Myelofibrosis

We describe the case of a 39-year-old man with idiopathic myelofibrosis, who developed histiocytic sarcoma (true histiocytic lymphoma) 6 months after diagnosis. The patient developed generalized lymphadenopathy. A lymph node biopsy snowed pronounced distension of the sinuses in the medulla and periphery, caused by the accumulation of large tumor cells. The tumor cells had abundant clear or eosinophilic cytoplasm. The nuclei were of various sizes and shapes, with condensed chromatin and prominent nucleoli. Some tumor cells displayed erythrophagocytosis. Immunohistochemically, the tumor cells were positive for CD68, α^sub 1^-antitrypsin, CD45, CD45RO, and S100 protein, and were negative for B- and T-cell markers, CD30, CD1a, lysozyme, myeloperoxidase, factor VIII-related antigen, CAM 5.2, and HMB-45. Despite multiagent chemotherapy, the patient died of disease 25 months after diagnosis. Although histiocytic sarcomas are very rare, their recognition may be important for clinical and prognostic reasons.

(Arch Pathol Lab Med. 2004;128:1167-1170)

Before the advent of immunophenotyping, large cell lymphomas were commonly designated "histiocytic" in accordance with the Rappaport classification.1 Rappaport also recognized a different form of histiocytic neoplasm, characterized by widespread intrasinusoidal dissemination, for which the term malignant histiocytosis was favored.1 Owing to advances in immunohistochemistry and molecular biology, it is now generally accepted that the majority of these neoplasms were B- or T-cell, nonHodgkin lymphomas.2 True histiocytic lymphomas, which are positive for histiocytic markers and negative for B- and T-cell markers, as well as being negative for Ki-I, have been reported.3-9 True histiocytic lymphoma (histiocytic sarcoma) remains very rare. We recently encountered histiocytic sarcoma associated with idiopathic myelofibrosis. In this article, the microscopic and immunohistochemical features of the tumor are described in detail, and its differential diagnoses are discussed.

REPORT OF A CASE

A 39-year-old man presented with general weakness and asthenia, which he had been experiencing for 3 months. The patient had pancytopenia and hepatosplenomegaly. His leukocyte count was 2.3 × 10^sup 3^/µL erythrocyte count, 1.15 × 10^sup 6^/×L; hemoglobin, 4.3 g/dL; hematocrit, 13.6%; and platelet count, 38 × 10^sup 3^/×L. Erythroblasts and myelocytes were observed in peripheral blood smears. A bone marrow biopsy and clot examinations indicated idiopathic myelofibrosis. A cytogenetic analysis of the bone marrow showed a normal karyotype (46,XY). Postembolization splenectomy and a needle biopsy of the liver were performed. His histogram had been within normal ranges for 3 months. The patient, however, showed anemia, thrombocytopenia (88 × 10^sup 3^/ ×L), and generalized lymphadenopathy. After the diagnosis on a supraclavicular lymph node biopsy, the patient was treated with multiagent chemotherapy. The patient died of disease 25 months after diagnosis. An autopsy was performed.

MATERIALS AND METHODS

Sections of formalin-fixed, paraffin-embedded tissues were stained with hematoxylin-eosin, Giemsa, and periodic acid-Schiff.

Immunohistochemical studies were performed using formalinfixed, paraffin-embedded sections. Antibodies used in this study are summarized in the Table. The avidin-biotin peroxidase complex technique and the peroxidase-antiperoxidase technique were used. Appropriate positive and negative control experiments were also performed.

Amplification of the T-cell receptor γ-chain gene was carried out using the methods of McCarthy et al10 at SRL Laboratory, Tokyo, Japan. DNA was extracted from fresh frozen tissue of the supraclavicular lymph node, purified, and digested with appropriate restriction enzymes (New England BioLabs, Beverly, Mass).

