Distinct Features of Angioimmunoblastic T-Cell Lymphoma With Bone Marrow Involvement

Transcription

1 Hematopathology / AITL With Bone Marrow Involvement Distinct Features of Angioimmunoblastic T-Cell Lymphoma With Bone Marrow Involvement Young-Uk Cho, MD, 1 Hyun-Sook Chi, MD, 2 Chan-Jeoung Park, MD, 2 Seongsoo Jang, MD, 2 Eul-Ju Seo, MD, 2 and Jooryung Huh, MD 3 Key Words: Angioimmunoblastic T-cell lymphoma; Bone marrow histomorphologic features; Immunophenotypic characteristics DOI: /AJCPQXKCHQH4VAJ5 Abstract We retrospectively reviewed the clinical and laboratory data and bone marrow (BM) histomorphologic features in 33 angioimmunoblastic T-cell lymphoma (AITL) cases. Paraffinembedded BM core biopsy specimens were reacted immunohistochemically with antibodies to pan T-cell markers, CD20, CD10, CD21, and bcl-6. The frequency of BM involvement was 70% (23/33). The following features were more significant in patients with than without BM involvement: fever, hepatosplenomegaly, pleural effusion, elevated lactate dehydrogenase level, hypoalbuminemia, hyponatremia, hypocalcemia, anemia, circulating atypical cells, hypercellular marrow, and plasmacytosis in the BM. Morphologic features included nodular or interstitial infiltration in a paratrabecular distribution, periodic acid Schiff positive intercellular materials, mixed infiltrates of T and B cells, presence of clear cells, and blood vessel proliferation. Immunohistochemical assays revealed that CD5, bcl-6, and CD10 were useful markers of BM infiltration. Seven cases with BM involvement were interpreted as negative for lymphoma initially, mainly owing to insufficient information in nodal biopsy specimens. Several clinical and laboratory features indicate BM involvement of AITL at diagnosis. Laboratory physicians should be more familiar with BM histomorphologic and immunophenotypic characteristics suggesting AITL infiltration to make accurate diagnoses even without prior nodal biopsies. Angioimmunoblastic T-cell lymphoma (AITL) is a peripheral T-cell lymphoma characterized by systemic disease, a polymorphous infiltrate involving the lymph nodes, and the prominent proliferation of high endothelial venules and follicular dendritic cells (FDCs). Extranodal involvement is observed in many patients at the time of diagnosis, with frequently involved sites that include the bone marrow (BM), spleen, skin, and lungs. AITL typically manifests with systemic illness, characterized by B symptoms and generalized lymphadenopathy, often mimicking an infectious process. At diagnosis, most patients have hepatosplenomegaly and skin rash. Laboratory investigations often show anemia and/or other cytopenias, hypergammaglobulinemia, elevated lactate dehydrogenase (LDH) levels, and circulating autoantibodies. 1 Immunohistochemical staining can aid in the diagnosis of AITL. Tumor cells can be recognized by their aberrant expression of CD10, a cell surface metallopeptidase normally expressed by a subset of lymphocytes, and by epithelial and stromal cells. 1-5 However, CD10 expression is also normally found on mature granulocytes, thus making its interpretation difficult in BM biopsy specimens. FDC proliferation is best noted by immunostaining for FDC markers, including CD21, CD23, and CD35. 1 FDCs are conspicuous, usually surrounding the high endothelial venules in lymph node biopsy specimens. Recent studies showed that expression of the follicular chemokine CXCL13 is strongly associated with AITL. 6,7 Neoplastic cells of most AITL cases also expressed bcl-6, a transcription factor that is preferentially expressed by follicular helper T cells. 7 In patients with AITL, BM biopsy is performed routinely to determine the clinical stage of malignant lymphoma. BM 640 Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5

