Hairy Cell Leukemia

HEMATOPATHOLOGY Original Article

Hairy Cell Leukemia Diagnosis of Bone Marrow Involvement in ParaffinEmbedded Sections with Monoclonal Antibody DBA.44 HELENE HOUNIEU, M.D.,1 SHASHIKANT M. CHITTAL, F.R.C.P.(C),' TALAL AL SAATI, M.D., PH.D., 1 ANTOINE DE MASCAREL, M.D.,2 ELENA SABATTINI, M.D.,5 STEFANO PILERI, M.D.,4 BRUNANGELO FALINI, M.D.,5 ELISABETH RALFKIAER, M.D.,6 AGNES LE TOURNEAU, M.D.,3 JANICK SELVES, M.D.,1 JEAN-JACQUES VOIGT, M.D.,1 GUY LAURENT, M.D.,1 JACQUES DIEBOLD, M.D.,3 AND GEORGES DELSOL, M.D.1

cell leukemia from the more common B-cell chronic lymphocytic leukemia and bone marrow infiltrates of typical lymph nodebased lymphomas by immunomorphologic criteria. DBA.44 was valuable to (1) confirm the diagnosis of hairy cell leukemia, (2) estimate the bone marrow density of hairy cell leukemia before and after treatment, and (3) make the diagnosis of hairy cell leukemia in ambiguous cases, which are all properties that indicate its usefulness in the practice of diagnostic hematopathology. (Key words: Hairy cell leukemia; Monoclonal antibody DBA.44; Immunohistochemistry; Paraffin sections) Am J Clin Pathol 1992; 98:26-33

In a previous report on the generation of new anti-B-cell monoclonal antibodies reactive with routinely fixed and paraffin-embedded tissues, we noted a significant reactivity of the antibody DBA.44 with B lymphocytes of follicular mantle zones in reactive lymph nodes. DBA.44 was found to recognize some low- and high-grade B-cell lymphomas.' Within the group of low-grade lymphomas, the antibody was found to be discriminating for hairy cell leukemia (HCL) in preliminary studies. The staining properties of the antibody outlined the cytoplasmic projections of hairy cells. The pattern of reactivity appeared promising for immunomorphologic diagnosis of the tissue infiltrates of HCL, especially in routine bone marrow biopsy speci-

mens. Therefore we conducted a multicenter retrospective study, placing emphasis on the tissue infiltrates of HCL in bone marrow biopsy specimens. MATERIALS AND METHODS Monoclonal Antibody

DBA.44

DBA.44 is one of four anti-B monoclonal antibodies, generated using a new centroblastic lymphoma cell line (Deau cell line).1 Preliminary estimates of the reactivity of this antibody (with normal lymphoid tissues, tumors of B-cell and T-cell origin, and nonlymphoid normal tissues and tumors) that led to the current investigation have been reported.' The isotype of DBA.44 is of the immuM heavy chain subclass, but the molecular From2 the Anatomical Pathology Departments of 'CHU-Purpan, noglobulin Tou3 weight of the antigen recognized is not yet known.' This louse, CHU-St. Andre, Bordeaux, and Holel-Dieu, Paris. France; the Hematology Departments of the Universities of ABologna and iPerugia, antibody is commercially available from Immunotech Italy; and the Departments of Pathology, Rigshospilalet and Genlofte (Marseille, France) and Dakopatts (Copenhagen, DenHospitals of the University of Copenhagen, Copenhagen, Denmark. mark). Supported by grants from ARC, France, and the Harboe Foundation, Denmark. Tissue Sources and Types Received July 24, 1991; received revised manuscript and accepted for publication October 1, 1991. Six institutions participated in the study. The tissues Address reprint requests to Dr. Delsol: Laboratoire Anatomie Pathinvolved by HCL were retrieved from the files of 166 paologique, CHU-Purpan, Place du Dr. Baylac, 31059 Toulouse, France. 26

Downloaded from http://ajcp.oxfordjournals.org/ by guest on September 26, 2016

The new monoclonal antibody DBA.44 recognizes an unknown fixation-resistant B-cell differentiation antigen expressed by mantle zone lymphocytes, reactive immunoblasts, monocytoid B cells, and a small proportion of high- and low-grade lymphomas. Among node-based lymphomas, the strongest membrane staining was observed in centroblastic, immunoblastic, and monocytoid B-cell lymphomas. In studying bone marrow biopsy specimens from 166 patients with hairy cell leukemia, strong positive staining of surface membrane 'hairy' features of leukemic cells was observed in routinely fixed and decalcified bone marrow biopsy specimens of nearly all cases. The antibody distinguished hairy

HOUNIEU ET AL.

