Towards the Segmentation and Classification of White Blood Cell Cancer Using Hybrid Mask-Recurrent Neural Network and Transfer Learning

Inside the bone marrow, plasma cells are created, and they are a type of white blood cells. They are made from B lymphocytes. Antigens are produced by plasma cells to combat bacteria and viruses and prevent inflammation and illness. Multiple myeloma is a plasma cell cancer that starts in the bone marrow and causes the formation of abnormal plasma cells. Multiple myeloma is firmly identified by examining bone marrow samples under a microscope for myeloma cells. To diagnose myeloma cells, pathologists have to be very selective. Furthermore, because the ultimate decision is based on human sight and opinion, there is a possibility of error in the result. The nobility of this research is that it provides a computer-assisted technique for recognizing and detecting myeloma cells in bone marrow smears. For recognizing purposes, we have used Mask-Recurrent Convolutional Neural Network, and for detection purposes, Efficient Net B3 has been used. There are already many studies on white blood cell cancer, but very few with both segmentation and classification. We have designed two models. One is for recognizing myeloma cells, and the other is for differentiating them from nonmyeloma cells. Also, a new data set has been made from the multiple myeloma data sets, which has been used in our classification model. This research focuses on hybrid segmentation models and increases the accuracy level of the classification model. Both of our models are trained pretty well, where the Mask-RCNN model gives a mean average precision (mAP) of 93% and the Efficient Net B3 model gives 94.68% accuracy. The result of this research indicates that the Mask-RCNN model can recognize multiple myeloma and Efficient Net B3 can distinguish between myeloma and nonmyeloma cells and beats most of the state of the art in myeloma recognition and detection.

Download Full-text

Amphiregulin Is Produced by Myeloma Cells and Is Involved in Tumor-Environment Interactions in Multiple Myeloma.

Blood ◽

10.1182/blood.v104.11.4306.4306 ◽

2004 ◽

Vol 104 (11) ◽

pp. 4306-4306

Author(s):

Karène Mahtouk ◽

Dirk Hose ◽

Thierry Reme ◽

John De Vos ◽

Michel Jourdan ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Growth Factors ◽

Plasma Cells ◽

Myeloma Cell ◽

Erbb Receptors ◽

Heparin Binding ◽

Myeloma Cells ◽

Heparin Binding Growth Factors ◽

Egf Family

Abstract Multiple myeloma (MM) is characterized by the accumulation of clonal malignant plasma cells in the bone marrow. One of the hallmarks of plasma cells is the expression of the heparan-sulfate proteoglycan syndecan-1. In epithelial cells, syndecan-1 plays a major role as a coreceptor for heparin-binding growth factors and chemokines. This stresses that heparin-binding growth factors may play a major role in the biology of MM cells. Recently we have demonstrated that heparin-binding EGF-like growth factor (HB-EGF), one of the ten members of the Epidermal Growth Factor (EGF) family, is produced by the tumor microenvironment and is able to trigger myeloma cell growth. As amphiregulin (AREG) is another member of the EGF family that also binds heparan-sulphate chains, we investigated its role in MM. We looked for AREG expression on a panel of 7 normal plasmablastic cells (PPCs), 7 normal bone marrow plasma cells (BMPCs), purified MM cells from 65 patients and 20 myeloma cell lines (HMCLs), with Affymetrix U133A+B microarrays. We showed that primary MM cells overexpress AREG compared to normal BMPCs and PPCs. We then investigated the expression of the ErbB receptors with real-time RT-PCR. Myeloma cells variably expressed the 4 ErbB receptors. Normal BMPCs also expressed ErbB1 and ErbB2 unlike PPCs that did not express any ErbB receptors. We demonstrated that the high AREG expression by primary myeloma cells may have a dual effect. On the one hand, AREG stimulated IL-6 production and growth of bone-marrow stromal cells that highly express the AREG ErbB1 receptor. On the other hand, AREG could promote HMCL proliferation, suggesting that a functional autocrine loop involving AREG and ErbB receptors is involved in MM cell growth. Finally, we looked for the effect of ErbB inhibitors on MM cells of 14 patients cultured for 6 days together with their bone marrow environment. A pan-ErbB inhibitor (PD-169540, Pfizer) and an ErbB1-inhibitor (IRESSA, Astrazeneca) induced strong MM cell apoptosis in respectively 71% of patients (10 of 14) and 29% of patients (4 of 14). Of major interest, when PD169540 or IRESSA were combined with dexamethasone, they induced a dramatic myeloma cell death (respectively 92% and 69% inhibition of MM cell survival), while non-myeloma cells were unaffected. Thus ErbB activation is critical to trigger MM-cell survival in short-term culture. In conclusion, our findings provide evidence for a major role of AREG and HB-EGF in the biology of multiple myeloma and identify ErbB receptors as putative therapeutic targets. These data emphasize the interest of clinical evaluation of specific-ErbB-inhibitors in patients with MM, either used alone or in combination with dexamethasone.

Download Full-text

CS1: A Potential New Therapeutic Target for the Treatment of Multiple Myeloma.