2 Hematopathology / Original Article biopsy, however, may precede lymph node biopsy when malignant lymphoma is not included in the differential diagnosis, or BM biopsy findings may be interpreted before information is available on the final diagnosis from the nodal biopsy specimen. BM involvement is observed in 60% to 80% of patients with AITL. To our knowledge, however, there have been only a few reports describing the clinicopathologic features of BM involvement in AITL. 6,8-11 We, therefore, comprehensively reviewed the clinical and laboratory data and BM histomorphologic features of AITL cases. We assessed distinct features in patients with BM involvement, which may be useful for identifying disseminated AITL. Materials and Methods A search of the medical database of Asan Medical Center, Seoul, Korea, identified 39 patients with AITL from January 1990 to January The diagnosis of AITL was confirmed by lymph node biopsy. The data for 5 patients were excluded because BM biopsy was not performed and for 1 because of the absence of a paraffin-embedded BM block. Slides of BM aspiration and biopsy were independently reviewed by 2 hematopathologists (Y.C. and H.C.), and a rereview was performed to reach consensus in cases with discrepant findings. We retrospectively reviewed the medical charts of these patients. One patient was transferred from another hospital, owing to the diagnosis of AITL, and was treated with several doses of chemotherapy for lymphoma at the time of admission to our hospital. The data for that patient, therefore, were excluded from comparative clinical and laboratory analysis. The study protocol was approved by the institutional review board of Asan Medical Center. Paraffin-embedded BM core biopsy specimens were incubated with antibodies to CD2 (clone AB75, Novocastra, Newcastle upon Tyne, England), CD3 (clone F7.2.38, DAKO, Carpinteria, CA), CD5 (clone 4C7, Novocastra), CD20 (clone L26, DAKO), CD10 (clone 56C6, Novocastra), CD21 (clone 2G9, Novocastra), and bcl-6 (clone GI191E/A8, Cell Marque, Rocklin, CA) using an automated system operated according to the manufacturer s instructions (BenchMark XT, Ventana Medical Systems, Tucson, AZ). Epstein-Barr virus encoded RNA in situ hybridization (ISH) was performed on an autostainer (Ventana Medical Systems) using the ISH iview Blue Detection Kit (Ventana Medical Systems) according to the manufacturer s instructions. All statistical analyses were performed with MedCalc, version (MedCalc Software, Mariakerke, Belgium). The χ 2 test or Fisher exact test was used to assess categorical variables, and the Mann-Whitney test was used to assess continuous variables. A P value less than.05 was considered statistically significant. Results BM involvement was observed in 23 (70%) of 33 cases. The median age at diagnosis of the 23 patients with BM involvement was 58.5 years, whereas that of the 10 patients without BM involvement was 56 years. The male/female ratios in the involved and noninvolved groups were 2.8:1 (17:6) and 4:1 (8:2), respectively. Median duration of follow-up was markedly shorter in the group showing BM involvement than in the group without involvement (2.7 and 25.8 months, respectively). Of the 19 patients whose terminal events were known, 12 died, making the overall death rate 63%. At diagnosis, most patients had a history of constitutional symptoms, often associated with generalized lymphadenopathy (29/32 [91%]), splenomegaly (22/32 [69%]), and B symptoms (21/32 [66%]). Other common features included pleural effusion (17/32 [53%]), hepatomegaly (16/32 [50%]), cough or dyspnea (12/32 [38%]), peritoneal effusion (11/32 [34%]), and skin rash (10/32 [31%]). Extranodal involvement except for BM was reported in 7 cases, with the involved sites being the spleen in 3, the lung in 2, pleural fluid in 2, and peritoneal fluid in 1. In 1 case, there was simultaneous involvement of the spleen and pleural fluid. Clinical features observed more frequently in patients with than without BM involvement included fever (P =.014), hepatomegaly (P <.001), splenomegaly (P =.001), and pleural effusion (P =.032). Extranodal involvement was exclusive to patients with BM involvement despite marginal statistical power. Former negative results for lymphoma were revised to include BM involvement by AITL after review in 7 cases. Of note, in 6 cases, BM interpretation preceded final diagnosis of nodal biopsy. Critical misinterpretations included multiple myeloma, chronic idiopathic myelofibrosis, and idiopathic thrombocytopenic purpura in 1 patient each. Other interpretations included hypercellular marrow with reactive plasmacytosis, erythroid hyperplasia (consistent with hemolytic anemia), and normocellular marrow in the remaining 3 patients. Leukocytosis with or without neutrophilia was the most common