27

Immunohistochemkal Diagnosis of Hairy Cell Leukemia

Comparison of the reactivity of DBA.44 in tissues other than HCL was performed at one institution (CHU-Purpan, Toulouse, France). Bone marrow biopsy specimens from 85 patients with reactive states and myeloproliferative disorders were immunostained with DBA.44. Routinely fixed and paraffin-embedded lymph node specimens from 354 patients with the diagnosis of malignant lymphoma were retrieved from the files for testing. The detailed immunophenotypes of these cases were known because of previous studies using cryopreserved tissue.6 In the cases of malignant lymphoma, 97 bone marrow biopsy specimens that showed involvement by lymphoma formed the critical source for comparison of the reactivity of DBA.44. Paraffin blocks of 32 reactive lymph nodes and 10 splenectomy specimens (spleens from patients with Hypersplenism, hemolytic anemia, or thrombocytopenic purpura) also were used to study the reactivity of this antibody. Fixatives and Staining

Methods

Tissues were fixed by different fixatives, depending on the custom of the individual laboratory. The fixatives were ethanol-based Bouin's fluid (Duboscq-Brasil) (n = 113), Bouin's (n = 65), B5 (n = 49), and 10% buffered formalin (n = 13). Bone marrow biopsy specimens had been subjected to different decalcification methods (2% nitric acid, 10% EDTA, or commercially available decalcification solution; RDO, Eurobio, Paris, France). The study was thus uncontrolled for fixation and decalcification. In three of six institutions, immunostaining was performed by a three-step immunoperoxidase method with a previous trypsinization step. The method was similar to that reported previously.8 In three institutions, im-

munostaining was performed by the alkaline phosphataseanti-alkaline phosphatase method. 9 In most cases, bone marrow biopsy specimens were immunostained with one or more other antibodies. Antileukocyte common antigen/CD45 (Dakopatts, Copenhagen, Denmark) and anti-B-cell antibody MB2 (Biolyon, Lyon, France) were the most frequently used antibodies in three institutions. LNl/CDw75 (Biotest, Buc, France) was used in a few cases. Other anti-B-cell antibodies were DNA.7, DBB.42, and DND.53 (Immunotech). UCHL1/ CD45RO (Dakopatts) was the most frequently used antiT-cell antibody when there was a question of T-cell lineage of the infiltrating lymphocytes. RESULTS Reactive Conditions in Lymph Nodes, and Bone Marrow

Spleens,

DBA.44 reacted primarily with mantle zone lymphocytes in variable numbers (50% to 100%) and scattered immunoblasts. Germinal center cells were negative. The staining pattern was membrane associated, but a paranuclear dot-like reaction usually was observed in immunoblasts. Weak positive staining was observed in monocytoid B cells, particularly in toxoplasma lymphadenitis and human immunodeficiency virus-associated lymphadenitis. Sometimes few histiocytes exhibited tiny cytoplasmic granules that were clearly positive for DBA.44. Antigen shedding by surrounding B lymphocytes and subsequent phagocytosis by macrophages was considered as a possible explanation for this phenomenon. Follicular dendritic cells and interdigitating cells were DBA.44-negative. Staining of some follicular center cells was observed in two lymph nodes. Because of the sharp definition of the mantle zones, striking revisions of the lymphocytic mantles could be observed easily in some disorders. The mantle zones surrounding hyperplastic germinal centers were markedly attenuated or absent in reactive lymph nodes of human immunodeficiency virusseropositive patients. Conversely, expanded B-lymphocyte populations of the mantle zones were seen in the nodules of progressively transformed germinal centers. Similar reactivity of mantle zone cells, with or without weak staining of lymphocytes in the marginal zones, was noted in the white pulp of the spleen. Some cells with hairy surface features were found at the periphery of the marginal zones. Cells sharing similar morphologic features and admixed with a few normal-appearing small lymphocytes also were noted in the red pulp. In all bone marrow biopsy specimens with reactive hyperplasia, rare DBA.44-positive small lymphocytes were present. The staining pattern was membrane associated,