Blood ◽

10.1182/blood.v108.11.3457.3457 ◽

2006 ◽

Vol 108 (11) ◽

pp. 3457-3457 ◽

Cited By ~ 3

Author(s):

Eric D. Hsi ◽

Roxanne Steinle ◽

Balaji Balasa ◽

Aparna Draksharapu ◽

Benny Shum ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Flow Cytometry ◽

B Cell ◽

Mononuclear Cells ◽

Plasma Cells ◽

Target Cells ◽

Dependent Manner ◽

Myeloma Cells ◽

Effector Cells

Abstract Background: To identify genes upregulated in human memory B and plasma cells, naïve B cell cDNA was subtracted from plasma cell and memory B cell cDNA. One gene that was highly expressed in plasma cells encodes CS1 (CD2 subset 1, CRACC, SLAMF7), a cell surface glycoprotein of the CD2 family. CS1 was originally identified as a natural killer (NK) cell marker. Monoclonal antibodies (mAbs) specific for CS1 were used to validate CS1 as a potential target for the treatment of multiple myeloma (MM). Methods: Anti-CS1 mAbs were generated by immunizing mice with a protein comprising of the extracellular domain of CS1. Two clones, MuLuc63 and MuLuc90, were selected to characterize CS1 protein expression in normal and diseased tissues and blood. Fresh frozen tissue analysis was performed by immunohistochemistry (IHC). Blood and bone marrow analysis was performed using flow cytometry with directly conjugated antibodies. HuLuc63, a novel humanized anti-CS1 mAb (derived from MuLuc63) was used for functional characterization in non-isotopic LDH-based antibody-dependent cellular cytotoxicity (ADCC) assays. Results: IHC analysis showed that anti-CS1 staining occurred only on mononuclear cells within tissues. The majority of the mononuclear cells were identified as tissue plasma cells by co-staining with anti-CD138 antibodies. No anti-CS1 staining was detected on the epithelia, smooth muscle cells or vessels of any normal tissues tested. Strong anti-CS1 staining was also observed on myeloma cells in 9 of 9 plasmacytomas tested. Flow cytometry analysis of whole blood from both normal healthy donors and MM patients showed specific anti-CS1 staining in a subset of leukocytes, consisting primarily of CD3−CD(16+56)+ NK cells, CD3+CD(16+56)+ NKT cells, and CD3+CD8+ T cells. Flow cytometry of MM bone marrow showed a similar leukocyte subset staining pattern, except that strong staining was also observed on the majority of CD138+CD45−/dim to + myeloma cells. No anti-CS1 binding was detected to hematopoietic CD34+CD45+ stem cells. To test if antibodies towards CS1 may have anti-tumor cell activity in vitro, ADCC studies using effector cells (peripheral blood mononuclear cells) from 23 MM patients and L363 MM target cells were performed. The results showed that HuLuc63, a humanized form of MuLuc63, induced significant ADCC in a dose dependent manner. Conclusions: Our study identifies CS1 as an antigen that is uniformly expressed on normal and neoplastic plasma cells at high levels. The novel humanized anti-CS1 mAb, HuLuc63, exhibits significant ADCC using MM patient effector cells. These results demonstrate that HuLuc63 could be a potential new treatment for multiple myeloma. HuLuc63 will be entering a phase I clinical study for multiple myeloma.

Download Full-text

Targeting the Serine-Threonine Kinase PIM2: Implications in IL-6 Mediated Survival in Myeloma

Blood ◽

10.1182/blood.v124.21.3401.3401 ◽

2014 ◽

Vol 124 (21) ◽

pp. 3401-3401

Author(s):

Jayakumar R Nair ◽

Tyger L Howell ◽

Justin Caserta ◽

Carmen M Baldino ◽

Gerald Fetterly ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Drug Resistance ◽

