The multiple myeloma bone eco-system and its relation to oncogenesis.

Pure lytic bone lesions are the hallmark of myeloma (MM). MM is the only hematological malignancy associated with lytic bone lesions and the mechanisms of bone destruction are well documented both at the cellular and molecular levels. An uncoupling bone process characterizes MM, with stimulation of bone resorption and inhibition of bone formation. The capacity of MM cells to directly or indirectly inhibit bone formation is specific of MM, although many carcinomas have the capacity to stimulate bone resorption, directly or indirectly in a similar way to MM. Few MM do not develop bone lesions, while true sclerotic MM remain exceptional. Inhibition of bone formation is the major event explaining the transition from MGUS to overt MM. It is now well documented that bone cells regulate MM cell growth, osteoclast stimulating MM cell growth and osteoblasts inhibiting it. Progression of MM from MGUS is characterized by the selection of MM clones able to inhibit osteoblasts, favoring tumor growth. These data underline the interest of new treatments able to regenerate bone.

Autocrine insulin-like growth factor 1 and stem cell factor but not interleukin 6 support self-renewal of human myeloma cells.

In this study, we have identified the growth factors supporting myeloma self-renewal in eight myeloma cell lines. All cell lines able to form self-colonies displayed constitutive P-AKT and P-ERK1,2 but not P-STAT3 and did not express CD45, suggesting the presence of an insulin-like growth factor 1 (IGF1) loop. We showed that a blocking anti-insulin-like growth factor 1 receptor (IGF1R) monoclonal antibody (mAb) inhibited colony formation in correlation with IGF1R expression and decreased P-AKT. Imatinib or a blocking anti-stem cell factor (SCF) mAb also inhibited colony formation of two cell lines expressing C-KIT and SCF, and decreased P-AKT. Moreover, the PI3K/AKT pathway inhibitor wortmannin inhibited colony formation. Blocking interleukin (IL)6R did not inhibit colony formation in good agreement with a lack of constitutive P-STAT3. We showed that primary cells frequently co-expressed IGF1R/IGF1 but not C-KIT/SCF or IL6R/IL6, suggesting that in vivo autonomous growth could be possible via IGF1R. Despite their similar role in clonogenic growth and shared signaling pathway, IGF1R and C-KIT had opposite prognostic values, suggesting that they were surrogate markers. Indeed, we showed that both C-KIT and IGF1R prognostic values were not independent of MMSET expression. This study highlights the autocrine role of IGF1 in myeloma cells and reinforces the interest in targeting IGF1R in IGFR1(+) CD45(+/-) patients, such as MMSET(+) patients.

Human tumor stem cells, the petri dish and the patient. A tribute to Sydney E. Salmon.

During the last ten years, much work has been devoted to the concept of tumor stem cells, a concept first introduced by Virchow in 1855. Despite the importance and the quality of these works, they ignore the major step forward made by SE Salmon and his group from the University ofArizona atTucson, USA, during the seventies'. The purpose of this review is to (i) contribute to the original work of SE Salmon as a pioneer in the field of cancer stem cell research (ii) emphasize the importance of his contribution in this field of research and (iii) underline the other fields of his research, especially in the domain of mathematical oncology. Finally, we would like to show that SE Salmon made Multiple Myeloma a model in Oncology, many study groups being engaged into Myeloma research now in the world.

BH3-only protein Bik is involved in both apoptosis induction and sensitivity to oxidative stress in multiple myeloma.

BACKGROUND: although gene expression profile of multiple myeloma (MM) patients shows a wide range of Bik/Nbk expression, varying from absent to high, its regulation and function in myeloma cells is poorly understood. Thus, we addressed these questions in MM. METHODS: human myeloma cell lines (HMCLs) and primary purified myeloma cells were studied for Bcl-2 family protein expression by western blot and further correlation analysis was performed. Correlative study between Bik and thyrotroph embryonic factor (TEF) transcription factor expression was analysed by PCR. Stress oxidative response was analysed by flow cytometry. RESULTS: a strong expression of Bik protein was found only in one out of three of HMCL and correlated to Bcl-2 expression (P=0.0006). We demonstrated that Bik could be regulated at the protein level by Bcl-2 and at the transcriptional level by TEF. Bik overexpression sensitises myeloma cells to oxidative stress whereas Bik silencing increases resistance to H(2)O(2) oxidative stress. Furthermore, Bik ectopic expression disrupts Bim/Bcl-2 and Bim/Bcl-xL endogenous complexes triggering Bim release that could induce Bax and Bak activation. CONCLUSIONS: ours results suggest that Bik has a role in both, apoptosis induction and sensitivity to oxidative stress in myeloma cells. Small BH3 mimetic molecules should be considered for further apoptosis-based therapy in myeloma cells expressing endogenous Bik/Bcl-2 complexes.

