, Houston.

The authors have no relevant financial interest in the products or companies described in this article.

Presented as a poster at the 44th Annual Meeting of the American Society of Dermatopathology, Baltimore, Maryland, October 18–21, 2007.