Hematopathology

Langerhans' Cell Sarcoma (LCS)

Ronald Jaffe
Children's Hospital of Pittsburgh
Pittsburgh, PA


History
A female patient first presented at age 42 with a 1.7 cm ulcerated, raised nodular lesion on the right knee. The process was diagnosed on biopsy as "Histiocytosis X". There was a high mitotic rate (30% on MIB-1) but no atypical mitoses. A year later, she developed a raised nodule on the cheek, 1.3 cm that proved to be the same process. Tumor necrosis was more obvious and the MIB – 1 count was 50%. Rare atypical mitoses were encountered. The diagnosis was again "Histiocytosis X" and no immediate further treatment was given. Over the next few years, she developed additional lesions in small and large bowel and was treated with surgery and chemotherapy. She died, 8 years after presentation, with progressive tumor that involved skin, soft tissue, thyroid, liver, lung, bone, and brain, including the posterior pituitary.

The patient requested that her tumor be used for study and teaching. Tumor was harvested 5 days after death. I am privileged to honor her wish.


Slide 1
Click to view with ImageScope
Click to view with a Web-Based Viewer



Figure 1
H&E x 20. Liver. Diffuse proliferation of cells that have a "histiocytic" appearance
Figure 2
H&E x 40. Liver. The nuclear morphology holds a clue. The nucleus is more than coffee-bean shaped or grooved but displays a complex folded appearance.
Figure 3
H&E x 40. Liver. There is variation in nuclear size, pleomorphism, and a high mitotic rate that includes atypical mitoses.

Figure 4
Leukocyte Common Antigen. Liver. There is little to no staining of the histiocytic population.
Figure 5
Vimentin. Liver. All cells stain strongly.
Figure 6
CD163. Liver. Interspersed macrophages stain strongly but cells with the signature nucleus are unstained.

Figure 7
S100. Liver. There is uneven cytoplasmic staining but little nuclear staining.
Figure 8
Ki-67. Liver. The very high proliferative rate is highlighted, including atypical forms.
Figure 9
CDla. Liver. The staining is universal. The quality of the membrane stain in this example is compromised by delayed fixation.

Figure 10
Langerin. Liver. Many of the cells in this field, but not all, stain for the presence of Langerin.
Figure 11
H&E x 10. Lymph node. The node is largely replaced with only residual follicles but the sinus pattern is not evident on H&E.
Figure 12
H&E x 40. Lymph node. Sinuses are filled with the Langerhans type cells that have cytologic pleomorphism and evident mitoses.

Figure 13
S100. Lymph node. The pattern of staining highlights the sinus component and paracortical spill-over.
Figure 14
CDla. Lymph node. The sharp membrane quality of the staining is evident.
Figure 15
Langerin. Lymph node. The lesional cells are universally positive.


Diagnosis
Langerhans' cell sarcoma (LCS). Review of the prior lesions are also LCS.

Introduction
Langerhans' cell sarcoma (LCS) is a high-grade neoplasm that has overtly malignant cytologic features in most instances, and the Langerhans' cell phenotype, S100+/CDla+/langerin+ including the ultrastructural Birbeck granule. The tumor develops de novo and is not a consequence of recurrent Langerhans' cell histiocytosis (LCH). [1]

Clinical Findings & Biological Behavior
Langerhans' cell sarcoma is generally an adult disease, the age range being 8-81 years with an unconfirmed mention of a congenital case. The mean age is 44 years when 6 unpublished cases from my files are included with a median of 46 years and unlike prior smaller series that suggested a female predominance, the M:F ratio is 1:1. The most frequent sites of presentation are ranked as skin, soft tissue, lymph node, lung, tonsil, bone, and spleen and the usual clinical presentation is that of a mass lesion. [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] The clinical course is aggressive with limited response to chemotherapy and half of the published patients have died of progressive disease. [18]