Validation of UBE2C protein as a prognostic marker in node-positive breast cancer.

BACKGROUND: We recently identified and validated UBE2C RNA as a prognostic marker in 252 node-positive (N+) breast cancers by means of a microarray study. The aim of this study was to validate UBE2C protein as a prognostic marker in N+ breast cancer by immunohistochemistry (IHC). METHODS: To this end, 92 paraffin-embedded blocks were used. The impact of UBE2C IHC value on metastasis-free survival (MFS) and overall survival (OS) was evaluated and compared with Ki-67 and Nottingham prognostic index (NPI) performances. RESULTS: In accordance with genomic data, UBE2C IHC had a significant impact both on MFS and OS (hazard ratio=6.79 - P=0.002; hazard ratio=7.14 - P=0.009, respectively). Akaike information criterion proved that the prognostic power of UBE2C IHC was stronger than that of Ki-67 (and close to that of NPI). Furthermore, multivariate analyses with NPI showed that, contrary to Ki-67 IHC, UBE2C IHC remained an independent factor, both for MFS (adjusted P=0.02) and OS (adjusted P=0.04). CONCLUSION: We confirmed that UBE2C protein measured by IHC could be used as a prognostic marker in N+ breast cancer. The potential predictive interest of UBE2C as a marker of proteasome activity needs further investigations.

A humanised anti-IGF-1R monoclonal antibody (AVE1642) enhances Bortezomib-induced apoptosis in myeloma cells lacking CD45.

The humanised form of an antagonistic anti-IGF-1R mAb (AVE1642) selectively inhibits the growth of CD45(neg) myeloma cells. AVE1642 strongly increased bortezomib-induced apoptosis, correlated with an increase of Noxa expression. These results support the therapeutic use of anti-IGF-1R/bortezomib in CD45(neg) Myeloma patients, particularly those with the most aggressive form, t(4,14).

Significant impact of survivin on myeloma cell growth.

Survivin is a fascinating member of the inhibitor of apoptosis protein (IAP) family with its dual roles in mitosis and apoptosis, and emerges as an attractive target for cancer therapy. Multiple myeloma (MM) is a plasma cell malignancy, characterized by deregulated proliferation, cell-death processes and fatal outcome. We thus investigated survivin expression in myeloma cells and its role in MM biology to evaluate its potential interest as a target in MM treatment. Our results describe the cancer-specific overexpression of survivin in myeloma cells and show a significant correlation between survivin expression at protein level and clinical course of MM. Moreover, survivin knockdown by RNA interference led to growth rate inhibition of myeloma cells related to apoptosis induction and deep cell-cycle disruption. Finally, survivin knockdown sensitized myeloma cells to conventional anti-myeloma agents. Altogether, these data argue for the interest to evaluate survivin antagonists in MM treatment.

Pathogen-associated molecular patterns are growth and survival factors for human myeloma cells through Toll-like receptors.

Multiple myeloma (MM) patients are strongly vulnerable to infections, which remain a major cause of death. During infection, human immune cells sense the presence of invading pathogens through the Toll-like receptor family (TLR), which recognizes pathogen-associated molecular patterns (PAMP). We hypothesized that MM cells also could sense the presence of microorganisms, thus promoting myeloma disease progression. Here, we report that human myeloma cell lines (HMCL) and primary myeloma cells express a broad range of TLR, and are sensitive to the corresponding PAMP. Toll-like receptor 1, 7 and 9 are most frequently expressed by HMCL. The expression pattern of TLR does not correlate with the one of B cells, as TLR2 and 10 are lost while TLR3, 4 and 8 are acquired by some HMCL. Culture with TLR7- and TLR9-ligands saves HMCL from serum-deprivation or dexamethasone-induced apoptosis. Similarly, both ligands increase myeloma cell growth. These effects are mediated by an autocrine secretion of interleukin-6 (IL-6) since the neutralization of IL-6 blocks the growth and survival of HMCL. Thus, TLR expression and function are not restricted to the cells of the immune system and could be of advantage for cancer cells. In MM, recurrent infections could promote tumor growth and favor escape from standard therapies.

Mcl-1 is overexpressed in multiple myeloma and associated with relapse and shorter survival.