PATHOLOGIC FINDINGS

In the bone marrow biopsy and clots, the marrow was hypercellular and showed loss of fat spaces and hyperplasia of erythroblasts and megakaryocytes. The megakaryocytes showed a spectrum of nuclear morphology with both polylobated and small hypolobated forms and abundant amphophilic cytoplasm (Figure 1). Some of the megakaryocytes showed erythrophagocytosis. There was minimal fibrosis.

The spleen weighed 650 g and measured 22 × 12 × 8 cm. The infarction, mainly due to the embolization, was so prominent that only about 10% of the parenchyma was preserved. There were no nodular lesions. Most of the tissue available for the histologic examination was composed of the red pulp. Megakaryocytes with hyperchromatic polylobated nuclei and abundant clear to amphophilic cytoplasm with small vacuoles and erythrophagocytosis were scattered throughout the spleen (mainly in the sinuses) and were seen focally in small aggregates (Figure 2). The megakaryocytes revealed anisocytosis, abnormal nuclearcytoplasmic ratios, and plump lobulation of some nuclei. Myeloblasts with medium-sized nuclei with vesicular chromatin and small clusters of erythroblasts were also found, indicating extramedullary hematopoiesis. Reactive histiocytes with prominent erythrophagocytosis were abundant in the sinuses. The needle biopsy of the liver also showed extramedullary hematopoiesis. Immunohistochemically, some megakaryocytes were positive for CD45, CD68, and factor VIII-related antigen. They were negative for CD1a, CD45RO, CD20, anaplastic lymphoma kinase, lysozyme, α^sub 1^-antitrypsin, S100 protein, and epithelial membrane antigen (EMA). Some reactive histiocytes with erythrophagocytosis were positive for S100 protein.

The lymph node showed pronounced distension of the sinuses in the medulla and periphery, which was caused by the accumulation of large tumor cells (Figure 3). The paracortical areas were also invaded by the tumor cells. The tumor cells had abundant clear or eosinophilic cytoplasm. The nuclei were of various sizes and shapes (round, oval, kidney-shaped, or very irregular) with moderately condensed chromatin and medium-sized to large nucleoli (Figure 4). Some tumor cells displayed erythrophagocytosis and small vacuoles. The lymphoid tissues of the cortex and medulla were atrophie. The results of the immunohistochemical study are summarized in the Table. The tumor cells were positive for CD45, CD45RO, CD68 (Figure 5), α^sub 1^-antitrypsin, S100 protein, and EMA (cytoplasmic staining). They were negative for CD1a, CD3, CD4, CD15, CD20, CD21, CD30, CD34, CD35, CD43, CD56, CD79, lysozyme, myeloperoxidase, factor VIII-related antigen, anaplastic lymphoma kinase, CAM 5.2, and HMB-45.

No rearrangement of the T-cell receptor γ-chain gene was identified.

Autopsy revealed that the tumor involved the systemic lymph nodes, liver, bone marrow, and lung. The histology was identical to that of the lymph node biopsy. In the bone marrow and liver, the tumor displayed multinodular lesions. We identified no other primary tumor.

COMMENT

The malignant tumor in this case corresponded to the definition of histiocytic sarcoma according to the World Health Organization classification,11 that is, a malignant proliferation of cells showing morphologic and immunophenotypic features similar to those of mature tissue histiocytes. In the current case, the lymph node was characterized by the proliferation of large tumor cells in the sinuses and paracortical area. The tumor cells had abundant clear or eosinophilic cytoplasm and nuclei in various sizes and shapes, and showed erythrophagocytosis. There was expression of 2 histiocytic markers, CD68 and α^sub 1^-antitrypsin without accessory/dendritic markers. It lacked B-cell, T-cell, and myeloid markers, as well as CD30 expression. Since the large tumor cells proliferated mainly in the sinuses and expressed CD45, CD45RO, and EMA, anaplastic large cell lymphoma was initially considered. However, the expression of histiocytic markers was unusual for anaplastic large cell lymphoma, and CD45, CD45RO, S100 protein, and EMA were expressed in previously reported histiocytic sarcomas.