Vol. 98 • No. I


tients, in which the biopsies or splenectomies were performed with patient consent for diagnosis or during treatment. The files dated back to 1972.2 There were 209 marrow trephine biopsy samples, 15 liver biopsy samples (either needle or wedge), and 17 splenectomy specimens. The diagnosis of HCL was based on the clinical status and the presence of mononuclear cells in peripheral blood and/or bone marrow smears, which showed characteristic morphologic and enzyme cytochemical features.3"5 The median age of the patients was 57 years (range, 28 to 84 years). The male to female ratio was 5:1. In 50 treated patients (43 with alpha-interferon, 2 with 2'-deoxycoformycin, and 5 with other forms of chemotherapy), 75 bone marrow biopsy specimens were obtained between 3 and 60 months during and/or after completion of treatment. In 15 tissue sections, the diagnosis of HCL was corroborated with cryostat sections using a panel of monoclonal antibodies.6,7

28

HEMATOPATHOLOGY Original Article

but no cytoplasmic projections were seen. Many DBA.44positive small lymphocytes were present in three cases. The nature of these cells could not be established with certainty. Neither red nor white cell precursors or megakaryocytes were positive. Bone marrow macrophages also were negative. Tissues of Malignant Lymphomas Other than HCL

TABLE 1. REACTIVITY OF DBA.44 WITH MALIGNANT LYMPHOMAS OF B- OR T-CELL ORIGIN IN PARAFFINEMBEDDED SECTIONS (KIEL CLASSIFICATION)

Tumors Hairy cell leukemia

B cell, low grade Chronic lymphocytic leukemia Lymphoplasmacytic Centrocytic Centroblastic-centrocytic Plasmacytic Maltoma Monocytoid Mantle zone Unclassified Total B cell, high grade Centroblastic Immunoblastic Lymphoblastic Anaplastic (Ki-1 lymphoma) Unclassified Total Tcell Chronic lymphocytic leukemia Mycosis fungoides Lennert's lymphoma Angioimmunoblastic-like Pleomorphic Immunoblastic Lymphoblastic Anaplastic (Ki-1 lymphoma) Others Total Unknown lineage Total

Tissues Other than Bone Marrow N°+/N° Tested

Bone Marrows N°+/N° Tested

32/32 (Liver & spleen) (100%)

206/209

1/20 4/16 9/29 8/43 0/2 6/16 3/3 0/2 1/7 32/138(23%)

1/22 5/17 1/9 0/5 0/5

29/58 4/12 2/10 1/4 5/21 41/105(39%) 0/2 1/14 0/7 0/3 2/18 0/3 0/6 0/14 0/13 3/80 (3.8%) 5/31 (16%) 354

Blanks in the bone marrow column indicate nonavailability of tissue.

(98.5%)

6/27 13/85(15%) 1/2 0/2 1/5 2/9 (22%)

0/2 0/1 0/3 (0%) 97

Hairy Cell Leukemia Bone marrow specimens of untreated patients. DBA.44 reacted with 'hairy' cells in all bone marrow biopsy samples (206 of 209 cases, or 98.5%) (Table 1). The negative reaction seen in three cases could have resulted from artifacts caused by fixation and decalcification because isolated small lymphocytes (which serve as positive controls) were completely absent in these cases. Comparison of infiltrate as assessed with standard microscopic examination and after immunostaining with DBA.44 could be performed in 127 of 166 untreated patients. In 108 of 127 cases, the diagnosis of HCL could be made without immunostaining because of massive (n = 55) or moderate (n = 53) infiltration by typical cells, which was corroborated by clinical and hematologic features. In 19 of 166 cases, the diagnosis was difficult because the infiltrate was minimal (n = 6), atypical (n = 8), or presumably absent (n = 8) on conventional examination. In all cases (except one), all cells showed strong cellular staining intensities despite variations of the fixation and decalcification methods of the participating laboratories (Figs. 1 and 2). These cases of HCL also were positive for one or more of the following antibodies: LCA/CD45, MB2, DBB.42, or DND.53. The characteristic cytoplasmic projections, however, were highlighted and visible only with DBA.44 (Figs. 2 and 3). In 4 of 52 cases, the diagnosis of a moderate HCL infiltrate on conventional examination was revised to a massive infiltrate after DBA.44 immunostaining. Similarly, the infiltrate was upgraded to moderate in all six cases in which it was thought to be minimal on conventional examination. In eight cases, in which the infiltrating cells were not cytomorphologically characteristic, positively stained cells with hairy features indicated a diagnosis of HCL. In five cases, in which HCL was not

A.J.C.P. -July 1992


The reactivity of DBA.44 with various malignant lymphomas is shown in Table 1. The antibody reacted with neoplastic cells in a proportion of both low- and highgrade B-cell lymphomas. No significant staining differences were noted among bone marrow and other tissues. Among low-grade B-cell lymphomas that usually involve the bone marrow, positive staining was found in only 1 of 22 bone marrows of chronic lymphocytic leukemia.

