REVIEW Multiple Myeloma: the Cells of Origin – a Two-Way Street

REVIEW

Multiple myeloma: the cells of origin – A two-way street JR Berenson, RA Vescio and J Said

West Los Angeles VA Medical Center, UCLA School of Medicine, Los Angeles, CA, USA

Multiple myeloma results from an interplay between the mono- earlier precursor cells may be responsible for the proliferation clonal malignant plasma cells and supporting nonmalignant of the malignant population. The presence of a circulating cells in the bone marrow. Recent studies suggest that the final transforming event in this B cell disorder occurs at a late stage tumor component without obvious plasma cell morphology of B cell differentiation based on the characteristics of the also suggests that less mature lymphocytes may be part of the immunoglobulin genes expressed by the malignant clone as clone as well, which could explain the dissemination of the well as surface markers present on the tumor cells. Recently, disease throughout the bone marrow.4 In addition, evidence an increasing pathogenic role in this malignancy by the nonma- for tumor cells at even earlier stages of hematopoietic differen- lignant cells in the bone marrow has been suggested by several tiation came from studies showing the high rate of acute non- studies. Specific infection of these supporting cells by the lymphoblastic leukemia in these patients and the presence of recently identified Kaposi’s sarcoma-associated herpes virus 5,6 (KSHV) suggests a novel mechanism by which this nonmalig- non-lymphoid surface markers on malignant plasma cells. A nant population may lead to the development of this B cell variety of molecular biological techniques have subsequently malignancy and support its growth. been used to characterize further the cell types which are part Keywords: multiple myeloma; immunoglobulin genes; Kaposi’s of the malignant clone in this B cell malignancy. sarcoma-associated herpes virus Immunogenotypes Introduction Because of the monoclonal nature of the Ig synthesized by the malignant cells, the genes, which lead to the production of Multiple myeloma is characterized by the accumulation of this protein, can be used as molecular markers capable of malignant plasma cells in the bone marrow compartment.1 identifying surface markers on the malignant cells. In addition, These terminally differentiated B-lymphocytes all produce an specific characteristic changes in these genes occur at differ- identical immunoglobulin (Ig) known as a monoclonal pro- ent stages of B cell differentiation; and thus allow determi- tein, the laboratory hallmark of this malignancy. Although the nation of which cell types in this lineage are part of the predominant cell type in the bone marrow of these patients tumor clone. has the characteristics of a plasma cell, the low proliferative Each antibody-producing B cell produces a single type of activity of these cells has raised the possibility that less mature antibody. Early in normal B cell development, rearrangement B-lymphocytes represent the proliferating population, which of four separate gene segments leads to the development of give rise to the overabundant malignant plasma cell. Recent the heavy chain portion of a unique functional antibody.7 advances in molecular biological techniques have led to the These gene segments encode the constant region which deter- determination of the stage of B cell differentiation at which mines the class of antibody (eg C␮ for IgM, C␥ for IgG, C␣ this disease begins. for IgA), and three other gene segments which encode the

The tumor cells which reside within the bone marrow are variable region of the heavy chain; variable (VH), diversity (D), 7–10 supported by a nonmalignant population of stromal cells and joining (JH). Since these latter three joined segments which produce cytokines that enhance myeloma cell growth encode for the specific antigen recognition site of the final and prevent apoptosis. Moreover, the tumor cells themselves antibody, the recombination of genes which comprise this orchestrate the production of cytokines by these nonmalignant region must be unique to each antibody-producing cell. Thus, bone marrow cells which can lead to a further increase in specific changes in these genes occur during the maturation tumor cell burden. Recent demonstration of the Kaposi’s sar- of the antibody-producing B cell. Most of a single functional coma-associated herpesvirus in some of these nonmalignant heavy chain variable region is encoded by one of ෂ50 func- cells provides an additional novel mechanism by which the tional VH genes whereas shorter stretches of amino acids are bone marrow microenvironment may support this B cell derived from one or more D and one of six functional JH malignancy. genes.7–10 In addition, increased antibody specificity is accomplished by the addition of non-germline nucleotides (N segments).11 Once a functional heavy chain rearrangement The malignant cell of origin has occurred, the ␬ (and, if unsuccessful, ␭) light chain 12 undergoes a similar rearrangement of VL,JL, and C genes. The low proliferative rate of the phenotypically identified Additional specificity of the antibody is rendered by the malignant cell2 and inability of these cells to sustain tumor mutation of nucleotides within the specific areas of the vari- growth in vivo as demonstrated by kinetic studies3 imply that able region which bind antigen directly known as comp- lementarity determining regions (CDRs) leading to enhanced avidity of the antibody for antigen.13 This process of somatic Correspondence: JR Berenson, West Los Angeles VA Medical Center, UCLA School of Medicine, 11301 Wilshire Blvd (111H), Los Angeles, mutation occurs late in B cell development in germinal cen- CA 90073, USA; Fax: 310 268 4908 ters. Following the secretion of a functional active antibody, Received 20 November 1997; accepted 21 November 1997 this process ceases within this terminally differentiated B cell. Review JR Berenson et al 122 In multiple myeloma, the properties of the Ig genes allow of normal plasma cells have not been completed, it is precise determination of the stage of B cell development dur- unknown whether this increased frequency of D-D fusion and ing which malignant transformation occurred. Specifically, lack of N-region addition is specific to myeloma. Use of spe-