hematologic abnormality in 16 (50%) of 32 cases. Other common features included anemia (14/32 [44%]), eosinophilia (11/32 [34%]), and thrombocytopenia (10/32 [31%]). Blood smears in some cases showed distinct findings, including rouleaux formation (9/32 [28%]) and circulating atypical cells (9/32 [28%]) such as plasmacytoid lymphocytes or lymphomatous cells. Anemia (P =.003) and circulating atypical cells (P =.049) occurred more frequently in disease involving than not involving the BM. Red cell agglutination was seen exclusively in BM-involved cases (2/22 [9%]), which may reflect the mechanism of autoimmune hemolytic anemia. Common laboratory features included elevated levels of C-reactive protein (20/21 [95%]), β 2 -microglobulin (13/15 [87%]), and LDH (27/32 [84%]). However, biochemical Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5 641

3 Cho et al / AITL With Bo n e Marrow Involvement findings significantly higher in cases with than without BM involvement included hypoalbuminemia (P <.001), hypocalcemia (P =.011), hyponatremia (P =.027), and an elevated LDH level (P =.042). Serum protein electrophoresis revealed polyclonal gammopathy in 16 (84%) of 19 tested cases. In addition, a positive reaction to the direct Coombs test was seen in 9 (75%) of 12 tested cases, and autoimmune hemolytic anemia was found in 7 (58%) of 12 cases that had direct Coombs testing. Secondary BM changes were noted in a significant number of cases, the most prominent being reactive plasmacytosis in 20 (61%) of 33 cases. Plasmacytosis was significantly more frequent in cases with than without BM involvement (P =.047) and was associated with circulating plasmacytoid lymphocytes in this group. Hypercellular marrow was observed in 14 (42%) of 33 cases and was more frequent in cases with than without BM involvement (P =.036). Other BM changes included eosinophilia (12/33 [36%]), granulocytic hyperplasia (11/33 [33%]), myelofibrosis (4/33 [12%]), hemophagocytosis (4/33 [12%]), erythroid hyperplasia (3/33 [9%]), and megakaryocytic hyperplasia (3/33 [9%]). The latter 4 findings were observed exclusively in cases with BM involvement, although the small numbers of these events might dilute their statistical significances. Four cases demonstrating hemophagocytosis were also positive for the Epstein-Barr virus encoded RNA ISH stain. Histomorphologic analysis revealed that nodular infiltration was most frequently detected (13/23 [57%]) zimage 1Az. Focal or interstitial infiltration was predominant in the remaining cases zimage 1Bz and zimage 1Cz. A significant number of cases showed characteristic findings in AITL, including periodic acid Schiff positive intercellular materials (12/23 [52%]), blood vessel proliferation (9/23 [39%]), and the presence of clear cells (8/23 [35.8%]). A B C zimage 1z A, Nodular bone marrow infiltrated by angioimmunoblastic T-cell lymphoma (AITL) showing a polymorphous mixture of atypical lymphoid cells, eosinophils, and histiocytes (H&E, 200). B, Focal infiltration of AITL. Note a few clear cells in the lesion (H&E, 200). C, Interstitial infiltration of AITL in the paratrabecular area. Note the clear cells that were distributed widely in a whole area (H&E, 200). 642 Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5

4 Hematopathology / Original Article Expression of pan T-cell markers such as CD2, CD3, and CD5 was found in all cases zimage 2Az, zimage 2Bz, and zimage 2Cz. In contrast with strong intensity of CD3 and CD5, staining for CD2 was often weak. Of note, almost half (11/23 [48%]) showed mixed infiltration by neoplastic T cells and reactive B cells zimage 2Dz. Aberrant expression of CD10 was observed in 15 (65%) of 23 cases zimage 3z, but few had CD21+ FDCs and only 1 had individual CD21+ cells. After careful comparison with morphologic features on H&E-stained slides and immunostaining with T- and B-cell markers, these CD10+ cells were considered to be neoplastic T cells. Expression of bcl-6 was seen in 14 (61%) of 23 cases zimage 4z. Epstein-Barr virus encoded RNA was detected by ISH in 19 (58%) of nodal or BM samples obtained from 33 patients zimage 5z. Chromosomal aberrations were observed in 5 (24%) of 21 tested cases, including trisomy 3, trisomy 10, t(17;20), and t(2;14), alone or in association. The main features in our series are summarized in ztable 1z. Discussion AITL is diagnosed primarily by histologic examination of lymph node biopsy specimens and requires the assessment of architectural, cytomorphologic, and immunophenotypic features. However, morphologic features are rather nonspecific and are shared by other types of lymphomas and A B C D zimage 2z A, An immunostain for CD2. Note weak intensity in the T-cell population ( 400). B, C, and D, Same region as in A. B, Strong CD3 expression in the T-cell population ( 400). C, Strong CD5 expression in the neoplastic T cells ( 400). D, Scattered B cells (CD20+) in the infiltrate ( 400). Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5 643

5 Cho et al / AITL With Bo n e Marrow Involvement by various reactive conditions. 3 Because patients with AITL often have systemic manifestations and hematologic abnormalities, BM biopsy may precede the lymph node biopsy. In a recent review of 77 patients with AITL, for 28 patients (36%), the disease first diagnosed was not AITL but reactive lymphadenopathy or another subtype of lymphoma. 11 The diagnosis of AITL was established secondarily by reviewing the initial node biopsy because of clinical or laboratory atypical features or at relapse. In our review, 7 cases were initially misdiagnosed. Therefore, better knowledge of clinicopathologic features in patients with AITL is essential for early and accurate diagnosis. In our series of 33 patients with AITL, the frequency of BM involvement was 70%, a finding comparable with the 60% to 80% reported previously. 3,8-11 The distinct clinical and laboratory features of patients with BM involvement included hepatosplenomegaly, fever, pleural effusion, anemia, circulating atypical cells, hypoalbuminemia, an elevated LDH level, hypocalcemia, and hyponatremia. These features indicate that systemic manifestations are more frequent in patients with than without BM involvement. Extranodal involvement other than BM was noted exclusively in patients with BM involvement, again indicating the systemic nature of AITL. A recent study found that the most frequent secondary BM changes included hyperplasia of the hematopoietic precursors and polyclonal plasmacytosis. 6 Our results showed that hypercellular marrow and plasmacytosis were significant features in cases with BM involvement. These changes may be caused by elevated systemic levels of cytokines expressed in AITL. For example, interleukin 6, which was expressed by lymphoma cells in AITL cases, has been zimage 3z An immunostain for CD10 demonstrating a focal infiltrate of lymphoma ( 400). found to facilitate the proliferation of B cells, plasma cells, and T cells or endothelial cells in the lymph nodes of AITL cases. 12 If the level of interleukin 6 were greater in cases with than without BM involvement, it could lead to a higher degree of systemic illness and more prominent secondary BM changes. Two recent case reports highlight the notion that AITL could manifest with systemic plasmacytosis of the BM and peripheral blood, mimicking multiple myeloma. 13,14 Regardless of the underlying mechanisms, however, it is critical to recognize the occurrence of these reactive changes zimage 4z An immunostain for bcl-6 highlighting prominent nuclear positivity of neoplastic T cells ( 400). zimage 5z Nuclear positivity was evident in many cells within the infiltrate by Epstein-Barr virus encoded RNA in situ hybridization ( 200). 644 Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5

6 Hematopathology / Original Article in AITL to avoid misinterpretation as chronic myeloproliferative disease or multiple myeloma. 6 A few previous reports have described pathologic findings in AITL with BM involvement. Among the noteworthy findings of BM involvement in AITL are vascular proliferation, paratrabecular location, and myelofibrosis Perivascular collections of atypical medium to large lymphoid cells with clear or pale cytoplasm, called clear cells are another characteristic finding in lymph nodes of patients with AITL. 1 In our study, vascular proliferation and paratrabecular location were not as frequent as expected. Myelofibrosis was detected in a small number of cases with BM involvement. Most fibrotic lesions were minimal, except in 1 case, in which florid myelofibrosis and coarse bundles of collagen identified