Cell Death ◽

Plasma Cells ◽

Serine Threonine Kinase ◽

Myeloma Cells ◽

Threonine Kinase ◽

Equity Ownership ◽

Microarray Expression

Abstract Despite major advances in chemotherapy, multiple myeloma remains incurable and in need of new therapies that target novel pathways. Insufficient understanding of the molecular pathways that regulate survival in myeloma is a major impediment towards designing better therapies to prolong survival in patients or even cure the disease. This necessitates the identification of new protein targets that are crucial for the growth and survival of multiple myeloma. Just like normal plasma cells, MM cells also depend on their interactions with bone marrow stromal cells (BMSC) for survival and production of essential growth factors. We have previously shown that MM cells interact with dendritic cells (DC) in the microenvironment and in vitro can stimulate DC to produce IL-6 (ASH2010#132, ASH2011 #147, ASH2012#722). Our recent publications show that when MM cells are not in direct contact with DC, the IL-6 produced by DC can protect MM cells against dexamethasone induced cell death, while neutralizing the IL-6 with antibodies can reverse that effect (Nair et al., 2011). Unfortunately, exactly how this survival response is mediated in MM is not very clear. PIM2, a serine threonine kinase, part of the proto-oncogene group of PIM kinases has been implicated in survival in several types of cancers including prostate cancer and multiple myeloma. In our lab, microarray gene expression analysis of publicly available datasets (Figure 1) show a trend towards increased expression of PIM2 in plasma cells from myeloma patients (left panel), and significantly in the poor prognosis subgroup MAF (Zhan et al., 2006) (right panel). For the first time we show that IL-6 produced by DC may be protecting myeloma cells by up regulating PIM2 and inactivating a major protein translation inhibitor 4EBP1, which also happens to be a PIM2 target. We show that silencing PIM2 with siRNA down regulates PIM2 activity and reverses the inactivation of 4EBP1, while the latter is known to cause cell death in myeloma. We also demonstrate that neutralizing IL-6 in MM cells that either don’t produce IL-6 on their own (MM.1S) or those that do (U266), abrogates extraneous DC-IL6 ability to induce PIM2 and its downstream target 4EBP1. Recombinant IL-6 also provided similar induction of PIM2 in myeloma and increased 4EBP1 phosphorylation, which was again reversed by neutralizing the antibody against IL-6. In myeloma patients, the use of dexamethasone in frontline therapies is often complicated by the ability of the bone marrow environment to produce IL-6 that not only induce increased proliferation of MM but also help resist dexamethasone mediated cell death in myeloma. Interestingly, when we used a novel PIM2 inhibitor, JP_11646 (kindly provided by Jasco Pharmaceuticals, LLC), it not only arrested IL-6 induced proliferation even at sub-lethal doses, but also prevented IL-6 mediated rescue of myeloma cells (Figure 2). This suggests that PIM2 might be a major player in IL-6 mediated drug resistance in myeloma and targeting it may help to subvert IL-6 mediated survival in myeloma. Through RT-PCR and westerns, we also show that IL-6 modulates PIM2 expression and activity resulting in increased 4EBP1 phosphorylation (Figure 3). This was abrogated when PIM2 activity was inhibited by JP_11646 (Figure 3). We also present data that suggests IL-6 via PIM2 may be regulating other anti-apoptotic molecules downstream of IL-6 receptors including MCL-1, that is vital to MM survival. Developing PIM2 targeted therapies provides an exciting opportunity to affect the myeloma tumor microenvironment where MM induced IL-6 production from BM could be inducing drug resistance. Figure 1: Microarray expression ofPIM2 in myeloma and MAF Figure 1:. Microarray expression ofPIM2 in myeloma and MAF Figure 2: PIM2 inhibition abrogates IL-6 induced MM proliferation (A) and protection (B). Figure 2:. PIM2 inhibition abrogates IL-6 induced MM proliferation (A) and protection (B). Figure 3: Inhibiting PIM2 activity prevents PIM2 induced phosphorylation of 4EBP1 by IL-6 in myeloma Figure 3:. Inhibiting PIM2 activity prevents PIM2 induced phosphorylation of 4EBP1 by IL-6 in myeloma Disclosures Caserta: Jasco Pharmaceuticals LLC: Equity Ownership. Baldino:Jasco Pharmaceuticals LLC: Equity Ownership.

Download Full-text

Targeting EZH2 in Multiple Myeloma Could be Promising for a Subgroup of MM Patients in Combination with IMiDs

Blood ◽

10.1182/blood.v128.22.311.311 ◽

2016 ◽

Vol 128 (22) ◽

pp. 311-311 ◽

Cited By ~ 4

Author(s):

Laurie Herviou ◽

Alboukadel Kassambara ◽

Stephanie Boireau ◽

Nicolas Robert ◽

Guilhem Requirand ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Board Of Directors ◽