Histologic & Immunophenotype Findings
The illustrated case is prototypical though not as anaplastic as some examples. The tumor is commonly first recognized as being malignant or high-grade and only of Langerhans' cell type when the phenotypic markers are revealed. The lesional cells can be oval, 'epithelioid', or slightly spindled. When lymph nodes are involved, the infiltrate is largely confined to the sinuses, although this pattern may be obscured if the infiltrate is advanced. A morphologic clue to the diagnosis is the large nuclear size that includes giant forms and the extensive and very complex nuclear folding – the term "grooved" does not do justice to the complexity. Interspersed cells can appear more monocytoid with a nuclear pattern that resembles a paper folded over from the corners. Mitoses can be common and large, atypical, or bizarre forms can occur. The MIB-1 count in this instance and others was over 50% of all the large cells. The phenotype of the cells is classical for the Langerhans' cell; Leukocyte Common Antigen- (or trace only), vimentin +, S100+ nuclear and cytoplasmic, CDla+ sharp membrane pattern and langerin +. In contrast to Langerhans' cell histiocytosis (LCH), the staining of LCS may be less uniform, more patchy, and inconsistent. In the current autopsy example, langerin was present on less than 20% of cells overall but more widespread in earlier lesions. Birbeck granules have generally (but not always) [17] been found on ultrastructural examination. A loss of CDla and/or langerin has been seen on late recurrences. CD14, CD68, and CD163 highlight a high content of intra-tumoral macrophages. Giant cells can be of two types, tumoral, that bear grooved, folded nuclei like those of the LCS cells and also have a small amount of patchy Langerhans' cell-type staining, and macrophage or osteoclast type with high cytoplasmic CD68. There has been a report of CD56 expression in LCS with the suggestion that it marks a more refractory population of cases. [19] The current example was not positive, nor were the five other cases on file. CD30 may be positive and a source of consternation because it mimics the sinus growth pattern of someanaplastic large cell lymphomas.

Differential Diagnosis
Langerhans' cell sarcoma can, at presentation and depending on the site, be taken for a broad differential that includes undifferentiated carcinomas, large cell lymphomas, metastatic melanoma, and sarcomas. The tumor class is cytokeratin -,vimentin +, and LCA is generally absent to trivial. The demonstration of Langerhans' cell affiliation by virtue of the nuclear complexity, S100+, CDla+, langerin + with or without Birbeck granules is convincing [20] and serves to exclude most other hematopoietic or soft-tissue candidates. The differential diagnosis, once the phenotype is established, revolves around three possible confounders; Langerhans' cell histiocytosis, hyperplastic collections of Langerhans' cells and other neoplastic conditions that can express CDla. "Indeterminate cell histiocytosis" will also be briefly discussed. Most instances of LCH present with skin/bone/soft tissue and/or lymph node involvement, singly or in combination. Is there a clear and unequivocal separation between LCH and LCS? For the most part the answer is affirmative based on the histology. LCH has a very bland nuclear morphology without nuclear pleomorphism (though giant cells may be a large component) and devoid of atypical mitoses. The MIB-1 fraction in the CD1a+ LCH cells, however, can be as high as 30% in instances. Systemic disease in childhood, especially under age 2 years still carries a substantial mortality (20%). The histopathology, cytology, and phenotype of these fatal systemic lesions are no different from those of localized LCH, and even at autopsy, disseminated LCH lacks the high-grade nuclear features of Langerhans' cell sarcoma.

Are there transitions from LCH to sarcoma?

The literature has only very rare anecdotal instances that appear to be well documented and in which the original lesion was unequivocally LCH, followed by "recurrences" that were anaplastic ("progression to sarcoma") [4, 21, 22] In at least one case reviewed here it was clear that the high degree of pleomorphism and high local turnover in the original, labeled "LCH" had been under-appreciated. In this case, the same was true; review of the original lesions revealed a high-grade tumor from the outset.

Is there a group of cases, especially in adults, in which it is hard to tell if the presenting lesion is likely to behave as sarcoma? Since pleomorphism and anaplasia may be of varying degrees and in the eye of the beholder, there are rare adult patients who have lesions that fail the pleomorphism/anaplasia criterion for the diagnosis of Langerhans' cell sarcoma, but in whom the local cell turnover, judged by MIB-1 content, is over 50%. Clinical follow-up in this small group of patients is not yet clear-cut.

Langerhans' Cell Hyperplasia
Occasionally sheets of Langerhans' cells can be encountered within other lesions, mostly neoplastic, such as nodal leukemias, lymphomas, soft tissue sarcomas, thymus, thyroid, or non-Langerhans' cell histiocytic lesions like Rosai-Dorfman disease. The biological outcome of these Langerhans cell aggregates is determined solely by the lesions in which they appear. There are no reports of Langerhans' cell histiocytosis at any other sites in any of these patients. Christie et al demonstrated that in some of these lesions the proliferation, unlike LCH, is not clonal and they suggested that these Langerhans' cell aggregates occurring within other neoplasms are reactive "hyperplastic" foci rather than Langerhans' cell histiocytosis and certainly not sarcomatous. [23] Florid dendritic cell responses to other neoplastic lesions, predominantly in the skin and soft tissues, can mimic LCH, rarely LCS. [24] Langerin is absent from the reactive CD1a+ population.