by Masson-trichrome stain led to a misdiagnosis of chronic idiopathic myelofibrosis. Clear cells were found in fewer than half of the cases with BM involvement. In a previous study describing the histologic features of extranodal biopsy specimens in AITL cases, these sites showed a similar spectrum of histologic changes, ranging from those that could be mistaken for reactive lymphoid proliferation to features suggestive of lymphoma, but not specifically AITL. 3 Thus, it is important to recognize features specific for BM involvement in AITL and to perform immunohistochemical assays to identify the nature of the lymphomatous lesions. As expected, the neoplastic T cells expressed all pan Tcell antigens in our study. However, CD3 and CD5 were more useful markers than CD2 because the involved and the uninvolved areas were easily distinguished by the strong intensity of immunostaining. Aberrant loss of 1 or more pan T-cell antigens is a feature generally encountered in T-cell neoplasms. For example, peripheral T-cell lymphoma, not otherwise specified, is usually characterized by an aberrant T-cell phenotype with frequent down-regulation of CD5 and CD7. In a recent study of AITL cases, subsets of neoplastic cells had loss of or weak surface expression of CD3. 5 Therefore, we emphasize the critical role of immunostaining for CD5 rather than CD3 in the detection of lymphomatous BM lesions ztable 1z Summary of Main Clinical, Laboratory, and Histomorphologic Features for Patients With Angioimmunoblastic T-Cell Lymphoma * Bone Marrow Involvement All Patients (n = 33) Negative (n = 10) Positive (n = 23) P Clinical features Fever 21 (66) 3 (30) 18 (82).014 Hepatomegaly 16 (50) 0 (0) 16 (73) <.001 Splenomegaly 22 (69 2 (20) 20 (91) <.001 Pleural effusion 17 (53) 2 (20) 15 (68).032 Extranodal involvement 7 (22) 0 (0) 7 (32).069 Laboratory features Anemia 14 (44) 0 (0) 14 (64).003 Circulating atypical cells 9 (28) 0 (0) 9 (41).049 Hypoalbuminemia 19 (59) 1 (10) 18 (82) <.001 LDH level above normal 27 (84) 6 (60) 21 (95).042 Hypocalcemia 12 (38) 0 (0) 12 (55).011 Hyponatremia 14 (44) 1 (10) 13 (59).027 Positive direct Coombs test 9/12 (75) ND 9/12 (75) Autoimmune hemolytic anemia 7/12 (58) ND 7/12 (58) Secondary bone marrow changes Hypercellular marrow 14 (42) 1 (10) 13 (57).036 Plasmacytosis 20 (61) 3 (30) 17 (74).047 Eosinophilia 12 (36) 1 (10) 11 (48).093 Morphologic features of bone marrow involvement Infiltration pattern No. of nodular/focal/interstitial 13/3/7 Paratrabecular 7 (30) Immunohistochemical stains CD2+ 23 (100) CD3+ 23 (100) CD5+ 23 (100) CD (65) CD21+ 1 (4) bcl (61) LDH, lactate dehydrogenase; ND, not done. * Data are given as number (percentage) or unless otherwise indicated. The clinical and laboratory features are for 32 patients, except for the direct Coombs test and autoimmune hemolytic anemia, which are for 12 patients. The clinical and laboratory features are for 22 patients, except for the direct Coombs test and autoimmune hemolytic anemia, which are for 12 patients. Many patients had complex infiltration patterns, including interstitial and small nodular. The classification was primarily based on the predominant infiltration pattern. Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5 645

7 Cho et al / AITL With Bo n e Marrow Involvement of AITL, although the intensity of CD3 expression was not attenuated significantly in our study. In addition, bcl-6 could be a surrogate marker for detecting BM involvement in AITL because bcl-6 expression was found in 61% of the cases in our study. This is in line with recent findings in which neoplastic cells of tissue sections expressed bcl-6 in 76% of AITL cases. 7 Although neoplastic T cells can be identified by the aberrant expression of CD10 in lymph nodes of most AITL cases, 2 and CD10 expression was observed in a fraction of peripheral T cells in AITL cases, 4 CD10 expression in BM biopsy specimens was not as prominent as in nodal biopsy specimens. 3,11 In our study, CD10 was expressed in 65% of cases, with findings ranging from scattered distribution to localized positivity. A BM biopsy specimen of a Japanese patient with AITL and plasmacytosis also showed clusters of CD10+ cells, supporting the usefulness of CD10 immunostaining in BM interpretation in AITL. 13 In contrast, CD21+ cells were present in a scattered pattern in only 1 case. In another study, the degree of CD21+ FDC meshwork at extranodal sites showed a spectrum, from milder expansion than in nodal biopsy specimens to no evidence of FDC meshwork in BM biopsy specimens. 3 Therefore, CD10 is a more useful marker than CD21 for assessing BM findings in AITL. One of the characteristic findings of AITL was the coexistence of neoplastic T cells and admixed B cells in the lymph nodes. 1 We noted that significant numbers of BM specimens contained small to large B cells, as assessed by CD20 immunostaining. These findings can lead to misinterpretation of BM infiltrates as benign lymphoid aggregates, which are not uncommon in older populations, or as T cell rich large B-cell lymphomas, a variant of diffuse large B-cell lymphoma characterized by fewer than 10% large neoplastic B cells. 6 This underlines once again the need for nodal biopsy at the time of BM interpretation and the requirement of familiarity with the distinct features of BM involvement by AITL. We have identified the distinct clinical and laboratory features of AITL with BM involvement. Although BM is commonly involved in AITL, its histomorphologic features differ from those of nodal biopsy specimens and may reflect various secondary changes. Misinterpretation of BM biopsy specimens may occur for various reasons, including frequent hematologic abnormalities and secondary BM changes. It is important to recognize the histomorphologic and immunophenotypic features of BM specimens and the associated clinical and laboratory features for the accurate diagnosis of AITL. Address reprint requests to Dr Chi: Dept of Laboratory Medicine, University of Ulsan, College of Medicine and Asan Medical Center, Poongnap-dong, Songpa-gu, , Seoul, Korea. References 1. Dogan A, Attygalle AD, Kyriakou C. Angioimmunoblastic T-cell lymphoma. Br J Haematol. 2003;121: Attygalle A, Al-Jehani R, Diss TC, et al. Neoplastic T cells in angioimmunoblastic T-cell lymphoma express CD10. Blood. 2002;99: Attygalle AD, Diss TC, Munson P, et al. CD10 expression in extranodal dissemination of angioimmunoblastic T-cell lymphoma. Am J Surg Pathol. 2004;28: Baseggio L, Berger F, Morel D, et al. Identification of circulating CD10 positive T cells in angioimmunoblastic T-cell lymphoma. Leukemia. 2006;20: Merchant SH, Amin MB, Viswanatha DS. Morphologic and immunophenotypic analysis of angioimmunoblastic T-cell lymphoma: emphasis on phenotypic aberrancies for early diagnosis. Am J Clin Pathol. 2006;126: Grogg KL, Morice WG, Macon WR. Spectrum of bone marrow findings in patients with angioimmunoblastic T-cell lymphoma. Br J Haematol. 2007;137: Dupuis J, Boye K, Martin N, et al. Expression of CXCL13 by neoplastic cells in angioimmunoblastic T-cell lymphoma (AITL): a new diagnostic marker providing evidence that AITL derives from follicular helper T cells. Am J Surg Pathol. 2006;30: Pangalis GA, Moran EM, Rappaport H. Blood and bone marrow findings in angioimmunoblastic lymphadenopathy. Blood. 1978;51: Diebold J, Tulliez M, Vercelli-Retta G, et al. Histopathologic aspects of bone marrow in angioimmunoblastic lymphadenopathy. Ann Pathol. 1984;4: Ghani AM, Krause JR. Bone marrow biopsy findings in angioimmunoblastic lymphadenopathy. Br J Haematol. 1985;61: Lachenal FL, Berger F, Ghesquieres H, et al. Angioimmunoblastic T-cell lymphoma: clinical and laboratory features at diagnosis in 77 patients. Medicine. 2007;86: Yamaguchi S, Kitagawa M, Inoue M, et al. Cell dynamics and expression of tumor necrosis factor (TNF)-alpha, interleukin-6, and TNF receptors in angioimmunoblastic lymphadenopathytype T cell lymphoma. Exp Mol Pathol. 2000;68: Sakai H, Tanaka H, Katsurada T, et al. Angioimmunoblastic T-cell lymphoma initially presenting with replacement of bone marrow and peripheral plasmacytosis. Intern Med. 2007;46: Yamane A, Awaya N, Shimizu T, et al. Angioimmunoblastic T-cell lymphoma with polyclonal proliferation of plasma cells in peripheral blood and marrow. Acta Haematol. 2007;117: From the Departments of 1 Laboratory Medicine, Eulji University School of Medicine, Eulji General Hospital, Seoul; and 2 Laboratory Medicine and 3 Pathology, University of Ulsan, College of Medicine and Asan Medical Center, Seoul, Korea. 646 Am J Clin Pathol 2009;131: DOI: /AJCPQXKCHQH4VAJ5