Hyperplasia of Mantle/Marginal Zone B Cells With Clear Cytoplasm in Peripheral Lymph Nodes A Clinicopathologic Study of 35 Cases

Transcription

1 Hematopathology / HYPERPLASIA OF MANTLE/MARGINAL ZONE B CELLS WITH CLEAR CYTOPLASM Hyperplasia of Mantle/Marginal Zone B Cells With Clear Cytoplasm in Peripheral Lymph Nodes A Clinicopathologic Study of 35 Cases John P. Hunt, MD, 1 Joel A. Chan, MD, 2 Michael Samoszuk, MD, 3 Russell K. Brynes, MD, 4 Antonio M. Hernandez, MD, 2,5 Randall Bass, MD, 2 Dennis D. Weisenburger, MD, 6 Konrad Müller-Hermelink, MD, 7 and Bharat N. Nathwani, MD 1,8 Key Words: Mantle cells with clear cytoplasm; Marginal zone B cells; Primary follicles; Mantle cell hyperplasia; Marginal zone B-cell hyperplasia; Non-Hodgkin lymphoma Abstract We describe 35 peripheral lymph nodes classified as mantle cell/marginal zone B-cell hyperplasia with clear cells using morphologic and immunologic findings. For the purpose of this study, we obtained clinical follow-up information and performed immunoglobulin gene rearrangement studies on paraffin sections by polymerase chain reaction. Architecturally, the nodes were suggestive of a benign process: no pericapsular infiltration, sinuses readily identified, scattered reactive follicles present, and paracortical nodular hyperplasia present. No monocytoid B cells were present. Focally, small lymphoid cells with round nuclei and clear cytoplasm (clear cells) formed monomorphic nodular, inverse follicular, and/or marginal zone patterns. Flow cytometry and immunohistochemical analysis revealed neither light chain restriction nor an aberrant B-cell phenotype. Immunoglobulin gene rearrangement studies showed a clonal band in 1 of 26 cases in which DNA was amplified. To ascertain the clinical relevance of this positive case, follow-up information was obtained 30 months after the initial biopsy; the 83-year-old woman was alive without treatment but had splenomegaly and bone marrow involvement by marginal zone B-cell lymphoma. The morphologic and immunologic criteria used for diagnosis of mantle cell/marginal zone B-cell hyperplasia with clear cytoplasm are valid; however, to rule out the possibility of occult lymphoma, immunoglobulin gene rearrangement studies and clinical follow-up are necessary. In the World Health Organization classification of malignant lymphomas, there are 3 separate categories for the primary marginal zone B-cell lymphomas: splenic, extranodal, and nodal. 1 These are considered separate disease entities, since there are differences in their clinical, immunologic, and cytogenetic features. The malignant cells of the primary splenic type are IgD positive, whereas those of primary extranodal marginal zone B-cell lymphoma are IgD negative. 2 With primary nodal cases, most are IgD negative but a minority are IgD positive. 3 However, IgD is negative in normal marginal zone B cells seen in the Peyer patches and the spleen Regarding benign marginal zone B cells, they are normally seen in spleen, often in mesenteric lymph nodes, and have been reported in nodes whose biopsy sites 4 are not specified. On the other hand, in peripheral lymph nodes, they are reported to be absent 5-8 or only rarely 9,10 or occasionally seen. 11 In 1990, van Krieken and Lennert 10 reported 1 case of marginal zone B-cell hyperplasia in a peripheral lymph node. The marginal zone B cells formed, in this case, an inverse follicular pattern (mantle cell nodules surrounded by marginal zone B cells) and monomorphic nodules. 10 In a letter, Stein 12 stated that these observations had not been previously reported or were exceptional. Isaacson 9 wrote in 2001 that a marginal zone B-cell zone is often present in mesenteric lymph nodes, whereas this is rarely the case in peripheral lymph nodes. Further clouding the issue is the fact that mantle cells may have moderate quantities of pale to clear cytoplasm and, thus, resemble marginal zone B cells. 13 In view of the paucity of published information on hyperplasia of mantle cells with clear cytoplasm or marginal zone B cells in peripheral lymph nodes, we present our experience with Am J Clin Pathol 2001;116: American Society of Clinical Pathologists

2 Hematopathology / ORIGINAL ARTICLE peripheral lymph nodes that we classified as mantle cell/marginal zone B-cell hyperplasia with clear cells using morphologic and immunophenotypic criteria. For the purpose of this study, these 35 peripheral lymph nodes were further studied by obtaining clinical follow-up and by doing additional immunohistochemical stains and immunoglobulin gene rearrangement studies. Materials and Methods Case Selection Over a 3-year period (July 1, 1997-June 30, 2000), one of us (B.N.N.) received 3,278 cases in consultation. From these, 1,556 were lymph nodes, of which 447 were benign. Of the benign nodes, 385 were peripheral and 35 of these were classified as mantle zone and/or marginal zone B-cell hyperplasia, with clear cells (see Definitions ), and thus form the basis for the present study. Abdominal and mediastinal lymph nodes were not included in the study because central lymph nodes are reported to have benign marginal B cells. Control Group