Tumor cells in 1 of 9 bone marrow specimens involved by centrocytic lymphoma and 5 of 17 bone marrow specimens of lymphoplasmacytic lymphoma were positive for DBA.44. Most of these bone marrow samples involving B-cell lymphomas tested uniformly positive for LCA/ CD45, MB2, DBB.42, and DND.53 antibodies. The neoplastic cells in follicular lymphomas were negative for DBA.44. The three cases of monocytoid B-cell lymphomas were strongly positive for DBA.44. Reactivity with highgrade lymphoma was noted primarily with centroblastic and immunoblastic B-cell lymphomas (approximately 50% of cases). It must be noted that the Deau cell line used as immunogen to produce the antibody was a centroblastic lymphoma. The antibody was nonreactive in most T-cell lymphomas (Table 1). Some cases of pleomorphic lymphomas and mycosis fungoides were positive, but the staining was restricted to the paranuclear area in only a few large cells.

HOUNIEU ET AL. lmmunohistochemical Diagnosis of Hairy Cell Leukemia

29

>•>

suspected by clinical or hematologic criteria, the bone marrow was found to contain DBA.44-positive hairy cells in significant numbers. The diagnosis of HCL was later confirmed in these five cases by the demonstration of hairy cells in peripheral blood and/or bone marrow aspirates. Some other observations are noteworthy. First, immunostaining with DBA.44 disclosed topographic variations in HCL infiltrates in bone marrows; some areas were massively involved, whereas other areas showed minimal to moderate involvement (Fig. 2). Second, in two cases with an intense fibroblastic reaction of the marrow, hairy cells were clearly detected among fibroblasts (Fig. 4). Third, in two other cases, non-neoplastic nodular lymphocytic infiltrates were unstained byJDBA.44. Last, the detection of hairy cells in the vascular channels of most cases was important, as one case was diagnosed on the basis of almost exclusive intravascular involvement.

Bone marrow specimens of treated patients. Seventyfive bone marrow biopsy samples were available in 50 treated patients. Hairy cell leukemia infiltrates decreased from massive to moderate after treatment, but minimal infiltrates often were undetectable by conventional examination (Table 2 and Fig. 3). In 41 bone marrow biopsy specimens that were considered as uninvolved or equivocally involved, immunostaining with DBA.44 resulted in the revised diagnosis of minimal to moderate infiltrate (Table 2). Hyperplastic bone marrow, with erythroblastic islands rich in immature cells, was a common finding after alpha-interferon treatment. No significant increase of reticulin fibers was noted in these cases, but DBA.44 stained a significant number of typical hairy cells obscured by hematopoietic cells. In 34 treated bone marrow samples, conventional examination showed persistent infiltration ranging from

Vol. 98 • No. 1


FIG. 1. Bone marrow trephine biopsy specimen in a case of hairy cell leukemia in which the infiltrate was massive, a.—Usual cytomorphology of hairy cells showing round or slightly indented nuclei (hematoxylin and eosin, X500). b.—Strong membrane staining of hairy cells with DBA.44. Negative cells are residual hematopoietic elements, which were not obvious in conventional examination (immunoperoxidase, X500).

30

HEMATOPATHOLOGY f Article were absent. Cross-reactivity of a few immature erythroblasts was found in 2 of 32 bone marrow biopsy specimens of myelodysplasia. DISCUSSION

FIG. 2. Topographic variations in the density of DBA.44-positive cells in a case of hairy cell leukemia. Only a few positive cells were present in some areas (arrows) (X200). Inset: Strong membrane and cytoplasmic staining of hairy cells in the massively involved area (immunoperoxidase, X800).

massive (n = 7) or moderate (n = 13) to minimal (n = 14). In two and three cases, respectively, assessment of the infiltrate was changed from moderate to massive and from minimal to moderate after DBA.44 staining (Table 2). Bone Marrow of Myelodysplastic and Myeloproliferative Disorders In other hematopoietic disorders (n = 85), rare DBA.44positive small lymphocytes were found in many bone marrow samples, a finding that was similar in bone marrow biopsy specimens in reactive states. The staining was membrane associated, but the cytoplasmic projections