comparison of the myeloma VH gene sequence to the most ciﬁc JH and D genes is quite comparable to usage in mature homologous germline gene has shown marked somatic B-lymphocytes further supporting the late B cell origin of mutation (median 8%) of this gene segment.14 This frequency myeloma. of somatic mutation is only found in the most differentiated B cells after antigenic stimulation has occurred. In further support of an antigenically driven process leading to these malig- Immunophenotype nant clones, analysis of regions which bind antigen within these genes, the CDRs, demonstrates a marked predilection Early studies using aneuploidy as a basis of malignant plasma for somatic mutations compared to the other parts of the gene cell determination suggested the existence of myeloid, mega- which are primarily responsible for maintaining the structural karyocytic, and T cell markers on the surface of the malignant integrity of the molecule, the so-called framework regions clone in myeloma.4 However, the use of these so-called ‘lin- (FRs).15 Moreover, in antigenically driven cells, mutations eage-speciﬁc markers’ was problematic since it was learned occurring within the CDRs should be more often associated that normal plasma cells could express these same antigens.32 with a change in the encoded amino acid compared to those The presence of a unique molecular marker, the Ig gene changes occurring in the FRs. Indeed, the ratio of replacement expressed by the malignant clone, has allowed a more precise to silent mutations is double in the CDRs compared to the FRs determination of surface markers present on the malignant in multiple myeloma.14 clone. Importantly, analysis of multiple clones shows no clonal

diversity in VH genes from myeloma patients, and determination of these sequences during the course of their disease CD34 shows no clonal evolution.14,16,17 These results differ from other B cell tumors originating in germinal centers at earlier CD34 is expressed on early hematopoietic precursors includ- stages of B cell differentiation.18 Analysis of CDR3 genes com- ing the pluripotent stem cell.33 However, some B cells also 34 prised of the D, JH and N segments further supports the high express this surface marker. Reinfusion of hematopoietic rate of somatic mutation with a high replacement to silent cells selected for CD34 can lead to rapid and sustained ratio.19 These segments also show a lack of clonal diversity.19 engraftment following myeloablative chemotherapy.35 Since Thus, these results suggest that the final oncogenic event in we and others had previously demonstrated the presence of myeloma occurs very late in B cell differentiation. Although tumor cells in not only the bone marrow but also the blood some studies have suggested the existence of a pre-class of myeloma patients, it became of both biological and clinical switched (C␮-containing) monoclonal cell in this disease,20–22 importance to determine whether any malignant cells the frequency of these cells and their contribution to the expressed this stem cell marker in myeloma patients. Initially, malignant process has not been established. Using colony CD34+ cells were purified from myeloma patients’ bone mar- hybridization techniques, our laboratory has not been able to row using a combination of an immunoadsorption column demonstrate C␮-containing clonal cells in myeloma.23 and flow sorting. Using a sensitive PCR-based assay with primers derived from the unique CDRs expressed by the patient’s malignant clone, no tumor cells were detected in the CD34- 36 Use of specific VH, D, and JH genes expressing population. Several other groups have confirmed this initial report37,38 although two reports have suggested the + 39,40 The approximately 50 functional VH genes have been divided existence of a CD34 tumor cell in myeloma. However, into seven families based on sequence homology ranging in CD34 selection with an immunoadsorption column (Ceprate; 24,25 size from one (VH6) to 22 (VH3). In mature B cells, VH Cellpro, Bothell, WA, USA) on autologous peripheral stem gene family use is generally proportional to the number of cell autografts resulted in a reduction of autograft tumor bur- functional genes.26 However, specific genes have been shown den by 2.7 to Ͼ4.5 logs as quantified using a PCR assay with 27 41 to be overused by immature B cells including VH4.21. This patient-specific Ig gene primers. Results of a recently com- same gene has also been frequently expressed by the auto- pleted randomized phase III trial using the Ceprate device42 reactive and malignant B cells from patients with autoimmune confirmed this marked reduction in autograft tumor burden disease and several B cell neoplasms, respectively.28–30 In fact, (median, 3.3 logs) with the majority of autografts rendered diffuse large cell lymphomas, which develop from relatively ‘tumor-free’ to the sensitivity of the assay (1:700 000). Despite