Research Funding ◽

Plasma Cells ◽

Newly Diagnosed ◽

Advisory Committees ◽

Myeloma Cells ◽

Ezh2 Expression ◽

Bone Marrow Plasma

Abstract Multiple Myeloma is a B cell neoplasia characterized by the accumulation of clonal plasma cells within the bone marrow.Epigenetics is characterized by a wide range of changes that are reversible and orchestrate gene expression. Recent studies have shown that epigenetic modifications play a role in multiple myeloma (MM) by silencing various cancer-related genes. We investigated the epigenetic genes differentially expressed between normal bone marrow plasma cells (BMPC ; N=5) and MM plasma cells from patients (N=206). Using SAM (Significance Analysis of Microarrays) analysis, only 12 genes significantly differentially expressed between BMPC and MM cells (ratio > 2 and FDR (false discovery rate) < 5%) were identified, including the EZH2 histone methyltransferase. EZH2, the enzymatic subunit of Polycomb Repressive Complex 2, is a histone methyltransferases able to repress gene expression by catalyzing H3K27me3 histone mark. EZH2 overexpression has been associated with numerous hematological malignancies, including MM. We thus studied EZH2 role in MM physiopathology and drug resistance. EZH2 expression was analyzed in normal bone marrow plasma cells (BMPCs; N=5), primary myeloma cells from newly diagnosed patients (MMCs; N=206) and human myeloma cell lines (HMCLs; N=40) using Affymetrix microarrays. EZH2 gene is significantly overexpressed in MMCs of patients (median 574, range 105 - 4562) compared to normal BMPCs (median = 432; range: 314 - 563) (P < 0.01). The expression is even higher in HMCLs (median 4481, range 581 - 8455) compared to primary MMCs or BMPCs (P < 0.001). High EZH2 expression is associated with a poor prognosis in 3 independent cohorts of newly diagnosed patients (Heidelberg-Montpellier cohort - N=206, UAMS-TT2 cohort - N=345 and UAMS-TT3 cohort - N =158). Furthermore, GSEA analysis of patients with high EZH2 expression highlighted a significant enrichment of genes involved in cell cycle, downregulated in mature plasma cells vs plasmablasts, and EZH2 targets. Specific EZH2 inhibition by EPZ-6438 EZH2 inhibitor induced a significant decrease of global H3K27me3 in all the HMCLs tested (P < 0.01) and inhibited MM cell growth in 5 out of the 6 HMCLs tested. The inhibitory effect of EZH2 inhibitor on MM cell growth appeared at day 6 suggesting that it is mediated by epigenetic reprogramming. To confirm that EZH2 is also required for the survival of primary MMCs from patients, primary MM cells (n = 17 patients) co-cultured with their bone marrow microenvironment and recombinant IL-6 were treated with EPZ-6438. As identified in HMCLs, EZH2 inhibition significantly reduced the median number of viable myeloma cells by 35% (P = 0.004) from a subset of patients (n=9) while the other group (n=8) was resistant. Of interest, EPZ-6438 induced a significant global H3K27me3 decrease in both groups of patient. RNA sequencing of 6 HMCLs treated with EPZ-6438 combined with H3K27me3 ChIP analyses allowed us to create an EZ GEP-based score able to predict HMCLs and primary MM cells sensitivity to EZH2 inhibitors. We also observed a synergy between EPZ-6438 and Lenalidomide, a conventional drug used for MM treatment. More interestingly, pretreatment of myeloma cells with EPZ-6438 significantly re-sensitize drug-resistant MM cells to Lenalidomide. Investigating the effect of EPZ-6438/Lenalidomide combination in MMC, we identified that IKZF1, IRF4 and MYC protein levels were significantly more inhibited by the combination treatment (65.5%, 63.9% and 14.8% respectively) compared with Lenalidomide (51.5%, 43% and 2.2%) or EPZ-6438 (45.2%, 38.7% and 6.2%) alone. Clinical trials are ongoing with EZH2 inhibitors in lymphoma and could be promising for a subgroup of MM patients in combination with IMiDs. Furthermore, the EZ score enables identification of MM patients with an adverse prognosis and who could benefit from treatment with EZH2 inhibitors. Disclosures Goldschmidt: Celgene: Membership on an entity's Board of Directors or advisory committees, Research Funding; Onyx: Honoraria, Membership on an entity's Board of Directors or advisory committees; Bristol-Myers Squibb: Membership on an entity's Board of Directors or advisory committees, Research Funding; Millennium: Membership on an entity's Board of Directors or advisory committees, Research Funding; Chugai: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding; Takeda: Membership on an entity's Board of Directors or advisory committees; Amgen: Membership on an entity's Board of Directors or advisory committees. Hose:EngMab: Research Funding; Takeda: Other: Travel grant; Sanofi: Research Funding.

Download Full-text

A Role for Syntenin-1 in Multiple Myeloma Cell Survival

Blood ◽

10.1182/blood-2018-99-113594 ◽

2018 ◽

Vol 132 (Supplement 1) ◽

pp. 1008-1008

Author(s):