The differential diagnosis includes Langerhans cell sarcoma, diffuse large B-cell lymphoma, peripheral T-cell lymphoma, anaplastic large cell lymphoma of T/null-cell type, metastatic undifferentiated carcinoma, and melanoma. Tumor cells in Langerhans cell sarcoma have nuclei that are grooved, indented, folded, or lobulated, with fine chromatin and a thin nuclear membrane, features that are absent in histiocytic sarcoma.12 Immunohistochemically, Langerhans cell sarcoma expresses CD1a, which is negative in histiocytic sarcoma. Pileri et al2 recently recommended immunostaining of CD68, lysozyme, CD1a, S100 protein, CD21, and CD35 for the differential diagnosis of histiocytic/dendritic cell neoplasms. A battery of immunostains can make it possible to determine whether the cells involved are of B- or T-cell origin. Because of the wide diversity in morphology of histiocytic sarcoma and the unreliability of histology in identifying them, positivity with histiocyte-associated antibodies along with the lack of B- or T-antigen expression was the first element of suspicion that the tumors were histiocytic in type and nature. Metastatic carcinoma and melanoma can be ruled out by the absence of cytokeratins and HMB-45 expression. Evaluation of a battery of antibodies in the context of morphology is essential in the workup of these neoplasms.

No rearrangement of the T-cell receptor γ-chain gene was identified in this tumor. The tumor was not analyzed by immunoglobulin gene rearrangement studies. The contribution of gene rearrangement studies to the diagnosis of histiocytic sarcoma still remains unclear. One of 3 cases by Kamel et al5 exhibited rearrangement of antigen receptor genes. In the series reported by Hanson et al,7 3 of 5 cases demonstrated T-cell receptor and/or immunoglobulin gene rearrangements. While T-cell receptor and immunoglobulin gene rearrangements usually demonstrate rumor clonality of T-cell and B-cell neoplasms, respectively, it has been well demonstrated that gene rearrangement is not always lineage specific.13 Thus, while gene rearrangement studies may support clonality, they do not always indicate lineage and should be interpreted in the context of morphologic and immunophenotypic data.

The intriguing point of this case is the preexistence of idiopathic myelofibrosis. The patient had weakness, anemia, and pancytopenia. The bone marrow examination showed hypercellularity with atypically increased numbers of megakaryocytes, which clustered together. Erythroid hyperplasia and minimal fibrosis were also observed and were believed to represent the cellular phase of idiopathic myelofibrosis.14 The patient presented with hepatosplenomegaly. In the massively enlarged spleen, extramedullary hemopoiesis was prominent, with clusters of atypical megakaryocytes with vacuoles and erythrophagocytosis. The liver biopsy showed extramedullary hemopoiesis. There was no evidence of invasion of histiocytic sarcoma. Association with other hematologic disorders, including acute leukemia, lymphoma, or myelodysplasia, has been observed in histiocytic sarcomas.2,9,14 The reported incidence of acute leukemia in patients with idiopathic myelofibrosis ranged from 5% to 30%.14 To the best of our knowledge, there has been no report associated with idiopathic myelofibrosis. Whether the association between histiocytic sarcoma and idiopathic myelofibrosis is a simple coincidence or indicates some etiopathogenetic relationship between these 2 unrelated disorders is unknown. Cytokines implicated in the development of myelofibrosis include transforming growth factor β, platelet-derived growth factor, epidermal growth factor, basic fibroblast growth factor, and calmodulin.15 Proliferating megakaryocytes in myelofibrosis might secrete some cytokines that promote malignant transformation of histiocytes or pluripotential stem cells. Analyses of more cases are required. Histiocytic sarcomas run a very aggressive clinical course in most patients. In a study by Kamel et al,5 6 of 11 patients died of disease within 0.5 to 36 months, and 58% of patients died of disease in the study by Pileri et al.2 Although histiocytic sarcomas are very rare, their recognition may be important for clinical and prognostic reasons.

