New Immunophenotypic Profile of Waldenstrom's Macroglobulinemia: Interest of CD80 and CD86 Staining

Abstract

The immunophenotypic characterization is an essential tool in the diagnosis of hematological malignancies but the immunophenotypic features in Waldenstrom's macroglobulinemia (WM) remain not clearly defined. We studied 96 cases of WM diagnosed by monoclonal IgM in the serum and morphological lymphoplasmacytic bone marrow infiltration, and we compared results to 33 cases of other chronic B-cell lymphoproliferative disorders (LPD), including marginal zone (MZL)(n=23), mantle cell (MCL)(n=8) and follicular (FL)(n=2) lymphomas. Patients with a Matutes score >3 (chronic lymphocytic leukemia) and with pathognomonic immunophenotype (hairy cell leukemia) were excluded. Immunophenotypic analysis was performed by flow cytometry using six-colour staining (FACS Canto II, Becton Dickinson). In WM and LPD groups, a monoclonal B-cell population was identified in blood (31 and 28 patients, respectively), blood and bone marrow (28 and 4 patients) or bone marrow samples (23 and 1 patients). Overall, 61% of WM patients showed a monoclonal B-cell population in blood. Neoplastic cells of WM and LPD patients with blood and/or bone marrow involvement expressed a monoclonal immunoglobulin light chain kappa (in 70% and 73% of cases respectively) or lambda (30% and 27%). The intensity of expression of the light chain was heterogeneous in both groups (high, normal or low expression in 43%, 27% or 30% of WM, and in 52%, 33% or 15% of LPD, respectively). All pan-B antigens (CD20, CD19, CD79b) were positive for at least 97% of patients. Results obtained with other antigens in WM compared to LPD were: CD10 = 10% vs 7% of patients, CD23 = 33% vs 56%, CD5 = 14% vs 26%, FMC7 = 76% vs 89%, CD38 = 56% vs 41%, CD25 = 86% vs 84%, CD43 = 12% vs 16%, and CD11c = 10% vs 36%. The intensity of expression of these antigens was heterogeneous in both groups. Among the antigens only tested in the WM group, CD1c and CD27 were positive for 70% of patients, IgM and IgD for 95% of patients, and CD103 as well as CD117 were negative in all cases. No difference was found between blood and bone marrow for all previous antigens. Plasma cells (CD38/CD138 positive cells) were found at low levels (less than 2.5% of B-cells) for 46% of WM in blood and/or bone marrow samples. Among the 10 WM patients tested for ZAP-70 expression, 9 were negative and 1 showed a low intensity expression. These results confirm that the immunophenotypic analysis usually performed with standard antigens does not allow defining a typical profile of WM. In order to tentatively identify the WM among the B-cell malignancies, we studied the expression of molecules known to be involved in B-cell development or in costimulatory pathways of antigenic activation, namely CD69, CD83, CD80 and CD86. We first analyzed blood samples of 24 WM patients showing a peripheral monoclonal B-cell population. CD80 was positive (> 20% of B-cells) in all cases and CD83, CD69 and CD86 were always negative. Among these WM patients, 13 were also studied for the bone marrow phenotype. No difference was found between blood and bone marrow phenotype in 11/13 WM cases. We then studied 11 LPD with blood tumoral involvement (MZL(n=7), MCL(n=2) and FL(n=2)). In these LPD, CD69 and CD83 were always negative and, in most cases (9/11 patients), CD80 and CD86 were also negative. Interestingly, CD80 was found positive in 2 patients with MZL, but the CD80 positivity was always associated to the CD86 positivity. Altogether, these data suggest that the inclusion of CD80 and CD86 in the panel of cytometric analysis allow to discriminate WM from other B-LPD with peripheral blood involvement.