To ascertain the phenotype of benign mantle zone cells with scant cytoplasm as seen in the mantle zones of secondary follicles, we selected 5 benign peripheral lymph nodes with reactive secondary follicles for detailed immunophenotypic analysis. Clinical Information The consultation report and original pathology report were reviewed for all cases. In addition, we obtained clinical follow-up information via questionnaires submitted to the referring pathologists and clinicians. Information requested included additional history about adenopathy or splenomegaly, development of lymphoma or leukemia, the results of subsequent biopsies, chemotherapy or radiation therapy, and survival. Definitions The following definitions were used. Clear Cells These small lymphocytes with round to slightly irregular nuclei and moderate or abundant amounts of pale to clear cytoplasm appeared loosely packed and produced areas of pallor on H&E-stained sections at low power. The pale appearance was due to the clear cytoplasm, which increased the space between individual nuclei. The presence of clear cytoplasm was an inherent feature of these cells and did not represent an artifactual increase in cytoplasm due to shrinkage of the nucleus since the chromatin details were well visualized. The clear cells were distributed in 3 distinct patterns 14 : 1. Monomorphic nodular pattern (spherical structures, 1 layer): Adjacent to follicles and elsewhere, there were multiple discrete, small, monomorphic nodules of clear cells that were pale staining at low power. In most cases, there were rare follicular dendritic reticulum cells in the nodules. 2. Inverse follicular pattern (spherical structures, 2 layers with an inverse arrangement): The clear cells formed a pale outer zone around darkly staining nodules composed of mantle cells with scant cytoplasm (an inverse follicular pattern due to dark color on the inside and light color on the outside). 3. Marginal zone pattern (spherical structures, 3 layers): The clear cells formed a third (outer) pale staining layer around a darkly staining second layer of mantle cells (mantle zone) that surrounded an inner pale zone of follicular center cells (germinal center). Monocytoid B Cells The term monocytoid B cells is used for medium-sized, B-lymphoid cells with abundant pale to clear cytoplasm present within and around sinuses and adjacent to follicles, as seen in toxoplasmosis. Admixed with these monocytoid B cells, there are frequently neutrophils and scattered transformed cells. Thus, the term monocytoid B cells is used as a morphologic descriptor without determining their origin or mutational profile. 15 Histopathologic Review The histopathologic features on H&E-stained sections were reviewed by 3 pathologists (J.P.H., J.A.C., and B.N.N.). The following features were evaluated in each case: (1) clear cells, (2) monocytoid B cells, (3) germinal centers, (4) mantle zones, (5) sinuses, (6) interfollicular areas, and (7) plasma cells. With regard to clear cells, their pattern of distribution (nodular, inverse follicular, or marginal zone) and whether there was overlap of these patterns, as well as the presence of pericapsular infiltration by the clear cells and clear cell clusters within sinuses, was recorded. The cytologic features of the clear cells that were evaluated included the following: nuclear size and shape, presence or absence of nucleoli, amount of cytoplasm, mitotic activity, and admixture of other cell types. In each case, the nuclear chromatin structure (heterochromatin and euchromatin) was assessed to ascertain whether the clear cytoplasm was due to nuclear retraction (fixation artifact) or was an inherent feature of the cells. American Society of Clinical Pathologists Am J Clin Pathol 2001;116:

3 Hunt et al / HYPERPLASIA OF MANTLE/MARGINAL ZONE B CELLS WITH CLEAR CYTOPLASM The percentages of surface areas occupied by the germinal center cell compartment, the mantle zones, clear cell nodules, interfollicular areas, and sinuses were estimated; their range, median, and mean also were calculated. The percentage of the surface area occupied by plasma cells was estimated, and the range, mean, and median were calculated. The sinuses also were evaluated for patency, distention, or obliteration, as well as for cell content (eg, histiocytes, lymphocytes, clear cells, or monocytoid B cells). Immunophenotypic Studies on Paraffin Sections and by Flow Cytometry Immunohistochemical stains were performed on formalin, zinc-formalin, or B-5 fixed, paraffin-embedded tissue. The sources of the antibodies and the stains performed were as follows: CD3, CD5, CD20, CD23, CD43, cyclin D1, bcl-2, Ki-67, kappa, lambda, IgD, and IgM. These stains were performed with standard antigenretrieval methods (microwave and/or Proteinase treatment) by various laboratories with suitable controls. Antibody to Ki-B3, an epitope of CD45RA that stains a subset of B cells that includes mantle cells, was a gift from R. Parwaresch, MD, Kiel, Germany. For Ki-B3 immunostaining, slides were deparaffinized with xylene and rehydrated with graded ethanols. Sections were placed in a 0.01-mol/L concentration of citrate