We and others have shown that Mcl-1 was essential for the survival of human myeloma cells in vitro. Furthermore, this antiapoptotic protein is upregulated by interleukin-6, which plays a critical role in multiple myeloma (MM). For these reasons, we have evaluated the expression of Mcl-1 in vivo in normal, reactive and malignant plasma cells (PC), that is, myeloma cells from 51 patients with MM and 21 human myeloma cell lines (HMCL) using flow cytometry. We show that Mcl-1 is overexpressed in MM in comparison with normal bone marrow PC. In total, 52% of patients with MM at diagnosis (P=0.017) and 81% at relapse (P=0.014 for comparison with diagnosis) overexpress Mcl-1. Of note, only HMCL but not reactive plasmacytoses have abnormal Mcl-1 expression, although both PC expansions share similar high proliferation rates. Of interest, Bcl-2 as opposed to Mcl-1, does not discriminate malignant from normal PC. Finally, the level of Mcl-1 expression is related to disease severity, the highest values at diagnosis being associated with the shortest event-free survival (P=0.002). In conclusion, Mcl-1, which has been shown to be essential for the survival of human myeloma cells in vitro, is overexpressed in vivo in MM in relation with relapse and shorter survival. Mcl-1 represents a potential therapeutical target in MM.

Normal and malignant human plasma cells: proliferation, differentiation, and expansions in relation to CD45 expression.

In this study, we have evaluated the proliferation and the phenotype of human plasma cells of different origins, i.e., from tonsil, peripheral blood, bone marrow as well as plasma cells generated in vitro from memory B cells. We have demonstrated that plasma cells from tonsil, peripheral blood, as well as those generated in vitro, were highly proliferating and presented a homogeneous CD45bright phenotype. In contrast, bone marrow plasma cells were heterogeneous for CD45 expression but their proliferation was restricted to the CD45bright compartment. Subsequently, their CD45 expression decreased with proliferation arrest and final maturation. We also studied the proliferation of abnormal plasma cells, i.e., peripheral blood reactive plasmacytoses and multiple myeloma (MM). All reactive plasmacytoses turned out to be homogeneous expansions of CD45bright plasma cells with unusually high labeling index. In contrast, CD45 expression was heterogeneous in MM as in normal bone marrow. However, a minor CD45bright population was also always the most proliferating one as opposed to a major population of less or non-proliferating myeloma cells characterized by a weaker or a lack of CD45 expression. In conclusion, proliferation is linked to plasma-cell generation and a CD45bright phenotype is the hallmark of the most proliferating normal, reactive as well as malignant plasma cells.

t(11;14) and t(4;14) translocations correlated with mature lymphoplasmacytoid and immature morphology, respectively, in multiple myeloma.

Recent studies have shown that two recurrent translocations, t(4;14)(p16;q32) and t(11;14)(q13;q32), define distinct entities with different prognosis in multiple myeloma (MM). We addressed the issue of whether these illegitimate IGH rearrangements could contribute to the morphological heterogeneity of the malignant plasma cells (PC). Bone marrow aspirates of 178 untreated MM cases with successful molecular cytogenetics analysis using fluorescence in situ hybridization were reviewed. PC of 25/48 (52%) patients with t(11;14) exhibited a lymphoplasmacytoid morphology. Moreover, 25/27 (93%) of the cases with this morphological profile bore the t(11;14). In addition, both cytogenetics and morphological subtypes shared higher incidence of nonsecretory MM. In contrast, 17 out of 28 cases (61%) with t(4;14) exhibited PC with diffuse chromatin pattern. Interestingly, both t(4;14) translocation and immature morphology correlated with higher incidence of high tumor mass and chromosome 13 abnormality. In conclusion, our results suggest that a particular morphology can be the signature of chromosomal abnormalities in MM.

Qualitative and quantitative analysis of human herpesviruses in chronic and acute B cell lymphocytic leukemia and in multiple myeloma.