The 'unique' antigenic profile of HCL can only be demonstrated in cryostat sections, and difficulty obtaining quality frozen sections of the bone marrow is an additional limiting factor. Most previous studies were performed on bone marrow aspirate,7,23,25,27,28 splenectomy,14,20"22,26,30 or liver biopsy specimens.29 Trephine bone marrow biopsy is often required for diagnosis and follow-up of HCL because of the high rate of "dry tap" caused by fibrosis.19,36 Thus the practical value of antibodies reactive on routinely fixed, decalcified, and paraffin-embedded tissues needs no emphasis. Tissue infiltrates of HCL have been tested by CD45 (leukocyte common antigen)37 and several monoclonal antibodies reactive with B-cell-associated antigens on routinely prepared sections in a few cases.38"42 Although useful in identifying a B-cell lymphoid infiltrate in tissues, antibodies such as MB2,40,41 4KB5/CD45R,39,41 Ki-B3,42 LN2/CD74,40,41 DND.53/CD74-like,' and DBB.421 do not show the restrictive pattern of DBA.44. Furthermore, the fixation-resistant antigens recognized by the operationally useful anti-B-cell antibodies, such as L26 (CD20),38 often are weakened during decalcification of the bone. Several findings suggest that DBA.44 antibody recognizes a new fixation-resistant differentiation antigen ex-

A.J.C.P. • July 1992


The clinicopathologic and biologic features of HCL have been well recorded.2"510"15 Hairy cell leukemia most frequently involves bone marrow and peripheral blood, with splenomegaly as a common feature as a result of sequestration of neoplastic cells. There is general agreement that HCL is a clonal B-cell disorder often characterized by immunoglobulin heavy chains and a single immunoglobulin light chain.716"18 The existence of true T-cell HCL is controversial.16,19 The phenotypic characteristics of HCL on cryopreserved tissue reside in the simultaneous expression of several B-cell antigens (CD 19+, CD20+, CD22+), interleukin-2 receptor (CD25), CD1 lc molecules of the LFA subfamily, and the nonexpression of CD5, thus constituting a distinctive immunophenotypic profile.717,20"29 More recently, reactivity with one of four monoclonal antibodies (B-ly7, HML1, Ber-Act8, and LF61)30"34 that identify the same HCL-associated trimeric molecule (150, 125, 105 kD), in addition to the CD22 antibody, have been proposed as a highly specific reagent combination for monitoring residual disease in HCL patients, treated with interferon or deoxycoformycin.35 However, these antibodies require fresh tissues.

H O U N I E U ET

AL.

31

is of Hairy Cell Leukemia

FIG. 4 (right). Trephine bone marrow biopsy specimen of a treated case of hairy cell leukemia, with persistent intense fibrosis, immunostained with DBA.44. Positive cells (arrows) were impossible to identify in routine sections (immunoperoxidase, X400).

TABLE 2. HCL—COMPARISON OF STANDARD MICROSCOPY OF BONE MARROW INFILTRATES WITH DBA.44 IMMUNOMORPHOLOGY IN TREATED CASES (ALPHA-INTERFERON)*

pressed by a subset of B lymphocytes. The difference from other anti-B antibodies became apparent by the staining pattern in lymph nodes and lymphomas. The characteristic pattern of reactivity with mantle zone lymphocytes, monocytoid B cells, and monocytoid B-cell lymphomas may have a specific bearing on the proposed theories of the origin of hairy cells. 19,2M3~45 It is possible that the two negative mantle zone lymphomas arose from DBA.44negative mantle zone lymphocytes, but so were most centrocyte lymphomas (derived from mantle zone lymphocytes) (Table 1). Within malignant lymphomas, DBA.44 reacted with all cases of HCL, regardless of the fixative used and the manner of decalcification of the bone marrow specimens. The antibody did not react with either Tlymphocytic infiltrate or with normal hematopoietic elements of the bone marrow, unlike many other B-cell-

Infiltrate Assessed by Standard Microscopy

DBA.44 N°+/N° Tested

Infiltrate Assessed After DBA.44 Staining

Persistent—massive Persistent—moderate

7/7 13/13

Persistent—minimal

14/14

Doubtful or none

41/41

Unchanged Revised to massive in 2/13, Revised to moderate in 3/14, Revised to persistent— minimal or persistent—moderate

Total

75/75

* Of Ihe 50 treated patients (some with several biopsies during or after treatment), 43 received alpha-interferon, 2 received 2'-deoxycoformycin, and 5 were treated with other forms of chemotherapy.

