The first speaker, Dr. Linda Pilarski of the University of Alberta, presented data generated in her laboratory on the origin of the myeloma cell. Myeloma cells are typically plasma cells, ie; the most differentiated B cells. Her hypothesis is that these malignant plasma cells derive from clonotypic B cells, which would represent the myeloma progenitors. The major questions regarding these B cells are: (1) do they really persist in myeloma patients? and (2) if so, what is their contribution to myeloma disease? These clonotypic B cells are characterized by a molecular rearrangement identical to that of the malignant plasma cells, but they express immunoglobulin (Ig)M rather than IgG or IgA. Dr. Pilarski reported on a clinical study with 29 patients that showed a significant correlation between the presence of clonotypic B cells and reduced patient survival. A correlation between the number of plasma cells within the bone marrow at diagnosis and a higher level of serum b2-microglobulin was observed in this small series. Moreover, these cells are detected all along the disease course, even in patients receiving high-dose therapies. Several hypotheses can explain the persistance of IgM clonotypic cells: (1) a single B or plasma cell may express both pre- and postswitch isotypes ; (2) small populations of pre- and post-switch memory myeloma B cells may circulate in the blood ; and (3) the myeloma progenitor may be a postgerminal center IgM-expressing memory B cell that undergoes persistent, directed isotype switching and terminal differentiation to myeloma plasma cells. In other words, these chemoresistant clonotypic B cells would represent a reservoire for the malignant clone and should be targeted by therapy to stop this process. It was suggested that these IgM transcripts could be produced by the malignant plasma cells themselves, since most malignant plasma cells retain one Cµ locus. These provocative results are still in debate.
The next speaker, Dr. Surinder Sahota, (University of Southamptom, U.K.) focused on molecular rearrangement of the Ig genes, esssentially IgH (encoding the Ig heavy-chain gene). During plasma cell ontogeny, this gene is rearranged through several molecular events. Within the bone marrow, the pro-B cell rearranges first the D (diversity) and J (junction) segments, and then the V (variable) with DJ regions. Analysis of selected VH segments in more than 100 patients revealed a bias in the selection. Whereas the V4-34 gene is selected in 5% to 10% of normal B cells, it has never been observed in myeloma patients. Analysis of the somatic mutation pattern revealed that myeloma cells are invariably mutated, without intraclonal variation, with a pattern providing evidence for antigen selection. However, molecular analyses of tumor cells from hairy-cell leukemia (an ontogenetically closely related malignancy) have shown that these cells are able to form different clonally related transcripts, including IgM transcripts. This situation is currently observed in myeloma, but it may represent another proof against the implication of clonotypic B cell precursors in myeloma pathogenesis.
Dr. Marleen Bakkus (Free University Brussels, Belgium) addressed the same topic (ie, molecular IgH rearrangement) but in a mouse model of multiple myeloma (MM). A specific mouse strain (C57BL/KaLwRij) spontaneously develops, in a small percentage of cases, a disease that closely resembles humanMM. Several clones (the 5T clones) have been generated and then propagated in this mouse strain. Two specific clones, the 5T2 and the 5T33, reflect different forms of the human disease, offering an experimental system for studying the molecular IgH rearrangements. These two clones express monoclonal IgG. Using sensitive PCR assays, researchers have isolated isotype switch variants (ie, IgA or IgM) in the bone marrow of mice grafted with the two clones. Of note, the analysis of 5T33 cells growing in vitro failed to detect such isotype switch variants, indicating that in vivo phenomena, such as interactions with the bone marrow microenvironment, are crucial for the appearance of these variant transcripts. This model clearly suggests that these isotype variants originate from malignant myeloma cells (and not the opposite) through complex molecular events (such as trans-switching or downstream-switching) and does not support the existence of pre-switched myeloma precursors.
The next speaker was Dr. Leif Bergsagel of Cornell Medical Center in New York. Based on analysis of the molecular IgH breakpoints observed in 40 myeloma cell lines, Dr. Bergsagel has developed a model for the oncogenesis ofMM. His group previously showed that most (if not all) human myeloma cell lines display illegitimate IgH rearrangements. Most of these breakpoints involve switch regions (located on the 5’ side of each constant domain except Cd), whereas in some cases, the breakpoints occur in the JH region. These results are consistent with the hypothesis that these translocations have been mediated by errors during the switch recombination and somatic mutation processes and may represent primary events in the transformation process. However, some myeloma cell lines present IgH rearrangements occurring in other sites. A detailed analysis of the 40 cell lines reveals striking features. One observation was that about 40% of the tumors retain the Cµ gene. This finding supports the hypothesis of possible trans-switching events, able to generate IgM transcripts, and constitutes experimental proof against the hypothesis of clonotypic IgM B-cell myeloma precursors. Analysis of the translocation partners in myeloma cell lines revealed three main regions: 4p16 (FGFR3/MMSET), 11q13 (CCND1), and 16q23 (c-maf). Cloning of the breakpoints in these cell lines has shown that they uniformly fall within JH or switch regions. In contrast, cloning of translocations with 8q24 (c-myc) or other sporadic chromosomal regions has shown that they almost systematically fall outside the JH and switch regions. The latter translocations would not be mediated by B-cell-specific recombination mechanisms but may result from other processes such as genomic instability. Dr. Bergsagel suggested that the former translocations would be primary oncogenetic events, whereas the second ones would represent random secondary events. His group also analyzed the cell lines for other recurrent genetic events such as mutations of N- and K-ras, observed in 40% of cases ; however, these events were never observed in cell lines harboring a mutated form of the translocation t(4;14).The group proposed an oncogenetic model based on analysis of the illegitimate IgH recombination breakpoints, separating primary and secondary events. This model needs to be demonstrated in patients’ primary tumors.
The final speaker, Dr. Malcom S. Moore (Memorial Sloan-Kettering Cancer Center) discussed telomerase activity in myeloma cells. Telomerase expression is normally restricted to a few cell types, such as fetal or germinal cells. In somatic cells, telomerase activity is rapidly repressed after birth. Without telomerase activity, somatic cells exhibit progressive loss of telomeric sequences with each cell division and finally enter senescence when a critical length of telomeres is reached. In contrast, malignant cells are capable of escaping the senescent checkpoint and undergoing unlimited proliferation. Indeed, activation of telomerase is widely observed in human malignancies. Dr. Moore reported on telomerase activity in highly purified myeloma cells. A wide range of telomerase activity was observed among 135 samples, reflecting the heterogeneity of myeloma. In contrast, in a few individuals with monoclonal gammopathy of undetermined significance (MGUS), telomerase activity was constantly low. Telomerase activity was linearly correlated with a number of disease variables, especially the number of plasma cells and poor-prognosis cytogenetic abnormalities. This parameter could represent a novel prognostic tool for patients with MM.
This session was extremely rich and highlighted the complexity of the oncogenesis ofMM. Rearrangements of the IgH locus appear to be very powerful for the understanding of these processes. However, most of the results obtained so far have been generated in mouse or in myeloma cell lines, which may not perfectly reflect the situation in patients. The next step will be to focus on patient tumors to gain further insight into the oncogenesis of MM.