Real-time quantitative polymerase chain reaction (qPCR) was used to quantify viral loads of human herpesviruses (HHVs) at diagnosis in 61 samples of malignant B cells: 21 chronic lymphocytic leukemia (B-CLL), 29 acute lymphoblastic leukemia (B-ALL) and 11 multiple myeloma (MM); control samples were blasts from 16 acute myeloid leukemia (AML) and 24 blood or bone marrow samples from healthy donors. The majority of samples from healthy donors and patients (B-ALL, B-CLL or AML, but not MM) was positive for EBV and contained <100 ebv copies/100 ng dna. ebv loads were occasionally high (>500 copies/100 ng DNA) in B-ALL (2/16) and in B-CLL (2/21) samples. The fractions of samples positive for HHV-8 and HHV-6A, less than 10% for MM patients, were high for adults with B-ALL (18.8% HHV-8+, 43.8% HHV-6A+) or B-CLL (28.6% HHV-8+, 52.4% HHV-6A+). B-ALL, B-CLL and MM samples were rarely positive for HHV-6B and HHV-7. Lastly, 75% of B-ALL samples were positive for CMV, and CMV loads were significantly higher in B-ALL samples than in MM, B-CLL or AML samples. We also used PCR with consensus-degenerate hybrid oligonucleotide primers (CODEHOP) to look for novel HHVs in B cell samples: no sequence indicative of a new HHV was detected. Altogether, the data indicate that the presence of multiple HHVs, including EBV and CMV at high loads, is not rare in B-ALL and B-CLL cell samples. Future prospective studies should determine whether patients with high EBV/CMV loads at diagnosis in tumor samples face a higher risk of delayed hematological recovery, virus-related complications or relapse.

Farnesyl transferase inhibitor R115777 induces apoptosis of human myeloma cells.

R115777, a nonpeptidomimetic farnesyl transferase inhibitor has recently demonstrated a significant antileukemic activity in vivo in acute myeloid leukemia. Multiple myeloma (MM) is a fatal hematological malignancy characterized by an accumulation of long-lived plasma cells within the bone marrow. In the present study, we have investigated the effect of the R115777 on growth and survival of myeloma cells. We have found that R115777 induced (1) a significant and dose-dependent growth inhibition of the three myeloma cell lines tested; and (2) a significant and time-dependent apoptosis. R115777 also induced apoptosis in the bone marrow mononuclear cell population of four MM patients, being almost restricted to the malignant plasma cells. Finally, we have investigated the effect of the R115777 in the Ras/MAPK and JAK/STAT pathways which are implicated in survival and/or proliferation in MM. The phosphorylation of both STAT3 and ERK1/2 induced by IL-6 was totally blocked at 15 microM of R115777 and partially blocked when R115777 was used at 10 and 5 microM. The induction of apoptosis by R115777 in myeloma cells and its implication in the regulation of JAK/STAT signalling suggest that R115777 might be an interesting therapeutical approach in MM.

Large deletions 5' to the ETO breakpoint are recurrent events in patients with t(8;21) acute myeloid leukemia.

Recurrent chromosomal rearrangements are observed in many leukemia subtypes. Recently, it has been shown that several of these translocations/inversions were associated with the loss of sequences located in the vicinity of the chromosomal breakpoints. So far, such deletions have not been described for the t(8;21) translocation. We have analyzed a series of 65 patients with t(8;21) using several probes specific for the ETO and AML1 regions. We have found six patients (9%) with deletion of the region 5' to ETO. In all six patients, the deletion encompassed at least 260 kb, and was even larger in two patients (up to 2 Mb). A similar analysis of the 21q22 region did not reveal any deletion of the 3'AML1 region. In conclusion, cytogenetically undetectable small deletions located immediately 5' to the ETO breakpoint were found to accompany the t(8;21) translocation in a significant percentage of cases. The clinical significance, if any, of these deletions remains to be determined.

Long-term results (12 years) of high-dose therapy in 127 patients with de novo multiple myeloma.

This report describes the long-term outcome of a cohort of 127 de novo multiple myeloma patients treated with at least one course of high-dose therapy (HDT) in a single institution between June 1985 and December 1995, for whom the minimum follow-up duration for survivors is 6 years. The 12-year overall survival (OS) and event-free survival (EFS) rates are 24.9% and 3.1%, respectively, and the median survival and EFS are 49 and 17 months, respectively. Only four patients are alive and disease-free 79, 90, 132 and 153 after the first HDT, respectively. Three of them received a subsequent allogeneic bone marrow transplantation. Three factors significantly influence OS in this series: B2M at diagnosis, age, and the completion of a second HDT. The 10-year survival is 18.9% for the group of patients with B2M level >3 mg/l at diagnosis as compared with 41% for patients with B2M < or =3, with a median survival of 31 months vs 73 (P = 0.01). The 10-year survival is 23.4% for the group of patients aged >55 years as compared with 36.5% for patients aged <55 years, with a median survival of 34.5 months vs 70.5 (P = 0.04). The 10-year survival is 20.4% for the group of patients who did not receive a second HDT as compared with 35.2% for patients who completed a second HDT, with a median survival of 29 months vs 70 (P = 0.02). In this study we show that some patients treated with HDT experience durable remission and prolonged survival. This survival is significantly influenced by age (< or =55 years), B2M at diagnosis (< or =3 mg/l) and by the completion of two cycles of HDT.