Tyler Moser-Katz ◽

Catherine M. Gavile ◽

Benjamin G Barwick ◽

Sagar Lonial ◽

Lawrence H. Boise

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Cell Death ◽

Cell Survival ◽

Cell Lines ◽

Plasma Cells ◽

Surface Expression ◽

Myeloma Cell ◽

Myeloma Cells ◽

Rpmi 8226

Abstract Multiple myeloma is the second most common hematological malignancy in the U.S. with an estimated 30,700 new diagnoses in 2018. It is a clonal disease of plasma cells that, despite recent therapeutic advances, remains incurable. Myeloma cells retain numerous characteristics of normal plasma cells including reliance on survival signals in the bone marrow for long term viability. However, malignant transformation of plasma cells imparts the ability to proliferate, causing harmful bone lesions in patients, and in advanced stages independence of the bone-marrow microenvironment. Therefore, we are investigating the molecular mechanisms of myeloma cell survival that allow them to become extramedullary. We identified syntenin-1 (SDCBP) as a protein involved in myeloma cell survival and a potential therapeutic target. Syntenin-1 is an adapter protein that has been shown to regulate surface expression of several transmembrane proteins by binding with membrane phospholipids and mediating vesicular trafficking of proteins throughout the cell. Syntenin-1 regulates the surface expression of CD138, a plasma/myeloma cell marker. Syntenin-1 has been shown to regulate apoptosis in numerous cancer cell lines including breast cancer, glioma, and pancreatic cancer but its role in multiple myeloma survival has not been studied. To determine if syntenin-1 expression has an effect on myeloma cell survival, we utilized the CoMMpass dataset (IA12), a longitudinal study of myeloma patients that includes transcriptomic analysis throughout treatment. We found that patients with the highest expression of syntenin-1 mRNA (top quartile) had significantly worse overall survival, progression-free survival, and a shorter response duration than those in the bottom quartile of expression. To determine if syntenin-1 has a role in myeloma cell survival, we used short hairpin RNA to knock down syntenin-1 (shsyn) in RPMI 8226 and MM1.s myeloma cell lines. We then determined the amount of cell death using Annexin-V staining flow cytometry four days following lentiviral infection. We found increased cell death in syntenin-1-silenced cells compared to our empty vector control in both RPMI 8226 (control=42.17%, shsyn=71.53%, p=0.04) and MM1.s cell lines (control=8.57%, shsyn=29.9%, p=0.04) suggesting that syntenin-1 is important for myeloma cell survival. Syntenin-1 contains two PDZ domains that allow it to bind to receptor proteins via their corresponding PDZ-binding motifs. We therefore wanted to look at correlation of syntenin-1 expression with CD138 and CD86, two PDZ-binding domain containing proteins expressed on the surface of myeloma cells. Using the CoMMpass dataset, we found patients with high expression of syntenin-1 had a median expression of CD86 that was twice as high as the total population (P<0.0001) while syntenin-1-low patients expressed CD86 at levels that were half as much as the population (P<0.0001). In contrast, there was no clear relationship between syntenin-1 and CD138 mRNA expression. Indeed if one takes into account all patients, there is a positive correlation between CD86 and syntenin-1 expression (r=0.228, P<0.0001) while there is a negative correlation between CD138 and syntenin-1 (r=-0.1923, P<0.0001). The correlation with CD86 but not CD138 suggests a previously undescribed role for syntenin-1 in myeloma cells. Our lab has previously shown that expression of CD86 is necessary for myeloma cell survival, and signals via its cytoplasmic domain to confer drug resistance. Silencing syntenin-1 results in a decrease in CD86 surface expression. However, there is no change in CD86 transcript or total cellular CD86 protein levels in our shsyn treated cells. Moreover, knockdown of CD86 resulted in increased protein expression and transcript levels of syntenin-1. Taken together, these data suggest that syntenin-1 may regulate CD86 expression on the cell surface. Our data supports a novel role for syntenin-1 in myeloma cell viability and as a potential regulator of CD86 surface expression. The role of syntenin-1 has not previously been explored in multiple myeloma and determining its molecular function is warranted as it may be an attractive target for therapeutic treatment of the disease. Disclosures Lonial: Amgen: Research Funding. Boise:AstraZeneca: Honoraria; Abbvie: Consultancy.

Download Full-text

Mcl-1 Is Overexpressed in Multiple Myeloma and Correlates with Disease Severity.

Blood ◽

10.1182/blood.v104.11.1419.1419 ◽

2004 ◽

Vol 104 (11) ◽

pp. 1419-1419

Author(s):

Soraya Wuilleme-Toumi ◽

Nelly Robillard ◽

Patricia Gomez-Bougie ◽

Philippe Moreau ◽

Steven Le Gouill ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Disease Severity ◽

Cell Lines ◽

Plasma Cells ◽

Myeloma Cell ◽

Lymphocytic Leukemia ◽

Myeloma Cells

Abstract Multiple Myeloma (MM) is a fatal malignancy of B-cell origin characterized by the accumulation of plasma cells within the bone marrow. The expression of the pro-survival members of the Bcl-2 family has been shown to be a key process in the survival of myeloma cells. More particularly, Mcl-1 expression turned out to be critical for their survival. Indeed, knockdown of Mcl-1 by antisenses induces apoptosis in myeloma cells. Finally, Mcl-1 was found to be the only anti-apoptotic Bcl-2 family member which level of expression was modified by cytokine treatment of myeloma cells. For these reasons, we have evaluated the expression of Mcl-1 in vivo in normal, reactive and malignant plasma cells (PC) i.e., myeloma cells from 55 patients with MM and 20 human myeloma cell lines using flow cytometry. We show that Mcl-1 is overexpressed in MM in comparison with normal bone marrow PC. Forty-seven percent of patients with MM at diagnosis (p=.017) and 80% at relapse (p=.014 for comparison with diagnosis) overexpress Mcl-1. Of note, only myeloma cell lines but not reactive plasmocytoses have abnormal Mcl-1 expression, although both plasmocyte expansion entities share similar high proliferation rates (>20%). Of interest, Bcl-2 as opposed to Mcl-1, does not discriminate malignant from normal PC. This shows that the overexpression of Mcl-1 is clearly related to malignancy rather than to proliferation. It will be important to know whether the overexpression of Mcl-1 is related to an abnormal response to cytokines like Interleukin-6 or to mutations of the promoter of the Mcl-1 gene as already described in B chronic lymphocytic leukemia. Finally, level of Mcl-1 expression is related to disease severity, the highest values being correlated with the shortest event-free survival (p=.01). 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 and correlates with disease severity. Mcl-1 represents a major therapeutical target in MM.

Download Full-text

Angiogenesis in Multiple Myeloma (MM): Angiogenic Switch or Reflexion of Plasma Cell Number? A Gene Expression Based Survey in Primary Myeloma Cells and the Bone Marrow Microenvironment.