immature B cells, use this VH gene in nearly two-thirds of this manipulation, engraftment was not different between the cases.30 patients receiving CD34-selected compared to -unselected 43 In myeloma, although VH gene family usage closely paral- autografts. lels results found in normal B cells, VH4.21 usage has only rarely been observed.31 Since this gene encodes antibodies capable of recognizing self-antigens, the elimination of B cells CD10 expressing this gene prior to plasma cell development would prevent production of possibly deleterious auto-antibodies. CD10, also known as the common acute lymphoblastic leu- Thus, this finding also supports that the final transforming kemia antigen (CALLA), was originally described as a surface event occurs at the stage of terminal B cell differentiation in antigen present on the cells of some human acute lymphoblas- myeloma. The CDR3 length in myeloma is comparable to nor- tic leukemias.44 Subsequently, it has been shown that CD10 mal B cells but fusion of multiple D genes is much more com- is expressed on a wide variety of hematopoietic cells45,46 mon than in normal B cells.19 In myeloma CDR3s, N-region including B cells at early and late stages of differentiation. insertion is minimal compared to normal B cells. Since studies Studies of CD10 expression in myeloma have revealed that Review JR Berenson et al 123 CD10-expressing myelomatous plasma cells are present in the malignant plasma cell. Specifically, nonmalignant stromal only a minority (average 25%; range, 10–60%) of cases, and cells in the bone marrow from myeloma patients have been that myeloma cells expressing this antigen comprised only a shown to promote the growth and prevent apoptosis of malig- small fraction of the malignant population.47–50 However, nant plasma cells.62–64 The cytokine IL-6 is produced in large these earlier studies were based on relatively insensitive quantities by these cells, and has been demonstrated to immunohistochemical studies. Using PCR with patient- enhance growth and prevent apoptosis of the tumor cells. The specific VH gene primers, our group has found a small CD10- production of this cytokine by stromal cells from myeloma bearing malignant population in all cases studied.51 patients has been shown to be enhanced by the adherence This malignant CD10-expressing population in myeloma of malignant plasma cells. Thus, there is a close ‘symbiotic’ may play an important role in the pathogenesis of this disease. relationship between bone marrow stromal cells and tumor Since CD10 is expressed on germinal center B cells with a cells which results in an increase in tumor burden in myeloma high proliferative activity, this malignant subpopulation in patients. Many years ago, Hamburger and Salmon65 suggested myeloma may represent the part of the clone which leads to the existence of a specific factor only present in bone marrow growth and disease progression. In support of this, Wearne et macrophages from myeloma patients, which was required for al52 have shown that the appearance of CD10-expressing supporting the malignant clone. tumor cells in the circulation occurs at the time of progression of myeloma. KSHV in other diseases

CD19 and CD20 A new member of the herpes virus family was discovered in a case of Kaposi’s sarcoma using representational difference CD19 and CD20 are markers which become expressed during analysis.66 In addition to being frequently found in KS associa- B cell development but are normally not present on terminally ted with or without HIV infection,67 this virus, Kaposi’s sar- differentiated plasma cells. Although malignant bone marrow coma-associated herpesvirus (KSHV) also known as herpesvi- 39 plasma cells also do not express these markers, controversy rus 8, has been identified in two other B cell disorders, exists regarding their presence on the circulating malignant primary effusion lymphoma and multicentric Castleman’s dis- 40,53–55 population. Most studies have suggested that these mar- ease.68,69 Interestingly, this virus contains an IL-6 homologue70 kers are rarely found on the circulating malignant cells in and all three of these diseases also share IL-6 as a growth myeloma patients. In fact, a recent study suggests not only factor. In these diseases, the virus has been localized to the that these circulating CD20-expressing cells are polyclonal malignant cells71–73 although some controversy still exists in but also that there is an inverse relationship between their Kaposi’s sarcoma.74–76 Recently, KSHV has been found in 54 numbers and prognosis in myeloma. cells bearing macrophage and dendritic cell markers cultured from the blood of Kaposi’s sarcoma patients.74 In one study, viral DNA has also been identified in peripheral blood B cells CD38 from nearly half of Kaposi’s sarcoma patients.75 Moreover, Previous studies have demonstrated the high expression of although latent virus was present in the Kaposi’s sarcoma CD38 on normal and malignant plasma cells.56,57 Although tumor itself, viral forms characteristic of KSHV replication have been localized within endothelial cells of KS lesions72 this antigen is not expressed on the pluripotent stem cell, it is 76 expressed weakly on early lineage-committed cells.58–60 If and were present in the peripheral blood of these patients. CD38 is expressed on all malignant cells, it may be possible to Thus although KSHV has been associated with the malignant select for CD38-negative cells in autografts and render these population, some studies support infection of nonmalignant products tumor-free. However, most studies of CD38 cells and suggest these cell types may disseminate the virus. expression in myeloma have relied only on insensitive immunohistochemical techniques, which would be incapable of identifying an infrequent CD38-negative tumor cell in mye- KSHV and multiple myeloma loma patients. Using flow sorting with anti-CD38 antibodies on peripheral blood mononuclear cells from myeloma Because of the IL-6-mediated support of malignant plasma patients, we identified a CD38-negative tumor cell population cells by stromal cells derived from the bone marrow of myel- with a PCR-based assay in all myeloma patients.61 However, oma patients, it was hypothesized that other growth factors for the frequency of these circulating CD38 cells was markedly myeloma may result from KSHV infection of myeloma bone less than the CD38-expressing tumor cells in all cases. Thus, marrow macrophages. The virus, with its own unique growth although tumor cells lacking CD38 exist in myeloma, they factor(s), could be capable of supporting these nonmalignant represent a relatively minor component of the malignant cells. To this end, we examined the long-term marrow cultures population. Since CD38 is not expressed on B cells prior to established from the adherent cells of patients with myeloma. Using both a PCR assay with the viral specific primers derived the plasma