Amplification of AML1 gene is present in childhood acute lymphoblastic leukemia but not in adult, and is not associated with AML1 gene mutation.

The AML1/CBFA2/RUNX1 gene is the target of many recurrent translocations seen in different leukemia subtypes. The t(12;21)(p13;q22) is the most frequent translocation observed in childhood B acute lymphoblastic leukemia (ALL), occurring in 20% to 25% of cases. In adult ALL this rearrangement is scarce. Another route of AML1deregulation could be point mutations in the runt domain. We now report on AML1amplification in two cases of childhood ALL, found in a series of 107 consecutive children with B-lineage ALL analyzed by fluorescence in situ hybridization (FISH). A parallel analysis of 42 adult B-ALL failed to detect any AML1 rearrangement by FISH. The two patients with AML1 amplification were further analyzed using molecular techniques. SSCP analysis did not detect any mutation. Furthermore, direct sequencing of the cDNA did not reveal any mutation. In conclusion, AML1amplification seems to be observed only in childhood ALL and is not associated with AML1 gene mutation. Other mechanisms, such as gene dosage effects could be hypothesized.

Myeloma cells induce imbalance in the osteoprotegerin/osteoprotegerin ligand system in the human bone marrow environment.

Although osteolysis is a common complication in patients with multiple myeloma (MM), the biologic mechanisms involved in the pathogenesis of MM-induced bone disease are poorly understood. Two factors produced by stromal-osteoblastic cells seem critical to the regulation of bone resorption: osteoprotegerin (OPG) and its ligand (OPGL). OPGL stimulates osteoclast differentiation and activity, whereas OPG inhibits these processes. The present study investigated whether myeloma cells affect physiologic OPG/OPGL balance in the bone marrow (BM) environment. Ten human myeloma cell lines and myeloma cells isolated from 26 consecutive patients with MM failed to express OPGL and only rarely produced a low amount of OPG. In a coculture system, human myeloma cells up-regulated OPGL expression but strongly down-regulated OPG production in preosteoblastic (preOB) or stromal cells (BMSCs) of primary human BM at the mRNA and protein levels. This effect, which was dependent on cell-to-cell contact between myeloma cells and BMSCs or preOB, partially involved the integrin VLA-4. In addition, overexpression of OPGL mRNA occurred in ex vivo BM cultures obtained from MM patients as compared with healthy donors, and immunohistochemical staining performed on BM biopsy specimens showed an increase of OPGL and a reduction of OPG expression in MM patients as compared with healthy subjects. In summary, these data indicate that myeloma cells affect the OPG/OPGL ratio in the BM environment and tend to confirm that the OPG/OPGL system is involved in the pathogenesis of MM-induced bone disease.

Rearrangements of the c-myc oncogene are present in 15% of primary human multiple myeloma tumors.

Rearrangements of the c-myc oncogene have been found in most plasmacytomas induced in mice and human myeloma cell lines (HMCLs) analyzed so far. However, neither induced mouse plasmacytomas nor HMCLs represent relevant models for human multiple myeloma (MM). To evaluate the incidence of c-myc rearrangements in human plasma cell dyscrasias, sets of probes were generated to allow direct assessment of c-myc translocations on interphase plasma cells by using fluorescence in situ hybridization. After validation of these probes, a large cohort of patients with either newly diagnosed MM (n = 529), relapsed MM (n = 58), primary plasma cell leukemia (PCL; n = 23), monoclonal gammopathy of undetermined significance (n = 65), or smoldering MM (n = 24) were analyzed. C-myc rearrangements were identified in 15% of patients with MM or primary PCL, independently of the stage of the disease (ie, diagnosis or relapse and MM or primary PCL). Analysis of the 2 main translocations observed on karyotyping, ie, t(8;14) and t(8;22), revealed that these specific translocations represented only 25% (23 of 91) of c-myc rearrangements. c-myc rearrangements were then correlated with several other patients' characteristics: illegitimate IgH recombinations, chromosome 13 deletions, and serum beta2-microglobulin levels. The only significant correlation was with a high beta2-microglobulin level (P =.002), although a trend for association with t(4;14) was observed (P =.08). Thus, c-myc rearrangement analysis in patients with MM revealed a strikingly lower incidence than that in HMCLs and plasmacytomas induced in mice, indicating that data obtained with these models cannot be directly extrapolated to human MM.