buffer, ph 6.0, and microwaved twice for 6 minutes per cycle. Reactivity was demonstrated by an avidin-biotin immunoperoxidase detection system using 3,3 diaminobenzidine-tetrahydrochloride dihydrate (Vector Laboratories, Burlingame, CA) as the chromogen. Semiquantitative evaluation of the staining characteristics of various cell compartments was performed by 3 pathologists (J.P.H., J.A.C., and B.N.N.). In 10 cases, flow cytometry results from the submitted lymph nodes also were available from the referring pathologists. Immunoglobulin Gene Rearrangement Studies To further evaluate the clonality of the lymphoid population, immunoglobulin heavy chain gene rearrangement studies were performed on DNA extracted from paraffin-embedded tissue by Quest Diagnostics, San Juan Capistrano, CA. DNA was extracted and amplified by previously published techniques Four primer sets were used: framework I/JH, mbr2/jh, framework III/JH, and GH26/GH27 as a DNA integrity control. Specimens were analyzed in duplicate, and those that yielded only a single weak polymerase chain reaction product with a particular primer set were interpreted as negative according to the laboratory s standard procedures. Specimens that did not yield any amplification product with the GH26/GH27 primer set were deemed to have insufficient amplifiable DNA. Results Clinical Findings at Referral Of the 35 patients, 9 were men and 26 were women, with a median age of 47 years (range, years). There were 20 axillary, 8 groin, and 7 cervical lymph node biopsy specimens for review. The lymph nodes ranged from 0.8 to 4.0 cm in greatest dimension (median, 1.6 cm), with 9 (26%) measuring 2.0 cm or larger. One patient had a concurrent breast carcinoma that required axillary dissection to ascertain the presence of metastases. Splenomegaly was absent in every case. Most patients were asymptomatic or had only localized symptoms, such as lymph node swelling, at referral. No patient had a history of small lymphocytic or marginal zone B-cell lymphoma or of leukemia. One patient was reported to have a history of diffuse large B-cell lymphoma of the tongue for which he had received local radiotherapy resulting in a complete remission. However, our diagnosis was mantle cell lymphoma of the tongue (CD20+, cyclin D1+, CD5+, CD43+, KiB3+, IgD+). Forty months later, this patient underwent a biopsy of an axillary node that was interpreted by us as mantle cell/marginal zone B-cell hyperplasia. Histopathologic and Immunophenotypic Studies Table 1 and Table 2 summarize the morphologic and cytologic findings in the 35 cases. To facilitate understanding of the cells seen in each compartment, we first describe the morphologic features of the cells in each compartment, followed by their phenotype. Table 1 shows the estimated surface areas occupied by each lymph node compartment and the plasma cells. Pericapsular infiltration by clear cells was not present in any case. Monocytoid B cells were absent in all cases. The sinuses were readily apparent in 34 cases (97%) and accounted for 16% of the surface area. The sinuses were distended in 21 cases (60%), patent in 13 cases (37%), and obliterated in only 1 case (3%). In 34 cases, the sinuses Table 1 Surface Area Occupied by Each Compartment/Cell Type in 35 Cases Compartment/Cell Type Range (%) Median (%) Sinuses Germinal centers Mantle zones Clear cells (nodular, inverse follicular, or marginal zone patterns) Paracortical nodules Other interfollicular areas Plasma cells Am J Clin Pathol 2001;116: American Society of Clinical Pathologists

4 Hematopathology / ORIGINAL ARTICLE Table 2 Morphologic Features of the Clear Cells in 35 Cases Cells With Clear Cytoplasm No. (%) Present Distribution/localization Nodular pattern 30 (86) Inverse follicular pattern 11 (31) Marginal zone pattern 9 (26) Fusion of nodules Focal 11 (31) Prominent 4 (11) Absent 20 (57) Amount of cytoplasm Moderate 14 (40) Moderate to abundant 18 (51) Abundant 3 (9) Nuclear shape Round 19 (54) Slightly irregular 12 (34) Irregular 4 (11) contained varying numbers of benign histiocytes, whereas 1 case showed small lymphocytes with scant cytoplasm filling the sinuses. In no cases were there clusters of clear cells within the sinuses. The germinal center cell compartment was seen in 34 cases (97%). The follicles (germinal centers surrounded by well-defined mantle zones) were well defined and separated from one another. The germinal centers were usually small and made up 10% of the surface area. They consisted of a mixture of centrocytes and centroblasts, with admixed mitosis, histiocytes, a few small lymphocytes, and follicular dendritic reticulum cells (FDRCs). Morphologically and immunohistochemically, the cells in the germinal centers were benign. Mantle zones were identifiable in 34 cases (97%), and occupied up to 20% of the surface area. Although in some cases a mantle zone pattern was seen, adjacent mantles showed no fusion. The mantle zones were composed of small lymphocytes densely packed together, with scant cytoplasm. Immunohistochemical stains showed the mantle cells to be positive for CD20, IgD, bcl-2, Ki-B3, and, in 9 (27%) of the 33 cases with