Blood ◽

10.1182/blood.v108.11.3405.3405 ◽

2006 ◽

Vol 108 (11) ◽

pp. 3405-3405

Author(s):

Dirk Hose ◽

John DeVos ◽

Christiane Heiß ◽

Jean-Francois Rossi ◽

Angela Rösen-Wolff ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Plasma Cell ◽

Empirical Bayes ◽

Plasma Cells ◽

De Novo ◽

Cell Number ◽

Bone Marrow Microenvironment ◽

Myeloma Cells ◽

Angiogenic Switch

Abstract BACKGROUND. Angiogenesis is a hallmark of active multiple myeloma. However, two etiologic hypotheses have been proposed: an angiogenic switch (i.e. differential or de novo expression of pro/antiangiogenic genes in MM), and, alternatively an effect of increased plasma cell number. AIM of this study was to investigate the angiogenic signature of multiple myeloma cells (MMC), normal bone marrow plasma cells (BMPC), the bone marrow microenvironment (BMME) and cellular subfractions therein. PATIENTS AND METHODS. 128 newly diagnosed MM-patients (65 training (TG) / 63 independent validation group (VG)) and 14 normal donors (ND) were included. Bone marrow aspirates were CD138-purified by activated magnetic cell sorting. Whole bone marrow (n=49) and FACSAria sorted subfractions thereof (n=5) were investigated. RNA was in-vitro transcribed and hybridised to Affymetrix HG U133 A+B GeneChip (TG) and HG U133 2.0 plus arrays (VG). Expression data were gcrma-normalised and the empirical Bayes algorithm used. p-Values were adjusted using the Benjamini-Hochberg method (Bioconductor). iFISH was performed on purified MM-cells using probesets for chromosomes 1q21, 9q34, 11q23, 11q13, 13q14, 15q22, 17p13, 19q13, 22q11 and the translocations t(4;14) and t(11;14). HGF expression was verified by real time RT-PCR and western blotting. Based on Medline review, we established a list of 89 pro- and 56 antiangiogenic genes and investigated their expression according to the stage of disease: BMPC vs. MGUS, SD stage I (asymptomatic myeloma) vs. SD stage II/III (symptomatic myeloma requiring therapy). RESULTS. BMPC express pro- (e.g. VEGFA) and antiangiogenic genes (e.g. TIMP2). Only one pro-angiogenic gene (hepatocyte growth factor, HGF) is significantly overexpressed in MMC compared to BMPC. HGF has previously been linked with myeloma progression and induction of angiogenesis. Six antiangiogenic genes (TIMP2, SERPINF1, COL18A1, PF4, THBS1, CXCL14) are downregulated in MMC compared with BMPC. Compared to healthy donors, the BMME of MM shows a significant downregulation of PLAU (urokinase, antiangiogenic) and upregulation of TNF(proangiogenic). CONCLUSION. Upregulation of HGF-expression, downregulation of TIMP2, SERPINF1, COLA18A1, PF4, THBS1 and CXCL14 expression in MMC as well as downregulation of PLAU and upregulation of TNFα in the BMME seem to indicate an “angiogenic switch”. However, given the relatively low number of differentially expressed genes (7/145) and the expression of angiogenic genes by BMPC, an effect caused by an increasing number of plasma cells might be evenly important.

Download Full-text

Survival Signals, Growth Cytokines, Immunosuppressive Factors and Their Cellular Perpetrators in the Myeloma Tumor Microenvironment.

Blood ◽

10.1182/blood.v112.11.1664.1664 ◽

2008 ◽

Vol 112 (11) ◽

pp. 1664-1664

Author(s):