Other Cdla-expressing Lesions
There are other neoplastic conditions that can express CDla and hence confound with Langerhans' cell sarcoma. Myelomonocytic leukemias and on occasion, myelodysplastic syndromes can give rise to skin, soft tissue, or organ infiltrates that are more differentiated than their marrow or blood counterparts. The maturation/differentiation may be in the direction of macrophage or dendritic cells and some express CDla. The lesion has been mistaken for LCH, even though the cells are more immature and lack the nuclear folding of the LCH cell. Mitoses, blastic appearance, and high proliferative index in this situation are also expected, making confusion with LCS possible. It is the panel combination of CDla, langerin, and S100 positivity that make the definitive diagnosis of Langerhans' cell lesions. Acute myeloid leukemia would variably express other antigens such as CD15, CD33, CD43, CD68, CD117, or myeloperoxidase. [25] A subset of T-lymphoblastic leukemia/lymphoma (cortical T-cell) is CD1a+ and would express other T-cell antigens and TdT. Uterine PEcomas (but not epithelioid leiomyomas) have been noted to express CDla, but the context is different. [26] Here too, langerin may be required for the absolute differential.

Indeterminate Cell Histiocytosis and Sarcoma
"Indeterminate cell lesions" are a mixed group, originally defined as being Langerhans' cell-like, expressing CDla but without Birbeck granules. There are many reasons for the failure to find Birbeck granules in a given lesion, and sampling variation is an important one. Historically, therefore, the diagnostic "entity" is suspect. The absence of langerin staining is to be expected, and has recently been documented. [27] Since the Langerhans' cell markers, CDla, langerin and the Birbeck granule are more variable in LCS than in LCH or even epidermal Langerhans' cells, and well documented the case under discussion, it is not unexpected that some instances will fail to express langerin or to show ultrastructural Birbeck granules and in some cases these markers that were originally present have been lost on the later recurrences. There seems to be little merit to the diagnosis of "Indeterminate cell" lesions including the term, "indeterminate cell sarcoma" [28] unless their biological behavior can be shown to be quite different from LCH or LCS, however, the current WHO classification includes this category in "other rare dendritic cell tumours". [29]

Summary
By virtue of their cytologic detail, the complex folded nuclei, high-grade neoplastic features and almost unique phenotypic panel, S100+ CDla+ langerin+, Langerhans cell sarcoma should not be misdiagnosed. Distinction from LCH, especially in adults, is usually obvious by the above criteria with only few exceptions. Half of the reported cases were lethal.

Take Home Points
  • The differential diagnosis of large cell tumors is broad and includes undifferentiated carcinoma, large cell lymphoma, metastatic melanoma, and sarcoma

  • Large cell tumors often represent an aggressive manifestation of more well-differentiated neoplasms

  • Attention to nuclear and cytoplasmic morphology can facilitate the diagnosis and choice of an immunologic panel

  • Phenotypic markers are useful in defining the cell of origin but are not unique and must be used in the context of a panel

  • Ultrastructural examination can still be informative in some tumors even though phenotypic markers may have superseded the utility.
References
  1. Jaffe R, Weiss LM, Facchetti F. Tumours derived from Langerhans cells. in WHO Classificiation of Tumours of the Haematopoietic and Lymphoid Tissues. Swerdlow SH, Campo E, Harris NL, et al., [Eds.] IARC: Lyons. 2008, 358-360.

  2. Imamura M, Sakamoto S, Hanazono H. Malignant histiocytosis: a case of generalized histiocytosis with infiltration of Langerhans' granule-containing histiocytes. Cancer 28: 467-475, 1971.

  3. Wood C, Wood GS, Deneau DG, et al. Malignant histiocytosis X. Report of a rapidly fatal case in an elderly man. Cancer 54: 347-352, 1984.

  4. Elleder M, Fakan F, Hula M. Pleiomorphous histiocytic sarcoma arising in a patient with histiocytosis X. Neoplasma 33: 117-128, 1986.

  5. Delabie J, De Wolf-Peeters C, De Vos R, et al. True histiocytic neoplasm of Langerhans' cell type. J Pathol 163: 217-223, 1991.

  6. Ben-Ezra J, Bailey A, Azumi N, et al. Malignant histiocytosis X. A distinct clinicopathologic entity. Cancer 68: 1050-1060, 1991.

  7. Tani M, Ishii N, Kumagai M, et al. Malignant Langerhans cell tumour. Br J Dermatol 126: 398-403, 1992.

  8. Favara BE, Feller AC, Pauli M, et al. Contemporary classification of histiocytic disorders. The WHO Committee On Histiocytic/Reticulum Cell Proliferations. Reclassification Working Group of the Histiocyte Society. Med Pediatr Oncol 29: 157-166, 1997.