adequate tissue for evaluation, for IgM. Ki-67 was positive in fewer than 5% of mantle cells. These mantle cells were negative for CD3, CD5, CD43, and cyclin D1, and there was no light chain restriction. Since admixed FDRCs were seen in most cases, the CD23 staining in the mantle cells was difficult to interpret; however, scattered mantle cells were positive in some cases. Clear cells were seen in every case and constituted 10% to 65% of the surface area (median, 20%). Clear cells occupied more than 50% of the surface area in only 4 cases (11%). Clear cells were present in 1 or more distinctive patterns Image 1A, Image 1B, and Image 1C : monomorphic spherical structures (nodules) in 30 cases (86%); an inverse follicular pattern in 11 cases (31%); and a marginal zone pattern in 9 cases (26%). Fusion of adjacent nodules was present focally in 11 (31%), with more extensive fusion in 4 (11%). In 23 cases (66%), clear cell nodules were seen abutting hyalinized sinus trabeculae. The clear cells showed moderate amounts of cytoplasm in 14 cases (40%), moderate to abundant amounts in 18 (51%), and more abundant cytoplasm in the remaining 3 (9%). The nuclear contours were round or slightly irregular in 31 cases (89%), with more irregularity in the remaining 4 (11%) Image 1D. In all cases, the nuclei had recognizable heterochromatin clumping and absent or very small, inconspicuous nucleoli. Rare transformed lymphoid cells were admixed among the clear cells in 29% of the cases, and rare neutrophils and plasma cells were admixed in 9% and 6%, respectively. No cases had prolymphocytes or paraimmunoblasts as seen in small lymphocytic lymphoma. Virtually all clear cells were positive for CD20 and bcl-2 in all cases Image 2A, Image 2B. In each case, a subset of the clear cells was positive for IgD and Ki-B3. In the Ki-B3 stains, 10% to 60% of the clear cells were positive (median, 30%) Image 2C, and 10% to 60% of the clear cells were positive for IgD (median, 20%) Image 2D. IgM was present in 9 (27%) of the 33 cases with adequate tissue for evaluation. CD23 showed membrane staining of scattered clear cells in 40% of the cases, and rare scattered FDRCs were admixed in some cases. Ki-67 was positive in fewer than 5% of the clear cells in each case. Moreover, scattered clear cells were weakly positive for kappa in 30 (86%) and lambda in 29 (83%) cases; the 1 case (3%) positive for kappa only is discussed under Correlation of Histologic Immunohistochemical, Flow Cytometric, and Gene Rearrangement Data. The clear cells were negative for CD3, CD5, CD43, and cyclin D1. Nodular expansion of the interfollicular areas was present in every case and occupied as much as 55% of the surface area (median, 25%). These areas showed mottling by Langerhans cells and histiocytes. There was melanin pigment in 28 cases (80%), consistent with dermatopathic lymphadenitis, whereas the remaining 7 (20%) showed no pigment and were classified as nodular T-zone hyperplasia. The remaining interfollicular areas contained mainly T lymphocytes and occupied an average of 11% of the surface area. Normal plasma cells were present in every case and occupied up to 15% of the surface area (median, 5%). A polytypic staining pattern was found with kappa and lambda in all cases, and a variable number of plasma cells stained for IgD or IgM in all cases. Flow Cytometry No monotypic B-cell population or aberrant B-cell phenotype was detected by flow cytometry in the 10 patients for whom such results were available. American Society of Clinical Pathologists Am J Clin Pathol 2001;116:

5 Hunt et al / HYPERPLASIA OF MANTLE/MARGINAL ZONE B CELLS WITH CLEAR CYTOPLASM A B C D Image 1 Formalin-fixed tissue sample. A, At scanning magnification (2 lens), the sinuses are patent, and there is no pericapsular infiltration. There are scattered small follicles with thick mantle zones and many small, pale-staining 1-layered nodules, some of which are closely packed together (H&E). B, At slightly higher magnification, using a 4 lens, the uniform, small, pale-staining nodules are apparent (H&E). C, Using a 10 lens, the pale color of the nodule due to the presence of clear cytoplasm is apparent (H&E). D, At high magnification, using a 40 lens, the cells are small and monotonous and have pale cytoplasm, absent to inconspicuous nucleoli, and readily identifiable chromatin structure. Note the absence of neutrophils, mitotic figures, and cellular atypia (H&E). Immunoglobulin Gene Rearrangement Studies In 9 of 35 cases, the GH26/GH27 primer set was negative, which was interpreted as inadequate DNA for gene rearrangement studies. In 25 of 26 cases, the DNA was adequate for polymerase chain reaction amplification and did not yield any clonal bands. However, in 1 case immunoglobulin gene rearrangement studies, run in duplicate, revealed a clonal band with the primer set FRIII/JH. Correlation of Histologic, Immunohistochemical, Flow Cytometric, and Gene Rearrangement Data In 29 cases, some clear cells showed weak polytypic staining for kappa and lambda. In 1 case, some clear cells were weakly positive for kappa but negative for lambda; however, flow cytometry revealed polyclonal B cells, and no clonal immunoglobulin gene rearrangement was detected. In an additional 5 cases, the clear cells were negative with both kappa and lambda stains. Flow cytometry was done in 3 of 554 Am J Clin Pathol 2001;116: American Society of Clinical Pathologists