Vol. 98 • No. I


FIG. 3 (left). A case of hairy cell leukemia treated with alpha-interferon, and considered negative for the presence of hairy cells by conventional examination. Significant numbers of DBA.44-positive cells with cytoplasmic projections (arrows) were present in hyperplastic marrow, with a positive cell in a venous channel. Note that hematopoietic cells were unreactive with this antibody (immunoperoxidase, X800).

32

HEMATOPATHOLOGY Original Article for HCL. Further experience will better define the limits of its utilty relative to expectations, as is usually the case with most monoclonal antibodies. Acknowledgments. The authors thank the many technologists from the participating laboratories. Dr. Chittal thanks the French Ministry of Education for support during leave from Memorials University Medical School, St. John's, Newfoundland, Canada.

Hairy cell leukemia may exhibit a variable subclinical course and the diagnosis may not be considered, especially in the absence of splenomegaly (up to 20%).19 We found such a setting ideal for suggesting the diagnosis of HCL because the antibody detected individual cells scattered in hyperplastic bone marrow, often revealing their characteristic cytoplasmic projections. The antibody also recognized hairy cells in cases in which the presence of an infiltrate by standard microscopic examination was considered equivocal, particularly after therapy. Complete elimination of neoplastic cells from the bone marrow may not be possible, but the finding of a persistent infiltrate may influence present or future therapeutic treatments.47"51 In conclusion, antibody DBA.44 appears to be a promising marker for the routine immunomorphologic diagnosis of the tissue infiltrates of HCL. Like other antibodies reactive with paraffin sections, DBA.44 is not "specific" A.J.C.P.-July

REFERENCES 1. Al Saati T, Caspar S, Brousset P, et al. Production of anti-B monoclonal antibodies (DBB. 42, DBA.44, DNA.7, and DND.53) reactive on paraffin-embedded tissues with a new B-lymphoma cell line grafted into athymic nude mice. Blood 1989;74:2476-2485. 2. Fabre J, Delsol G, Fedou R, Pris J. Etude des lesions hepatiques de la "hairy cell leukemia". Virchows Arch Abt B Zellpath 1974; 15: 139-156. 3. Bouroncle BA, Wiseman BK, Doan CA. Leukemic reticuloendotheliosis. Blood 1958;13:609-630. 4. Naiem F, Smith GS. Leukemic reticuloendotheliosis. Cancer 1974;34:1813-1821. 5. Golomb HE, Catovsky D, Golde DW. Hairy cell leukemia. A clinical review based on 71 cases. Ann Intern Med 1978;89:677-683. 6. Al Saati T, Laurent G, Caveriviere P, et al. Reactivity of Leu 1 and T101 monoclonal antibodies with B cell lymphomas (correlations with other immunological markers). Clin Exp Immunol 1984; 58: 631-638. 7. Falini B, Pulford K, Erber WN, et al. Use of a panel of monoclonal antibodies for the diagnosis of hairy cell leukemia. An immunohistochemical study of 36 cases. Histopathology 1986; 10:671687. 8. Delsol G, Al Saati T, Gatter ICC, et al. Coexpression of epithelial membrane antigen (EMA), Ki-1 and interleukin-2 receptor by anaplastic large cell lymphomas. Diagnostic value in so-called malignant histiocytosis. Am J Pathol 1988;130:59-70. 9. Cordell JL, Falini B, Erber WN, et al. Immunoenzymatic labeling of monoclonal antibodies complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP complexes). J Histochem Cytochem 1984;32:219-229. 10. Diebold J, Chomette G, Reynes M, Tricot G. La rate dans la leucemie a tricholeucocytes. Virchows Arch A (Pathol Anat and Histol) 1977;372:325-336. 11. Nanba K, Soban E, Bowling M, Berard C. Splenic pseudosinuses and hepatic angiomatous lesions. Am J Clin Pathol 1977; 67:415426. 12. Bartl R, Frisch B, Hill W, et al. Bone marrow histology in hairy cell leukemia. Identification of subtypes and their prognostic significance. Am J Clin Pathol 1983;79:531-545. 13. Yam L, Janckila A, Chan C, Li C. Hepatic involvement in hairy cell leukemia. Cancer 1983;51:1497-1504. 14. Burke J, Rappaport H. The diagnosis and differential diagnosis of hairy cell leukemia in bone marrow and spleen. Semin Oncol 1984;11:334-346. 15. Zafrani S, Degos F, Guigui B, et al. The hepatic sinusoid in hairy cell leukemia: An ultrastructural study of 12 cases. Human Pathol 1987;18:801-807. 16. Jansen J, Schuit HRE, Schreuder GMT, et al. Distinct subtype within the spectrum of hairy cell leukemia. Blood 1979;54:459-467. 17. Jansen J, Schuit H, Meijer C, et al. Cell markers in hairy cell leukemia studied in cells from 51 patients. Blood 1982;59:52-60. 18. Korsmeyer SJ, Greene WC, Cossman J, et al. Rearrangement and expression of immunoglobulin genes and expression of Tac antigen in hairy cell leukemia. Proc Natl Acad Sci USA 1983;80: 4522-4526. 19. Naiem F. Hairy cell leukemia: Characteristics of the neoplastic cells. Human Pathol 1988;19:375-388.

