Correlation with Bone Marrow B Cells and Plasma Cells

IWWM 2012 Proceedings

Immunoglobulin M Concentration in Waldenström Macroglobulinemia: Correlation With Bone Marrow B Cells and Plasma Cells

Ruth M. de Tute, Andy C. Rawstron, Roger G. Owen

Abstract Patients and Methods Serum immunoglobulin (Ig) M levels vary considerably among Eighty-seven patients (49 men, 38 women) fulfilling consensus patients with Waldenström macroglobulinemia, and previous stud- panel and World Health Organization diagnostic criteria1,2 for WM ies have failed to demonstrate a correlation with overall bone marrow were included in this analysis. The median age was 75 years (range, disease burden. In this study, bone marrow B cells and plasma cells 37-91 years) and the median IgM (M component concentration were enumerated by flow cytometry and correlated with serum IgM determined by densitometry) was 11 g/L (range, 1-51 g/L). All the concentrations. Monotypic B cells comprised a median of 6% of patients were assessed at initial presentation and were untreated. ϭ bone marrow leukocytes but did not correlate with IgM levels (r Multiparameter flow cytometry was performed in a single suprare- ϭ 0.071, P .5). Plasma cells, although typically present in lower gional hematopathology laboratory (HMDS Laboratory, Leeds, ϭ numbers (median, 0.52%) did show a correlation with IgM (r UK), and BM samples were received from 25 individual clinical 0.452, P ϭ .01). IgM levels in Waldenström macroglobulinemia, at hematology centers. IgM concentrations were determined in local least in part, correlate with the degree of plasma cell differentiation laboratories. Leukocytes were prepared from fresh BM aspirate sam- seen within the tumor. ples by incubation in a 10-fold excess of ammonium chloride (8.6 Introduction g/L in distilled water; Vickers Laboratories, Pudsey, UK) for 5 minutes at 37°C. Cells were then washed twice in buffer (FACSFlow, Waldenström macroglobulinemia (WM) is a B-cell lymphoprolif- which contained 0.3% bovine serum albumin; Sigma-Aldrich, Dor- erative disorder characterized by immunoglobulin (Ig) M monoclo- set, UK). nal gammopathy and bone marrow (BM) infiltration by lymphop- 6 1,2 Leukocytes (approximately 1 ϫ 10 ) were stained with 30 ␮L lasmacytic lymphoma. It is well recognized that there is considerable clinical heterogeneity, and, in particular, IgM concen- of a preprepared cocktail of antibodies. For plasma cell enumer- trations can vary widely from patient to patient. Previous studies ation, this comprised the following: CD27 fluorescein isothiocya- failed to demonstrate a correlation between IgM concentration and nate (FITC), CD56 phycoerythrin (PE), CD19 peridin chloro- overall BM disease burden.3,4 From a pathologic perspective, it is phyll protein-cyanine 5.5 (PerCP-Cy5.5), CD38 phycoerythrin- known that lymphoplasmacytic lymphoma shows a spectrum of cyanine (PE-Cy7), CD45 allophycocyanin-cyanine (APC-H7) morphologies from mature lymphocytes to plasma cells and that (all from Becton Dickinson, Oxford, UK), and CD138 allophy- there can be considerable interpatient variability in this cellular cocyanin (APC) (Miltenyi Biotec, Bisley, UK). The panel used for content.1-3,5,6 B-cell enumeration comprised Lambda FITC, Kappa PE, CD19 In this study, we chose to explore the relationship between BM PerCP-Cy5.5, CD5 PE-Cy7, CD20 APC, and CD45 APC-H7 cellular heterogeneity and IgM concentration by accurately enumer- (all from Becton Dickinson). After incubation in the dark at 4°C ating both B-cell and plasma cell components with multiparameter for 20 minutes, the cells were washed twice in buffer and resus- flow cytometry and correlating this with serum IgM concentrations. pended in FACSFlow. Samples were acquired on a FACSCanto II analyzer by using FACSDiva software (BD Biosciences, Oxford, UK), and a minimum of 100,000 events were acquired for each HMDS Laboratory, Leeds Teaching Hospitals NHS Trust, Leeds, UK screening tube. The Pearson correlation coefficient was used to Address for correspondence: Roger G. Owen, MD, HMDS Laboratory, Level 3, investigate the relationship between total BM B-cell and plasma Bexley Wing’s University Hospital, Beckett Street, Leeds LS9 7TF, UK cell numbers (expressed as a percentage of BM leukocytes) and Fax: ϩ44 113 206 7883; e-mail contact: [email protected]. serum IgM concentration.

2152-2650/$ - see frontmatter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clml.2013.02.018 Clinical Lymphoma, Myeloma & Leukemia April 2013 211 IgM Concentration and BM Plasma Cells in WM

Figure 1 Correlation of Bone Marrow B Cells and Plasma Cells With Serum Immunoglobulin (Ig) M Concentration in Waldenström B) Whereas Some) ;(5. ؍ P ,0.071 ؍ Macroglobulinemia. (A) B-cell Numbers Do Not Appear to Correlate With IgM (r (01. ؍ P ,0.452 ؍ Correlation Is Noted With Bone Marrow Plasma Cell Numbers (r

A 80 70

30 B-cell percentage 20

0 010203040 50 60 IgM level

B 9 8

3 Plasma cell percentage 2

0 010203040 50 60 IgM level

Results and Discussion levels, comprising a median 0.52% of BM leukocytes (range, 0.05%- WM is a heterogeneous disorder in which there is signiﬁcant vari- 8.4%). In contrast to B cells, plasma cell numbers did show some ation in serum IgM levels and BM cellular content. Some previous correlation with IgM concentration (r ϭ 0.452, P ϭ .01) (Figure studies failed to demonstrate a correlation between serum IgM levels 1B). and overall BM cellular burden.3,4 Similarly, prognostic factor anal- These data would support the hypothesis that IgM levels in WM yses have failed to demonstrate a reproducible effect of IgM concen- are, at least in part, related to the degree of plasma cell differentiation tration on overall outcome except in the context of very high levels within the tumor population. This is further supported by 2 recent when additional factors, eg, cardiovascular may contribute to poor publications. Pasricha et al9 estimated BM B-cell and plasma-cell overall survival.7,8 Given this cellular heterogeneity, we chose to eval- inﬁltration by using CD20 and CD138 immunostaining on tre- uate the relationship between both BM B cells and plasma cells, and phine biopsy sections, and found that IgM levels correlated with the serum IgM on the assumption that the plasma cells would be respon- plasma cell burden but not with the B-cell burden. Similarly, Kyrt- ϩ sible for the majority of IgM secretion. Monotypic B cells were de- sonis et al10 found that the overall level of CD138 expression, and, monstrable in all patients and comprised a median of 6% of BM by implication, the proportion of plasma cells as assessed by immu- leukocytes (range, 0.19%-67%) but did not appear to correlate with nohistochemistry also correlated with IgM levels. serum IgM concentration (r ϭ 0.071, P ϭ .5) (Figure 1A). Plasma It is interesting to that note that plasma cells are found in relatively cells were demonstrable in all cases but were typically present at lower small numbers in WM but still appear to correlate with IgM levels. In

212 Clinical Lymphoma, Myeloma & Leukemia April 2013 Ruth M. de Tute et al

the current study, we demonstrated a median of 0.52% plasma cells, BM disease burden. Appreciation of the plasma cell component in which is almost identical to that reported by Morice et al6 (median, WM is becoming increasingly relevant in the routine clinical man- 0.7%) who used similar flow cytometric methodology. These levels agement of patients and may indeed be a legitimate therapeutic target are consistently lower than that reported for CD138 staining of tre- in certain clinical scenarios. phine biopsy sections (median, 5%-10%),6,9 and there are likely to be a number of methodologic factors to account for this apparent Disclosure discrepancy. It is well established that flow cytometry consistently The authors have stated that they have no conflicts of interest. underestimates the overall plasma cell burden in myeloma, princi- pally as a consequence of sample dilution, but, despite this, it appears References to be a more relevant prognostic factor than standard morphologic 1. Owen RG, Treon SP, Al-Katib A, et al. Clinicopathological definition of Walden- 11,12 assessment. Plasma cell distribution may also be a factor in WM strom’s macroglobulinemia: consensus panel recommendations from the Second because clonal plasma cells have been reported distinct from areas of International Workshop on Waldenstrom’s Macroglobulinemia. Semin Oncol 6 2003; 30:110-5. lymphoid infiltration by some investigators. 2. Swerdlow SH, Campo E, Harris NL, et al. World Health Organization classification There now are some emerging data regarding the clinical signifi- of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC; 2008. 3. Bartl R, Frisch B, Mahl G, et al. Bone marrow histology in Waldenstrom’s macro- cance of the heterogeneity of cellular content in WM. It is clear that globulinaemia. Clinical relevance of subtype recognition. Scand J Haematol 1983; Ϫ ϩ monotypic plasma cells, typically CD20 and CD138 are demon- 31:359-75. strable in virtually all patients with WM, but they usually are present 4. Treon SP. How I treat Waldenström macroglobulinemia. Blood 2009; 114:2375-85. 5. Remstein ED, Hanson CA, Kyle RA, et al. Despite apparent morphologic and 5,6,9 in lower numbers than monotypic B cells. Plasma cells, however, immunophenotypic heterogeneity, Waldenstrom’s macroglobulinemia is consis- may be the dominant cell type in some patients, and, in this setting, tently composed of cells along a morphologic continuum of small lymphocytes, plasmacytoid lymphocytes and plasma cells. Semin Oncol 2003; 30:182-6. the rare entity of IgM myeloma becomes part of the differential 6. Morice WG, Chen D, Kurtin PJ, et al. Novel immunophenotypic features of mar- diagnosis.6,9 The latter, however, is readily distinguished from WM row lymphoplasmacytic lymphoma and correlation with Waldenstrom’s macroglobulinemia. Mod Pathol 2009; 22:807-16. on the basis of an aberrant plasma cell phenotype (plasma cells are 7. Morel P, Duhamel A, Gobbi P, et al. International prognostic scoring system for ϩ typically CD19– in IgM myeloma but CD19 in WM) and the Waldenström macroglobulinemia. Blood 2009; 113:4163-70. presence of IGH translocations and t(11;14) in particular.6,13 Recent 8. Morel P, Merlini G. Risk stratification in Waldenström macroglobulinemia. Expert ϩ Rev Hematol 2012; 5:187-99. studies have also demonstrated the significance of the CD138 9. Pasricha SR, Juneja SK, Westerman DA, et al. Bone-marrow plasma cell burden plasma cell component in the context of response assessment because correlates with IgM paraprotein concentration in Waldenström macroglobulinemia. J Clin Pathol 2011; 64:520-3. it appears that some therapies, particularly monoclonal antibody and 10. Kyrtsonis MC, Levidou G, Korkolopoulou P, et al. CD138 expression helps distin- purine analogue–based therapies appear to selectively deplete the guish Waldenstrom’s macroglobulinemia (WM) from splenic marginal zone lymphoma. Clin Lymphoma Myeloma Leuk 2011; 11:99-102. B-cell component but spare the plasma cell component, which can 11. Rawstron AC, Orfao A, Beksac M, et al. Report of the European myeloma network result in delayed IgM responses.14-16 Furthermore, it also has been on multi-parametric flow cytometry in multiple myeloma and related disorders. noted that the cellular composition seen in some patients can vary Haematologica 2008; 93:431-8. 12. Paiva B, Vidriales MB, Pérez JJ, et al. Multi-parameter flow cytometry quantifica- over time because plasma cell–rich lesions that mimic plasmacytoma tion of bone marrow plasma cells at diagnosis provides more prognostic information have been described at progression with some patients.17,18 than morphological assessment in myeloma patients. Haematologica 2009; 94: 1599-602. Therapeutic strategies in WM have traditionally targeted the mo- 13. Feyler S, O’Connor SJ, Rawstron AC, et al. IgM myeloma: a rare entity character- notypic B-cell component of the disease, but it is clear that antigenic ized by a CD20-CD56-CD117- immuno-phenotype and the t(11;14). Br J Haema- tol 2008; 140:547-51. targets for monoclonal antibodies, such as CD20 and CD52, are 14. Owen RG, Hillmen P, Rawstron AC. CD52 expression in Waldenstrom’s macro- expressed at significantly lower levels in plasma cells.14 It may be globulinemia: implications for alemtuzumab therapy and response assessment. Clin prudent to consider plasma cell–directed therapies in certain clinical Lymphoma 2005; 5:278-81. 15. Varghese AM, Rawstron AC, Ashcroft AJ, et al. Assessment of bone marrow re- scenarios, such as the hyperviscosity syndrome in which rapid disease sponse in Waldenstrom’s macroglobulinemia. Clin Lymphoma Myeloma 2009; control is required, or IgM-related syndromes and amyloidosis in 9:53-5. 16. Barakat FH, Medeiros LJ, Wei EX, et al. Residual monotypic plasma cells in pa- which the clinical sequelae are a consequence of the M protein rather tients with Waldenström macroglobulinemia after therapy. Am J Clin Pathol 2011; than tumor bulk per se. In this context, it is clear that bortezomib- 135:365-73. 17. Bashir Q, Lee CK, Stuart RW, et al. Phenotypic evolution of Waldenstrom’s mac- based therapies can result in rapid IgM responses, whereas plasma roglobulinemia to extramedullary plasmacytoma. J Clin Oncol 2008; 26:2408-10. cell–specific monoclonal antibodies, eg, elotuzumab, may also have a 18. Owen RG, Bynoe AG, Varghese A, et al. Heterogeneity of histological transforma- role in these clinical situations.19,20 tion events in Waldenstrom’s macroglobulinemia (WM) and related disorders. Clin Lymphoma Myeloma Leuk 2011; 11:176-9. 19. Treon SP, Ioakimidis L, Soumerai JD, et al. Primary therapy of Waldenström Conclusion macroglobulinemia with bortezomib, dexamethasone, and rituximab: WMCTG clinical trial 05-180. J Clin Oncol 2009; 27:3830-5. We would conclude that IgM levels in WM appear to correlate 20. Hsi ED, Steinle R, Balasa B, et al. CS1, a potential new therapeutic antibody target more with the degree of plasma cell differentiation than the overall for the treatment of multiple myeloma. Clin Cancer Res 2008; 14:2775-84.

Clinical Lymphoma, Myeloma & Leukemia April 2013 213