6 Hematopathology / ORIGINAL ARTICLE A B C D Image 2 A, The B cells in the nodules stain intensely with CD20. In the interfollicular areas, the number of CD20+ cells is few. B, The follicular centers are few, very small, and negative with bcl-2, which suggests that they are benign. The mantle zones and the 1-layered monomorphic nodules are strongly positive with bcl-2. Most positive cells between the nodules are benign T cells. C, The mantle zones and the nodules have Ki-B3 positive and Ki-B3 negative cells. Few cells between the nodules are positive. D, The mantle zones and the nodules have IgD-positive and IgD-negative cells. Few cells between the nodules are positive. the 5 cases, and the B cells were polyclonal. The fourth case had amplifiable DNA but no clonal immunoglobulin gene rearrangement. In the fifth case, flow cytometry was not performed, and there was no amplifiable DNA, so a monotypic population could not be excluded in this case. The 1 patient with a clonal immunoglobulin gene rearrangement was an 80-year-old woman with breast carcinoma who had an axillary lymph node dissection. The lymph nodes showed no metastases. The sinuses occupied 10% of the surface area and contained histiocytes. Areas of nodular paracortical hyperplasia without melanin pigment constituted 35% of the surface area, and other interfollicular areas covered 10%, whereas the follicle center cell compartment and mantle zones occupied 10% and 5%, respectively, and plasma cells occupied 5% of the surface area. Nodules of clear cells occupied 25% of the surface area, and no fusion of adjacent nodules was seen Image 3A. The clear cells had small, round to slightly irregular nuclei with inconspicuous nucleoli and moderate amounts of clear cytoplasm Image 3B. Some clear cells were weakly stained and appeared to be polytypic with kappa and lambda stains. All clear cells were positive for CD20 and bcl-2; about 20% stained with Ki-B3, and 10% were positive for IgD Image 3C, Image 3D, and they were negative for CD5, CD23, American Society of Clinical Pathologists Am J Clin Pathol 2001;116:

7 Hunt et al / HYPERPLASIA OF MANTLE/MARGINAL ZONE B CELLS WITH CLEAR CYTOPLASM A B C D Image 3 Formalin-fixed tissue sample. A, At low magnification, using a 10 lens, there are several pale-staining monomorphic nodules, but sinuses are patent (H&E). B, At high magnification, using a 40 lens, most cells are generally round cells with pale to clear cytoplasm. Admixed follicular dendritic reticulum cells, histiocytes, and rare transformed cells are seen (H&E). C, Within the nodule, only a minority of cells stain with Ki-B3. D, Within the nodule in the center, a small minority of the cells stain with IgD, while most cells are negative. CD43, cyclin D1, and IgM. Flow cytometry was not performed on this lymph node biopsy specimen. Clinical Follow-up Clinical follow-up was available for all 35 patients for 6.5 to 41 months (median, 19 months). In 1 patient, because a clonal immunoglobulin gene rearrangement was detected 30 months after our diagnosis on the initial nodal biopsy specimen, we immediately informed the pathologist that clinical follow-up was necessary to ascertain whether this patient had a lymphoma. Subsequently, the patient s clinician detected splenomegaly and performed a bone marrow biopsy and aspirate and flow cytometry studies. The bone marrow clot and core biopsy showed approximately 10% involvement of the marrow space by a lymphoma composed of small nodules of small lymphocytes with generally round nuclei and scant cytoplasm. By flow cytometry, the neoplastic lymphoid cells were positive for bright CD20, dim CD23, and dim monotypic kappa, but negative for CD5 and CD10. By immunohistochemical analysis, the small lymphocytes in the bone marrow clot sections were positive for CD20, bcl-2, IgD (2%), Ki-67 (<5%), and Ki-B3 (30%), but negative for CD5, CD10, CD23, CD43, cyclin D1, kappa, lambda, CD25, and DBA.44. These findings confirmed 556 Am J Clin Pathol 2001;116: American Society of Clinical Pathologists

8 Hematopathology / ORIGINAL ARTICLE involvement by a nodal or an extranodal marginal zone B-cell lymphoma. This patient is alive and asymptomatic 32 months after the initial biopsy, without treatment for lymphoma. Follow-up information for the remaining 34 patients revealed that 33 are alive and well, without evidence of lymphoma. However, in 1 patient, peripheral T-cell lymphoma of the large cell type developed in a cervical lymph node 2 months after our initial diagnosis of mantle/marginal zone B-cell hyperplasia in an axillary lymph node. This patient had stage III disease, was treated with BACOP (bleomycin, Adriamycin, Cytoxan, Oncovin, and prednisone) with little response, and died of progressive lymphoma after 5 months. Comparison of Mantle Cells of Control Cases With Clear Cells of 35 Study Cases In the 5 control cases of peripheral lymph