Faculty Bibliography

We developed a simple and rapid method to enrich tumor cells within bone marrow (BM) aspirates from patients with multiple myeloma (MM). Thirty patients with a median of 50% (8-85%) MM cells by morphology and 55% (6--85%) MM cells identified by CD38+CD45-cell surface phenotype were studied. BM mononuclear cells (BMMCs) were isolated by Ficoll Hypaque sedimentation and incubated with a cocktail of mouse monoclonal antibodies (mAbs) directed against CD3 (T cells); CD11b and CD14 (monocytes); CD33 (myeloid cells), CD45 and CD45RA (leucocyte common antigen); CD32 as well as glycophorin A. After the addition of anti-mouse Fc Ig-coated immunomagnetic beads, mAb-bound cells were removed in a magnetic field. The residual cell populations were enriched for MM cells, evidenced by >95% plasma cell morphology and >95% CD38+CD45RA-cell surface phenotype. Since this method requires only two short incubations, cell losses were minimal and the yield of MM cells was therefore high (>95%). Viability of the MM-cell enriched fractions was 99%, and these cells were functional in assays of proliferation, cell cycle analysis and immunoglobulin secretion. This immunomagnetic bead depletion method therefore permits the ready isolation of homogeneous populations of patient MM cells for use in both cellular and molecular studies.

Studies utilizing flow cytometry and PCR have shown that the B-cell compartment in myeloma contains cells which are clonally related to the myelomatous plasma cells. Current data, however, remains inconclusive regarding the extent of this involvement. By combining fluorescent immunophenotyping, tyramine signal amplification and fluorescence in-situ hybridization (FICTION-TSA), we have used the presence of numerical chromosomal abnormalities within plasma cells as a clonal marker to examine the CD20+ B-cell compartment for the presence of aneuploidy. A series of 54 cases of myeloma were screened for the presence of numerical abnormalities of chromosomes 3 and 11. FICTION-TSA was performed on 13 cases with either trisomy 3 or 11 and on a control group of six cases known to be disomic for the two chromosomes. B-cell numbers were reduced in the myeloma cases compared to the normal controls (median 1.8% v 3.0%, P = 0.05). In the cases with a chromosomal marker, three signals were seen in a median of 1.88% of CD20+ B cells compared to 2.58% within the control group. Comparison of the two groups using a Wilcoxon-Mann-Whitney U test showed no statistical significant difference. Using this data set, it was possible to exclude a 3.03% expansion of clonally related B cells (95% confidence level). We conclude that the B-cell compartment in myeloma does not represent the major site of clonal expansion, and if clonally related cells are present then the numbers are few.

There is a wide variation in the degree of marrow and blood involvement between patients with multiple myeloma. Both of these parameters are known to be highly significant prognostic factors, and the differences between patients may be due to variable expression of adhesion molecules. To test this we used three-colour flow cytometry to study three adhesion molecules associated with myeloma, namely CD38, CD56 and CD138. The level of expression of these molecules was compared with the distribution of myeloma plasma cells in bone marrow (n=59) and peripheral blood (n=26) in patients at presentation or relapse. The extent of marrow infiltration on the trephine biopsy correlated inversely with CD56 expression (Mann-Whitney U Test, P=0.022); there was no difference in CD38 or in CD138 expression. CD56 expression also correlated inversely with the number of circulating plasma cells (linear regression, R2=0.4268, slope=-0.58, P=0.0003). Peripheral blood plasma cell numbers correlated weakly with bone marrow plasmacytosis, and inversely with CD38 expression. The level of CD56 expression by neoplastic plasma cells was assessed in 37 patients over a median of 11 months (range 6-25). There was no significant change in expression (Wilcoxon Signed Rank, P=0.6271). We conclude that plasma cell CD56 expression is constant over the course of the disease; unlike CD138 expression, it is significantly linked to the degree of both bone marrow and peripheral blood involvement.

Liposomal encapsulation of anthracyclines is a potential method of drug targeting, altering both the antitumour activity and side-effect profile of anthracyclines. Liposomal daunorubicin (daunoxome) shows both altered pharmacokinetics and a potential for reducing dose-limiting cardiotoxicity compared to conventional daunorubicin. Anthracyclines have a common role in the treatment of multiple myeloma, a prevalent and fatal haematological malignancy. Avoiding cumulative anthracycline toxicity in these patients is important. There is also a need for more effective relapse schedules given that many patients have chemosensitive disease at relapse. We have analysed daunoxome in vitro in myeloma cell lines using a thymidine-based cytotoxicity assay and show superior efficacy compared to a pegylated liposomal doxorubicin derivative. Subsequently we have treated seven relapsed myeloma patients with a regime consisting of oral CCNU 25-50 mg/m2 on day 1, 4 days of oral dexamethasone 10 mg/m2 and intravenous daunoxome (liposomal daunorubicin) given for 4 days (total 100 mg/m2). The main toxicity was myelosuppression but non-haematological toxicity was minimal and the regime was well tolerated. Four out of seven of these heavily pretreated patients responded. Together with the in vitro data on its cytotoxicity in myeloma and its favourable pharmacokinetic profile further studies of liposomal daunorubicin in myeloma would be warranted.

In order to delineate the specific morphological and immunophenotypic features of B-cell lymphoproliferative disorders associated with trisomy 12, 172 sequential unselected cases of CD19+CD5+ B-cell disorders, primarily affecting the peripheral blood and bone marrow, were studied. Trisomy 12 was found in 24 cases (13.9%), with all cases morphologically classified as either CLL-PL or CLL-mixed by FAB criteria. Trisomy 12 was not found in any cases of typical CLL. Trisomy 12 cases demonstrated a significant higher expression of CD11a (P<0.0001) and CD20 (P<0.0006) when compared to cases with the equivalent morphology and immunophenotype, but without the chromosomal abnormality. Trisomy 12 cases also demonstrated a higher frequency of FMC7, CD38 expression and moderate to strong surface immunoglobulin staining. However, no correlation was detected between the percentages of trisomy 12 cells and cells expressing CD11a, CD38, FMC7 or sIg mean fluorescent intensity. Cells from trisomy 12 positive cases were sorted according to their CD11a expression using fluorescent activated cell sorting. There was a significant increase in the percentage of trisomy 12 cells within the CD11a+ sorted fraction compared to the unsorted population (P < 0.05), implying that trisomy 12 is associated with increased expression of CD11a. With the highly specific morphological and immunophenotypic features demonstrated by trisomy 12 cases in this study, it is highly likely that these cases constitute a specific group of B-cell lymphoproliferative disorders.

The reduced levels of normal immunoglobulin in patients with myeloma may be due to suppression of normal B-cell differentiation. However, reports on the numbers of B cells vary, with some finding decreases consistent with immunoparesis, and others reporting expansions of phenotypically aberrant cells. We have therefore assessed the phenotype and levels of B lymphocytes in patients at presentation (n = 23), in plateau or complete remission (PB n = 42, BM n = 18), and in relapse (PB n = 17, BM n = 14), in comparison to normal individuals (n = 10). Phenotypic analysis was performed using five-parameter flow cytometry, with CD14 used to exclude monocytes where necessary. We found no evidence of a phenotypically distinctive blood or marrow B-cell population in patients with myeloma, nor of an increase in the levels of any B-cell subset. Numbers of blood CD19+ 38+ normal plasma cell precursors were significantly reduced in presentation/relapse patients, but not in patients in plateau/remission. Total CD19+ cells were significantly reduced only in patients with circulating myeloma cells, detected by IgH-PCR. In the marrow, CD19+ B cells expressing CD5, CD10, CD34, CD38, CD45(low) and Syndecan-1 were significantly decreased at presentation/relapse, but not in patients in plateau/remission. The majority of these antigens are expressed by normal B-cell progenitors, indicating that myeloma also affects the early stages of B-cell development. The suppression of progenitor cells was not restricted to B-lymphoid differentiation, as total CD34+ cells were also significantly reduced in the marrow of myeloma patients at presentation. These results indicate that, if neoplastic B cells are present in myeloma, they are low in number and have a phenotype similar to their normal counterparts. Furthermore, there is a reversible suppression of CD19+ B lymphocytes that correlates inversely with disease stage, and specifically affects the early and late stages of normal B-cell differentiation.

The aim of this study was to develop a flow cytometric test to quantitate low levels of circulating myeloma plasma cells, and to determine the relationship of these cells with disease stage. Cells were characterized using five-parameter flow cytometric analysis with a panel of antibodies, and results were evaluated by comparison with fluorescent consensus-primer IgH-PCR. Bone marrow myeloma plasma cells, defined by high CD38 and Syndecan-1 expression, did not express CD10, 23, 30, 34 or 45RO, and demonstrated weak expression of CD37 and CD45. 65% of patients had CD19- 56+ plasma cells, 30% CD19- 56(low), and 5% CD19+ 56+, and these two antigens discriminated myeloma from normal plasma cells, which were all CD19+ 56(low). Peripheral blood myeloma plasma cells had the same composite phenotype, but expressed significantly lower levels of CD56 and Syndecan-1, and were detected in 75% (38/51) of patients at presentation, 92% (11/12) of patients in relapse, and 40% (4/10) of stem cell harvests. Circulating plasma cells were not detectable in patients in CR (n = 9) or normals (n = 10), at a sensitivity of up to 1 in 10,000 cells. There was good correlation between the flow cytometric test and IgH-PCR results: myeloma plasma cells were detectable by flow cytometry in all PCR positive samples, and samples with no detectable myeloma plasma cells were PCR negative. Absolute numbers decreased in patients responding to treatment, remained elevated in patients with refractory disease, and increased in patients undergoing relapse. We conclude that flow cytometry can provide an effective aternative to IgH-PCR that will allow quantitative assessment of low levels of residual disease.

We have developed a competitor-based RT-PCR technique which will detect and quantitate the CBFbeta/MYH11 transcripts associated with inv(16)(q22;p13) and have used it to study presentation and follow-up samples of acute myeloid leukaemia (AML). The levels of the leukaemia-specific transcripts are expressed as a ratio to a ubiquitously expressed mRNA species (Abl) which controls for RNA degradation. This technique has been applied to 75 consecutive patients presenting with either de novo AML or tMDS; 6/75 patients analysed were positive for the inv(16), all were confirmed by conventional cytogenetics. The inv(16) has a strong association with M4Eo, but we found only 2/6-positive patients to have this diagnosis (two patients with M2, one patient M1 and one patient had MDS). At presentation the levels of CBFbeta/MYH11 transcripts were 0.1-10/Abl transcript (mean 3.3/Abl transcript). Seventeen follow-up samples were available on 5/6 of these patients, and on two further patients in whom stored material was available. Following the first cycle of chemotherapy the level of transcripts was at least 10(-2) lower (0.1-10 x 10(-2)/abl transcript) than their presentation sample. Subsequent samples on these patients when in remission gave transcript levels in the range (1.0 x 10(-4) - 2 x 10(-3)/abl transcript), and three long-term follow-up samples were negative. We have developed a quantitative test which opens the possibility of predicting relapse by detecting changes in the numbers of leukaemia-specific transcripts.