Jayakumar R Nair ◽

Louise M Carlson ◽

Noreen Ersing ◽

Asher Alban Chanan-Khan ◽

Kelvin P. Lee

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Tumor Microenvironment ◽

Cell Lines ◽

Plasma Cells ◽

Immune Escape ◽

Human Monocyte ◽

Fold Increase ◽

Myeloma Cell ◽

Myeloma Cells

Abstract Multiple myeloma (MM) is an incurable neoplasia of terminally differentiated plasma cells in the bone marrow. Essential interactions of MM cells with host bone marrow stromal cells (BMSC) induce growth factors essential for MM progression and pathogenesis, as well as induce an immunosuppressive environment that inhibits endogenous and therapeutically-induced immune responses against the MM cells. However, despite their importance, little is known about the identity of these BMSC cells or the molecular basis of their interaction with myeloma cells. A potential MM surface protein that could be involved in these interactions is CD28, based on its known pro-survival role in T cells. Clinical studies have shown that expression of CD28 in multiple myeloma highly correlates (p=0.006) with myeloma disease progression. Moreover, CD28+ MM cells invariably express the CD28 ligand CD86. A survival role for MM-CD28 might involve interactions with cellular partners that express the B7 (CD80/CD86) ligands. Potential candidates would include CD86+ myeloma cells themselves or B7+ dendritic cells (DC) that are known to be closely associated with myeloma cells in the patient bone marrow. When myeloma-myeloma interactions were disrupted by using the high affinity CD80/CD86 blocker CTLA4Ig (Abatacept®), increased sensitivity to arsenic trioxide (ATO) and melphalan (MEL) was observed in all the three MM cell lines U266, RPMI8226 and MM1S. For U266 viability was 93% in media alone, 84% with CTLA4Ig (100 μg/ml) alone, 86% with 2 μM ATO alone and was significantly reduced to 36% with CTLA4Ig + ATO. Similar drops in viability were observed with 25 μM MEL in combination with CTLA4Ig (33% as opposed to 71–74 % with CTLA4Ig or MEL alone). Our data suggests that this does not involve the downregulation of anti-apoptotic proteins Bcl-2, Bcl-xL or Mcl-1, commonly associated with drug resistance in myeloma. In the second part of the study, we demonstrate that myeloma cell lines or primary CD138+ myeloma cells can enhance via direct contact the ability of human monocyte derived immature DC to produce the immunosuppressive tryptophan depleting enzyme indoleamine 2,3 dioxygenase (IDO, as estimated by kynurenine (Kyn) (a tryptophan catabolite) levels in the supernatant) and also the pro-plasma cell survival cytokine IL-6. In co-cultures of IFNg treated immature DCs with either MM cell lines or with primary CD138+ myeloma cells from patient BM aspirates, the activity of IDO was enhanced ~ 2–8 fold (81 mM kyn with U266 and 20–43mM with primary cells) over that observed in control IFNg-treated DCs (9.7 mM Kyn). Western analysis also demonstrated increased IDO expression relative to IFNg activated DC controls. Blocking MM-CD28 with (Fab)2 fragments of anti-hCD28 mAb 9.3 downregulated IDO activity (9.3 mM) close to that of control, demonstrating the involvement of MM-CD28 in these interactions. We also demonstrated a significant up-regulation of the pro-myeloma survival cytokine IL-6 when immature DCs were co-cultured with CD28+ MM1S (90–300 pg/ml), a 4–9 fold increase over that of DC only control (25 – 35 pg/ml). This was further enhanced when immature DCs cultured with IL-10 (+ GM-CSF + IL-4) was used in co-cultures with MM-1S (800 – 1300 pg/ml), or with primary CD138+ myeloma cells from patient bone marrow aspirates (128–1142 pg/ml). In conclusion, our data demonstrates that blocking myeloma-CD28 - myeloma-CD86 “autocrine” interaction can enhance drug cytotoxicity, while interactions with DCs produce the essential growth cytokines IL-6 and immunosuppressive enzyme IDO with potential implications in MM survival and immune escape. Use of clinically approved agents (e.g. Abatacept®) to block myeloma-CD28 binding to its B7 ligands (increase chemotherapeutic efficacy), 1-MT to inhibit IDO and targeting DCs in the microenvironment to disrupt the tumor microenvironment could be viable therapeutic strategies for the future treatment of multiple myeloma.

Download Full-text

A Novel Role For Histone Deacetylase 11 (HDAC11) In Plasma Cell Differentation and Survival

Blood ◽

10.1182/blood.v122.21.1907.1907 ◽

2013 ◽

Vol 122 (21) ◽

pp. 1907-1907

Author(s):

Eva Sahakian ◽

Jason B. Brayer ◽

John Powers ◽

Mark Meads ◽

Allison Distler ◽

...

Keyword(s):

Gene Expression ◽

Multiple Myeloma ◽

Bone Marrow ◽

B Cells ◽

Plasma Cell ◽

Research Funding ◽

Plasma Cells ◽

Myeloma Cells ◽

Peripheral Blood Plasma

Abstract The role of HDACs in cellular biology, initially limited to their effects upon histones, is now appreciated to encompass more complex regulatory functions that are dependent on their tissue expression, cellular compartment distribution, and the stage of cellular differentiation. Recently, our group has demonstrated that the newest member of the HDAC family of enzymes, HDAC11, is an important regulator of IL-10 gene expression in myeloid cells (Villagra A Nat Immunol. 2009). The role of this specific HDAC in B-cell development and differentiation is however unknown. To answer this question, we have utilized a HDAC11 promoter-driven eGFP reporter transgenic mice (TgHDAC11-eGFP) which allows the monitoring of the dynamic changes in HDAC11 gene expression/promoter activity in B-cells at different maturation stages (Heinz, N Nat. Rev. Neuroscience 2001). First, common lymphoid progenitors are devoid of HDAC11 transcriptional activation as indicated by eGFP expression. In the bone marrow, expression of eGFP moderately increases in Pro-B-cells and transitions to the Pre- and Immature B-cells respectively. Expression of eGFP doubles in the B-1 stage of differentiation in the periphery. Of note, examination of both the bone marrow and peripheral blood plasma cell compartment demonstrated increased expression of eGFP/HDAC11 mRNA at the steady-state. These results were confirmed in plasma cells isolated from normal human subjects in which HDAC11 mRNA expression was demonstrated. Strikingly, analysis of primary human multiple myeloma cells demonstrated a significantly higher HDAC11 mRNA expression in malignant cells as compared to normal plasma cells. Similar results were observed in 4/5 myeloma cell lines suggesting that perhaps HDAC11 expression might provide survival advantage to malignant plasma cells. Support to this hypothesis was further provided by studies in HDAC11KO mice in which we observed a 50% decrease in plasma cells in both the bone marrow and peripheral blood plasma cell compartments relative to wild-type mice. Taken together, we have unveiled a previously unknown role for HDAC11 in plasma cell differentiation and survival. The additional demonstration that HDAC11 is overexpressed in primary human myeloma cells provide the framework for specifically targeting this HDAC in multiple myeloma. Disclosures: Alsina: Millennium: Membership on an entity’s Board of Directors or advisory committees, Research Funding. Baz:Celgene Corporation: Research Funding; Millenium: Research Funding; Bristol Myers Squibb: Research Funding; Novartis: Research Funding; Karyopharm: Research Funding; Sanofi: Research Funding.