nodes with benign secondary follicles, the immunophenotype of the mantle cells was positive for CD20 and bcl-2, 85% (median, 85%) of mantle cells were positive for Ki-B3 and 90% (median, 90%) for IgD, but they were negative with CD5, CD43, and cyclin D1, and there was no light chain restriction. Fewer than 5% of the mantle cells were positive for Ki- 67, and scattered CD23+ cells were seen in 2 cases. The bcl-2 stain also was positive in all clear cells of our 35 cases. However, with respect to Ki-B3 and IgD, there were significant differences between the percentages of positive mantle cells in the control cases and the clear cells in our 35 cases. In the Ki-B3 stains, 80% to 95% of the control mantle cells were positive (median, 85%), whereas only 26% of the clear cells were positive (median, 20%) (P <.01). Similarly, in the IgD stain, 65% to 95% of the control mantle cells were positive (median, 90%), whereas only 30% of the clear cells were positive (median, 30%) (P <.01, Student t test). Discussion The 35 cases that we report represent 9.1% (35 of 385) of the benign peripheral lymph nodes that were submitted in consultation during the last 3 years. There were 9 men and 26 women (M/F ratio, 1:3) with a median age of 47 years at referral. The patients had biopsy specimens of peripheral lymph nodes that measured 0.8 to 4.0 cm in greatest dimension (median, 1.6 cm). Most patients had no symptoms or only localized symptoms. No patients had splenomegaly, and none had a history of small lymphocytic lymphoma, marginal zone B-cell lymphoma, or leukemia. One patient had a localized mantle cell lymphoma of the tongue 40 months before this, with complete response to local radiotherapy and no relapse to date. The morphologic findings in our 35 cases did not fit with the histologic description of any single, well-defined benign or malignant disease entity. On the one hand, the cases were characterized by no pericapsular infiltration, preserved sinuses (97%), few scattered benign follicles, and nodular T-zone hyperplasia, all of which suggested a benign proliferation. On the other hand, there were small, multifocal areas of clear cells producing monomorphic nodules and, focally, inverse follicular and marginal zone patterns. We concluded that these clear cells were the most important pathology, even though they occupied less than 50% of the surface area in most cases (89%), and the major question was whether these clear cell nodules represented focal lymphomatous involvement. Although a focal marginal zone and/or inverse follicular pattern was seen in a minority of our cases, it was never prominent as is seen in most cases of marginal zone B-cell lymphoma. The morphologic differential diagnosis for the clear cells included 3 benign and 6 malignant conditions. The benign conditions were as follows: (1) primary follicles with clear cytoplasm, (2) hyperplasia of mantle cells with clear cytoplasm, and (3) marginal zone B-cell hyperplasia with clear cells. The possible malignant lymphomas included the following: (1) small lymphocytic lymphoma, (2) mantle cell lymphoma with clear cytoplasm, 13 (3) nodal involvement by extranodal marginal zone B-cell lymphoma of the mucosa-associated lymphoid tissue (MALT) type, (4) primary nodal marginal zone B-cell lymphoma, (5) nodal involvement by primary splenic marginal zone B-cell lymphoma, and (6) nodal involvement by hairy cell leukemia. Clinical information was useful for narrowing the differential diagnosis. There was no splenomegaly or leukemia at the time of referral, and all biopsies were of peripheral lymph nodes. Thus, primary splenic marginal zone B-cell lymphoma and hairy cell leukemia were very unlikely. At low magnification, the monomorphic nodules of clear cells in our cases resembled the pseudofollicles (proliferation centers) seen in small lymphocytic lymphoma. However, our cases lacked prolymphocytes or paraimmunoblasts. Immunohistochemically, the B cells did not coexpress CD5 and CD43 and did not show light chain restriction. Therefore, our cases were not small lymphocytic lymphoma. In about 5% of cases of mantle cell lymphoma, the neoplastic mantle cells may have moderate quantities of pale-staining cytoplasm. Such cells may be a subset of mantle cells with clear cytoplasm or may represent so-called marginal zone B-cell differentiation. 13 Cytologically, malignant mantle cells with clear cytoplasm show more nuclear atypia and mitotic figures, which were virtually absent in the clear cells in our study. Since the clear cells in our cases were CD20+ but were negative for CD5, CD43, and cyclin D1, we concluded that they were not malignant mantle cells. American Society of Clinical Pathologists Am J Clin Pathol 2001;116:

9 Hunt et al / HYPERPLASIA OF MANTLE/MARGINAL ZONE B CELLS WITH CLEAR CYTOPLASM In extranodal marginal zone B-cell lymphoma of the MALT type, peripheral lymph nodes may be secondarily involved. Although no cases in our study had a history of or a concurrent MALT-type lymphoma, a peripheral node may be secondarily involved by an undetected MALT-type lymphoma. Monotypic plasma cells are reported in about 35% of the cases of MALT-type lymphoma but, in all our cases, the plasma cells were polytypic. Moreover, such cases often contain sinusoidal marginal zone B cells, which were lacking in our cases. Finally, the neoplastic marginal zone B cells in MALT-type lymphoma are IgD negative 3 and Ki-B3 negative, 19 whereas scattered clear cells were always IgD and Ki-B3 positive in our cases. Thus, based on clinical, morphologic, and immunophenotypic data, it is unlikely that our cases were secondary MALT-type lymphomas. In most primary nodal marginal zone B-cell lymphomas, the neoplastic cells usually have features of monocytoid B cells, often showing sinusoidal involvement and follicular colonization, with admixed neutrophils within the nodules of clear cells, as well as scattered mitotic figures and transformed cells. Although rare transformed cells and mitotic figures were identified in our cases, all of the other features were absent. Immunophenotypically, most nodal marginal zone B-cell lymphomas are IgD negative, although in some cases, the neoplastic cells can be IgD positive. 3 The bcl-2 positivity of the clear cells in our cases was not helpful in deciding whether these cells were benign or malignant, since bcl-2 is positive in benign marginal zone B cells in the spleen and abdominal lymph nodes, as well as in marginal zone B-cell lymphoma. 3 However, no information is available regarding the bcl-2 positivity in benign marginal zone B cells in peripheral nodes. Of note, benign monocytoid B cells in peripheral nodes are bcl-2 negative. 20 Thus, based on these data, it is unlikely that our cases were nodal marginal zone B-cell lymphoma. The absence of light chain restriction in paraffin sections does not rule out the possibility that our cases were B-cell lymphomas because it is difficult to demonstrate immunoglobulin light chain restriction in small lymphoid cells. This limitation was evident in our study since 1 (4%) of 26 cases with amplifiable DNA had a clonally rearranged immunoglobulin gene, which suggested that this patient had a B-cell lymphoma and not hyperplasia. However, in the presence of benign histologic features and no immunophenotypic abnormalities, the presence of a clonal population does not necessarily indicate the presence of lymphoma. 21,22 In this case, follow-up information obtained 32 months after the initial biopsy was helpful in resolving the issue. The patient was asymptomatic without lymphadenopathy and without any treatment for lymphoma, but has splenomegaly and bone marrow involvement by marginal zone B-cell lymphoma. Thus, there was at least a 4% error rate in our diagnosis. To ascertain whether our error rate could be higher, it was necessary to obtain clinical follow-up information on the remaining 34 patients. In 1 patient, an unrelated aggressive peripheral T-cell lymphoma of the large cell type developed 2 months after her initial biopsy, and she died 5 months later. The follow-up information on the remaining 33 cases showed that all patients were alive and no patient had been diagnosed with or treated for lymphoma. This information suggests that the morphologic and immunophenotypic criteria used in this study are valid; however, molecular studies and close clinical follow-up are necessary to establish the correct diagnosis. Finally, regarding the histogenesis of the clear cells, it was not possible to distinguish whether these cells represented benign mantle cells with clear cytoplasm or benign marginal zone B cells on the basis of histology alone. A review of the literature shows that most benign mantle cells in the mantle zones of secondary follicles are IgD positive 23 and Ki-B3 positive, 24 whereas benign marginal zone B cells in the spleen and in abdominal lymph nodes are IgD negative 25 and Ki-B3 negative. 4 Since benign marginal zone B cells are rarely seen in peripheral lymph nodes, 5-11 little information is available regarding their immunophenotype, but 1 study states that marginal zone B cells in these lymph nodes are also negative for IgD and Ki-B3. 10 In our cases, 26% of the clear cells stained positively for Ki-B3 and 30% stained positively for IgD, compared with 85% and 90% of control mantle cells, respectively. In view of these differences, our results suggest the following possibilities regarding the histogenesis of clear cells: (1) that IgD and Ki- B3 lack the specificity and sensitivity to establish the histogenesis of clear cells; (2) that these clear cells are naive small B cells of primary follicles that have acquired clear cytoplasm; (3) that these clear cells are a special type of mantle cell; (4) that these clear cells are marginal zone B cells; (5) that the clear cells represent a mixture of mantle cells (IgD+, Ki-B3+) and marginal zone B cells (IgD, Ki-B3 ); or (6) that these clear cells are some other distinctive type of B cells that needs to be characterized further. Thus, it seems that until the precise immunophenotype of benign and malignant marginal zone B cells in peripheral lymph nodes is ascertained, it may be difficult to establish the histogenesis of the clear cells reported in this study. From the 1 Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA; 2 Department of Hematopathology, Los Angeles County+University of Southern California Healthcare Network, Los Angeles; 3 Hematology/Oncology Center, Quest Diagnostics, San Juan Capistrano, CA; 4 Special Hematology Laboratories of the Department of Hematopathology, Los Angeles County+University of Southern California Healthcare Network and University of Southern California, Los Angeles; 5 Division of 558 Am J Clin Pathol 2001;116: American Society of Clinical Pathologists