1992


associated antibodies reactive on paraffin sections.38"42'46 Cross-reactivity with a few erythroblasts was found in only two instances of myelodysplasia among 85 bone marrow biopsy specimens not involved by lymphoma or HCL. Despite the rarity of the disorder, the diagnosis of HCL may not be difficult when the patient's clinical status is characteristic and when the infiltrate in the bone marrow is typical and identified easily. Antibody DBA.44 not only recognized the individual hairy cells but also underlined the usually lower estimate of the nature of the infiltrate and topographic variations of its density. The immunomorphologic considerations after DBA.44 positivity in appropriate clinical circumstances are limited. The differential diagnosis of disorders other than HCL can be made by simultaneous cytologic observations usually not possible in cryostat sections. DBA.44 was especially helpful in identifying the more common disorder of B-cell chronic lymphocytic leukemia because the cells of the latter did not react with this antibody in 21 of 22 bone marrow samples and in 19 of 20 lymph nodes (Table 1). Although DBA.44 reacted with 30% of lymphoplasmacytic lymphomas, the nuclear and cytoplasmic features in this disorder and the presence of monotypic cytoplasmic immunoglobulins are unlikely to make distinction from HCL difficult. The other small cell lymphomas encountered in the marrow, such as centrocytic lymphoma (small cleaved cell), can be labeled with DBA.44 in approximately one of three lymph nodes, but this is uncommon in the bone marrow samples (1 of 9 were positive). Also the weak to moderate staining of lymphoma cells, the absence of hairy features of these cells, and the usual paratrabecular infiltrates (HCL is not confined to the paratrabecular area) facilitate this differential diagnosis.' 4 ' 936 The differential diagnosis of acute lymphoblastic or acute myeloid leukemia was not considered difficult by any of the participating hematopathologists.

HOUNIEU ET AL.

33

lmmunohistochemical Diagnosis of Hairy Cell Leukemia 35. Falini B, Pileri S, Flenghi L, et al. Selection of a panel of monoclonal antibodies for monitoring residual disease in peripheral blood and bone marrow of interferon-treated hairy leukaemia patients. Br J Haematol 1990;76:460-468. 36. Burke JS. The value of bone marrow biopsy in the diagnosis of hairy cell leukemia. Am J Clin Pathol 1978;70:876-884. 37. Martin J, Boras V, Houwen B, Francovich N. Hairy cell leukemia and anti-leukocyte common antigen. Am J Clin Pathol 1988; 90: 412-420. 38. Cartun R, Coles F, Pastuszak W. Utilization of monoclonal antibody L26 in the identification and confirmation of B-cell lymphomas. Am J Pathol 1987;129:415-421. 39. Davey F, Gatter K., Ralfkiaer E, et al. Immunophenotyping of nonHodgkin lymphomas using a panel of antibodies on paraffinembedded tissues. Am J Pathol 1987; 129:54-63. 40. Norton A, Isaacson P. Detailed phenotypic analysis of B-cell lymphoma using a panel of antibodies reactive in routinely fixed wax-embedded tissue. Am J Pathol 1987; 128:225-240. 41. Hall P, D'Ardenne A, Stansfeld A. Paraffin section immunohistochemistry. I. Non Hodgkin's lymphoma. Histopathology 1988; 13: 149-160. 42. Feller A, Wacker H, Moldenhauer G, et al. Monoclonal antibody Ki-B3 detects a formalin resistant antigen on normal and neoplastic B cells. Blood 1987;70:629-636. 43. Hsu SM. Phenotypic expression of B lymphocytes. III. Marginal zone B-cells in the spleen are characterized by the expression of Tac and alkaline phosphatase. J Immunol 1985; 135:123-130. 44. Sheibani K, Sohn C, Burke J, et al. Monocytoid B-cell lymphoma: A novel B-cell neoplasm. Am J Pathol 1986; 124:310-318. 45. Sheibani K, Burke J, Swartz W, et al. Monocytoid B-cell lymphoma. Clinicopathologic study of 21 cases of a unique type of low-grade lymphoma. Cancer 1988;62:1531-1538. 46. Norton A, Isaacson P. Lymphoma phenotyping in formalin-fixed and paraffin wax-embedded tissues. II. Profiles of reactivity in the various tumour types. Histopathology 1989; 14:557-579. 47. Castaigne S, Sigaux F, Cantell K, et al. Interferon alpha in the treatment of hairy cell leukemia. Cancer 1986;57:1681-1684. 48. Quesada J, Gutterman J, Hersh E. Treatment of hairy cell leukemia with alpha-interferons. Cancer 1986;57:1678-1680. 49. Ratain MJ, Vardiman JW, Golomb HM. The role of interferon in the treatment of hairy cell leukemia. Semin Oncol 1986; 13:2128. 50. Spiers ASD, Moore D, Cassileth PA, et al. Remissions in hairy cell leukemia with Pentostatin (2'-Deoxycoformycin). N Engl J Med 1987;316:825-830. 51. Anonymous. Treatment of hairy cell leukemia. Lancet 1990,336: 149-150.