The authors thank the following members of our laboratory for their technical assistance: Yokihiro Takeuchi, Yukiko Shiomori, Takashi Honma, Kuniko Kobayashi, Kazuya Sakurai, and Toshihisa Toda.

References

1. Rappaport H. Tumours of the Hematopoietic System. Washington, DC: Armed Forces Institute of Pathology; 1966:9. Atlas of Tumor Pathology, section 3, fascicle 8.

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

3. Wilson MS, Weiss LM, Gatter KC, Mason DY, Dorfman FR, Warnke RA. Malignant histiocytosis: a reassessment of cases previously reported in 1975 based on paraffin section immunophenotyping studies. Cancer. 1990;66:530-536.

4. Copie-Bergman CC, Wotherspoon AC, Norton AJ, Diss TC, Issacson PG. True histiocytic lymphoma: a morphologic, immunohistochemical, and molecular genetic study of 13 cases. Am J Surg Pathol. 1998;22:1386-1392.

5. Kamel O, Gocke CD, Kell DL, Cleary ML, Warnke R. True histiocytic lymphoma: a study of 12 cases based on current definition. Leukemia Lymphoma. 1995;18:81-86.

6. Ralfkiaer E, Delsol G, O'Connor NT, et al. Malignant lymphomas of true histiocytic origin: a clinical, histological, immunophenotypic and molecular genetic study. J Pathol. 1990; 160:9-17.

7. Hanson CA, Jaszcz W, Kersey JH, et al. True histiocytic lymphoma: histopathology, immunophenotypic and geneotypic analysis. Br J Haematol. 1989;73: 187-198.

8. Milchgrub S, Kamel OW, Wiley E, Vuitch F, Cleary ML, Warnke RA. Malignant histiocytic neoplasms of the small intestine. Am J Surg Pathol. 1992;16:11-20.

9. Alexndrakis MG, Chatzivasili A, Stefanaki K, Eliopoulos CG. Co-existence of cutaneous true histiocytic lymphoma with refractory anaemia: report of a case. Haematologia. 2000;30:61-67.

10. McCarthy KP, Sloane JP, Kabarowski JHS, Matutes E, Wiedemann LM. A simplified method of detection of clonal rearrangements of the T-cell receptor-γchain gene. Diagn Mol Pathol. 1992;1:1 73-179.

11. Weiss L, Grogan TM, Muller-Hermelink HL, et al. Histiocytic sarcoma. In: laffe ES, Harris NL, Stein H, Vardiman JW, eds. Pathology and Genetics ot Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001:278-279. World Health Organization Classification of Tumours', vol 3.

12. Jaffe ES. Histiocytic and dendritic cell neoplasms. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001:273-277. World Health Organization Classification of Tumours; vol 3.

13. Ohyashiki JH, Ohyashiki K, Yoyama K, et al. T-cell receptor gene rearrangement and its expression in human myeloid leukemia cell lines. Cancer Genet Cytogenet. 1989:37:193-200.

14. Thielc J, Pierre R, lmbert M, Vardiman JW, Brunning RD, Flandrin G. Chronic idiopathic myelofibrosis. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001:35-38. World Health Organization Classification of Tumours; vol 3.

15. Romstein ED, Kurtin PJ, Nascimento AG. Sclerosing extramedullary hematopoietic tumor in chronic myeloproliferative disorders. Am J Surg Pathol. 2000;24:51-55.

Masaharu Fukunaga, MD; Hiroyuki Kato, MD

Accepted for publication May 25, 2004.

From the Department of Pathology, the Jikei University School of Medicine, Tokyo, Japan.

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

Reprints: Masaharu Fukunaga, MD, Department of Pathology, Jikei Daisan Hospital, 4-11-1, Izumihoncho, Komaeshi, Tokyo, 201-8601, Japan (e-mail: maasafu@jikei.ac.jp).

Copyright College of American Pathologists Oct 2004
Provided by ProQuest Information and Learning Company. All rights Reserved