Download Full-text

Identification Rate of Myeloma-Specific Clonotypes in Multiple Diagnostic Sample Types from Patients with Multiple Myeloma Using Next-Generation Sequencing Method

Blood ◽

10.1182/blood.v124.21.2036.2036 ◽

2014 ◽

Vol 124 (21) ◽

pp. 2036-2036 ◽

Cited By ~ 1

Author(s):

Hervé Avet-Loiseau ◽

Jill Corre ◽

Sabrina Maheo ◽

Jianbiao Zheng ◽

Malek Faham ◽

...

Keyword(s):

Multiple Myeloma ◽

Bone Marrow ◽

Next Generation Sequencing ◽

High Frequency ◽

Plasma Cells ◽

Employment Equity ◽

Myeloma Cells ◽

Sequencing Method ◽

Equity Ownership ◽

Generation Sequencing

Abstract Background: Recent reports support the prognostic importance of minimal residual disease (MRD) levels in multiple myeloma (MM) patients and suggest that novel methods for MRD assessment can play a role in the evolving MM treatment paradigm (Martinez-Lopez et al., Blood 2014). The application of next-generation sequencing (NGS)-based MRD assessment has been previously demonstrated in multiple lymphoid malignancies (Faham et al., Blood 2012; Ladetto et al., Leukemia 2013). NGS-based MRD assessment requires a diagnostic sample for initial identification of the myeloma clonotype. In order for this MRD assessment approach to be clinically practical, it must allow for analysis of a diverse set of diagnostic samples. In this study, we assessed the rate of myeloma clonotype identification in 6 sample types at diagnosis: bone marrow (BM) aspirate slides, RNA extracted from CD138+ plasma cells, methanol-fixed BM cells, BM mononuclear cells, RBC-lysed BM cells, and DNA extracted from small numbers of CD138+ plasma cells. Methods: Baseline samples were collected from 606 patients with MM. The following samples were provided at baseline: bone marrow aspirate (BMA) slides (164), RNA extracted from CD138+ plasma cells (402), methanol-fixed BM cells (30), BMA cell preparations using a Ficoll protocol (13), BMA cell preparations using an RBC lysis protocol (19), and DNA extracted from small numbers of CD138+ plasma cells (5). Samples with sufficient input DNA (>15ng) were included in the analysis, although this requirement was waived for samples from CD138+ cells. The Ficoll BMA cell preparations were divided into the mononuclear cell fraction and the lower Ficoll fraction, which is typically comprised of granulocytes and erythrocytes. Identification of myeloma clonotypes was performed using Sequenta's LymphoSIGHT™ method. Briefly, using universal primer sets, we amplified immunoglobulin heavy chain (IGH) and light chain (IGK) variable, diversity, and joining gene segments from genomic DNA. Amplified products were sequenced and analyzed using standardized algorithms for clonotype determination. Myeloma-specific clonotypes were identified for each patient based on their high-frequency (>5%) within the B-cell repertoire. Results: The NGS assay identified a high-frequency myeloma clonotype in 555/606 (92%) of patients with MM. Myeloma clonotype identification rates were 141/164 (86%) in BMA slides, 375/402 (93%) in RNA extracted from CD138+ plasma cells, 30/30 (100%) in methanol cell preparations, 13/13 (100%) in Ficoll cell preparations, 18/19 (95%) in RBC lysis cell preparations, and 5/5 (100%) using small amounts of input CD138+ DNA (approximately 5000 cells). These applicability rates are consistent with previous reports of sequencing applicability in MM patients. In thirteen patients, we investigated the potential loss of myeloma-specific clonotypes due to Ficoll cell preparation. The variation in myeloma cell loss was typically low but ranged from essentially no loss to the loss of more than 90% of the myeloma cells in the PBMC of one patient compared to the RBC lysis preparation. The myeloma cells were detected in the typically discarded lower layer of the Ficoll preparation which explained the loss. Conclusions: These results suggest that sequencing based MRD analysis is applicable in >90% of patients with MM. Multiple sample types, including archived BMA slides, can be used for identification of the myeloma clonotype. Further evaluation and optimization of sample processing methods is ongoing to enable application of the sequencing method for clinical MRD assessment in MM patients. Disclosures Zheng: Sequenta, Inc.: Employment, Equity Ownership. Faham:Sequenta, Inc.: Employment, Equity Ownership, Membership on an entity's Board of Directors or advisory committees. Munshi:Celgene: Consultancy; Onyx: Consultancy; Janssen: Consultancy; Sanofi-Aventi: Consultancy; Oncopep: Consultancy, Equity Ownership, Patents & Royalties.

Download Full-text