Vol. 98 • No. I


20. Hsu S, Yang K, Jaffe E. Hairy cell leukemia: A B-cell neoplasm with a unique antigenic phenotype. Am J Clin Pathol 1983; 80: 421-428. 21. Tubbs R, Savage R, Sebek B, et al. Antigenic phenotype of splenic hairy cells. Am J Med 1984;76:199-205. 22. Meijer C, Albeda F, Van der Valk P, et al. lmmunohistochemical studies of the spleen in hairy cell leukemia. Am J Pathol 1984; 115: 266-274. 23. Falini B, Schwarting R, Erber W, et al. The differential diagnosis of hairy cell leukemia with a panel of monoclonal antibodies. Am J Clin Pathol 1985;83:289-300. 24. Schwarting R, Stein H, Wang CY. The monoclonal antibodies alpha S-HCL 1 (alpha Leu-14) and alpha S-HCL 3 (alpha Leu-M5) allow the diagnosis of hairy cell leukemia. Blood 1985,65:974983. 25. Al-Katib A, Berman E, Black P, Koziner B. Hairy cell leukemia: A model for studying the B cell family of diseases. Semin Oncol 1986;13:48-54. 26. Burke J, Sheibani K. Hairy cells and monocytoid B lymphocytes: are they related? Leukemia 1987; 1:298-300. 27. Vardiman JW, Gilewski TA, Ratain MJ, et al. Evaluation of LeuM5 (CD1 lc) in hairy cell leukemia by the alkaline phosphatase anti-alkaline phosphatase technique. Am J Clin Pathol 1988; 90: 250-256. 28. Thaler J, Denz H, Dietze O, et al. Immunohistological assessment of bone marrow biopsies from patients with hairy cell leukemia: Changes following treatment with alpha-2-interferon and deoxycoformycin. Leukemia Res 1989;13:377-383. 29. Voigt J-J, Vinel JP, Caveriviere P, et al. Diagnostic immunohistochimique des localisations hemopatiques des hemopathies lymphoides malignes. Etude de 80 cas. Gastroenterol Clin Biol 1989;13:343-352. 30. Visser L, Shaw A, Slupsky J, et al. Monoclonal antibodies reactive with hairy cell leukemia. Blood 1989;74:320-325. 31. Flenghi L, Spinozzi F, Stein H, et al. LF61: A new monoclonal antibody directed against a trimeric molecule (150kDa, 125 kDa, 105 kDa) associated with hairy cell leukemia. Br J Haematol 1990;76:451-459. 32. Schwarting R, Dienemann D, Kruschwitz M, et al. Specificities of monoclonal antibodies B-ly-7 and HML-1 are identical. Blood 1990;75:320-321. 33. Mulligan S, Travade P, Matutes E, et al. B-ly-7, a monoclonal antibody reactive with hairy cell leukemia, also defines an activation antigen on normal CD8+ T cells. Blood 1990;76:959-964. 34. Moldenhauer G, Mielke B, Dorken B, et al. Identity of HML-1 antigen on intestinal intraepithelial T cells and of B-ly-7 antigen on hairy cell leukemia. Scand J Immunol 1990;32:77-82.