cell stage, however, these CD38-negative malig- 77 nant cells may represent less mature malignant cells with a from the KS330233 fragment of ORF26 and in situ hybridiz- more proliferative capability. ation (ISH) with a KSHV-specific probe, we demonstrated KSHV in all patients with myeloma78 except for one recently evaluated patient with early stage disease (Durie–Salmon The other cell of origin: the KSHV-infected dendritic cell stage IA). In addition, similar studies of patients with monoclonal gammopathy of undetermined significance (MGUS) Stromal cells in myeloma have shown the presence of KSHV in a quarter of these individuals. MGUS patients also have a monoclonal Ig protein but In myeloma, the role of the bone marrow microenvironment do not have the clinical manifestations of myeloma although has been shown to be of increasing importance in supporting approximately one quarter will ultimately develop myeloma Review JR Berenson et al 124 or a related B cell malignancy within 25 years.79 By contrast, portion of cells bear lytic virus.85 The expression of other viral normal subjects and patients with other hematological malig- proteins (MIPs) has been observed in preliminary studies from nancies did not contain KSHV in their cultured cells. our group. Several recent reports have not found evidence of Characterization of cells infected with virus using immuno- KSHV infection using serological assays.86,87 In addition, since histochemical techniques showed the presence of CD68, Kaposi’s sarcoma and myeloma have now both been linked CD83, and fascin but lack of CD1a, CD31 and CD34. Thus, to the same virus, it is puzzling that myeloma and Kaposi’s these results are consistent with a dendritic cell phenotype sarcoma are not linked epidemiologically.88 However, recent without characteristics of Langerhans cells80 which is similar studies from our group demonstrate the presence of KSHV to the cell type found infected in cultures from the peripheral sequence variations from those frequently identified in Kapo- blood of Kaposi’s sarcoma patients in one study. Recently, the si’s sarcoma patients. Specifically, KSHV has been divided human cytomegalovirus has been shown to be similarly into three strain types (A, B and C) with most Kaposi’s sarcoma derived from long-term cultures of PBMCs bearing macro- patients showing A strain89 whereas all of the myeloma phage and dendritic cell markers.81 Thus, dendritic cells may patients show a B strain pattern. However, of even more serve as a reservoir for a wide variety of different viruses. significance, consistent mutational differences including deletions capable of causing frameshifts are found in KSHV from myeloma patients. In addition, other B cell disorders KSHV in fresh bone marrow and blood associated with KSHV also show a reduced frequency of A strain90 consistent with our results in myeloma. Whether sero- Although the initial study showed the presence of KSHV in logical responses differ between KSHV strains is unknown. Of

cells obtained by long-term culture, it was important to estab- note, specific differences in KS330233 sequences exist between lish the presence of KSHV in fresh bone marrow and peri- our patients (3–8 bp), making PCR artifact an unlikely possi- pheral blood samples from myeloma and MGUS patients. bility. In five patients, sequencing of KSHV derived from the Using in situ hybridization on serial sections, KSHV was found same patient’s long-term cultures from bone marrow aspirate, in bone marrow core biopsies of 18/21 myeloma patients82 fresh bone marrow biopsy, and dendritic cell-enriched peri- and one in six MGUS patients, but not in patients with other pheral blood shows identical sequences or a single base pair hematological malignancies or normal subjects. Immunohis- difference. Taken together, these data suggest that the virus in tochemical staining on serial sections suggested that the virus myeloma patients may be a variant of the one originally ident- was present in dendritic cells which contained many long foot ified in Kaposi’s sarcoma.77 processes and comprised approximately 2% of the nucleated Another explanation for the lack of a serological response cells in these bone marrow biopsies. PCR confirmed the pres- may be due to an immunological defect. Although immuno- ence of KSHV in these bone marrow biopsies whereas normal deficiency is one of the hallmarks of myeloma, the presence of subjects contained no detectable virus. serological responses to EBV in these patients makes a general Studies of peripheral blood samples have rarely demon- humoral defect less likely.86,87 However, these patients may strated KSHV in unenriched PBMCs (1/15) from myeloma harbor a specific inability to respond to KSHV; and thus, are patients whereas enrichment of these cells using immunomag- susceptible to its infection. In addition, viral infection of den- netic selection with CD68 and CD83 antibodies led to detec- dritic cells by other herpesviruses has been associated with tion of KSHV in 33/40 myeloma patients and 5/20 MGUS functional defects of these cells. patients. However, KSHV has not been identified in patients Recently, measles virus infection of dendritic cells has been with other malignancies or Paget’s disease of the bone, and shown to disseminate the infections and induce apoptosis and only rarely (1/16) in normal volunteers.83 Interestingly, in reduce cytokine production by these cells leading to a poor these studies, most myeloma patients lacking the presence of immune response.91–93 Similarly, KSHV infection of dendritic virus in blood and/or bone marrow were in remission follow- cells in myeloma may allow dissemination of the infection ing high-dose or conventional chemotherapy or were at an throughout the bone marrow and also contribute to the early clinical stage. impaired immune response. In support of this, preliminary Preliminary studies have also suggested the presence of studies show a diminished expression of class II molecules virus in several other disorders of terminal B cell differen- and reduced ability