10 Hematopathology / ORIGINAL ARTICLE Hematopathology and Flow Cytometry, Kaiser Hospital, Los Angeles, CA; 6 Hematopathology, Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha; 7 Department of Pathology, University of Würzburg, Würzburg, Germany; and 8 Department of Hematopathology, Los Angeles County+University of Southern California, Los Angeles. Address reprint requests to Dr Nathwani: 1200 N State St, Room 2422, Los Angeles, CA References 1. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization Classification of Neoplastic Diseases of the Hematopoietic and Lymphoid Tissues: report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November J Clin Oncol. 1999;17: Isaacson PG. Splenic marginal zone lymphoma [letter]. Blood. 1996;88: Campo E, Miquel R, Krenacs L, et al. Primary nodal marginal zone lymphomas of splenic and MALT type. Am J Surg Pathol. 1999;23: van Krieken JH, von Schilling C, Kluin M, et al. Splenic marginal zone lymphocytes and related cells in the lymph node: a morphologic and immunohistochemical study. Hum Pathol. 1989;20: Stein H, Bonk A, Tolksdorf G, et al. Immunohistologic analysis of the organization of normal lymphoid tissue and non-hodgkin s lymphomas. J Histochem Cytochem. 1980;28: Hsu S-M, Cossman J, Jaffe ES. Lymphocyte subsets in normal human lymphoid tissues. Am J Clin Pathol. 1983;80: Spencer J, Finn T, Pulford AF, et al. The human gut contains a novel population of B-lymphocytes which resemble marginal zone cells. Clin Exp Immunol. 1985;62: Van den Ord JJ, de Wolf-Peeters C, Desmet VJ. The marginal zone in the human reactive lymph node. Am J Clin Pathol. 1986;86: Isaacson PG. Gastrointestinal lymphoma and lymphoid hyperplasia. In: Knowles DM, ed. Neoplastic Hematopathology. 2nd ed. New York, NY: Lippincott Williams & Wilkins; 2001: van Krieken JH, Lennert K. Proliferation of marginal zone cells mimicking malignant lymphoma. Pathol Res Pract. 1990;186: Wolf-Peeters CD, Tierens A, Achten R. Normal histology and immunoarchitecture of the lymphohematopoietic system. In: Knowles DM, ed. Neoplastic Hematopathology. 2nd ed. New York, NY: Lippincott Williams & Wilkins; 2001: Stein H. Proliferation of marginal zone cells mimicking malignant lymphoma: letter to the case [letter]. Pathol Res Pract. 1990;186: Swerdlow SH, Zukerberg LR, Yang W-I, et al. The morphologic spectrum of non-hodgkin s lymphomas with BCL1/cyclin D1 gene rearrangements. Am J Surg Pathol. 1996;20: Nathwani BN, Hernandez AM, Drachenberg MR. Diagnostic significance of morphologic patterns of lymphoid proliferations in lymph nodes. In: Knowles DM, ed. Neoplastic Hematopathology. 2nd ed. New York, NY: Lippincott Williams & Wilkins; 2001: Stein K, Hummel M, Korbjuhn P, et al. Monocytoid B cells are distinct from splenic marginal zone cells and commonly derive from unmutated naive B cells and less frequently from postgerminal center B cells by polyclonal transformation. Blood. 1999;94: Lombardo JF, Hwang TS, Maiese RL, et al. Optimal primer selection for clonality assessment by polymerase chain reaction, III: intermediate and high grade B-cell lymphomas. Hum Pathol. 1996;27: Segal JH, Jorgensen T, Masih AS, et al. Optimal primer selection for clonality assessment by polymerase chain reaction analysis, I: low grade B-cell lymphoproliferative disorders of nonfollicular center cell type. Hum Pathol. 1994;25: Segal GH, Jorgensen T, Scott M, et al. Optimal primer selection for clonality assessment by polymerase chain reaction analysis, II: follicular lymphomas. Hum Pathol. 1994;25: Lennert K. The proposal for a Revised European American Lymphoma Classification: a new start of a transatlantic discussion. Histopathology. 1995;26: Hernandez AM, Nathwani BN, Nguyen D, et al. Nodal benign and malignant monocytoid B-cells with and without follicular lymphomas: a comparative study of follicular colonization, light chain restriction, bcl-2 and t(14;18) in 39 cases. Hum Pathol. 1995;26: Collins RD. Is clonality equivalent to malignancy: specifically, is immunoglobulin gene rearrangement diagnostic of malignant lymphoma [editorial]? Hum Pathol. 1997;28: Kurtin PJ. How do you distinguish benign from malignant extranodal small B cell proliferations? Am J Clin Pathol. 1999;111(suppl 1):S119-S van der Valk P, Meijer CJLM. The histology of reactive lymph nodes. Am J Surg Pathol. 1987;11: Lai R, Visser L, Poppema S. Tissue distribution of restricted leukocyte common antigens: a comprehensive study with carbohydrate and protein specific CD45R antibodies. Lab Invest. 1991;64: Spencer J, Perry ME, Dunn-Walters DK. Human marginalzone B cells. Immunol Today. 1998;19: American Society of Clinical Pathologists Am J Clin Pathol 2001;116: