L Richter

Vascular endothelial growth factor (VEGF) plays an important role in angiogenesis by acting as a potent inducer of vascular permeability as well as serving as a specific endothelial cell mitogen. The importance of angiogenic factors such as VEGF, although clearly established in solid tumors, has not been fully elucidated in human hematopoietic neoplasms. We examined the expression of mRNA and protein for VEGF in 12 human hematopoietic tumor cell lines, representing multiple lineages and diseases, including leukemia, lymphoma, and multiple myeloma. Our results revealed that VEGF message was expressed in these cells and that the corresponding protein was secreted into the extracellular environment. Five of the 12 cell lines were also found to express the Flt-1 receptor for VEGF at a moderate to strong level, suggesting an autocrine pathway. When human vascular endothelial cells were exposed to recombinant human VEGF, there was an increase in the mRNA for several hematopoietic growth factors including macrophage colony-stimulating factor, granulocyte colony-stimulating factor and interleukin 6. Plasma cells in the bone marrow from patients diagnosed with multiple myeloma were found to express VEGF, whereas both the Flt-1 and KDR high affinity VEGF receptors were observed to be markedly elevated in the normal bone marrow myeloid and monocytic cells surrounding the tumor. These data raise the possibility that VEGF may play a role in the growth of hematopoietic neoplasms such as multiple myeloma through either a paracrine or an autocrine mechanism.

Bone marrow samples from 55 patients with multiple myeloma (MM) and 23 patients with monoclonal gammopathy of undertermined significance (MGUS) were evaluated with a broad panel of monoclonal antibodies. Plasma cells from 78% (43/55) of patients with MM strongly expressed the natural killer cell antigen CD56 (NKH-1, Leu-19). Of the 23 patients with MGUS, none showed strong CD56 reactivity, although three had weak reactivity in less than 20% of plasma cells. Myeloma cells expressing CD56 did not coexpress the CD57 or CD16 antigens. Patients with CD56-positive plasma cells had both indolent and aggressive disease. However, the 12 CD56-negative patients had predominantly aggressive disease with an unexpected preponderance of kappa Bence Jones only myeloma (5/10[50%] evaluable patients). Polyclonal plasma cells from non-neoplastic tissue sites (normal bone marrows, lymph nodes, tonsillar biopsies, and gut-mucosa biopsies) showed a near absence of CD56. We conclude that isolated, strong CD56 expression is common in MM, but not in MGUS or reactive plasma cells. The potential biologic importance of CD56 positivity in myeloma is reviewed.

To assess the prognostic significance of the growth fraction in diffuse large cell lymphoma (DLCL), we studied 105 DLCL patients with the monoclonal antibody Ki-67 applied to frozen tissue sections. Ki-67 detects a nuclear antigen associated with cell proliferation not found in resting cells. Ki-67 findings and other clinical prognostic factors were correlated with outcome using univariate and multivariate analyses in the proportional hazards model. High proliferative activity, defined as nuclear Ki-67 expression in greater than 60% of malignant cells (Ki-67 greater than 60), was found to be a strong predictor of poor survival among these patients (P = .003, log-rank). The 19 patients with Ki-67 greater than 60% had a median survival of 8 months compared with a median survival of 39 months for the 86 patients with Ki-67 less than or equal to 60%. Examination of pretreatment clinical variables indicated the patient groups were similar with regard to age, sex, stage, B symptoms, tumor bulk, and lactate dehydrogenase (LDH). Both patient groups received comparable curative intent therapy and showed comparable complete response rate precluding treatment differences as modifying outcome. Multivariate analysis indicated Ki-67 is an independent predictor of survival (multivariate P = .006). Further statistical analysis using only B-cell DLCL patients treated with CHOP (63 patients) indicated that Ki-67 greater than 60 retained strong prediction of poor outcome (P = .002, log-rank) among this homogeneous group. We conclude that high proliferative activity (Ki-67 greater than 60) is an independent factor allowing laboratory prediction of probable poor outcome of DLCL.

Using several multiple drug-resistant human myeloma cell lines as standards, we developed an immunohistochemical staining technique and means of quantitating P-glycoprotein in individual myeloma cells. The level of staining intensity for P-glycoprotein in individual myeloma cells was quantitated by measuring the average optical density of each cell with a microscopic computerized cell analysis system. Using this system, we observed that the level of P-glycoprotein for individual cells within a cell population of known drug sensitivity was very homogeneous (coefficient of variation less than or equal to 13%). Analysis of cell lines with gradually increasing levels of multidrug resistance (8226/S, 8226/Dox6 and 8226/Dox40) demonstrated a close association between the level of resistance to doxorubicin, defined by the mean lethal dose (D0) and the amount of P-glycoprotein on individual cells determined by the optical density (r = 0.82, P less than 0.0005). Intracellular doxorubicin (DOX) accumulation in the individual cell lines was inversely related to the level of drug resistance expressed as D0. P-glycoprotein was also detected in the marrow-derived myeloma cells of patients with drug refractory disease using immunohistochemical staining. The amount of P-glycoprotein in the cells of one patient was directly compared to the amount found in the simultaneously stained standard cell lines (8226/Dox6 and 8226/Dox40) by comparing the optical densities for individual cells. Using this immunohistochemical technique to detect and quantitate P-glycoprotein in patient myeloma cells and comparing it to standard multidrug resistant myeloma cell lines may be of value in determining the contribution of P-glycoprotein to clinical drug resistance in patients with multiple myeloma.

A novel pre-B cell component in direct and cultured myeloma bone marrow material has been delineated by using immunochemistry and flow cytometry techniques. Our phenotypic studies suggest a novel hybrid expression of pre-B and plasma cell antigens with coexpression of cytoplasmic mu, common acute lymphoblastic leukemia antigen, terminal deoxynucleotidyl transferase, and plasma cell antigens (PCA-1 and PC-1). This suggests that myeloma pre-B-like cells are aberrant malignant cells and not normal pre-B lymphocytic counterparts. With the advantage of a pure and stable source of these cells from M3 culture to allow molecular characterization, we performed one- and two-dimensional gel electrophoresis and Western blotting. We found that the cytoplasmic mu in myeloma pre-B-like cells has a molecular weight of 74,000 daltons and an isoelectric point of 6.3 and that it is strikingly homogeneous and discrete in size and charge compared with standard secretory mu, which suggests an aberrant, mutant, or monoclonal form of mu. Monoclonality was further evidenced by heavy- and light-chain immunoglobulin gene rearrangements demonstrated with JH and C kappa probes. We also established that this novel myeloma pre-B component is a major proliferative element as determined by double-labeling experiments with phenotype coupled to labeling/proliferative indexes. Our stimulatory studies indicate some capacity of these cells to mature on exposure to phorbol esters. These myeloma pre-B cells may represent the stem cell or self-renewal component in myeloma. Our establishment of these cells in long-term culture offers a considerable asset in studying the immature cells, which may be critical to the immortalization of myeloma.