of these KSHV-infected bone marrow den- tiation including amyloidosis, Waldenstrom’s macroglobuli- dritic cells from myeloma patients to stimulate allogeneic nemia, POEMS, and solitary plasmacytoma.84 Regarding the responder lymphocytes compared to similarly derived cells latter B cell malignancy, KSHV has been found in the plasma from normal subjects and patients with other B cell malig- cell tumor but not in uninvolved bone marrow. More cases nancies not infected with KSHV.94 are necessary to confirm this result which suggests the importance of the KSHV-infected dendritic cell in the development of disseminated multiple myeloma. Ongoing studies are also The malignant cell and the KSHV-infected dendritic determining whether the presence of KSHV predicts which cell: is there a relationship? MGUS patients will ultimately develop myeloma. It is intriguing to note that approximately one quarter of patients Since KSHV infection is known to occur in the general popu- with MGUS harbor virus, and a similar proportion transform lation; and, in fact, we have identified this virus in one normal to a B cell malignancy. volunteer’s PBMCs, other factors must be required for the Determination of expression of KSHV-specific genes will development of myeloma. Myeloma is frequently associated help elucidate the state of the virus (latent vs lytic). Studies on with translocations involving the immunoglobulin heavy cultured cells have shown the presence of viral IL-6,78 suggest- chain gene locus,95 and this has been recently demonstrated ing the possible existence of some cells bearing lytic virus. to involve DNA sequences responsible for heavy chain class However, the relevance of this data derived from long-term switching.96,97 Recent cloning of non-immunoglobulin DNA cultures to the situation in vivo is questionable. In Kaposi’s near these breakpoints has identified new putative onco- sarcoma, although most cells are latently infected, a small pro- genes.97–99 Interestingly, preliminary analysis of one of these Review JR Berenson et al 125 genes involved in a translocation involving 11q13 and 14q32 11 Desiderio SV, Yancopoulos GD, Paskind M, Thomas E, Boss MA, identifies a possible new member of the FGF family. Interest- Landau N, Alt FW, Baltimore D. Insertion of N regions into ingly, also localized at 11q13 are two other FGF members heavy-chain genes is correlated with expression of terminal deoxytransferase in B cells. Nature 1984; 311: 752–755. including one, KS3, which was originally cloned from a Kapo- 12 Yancopoulos GD, Alt FW. Developmentally controlled and si’s sarcoma lesion and acts as a growth factor for this tissue-specific expression of unrearranged VH gene segments. tumor.100 Thus, it is possible that this new FGF member may Cell 1985; 40: 271–281. provide the myeloma cells with the means of helping support 13 Tonegawa S. Somatic generation of antibody diversity. Nature the KSHV-infected dendritic cell. This would allow a cytokine 1983; 302: 575–581. loop to occur between the malignant cell and its virally 14 Vescio RA, Cao J, Hong CH, Lee JC, Wu CH, Der Danielanian infected nonmalignant partner. M, Wu V, Newman R, Lichtenstein AK, Berenson JR. Myeloma Ig heavy chain V region sequences reveal prior antigenic selection and marked somatic mutation but no intraclonal diversity. J Immunol 1995; 155: 2487–2497. Conclusions 15 Amit AG, Mariuzza RA, Phillips SEV, Poljak RJ. Three-dimen- sional structure of an antigen-antibody complex at 6 A resolution. These early studies suggest the importance of KSHV in the Nature 1985; 313: 156–158. development of myeloma. However, much work is required 16 Bakkus MHC, Heirman C, Van Riet B, Van Camp B, Thielemans K. Evidence that multiple myeloma Ig heavy chain VDJ genes to more firmly establish the role of this new herpesvirus in contain somatic mutations but show no intraclonal variation. this B cell malignancy. Viral gene expression studies will help Blood 1992; 80: 2326–2335. determine the viral state and provide information which may 17 Ralph QM, Brisco MJ, Joshua DE, Brown R, Gibson J, Morley hint at possible therapeutic approaches. Studies both in vivo AA. Advancement of multiple myeloma from diagnosis through and in vitro demonstrating the effect of these virally infected plateau phase to progression does not involve a new B-cell clone: cells on malignant plasma cells are critical. Infection of nor- evidence from the Ig heavy chain gene. Blood 1993; 82: 202– mal stroma with the virus from myeloma patients will allow 206. 18 Zelenetz AD, Chen TT, Levy R. Clonal expansion in follicular a better understanding of the specific role of this virus in the lymphoma occurs subsequent to antigenic selection. J Exp Med pathogenesis of myeloma. Little is known about the effect of 1992; 176: 1137–1148. antiviral agents on KSHV in Kaposi’s sarcoma in vivo and in 19 Vescio RA, Kunkel L, Wu CH, Cao J, Han EJ, Belson DJ, Rettig M, vitro.101,102 Models allowing studies of virally infected samples Schiller GJ, Berenson JR. Multiple myeloma heavy chain CDR3 from myeloma patients must be developed rapidly. This may analysis reveal mutational hot spots, frequent D-D fusions, and lead to early clinical trials allowing a novel therapeutic prior antigenic selection. Blood 1996; 88: 294a. approach to this incurable B cell malignancy. 20 Corradini P, Boccadoro M, Voena C, Pileri A. Evidence for a bone marrow B cell transcribing malignant plasma cell VDJ joined to Cm sequence in immunoglobulin (IgG)- and IgA- secreting multiple myelomas. J Exp Med 1993; 178: 1091–1096. References 21 Billadeau D, Ahmann G, Greipp P, Van Ness B. The bone marrow of multiple myeloma patients contains B cell populations at 1 Durie BGM, Salmon SS. Staging, kinetics and flow cytometry of different stages of differentiation that are clonally related to the multiple myeloma. In: Wiernik P, Canellos G, Kyle R, Schiffer C malignant plasma cell. J Exp Med 1993; 178: 1023–1031. (eds). Neoplastic Diseases of the Blood, (vol 2). Churchill 22 Bakkus MH, Van Riet I, Van Camp B, Thielemans K. Evidence Livingstone: New York, 1985, pp 513–528. that the clonogenic cell in multiple myeloma originates from a 2 Drewinko B, Alexanian R, Boyer H, Barlogie B, Rubinow SI. The pre-switched but somatically mutated B-cell. Br J Haematol growth fraction of human myeloma cells. Blood 1981; 57: 1994; 87: 68–74. 333–338. 23 Vescio R, Hong C, Cao J, Lee C, Der Danielian M, Lichtenstein 3 Pileri A, Tarocco RP. In vivo kinetic studies in human myeloma. AK, Berenson JR. Multiple myeloma clones are derived from Haematologica 1974; 59: 10–24. post-class switch precursor cells. Blood 1993; 82 (Suppl. 1): 4 Berenson J, Wong R, Kim K, Brown N, Lichtenstein A. Evidence 259a. for peripheral blood B lymphocyte but not T lymphocyte involve- 24 Cook GP, Tomlinson IM, Walter G, Riethman H, Carter NP, ment in multiple myeloma. Blood 1987; 70: 1550–1553. Buluwela L, Winter G, Rabbitts T. A trap of the human immuno- 5 Cuzick J, Erskine S, Edelman D, Galton DAG. A comparison of globulin VH locus completed by analysis of the telomeric region the incidence of the myelodysplastic syndrome and acute of chromosome 14q. Nature Genet 1994; 7: 162–168. myeloid leukemia following melphalan and cyclophosphamide 25 Saz I, Kelly P, Williams C, Scholl S, Tucker P, Capra JD. The treatment for myelodysplastic syndrome and acute myeloid leu- smaller human VH gene families display remarkably little poly- kemia following melphalan and cyclophosphamide treatment for morphism. EMBO J 1989; 8: 3741–3748. myelomatosis. Br J Cancer 1987; 55: 523–529. 26 Brezinschek HP, Brezinschek RI, Lipsky PE. Analysis of the heavy 6 Epstein JH, Xiano-Yan H. Markers of multiple hematopoietic-cell chain repertoire of human peripheral B cells using single-cell lineages in multiple myeloma. New Engl J Med 1990; 322: polymerase chain reaction. J Immunol 1995; 155: 190–202. 664–668. 27 Kraj P, Friedman DF, Stevenson F, Silberstein LE. Evidence for 7 Berman JE, Millis SJ, Pollack R, Smith CL, Suh H, Heinke B, Surti U, Kowal C, Chess L, Cantor CR, Alt FW. Content and organiza- the overexpression of the VH4–34(VH4.21) Ig gene segment in tion of the human Ig VH locus: definition of three new VH famil- the normal adult human peripheral blood B cell repertoire. J ies and linkage to the Ig CH locus. EMBO J 1988; 73: 727–738. Immunol 1995; 154: 6406–6420. 8 Tomlinson IM, Walter G, Marks JD, Llewelyn MB, Winter G. The 28 Stevenson FK, Speilerberg MB, Treasure J, Shapman CJ, Silb- repertoire of human germline VH sequences reveals about 50 erstein LE, Hamblin TJ, Jones DB. Differential usage of an Ig groups of VH segments with different hypervariable loops. J Mol heavy chain region gene by human B cell tumors. Blood 1993; Biol 1992; 227: 776–798. 82: 224–230. 9 Ravetch JV, Siebenlist T, Korsmeyer S, Waldmann T, Leder P. 29 Pascual V, Capra JD. VH4–21, a human VH gene segment over- Structure of the human immunoglobulin m locus: characteriz- represented in the autoimmune repertoire. Arth Rheum 1992; ation of embryonic and rearranged J and D genes. Cell 1981; 27: 35: 11. 583–591. 30 Hsu FJ, Levy R. Preferential use of the VH4 immunoglobulin gene 10 Ichihara Y, Matsuoka H, Kurosawa Y. Organization of human family by diffuse large cell lymphoma. Blood 1995; 86: 3072– immunoglobulin heavy chain diversity gene loci. EMBO J 1988; 3082. 7: 4141–4150. 31 Rettig MB, Vescio RA, Cao J, Wu CH, Lee JC, Han E, Der Daniel- Review JR Berenson et al 126 anian M, Newman R, Hong C, Lichtenstein AK, Berenson JR. 50 Durie BGM, Grogan TM. CALLA-positive myeloma: an aggress- Blood 1996; 87: 2846–2852. ive subtype with poor survival. Blood 1985; 66: 229–232. 32 Terstappen LWMM, Johnsen S, Segers-Nolten IMJ, Loken MR. 51 Cao J, Vescio RA, Rettig MB, Hong CH, Kim A, JC Lee, Licht- Identification and characterization of plasma cells in normal enstein AK, Berenson J. A CD-10 positive subset of malignant bone marrow by high-resolution flow cytometry. Blood 1990; 76: cells is identified in multiple myeloma using PCR with patient- 1739–1747. specific primers. Leukemia 1995; 9: 1948–1953. 33 Andrews RG, Singer JW, Bernstein ID. Precursors of colony- 52 Wearne AJ, Joshua DE, Brown RD, Kronenberg H. Multiple myel- forming cells in humans can be distinguished from colony- oma: the relationship between CALLA-(CD10)-positive lympho- forming cells by expression of the CD33 and CD34 antigens and cytes in the peripheral blood and light chain isotype suppression. light scatter properties. J Exp Med 1989; 169: 1721–1731. Br J Haematol 1987; 67: 39–44. 34 Schmitt C, Eaves CJ, Lansdorp PM. Expression of CD34 on human 53 Pilarski LM, Jensen GS. Monoclonal circulating B cells in mul- B cell precursors. Clin Exp Immunol 1991; 85: 168–173. tiple myeloma: a continuously differentiating, possibly invasive, 35 Berenson RJ, Bensinger WI, Hill RS, Andrews R, Garcia-Lopez J, population as defined by expression of CD45 isoforms and Kalamasz DF, Still BJ, Spitzer G, Buckner CD, Bernstein ID, adhesion molecules. Hematol Oncol Clin N Am 1992; 2: 297– Thomas ED. Engraftment after infusion of CD34+ marrow cells in 322. patients with breast cancer or neuroblastoma. Blood 1994; 77: 54 Kay NE, Leong T, Kyle RA, Greipp P, Billadeau D, Van Ness B, 1717–1722. Bone N, Oken MM. Circulating blood B cells in multiple myel- 36 Vescio RA, Hong CH, Cao J, Kim A, Schiller GJ, Lichtenstein AK, oma: analysis and relationships to circulating clonal cells and Berenson RJ, Berenson JR. The hematopoietic stem cell antigen, clinical parameters in a cohort of patients entered on the Eastern CD34, is not expressed on the malignant cells in multiple myel- Cooperative Oncology Group phase III E9486 clinical trial. oma. Blood 1994; 84: 3283–3290. Blood 1997; 90: 340–345. 37 Owen RG, Child JA, Rawson A, Smith GM, Johnson R, Wood- 55 Chen BJ, Epstein J. Circulating clonal lymphocytes constitute a head V, Elsworth A, Morgan GJ. Detection of contaminating cells minor subpopulation of B cells. Blood 1996; 87: 1972–1976. in PBPC harvests and the efficacy of CD34 selection in patients 56 Kawano MM, Huang N, Harada H, Harada Y, Sakai A, Tanaka with multiple myeloma. Blood 1994; 84 (Suppl. 1): 352a. H, Iwato K, Kuramoto A. Identification of immature and mature 38 Goldschmidt H, Moos M, Cremer F, Atzberger A, Witt B, Haas myeloma cells in the bone marrow of human myelomas. Blood R, Hunstein W. Antigenic profile and PCR-analysis of peripheral 1993; 82: 564–570. blood progenitor cell autografts in multiple myeloma. Blood 57 Leo R, Boeker M, Peest D, Hein R, Bartl R, Gessner JE, Selbach 1994; 84 (Suppl. 1): 385a. J, Wacker G, Deicher H. Multiparameter analyses of normal and 39 Takishita M, Kosaka M, Goto T, Saito S. Cellular origin and extent malignant human plasma cells: CD38++, CD56+, cIg+ is the com- of clonal involvement in multiple myeloma. Br J Haematol 1994; mon phenotype of myeloma cells. Ann Hematol 1992; 64: 87: 735–742. 132–139. 40 Szczepek AJ, Bergsagel PL, Axelsson L, Brown CB, Belch AR, 58 Muench MO, Cupp J, Polakoff J, Roncarolo MG. Expression of Pilarski LM. CD34+ cells in the blood of patients with multiple CD33, CD38 and HLA-DR on CD34+ human fetal liver progeni- myeloma express CD19 and IgH mRNA and have patient-spe- tors with high proliferative potential. Blood 1994; 83: 3170– cific IgH VDJ gene rearrangements. Blood 1997; 89: 1824–1833. 3181. 41 Schiller G, Vescio R, Freytes C, Spitzer G, Sahebi F, Lee M, Wu 59 Huang S, Terstappen LWMM. Formation of hematopoietic CH, Cao J, Lee JC, Hong CH, Lichtenstein A, Lill M, Hall J, microenvironment and hematopoietic stem cells from single Berenson R, Berenson J. Transplantation of CD34+ peripheral human marrow stem cells. Nature 1992; 360: 745–749. blood progenitor cells after high-dose chemotherapy for patients 60 Terstappen LWMM, Huang S. Lymphoid and myeloid differen- with advanced multiple myeloma. Blood 1995; 86: 390–397. tiation of single human CD34+, HLA-DR+, CD38− hematopoietic 42 Vescio R, Stewart A, Ballester O, Noga S, Rugo H, Freytes C, stem cells. Blood 1994; 83: 1515–1526. Stadtmauer E, Kogut N, Tarantolo S, Stiff P, Pinero L, Dalton W, 61 Rettig M, Hong C, Vescio R, Kim A, Han E, Shin J, Lichtenstein Ho A, Schiller G, Benyunes M, Jacobs C, Dipersio J, Anderson A, Berenson J. Reduction of tumor cells in multiple myeloma K, Berenson J. Myeloma cell tumor reduction in PBPC autografts (MM) peripheral blood using a CD38 immunoadsorption col- following CD34 selection: the results of a phase III trial using the umn. Proc Am Soc Clin Oncol 1995; 14: 451. Ceprate device. Blood 1997 (submitted). 62 Lichtenstein A, Tu Y, Fady C et al. Interleukin-6 inhibits apoptosis 43 Schiller G, Stewart AK, Ballester O, Noga S, Rugo H, Freytes E, of malignant plasma cells. Cell Immunol 1995; 162: 271–281. Stadtmauer E, Kogut N, Taranfolo S, Stiff P, Pineiro L, Dalton W, 63 Bataille R, Harousseau J-L. Multiple myeloma. New Engl J Med Ho A, Vescio R, Benyunes M, Jacobs C, DiPersio J, Anderson 1997; 336: 1657–1664. K, Berenson J. A phase III study evaluating CD34+-selected vs 64 Uchiyama H, Barut BA, Mohrbacher AF et al. Adhesion of human unselected autologous peripheral blood progenitor cell trans- myeloma-derived cell lines to bone marrow stromal cells stimu- plantation for patients with advanced multiple myeloma: lates interleukin-6 secretion. Blood 1993; 82: 3712–3720. engraftment results. Blood 1997 (submitted). 65 Salmon SE, Hamburger AW. Immunoproliferation and cancer: a 44 Greaves MF, Brown G, Rapson NT, Lister TA. Antisera to acute common macrophage-derived promoter substance. Lancet 1978; lymphoblastic leukaemia cells. Clin Immunol Immunopathol 1: 1289–1290. 1975; 4: 67–84. 66 Moore PS, Boshoff C, Weiss R, Chang Y. Molecular mimicry of 45 Hsu SM, Jaffe ES. Phenotypic expression of B lymphocytes: human cytokine and cytokine response pathway genes by KSHV. identification with monoclonal antibodies in normal lymphoid Science 1996; 274: 1739–1744. tissues. Am J Pathol 1984; 114: 387–395. 67 Chang Y, Cesarman E, Pessin MS et al. Identification of Herpesvi- 46 Tomlinson IM, Walter G, Marks JD, Llewelyn MB, Winter G. The rus-like DNA sequences in AIDS-associated Kaposi’s sarcoma. repertoire of human germline VH sequences reveals about fifty Science 1994; 266: 1865–1869. groups of VH segments with different hypervariable loops. J Mol 68 Moore PS, Chang Y. Detection of herpesvirus-like DNA Biol 1992; 227: 776–798. sequences in Kaposi’s sarcoma in patients with and those without 47 Jensen GS, Mant MJ, Belch AJ, Berenson JR, Ruether BA, Pilarski HIV infection. New Engl J Med 1995; 332: 1181–1185. LM. Selective expression of CD45 isoforms defines CALLA+ 69 Cesarman E, Chang Y, Moore PS, Said JW, Knowles DM. Kaposi’s monoclonal B-lineage cells in peripheral blood from myeloma sarcoma-associated herpesvirus-like DNA sequences in AIDS- patients as late stage B cells. Blood 1991; 78: 711–719. related body cavity-based lymphomas. New Engl J Med 1995; 48 Ruiz-Arguelles GJ, Katzmann JA, Greipp PR, Gonchoroff NJ, Gar- 332: 1186–1191. ton KP, Kyle RA. Multiple myeloma: circulating lymphocytes that 70 Soulier J, Grollet L, Oksenhendler E et al. Kaposi’s sarcoma- express plasma cell antigens. Blood 1984; 64: 352–356. associated herpesvirus-like DNA sequences in multicentric 49 Caligaris-Cappio F, Bergui L, Tesio L, Pizzolo G, Malavasi F, Chi- Castleman’s disease. Blood 1995; 86: 1276–1280. losi M, Campana D, van Camp B, Janossy G. Identification of 71 Said JW, Chien K, Tasaka T, Takeuchi S, Asou H, Cho SK, Cesar- malignant plasma cell precursors in the bone marrow of multiple man E, Knowles DM, Koeffler HP. Kaposi’s sarcoma associated myeloma. J Clin Invest 1985; 76: 1243–1251. herpesvirus (KSHV or HHV8) in primary effusion lymphoma: Review JR Berenson et al 127 ultrastructural demonstration of herpesvirus in lymphoma cell associated herpesvirus gene expression in endothelial (spindle) lines. Blood 1996; 87: 4937–4943. tumor cells. J Virol 1997; 71: 715–719. 