  9. Bohn OL, Ruiz-Arguelles G, Navarro L, et al. Cutaneous Langerhans cell sarcoma: a case report and review of the literature. Int J Hematol 85: 116-120, 2007.

  10. Diaz-Sarrio C, Salvatella-Danes N, Castro-Forns M, et al. Langerhans cell sarcoma in a patient who underwent transplantation. J Eur Acad Dermatol Venereol 21: 973-976, 2007.

  11. Ferringer T, Banks PM, Metcalf JS. Langerhans cell sarcoma. Am J Dermatopathol 28: 36-39, 2006.

  12. Itoh H, Miyaguni H, Kataoka H, et al. Primary cutaneous Langerhans cell histiocytosis showing malignant phenotype in an elderly woman: report of a fatal case. J Cutan Pathol 28: 371-378, 2001.

  13. Julg BD, Weidner S, Mayr D. Pulmonary manifestation of a Langerhans cell sarcoma: case report and review of the literature. Virchows Arch 448: 369-374, 2006.

  14. Misery L, Godard W, Hamzeh H, et al. Malignant Langerhans cell tumor: a case with a favorable outcome associated with the absence of blood dendritic cell proliferation. J Am Acad Dermatol 49: 527-529, 2003.

  15. Pileri SA, Grogan TM, Harris NL, et al. Tumours of histiocytes and accessory dendritic cells: an immunohistochemical approach to classification from the International Lymphoma Study Group based on 61 cases. Histopathology 41: 1-29, 2002.

  16. Sumida K, Yoshidomi Y, Koga H, et al. Leukemic transformation of Langerhans cell sarcoma. Int J Hematol 87: 527-531, 2008.

  17. Zhao G, Luo M, Wu ZY, et al. Clinical and Pathological Features of Langerhans Cell Sarcoma: A Case Report of Langerhans Cell Sarcoma Involving Gallbladder and Peritoneal Lymph Nodes. Int J Surg Pathol, 2008.

  18. Yoshimi A, Kumano K, Motokura T, et al. ESHAP therapy effective in a patient with Langerhans cell sarcoma. Int J Hematol 87: 532-537, 2008.

  19. Kawase T, Hamazaki M, Ogura M, et al. CD56/NCAM-positive Langerhans cell sarcoma: a clinicopathologic study of 4 cases. Int J Hematol 81: 323-329, 2005.

  20. Lau SK, Chu PG, Weiss LM. Immunohistochemical expression of Langerin in Langerhans cell histiocytosis and non-Langerhans cell histiocytic disorders. Am J Surg Pathol 32: 615-619, 2008.

  21. Lee JS, Ko GH, Kim HC, et al. Langerhans cell sarcoma arising from Langerhans cell histiocytosis: a case report. J Korean Med Sci 21: 577-580, 2006.

  22. Saribeyoglu ET, Atay D, Bilgic B, et al. A patient with Langerhans cell sarcoma arising from Langerhans cell histiocytosis. at 24th Annual Meeting of the Histiocyte Society. October 1-2, 2008. Berlin, Germany.

  23. Christie LJ, Evans AT, Bray SE, et al. Lesions resembling Langerhans cell histiocytosis in association with other lymphoproliferative disorders: a reactive or neoplastic phenomenon? Hum Pathol 37: 32-39, 2006.

  24. Jokinen CH, Wolgamot GM, Wood BL, et al. Lymphomatoid papulosis with CD1a+ dendritic cell hyperplasia, mimicking Langerhans cell histiocytosis. J Cutan Pathol 34: 584-587, 2007.

  25. Lauritzen AF, Delsol G, Hansen NE, et al. Histiocytic sarcomas and monoblastic leukemias. A clinical, histologic, and immunophenotypical study. Am J Clin Pathol 102: 45-54, 1994.

  26. Adachi Y, Horie Y, Kitamura Y, et al. CD1a expression in PEComas. Pathol Int 58: 169-173, 2008.

  27. Rezk SA, Spagnolo DV, Brynes RK, et al. Indeterminate cell tumor: a rare dendritic neoplasm. Am J Surg Pathol 32: 1868-1876, 2008.

  28. Deng A, Lee W, Pfau R, et al. Primary cutaneous Langerhans cell sarcoma without Birbeck granules: indeterminate cell sarcoma? J Cutan Pathol 35: 849-854, 2008.

  29. Weiss LM, Chan JCK, Fletcher CDM. Other rare dendritic cell tumours. in WHO Classificiation of Tumours of the Haematopoietic and Lymphoid Tissues. Swerdlow SH, Campo E, Harris NL, et al., [Eds.] IARC: Lyons. 2008, 365.