72 Said JW, Chien K, Tasaka T, Koeffler HP. Ultrastructural charac- 86 Marcelin A-G, Dupin N, Bouscary D, Bossi P, Cacoub P, Ravaud terization of human herpesvirus 8 (Kaposi’s sarcoma-associated P, Calvez V. HHV-8 and multiple myeloma in France. Lancet Herpesvirus) in Kaposi’s sarcoma lesions: electron microscopy 1997; 350: 1144. permits distinction from cytomegalovirus (CMV). J Pathol 1997; 87 MacKenzie J, Sheldon J, Morgan G, Cook G, Schultz TJ, Jarrett 182: 273–281. RF. HHV-8 and multiple myeloma in the UK. Lancet 1997; 350: 73 Boshoff C, Schultz TF, Kennedy MM, Graham AK, Fisher C, 1144–1145. Thomas A, McGee JO’D, Weiss RA, O’Leary JJ. Kaposi’s sar- 88 Hjalgrim H, Frisch M, Melbye M. Incidence rates of multiple coma-associated herpesvirus infects endothelial and spindle myeloma and Kaposi’s sarcoma do not correlate. Eur J Cancer cells. Nature Med 1995; 1: 1274–1278. (in press). 74 Uccini S, Sirianni MC, Vincenzi L, Topino S, Stoppacciaro A, 89 Zong J-C, Metroka C, Reitz MS, Nicholas J, Hayward GS. Strain Parola ILL, Capuano M, Masin C, Cerimele D, Cella M, Lanza- variability among Kaposi sarcoma associated herpesvirus (human vecchia A, Allarena P, Mantovani A, Baroni CD, Rucco LP. Kapo- herpesvirus 8) genomes: evidence that a large cohort of United si’s sarcoma cells express the macrophage-associated antigen States AIDS patients may have been infected by a single common mannose receptor and develop in peripheral blood cultures of isolate. J Virol 1997; 71: 2505–2511. Kaposi’s sarcoma patients. Am J Pathol 1997; 150: 929–938. 90 A letter on KSHV-strain type. 75 Whitby D, Howard M, Tenant-Flowers M, Brink N, Copas A, 91 Bhardwaj N. Interactions of viruses with dendritic cells: a double- Boshoff C, Hatziannou T, Suggett F, Aldam D, Denton A. Detec- edged sword. J Exp Med 1997; 186: 795–799. tion of Kaposi sarcoma associated herpesvirus in peripheral 92 Fugier-Vivier I, Servet-Delpret C, Rivailler P, Rissoan M-C, Liu Y- blood of HIV-infected individuals and progression to Kaposi’s J, Rabourdin-Combe C. Measles virus suppresses cell-mediated sarcoma (see comments) Lancet 1995; 346: 799–802. immunity by interfering with the survival and functions of den- 76 Decker LL, Shankar P, Khan G, Freeman RB, Dezube BJ, Lieber- dritic cells and T cells. J Exp Med 1997; 186: 813–823. man J, Thorley-Lawson DA. The Kaposi’s sarcoma-associated 93 Grosjean I, Caux C, Bella C, Berger I, Wild F, Bancherau J, Kai- serlian D. Measles virus infects human dendritic cells and blocks herpesvirus (KSHV) is present as an intact latent genome in KS + tissue but replicates in the peripheral blood mononuclear cells their allostimulatory properties for CD4 T cells. J Exp Med 1997; of Kasposi’s sarcoma patients. J Exp Med 1996; 184: 283–288. 186: 801–812. 77 Chang Y, Cesarman E, Pessin MS, Lee F, Culpepper J, Knowles 94 Nepom GT, Masewicz S, Rettig M, Belson D, Vescio R, Berenson DM, Moore DS. Identification of herpesvirus-like DNA J. Functional deficits in bone marrow-derived dendritic cells sequences in AIDS-associated Kaposi’s sarcoma. Science 1994; infected with KSHV in multiple myeloma patients. Blood 1997 266: 1865–1869. (submitted). 78 Rettig MB, Ma HJ, Vescio RA, Pold M, Schiller G, Belson D, 95 Calasanz MJ, Cigusoda JC, Odero MD, Ferreira C, Ardanaz MT, Savage A, Nishikubo C, Wu C, Fraser J, Said JW, Berenson JR. Fraile A, Carrasco JL, Sole F, Cuesta B, Guillon A. Cytogenetic Kaposi’s sarcoma-associated herpesvirus infection of bone mar- analysis of 280 patients with multiple myeloma and related dis- row dendritic cells from multiple myeloma patients. Science orders: primary breakpoints and clinical disorders. Genes Chromos Cancer 1997; 18: 84–93. 1997; 276: 1951–1954. 96 Rettig MB, Pold M, Nguyen V, Cao J, Vescio RA, Berenson JR. 79 Kyle RA. ‘Benign’ monoclonal gammopathy – after 20 to 35 years Translocations into the immunoglobulin heavy chain gene (IgH) of follow-up. Mayo Clin Proc 1993; 68: 26–36. in multiple myeloma (MM). Blood 1996; 88: 294a. 80 Hart DNJ. Dendritic cells: unique leukocyte populations which 97 Chesi M, Nardini E, Brents LA, Schrock E, Ried T, Kuehl WM, control the primary immune reponse. Blood 1997; 90: 3245– Bergsagel PL. Frequent translocation t(4;14)(p16.3;q32.3) in mul- 3287. tiple myeloma is associated with increased expression and activ- 81 Soderberg-Naucler C, Fish KN, Nelson JA. Reactivation of latent ating mutations of fibroblast growth factor receptor 3. Nat Genet human cytomagalovirus by allogeneic stimulation of blood cells 1997; 16: 260–264. from healthy donors. Cell 1997; 91: 119–126. 98 Pold M, Rettig M, Munroe D, Vescio R, Hall J, Berenson J. 82 Said JW, Rettig MR, Heppner K, Vescio R, Schiller G, Belson D, Expression of novel genes in human chromosomal locus 11Q13. Savage A, Shintaku PI, Koeffler PH, Asou H, Pinkus G, Pinkus J, Blood 1997 (submitted). Schrage M, Green E, Berenson J. Localization of Kaposi’s sar- 99 Iida S, Rao PH, Butler M, Corradini P, Boccadoro M, Klein B, coma-associated herpesvirus in bone marrow biopsy samples Chaganti RSK, Dalla Favera R. Deregulation of MUM1/IRF4 by from patients with multiple myeloma. Blood 1997; 90: 4278– chromosomal translocation in multiple myeloma. Nat Genet 4282. 1997; 17: 226–230. 83 Rettig MB, Vescio RA, Moss TJ, Ma HJ, Schiller G, Berenson JR. 100 Delli Bovi P, Curatola AM, Kern FG, Greco A, Ittmann M, Detection of Kaposi’s-sarcoma-associated herpesvirus in the peri- Basilico C. An oncogene isolated by transfection of Kaposi’s sar- pheral blood of multiple myeloma patients. Blood 1997 coma DNA encodes a growth factor that is a member of the FGF (submitted). family. Cell 1987; 50: 729–737. 84 Rettig M, Vescio R, Ma H, Moss T, Schiller G, Solomon A, Said 101 Kedes D, Ganem D. Sensitivity of Kaposi’s sarcoma-associated J, Berenson J. Detection of Kaposi’s sarcoma-associated herpesvi- herpesvirus replication of antiviral drugs: implications for poten- rus in the dendritic cells of Waldenstrom’s macroglobulinemia tial therapy. J Clin Invest 1997; 99: 2082–2086. and primary amyloidosis patients. Blood 1997 (submitted). 102 Rizzeri DA, Liu J, Traweek ST, Miralles GD. Clearance of HHV- 85 Staskus K, Zhong W, Gebhardt K, Herndier B, Wang H, Renne 8 from peripheral blood mononuclear cells with a protease R, Pudney J, Anderson D, Ganem D, Haase A. Kaposi’s sarcoma- inhibitor. Lancet 1997; 349: 775–776.