ALBERT

Description: Gene expression profiling (GEP) of purified plasma cells 48 hours after thalidomide and dexamethasone test doses showed these agents' mechanisms of action and provided prognostic information for untreated myeloma patients on Total Therapy 2 (TT2). Bortezomib was added in Total Therapy 3 (TT3), and 48 hours after bortezomib GEP analysis identified 80 highly survival-discriminatory genes in a training set of 142 TT3A patients that were validated in 128 patients receiving TT3B. The 80-gene GEP model (GEP80) also distinguished outcomes when applied at baseline in both TT3 and TT2 protocols. In context of our validated 70-gene model (GEP70), the GEP80 model identified 9% of patients with a grave prognosis among those with GEP70-defined low-risk disease and 41% of patients with favorable prognosis among those with GEP70-defined high-risk disease. PMSD4 was 1 of 3 genes common to both models. Residing on chromosome 1q21, PSMD4 expression is highly sensitive to copy number. Both higher PSMD4 expression levels and higher 1q21 copy numbers affected clinical outcome adversely. GEP80 baseline-defined high risk, high lactate dehydrogenase, and low albumin were the only independent adverse variables surviving multivariate survival model. We are investigating whether second-generation proteasome inhibitors (eg, carfilzomib) can overcome resistance associated with high PSMD4 levels.

Description: Abstract 1804 Poster Board I-830 Introduction MicroRNAs (miRNAs) are non-coding RNAs that regulate mRNA expression. miRNAs often act synergistically to repress target genes and their deregulation can contribute to the initiation and progression of a variety of cancers. The clinical relationship between global expression miRNA and mRNA in cancer has not been studied in detail. Methods We used whole genome microarray analyses of CD138-enriched plasma cells from 52 newly diagnosed cases of multiple myeloma (MM) to correlate miRNA expression profiles with a validated mRNA-based risk stratification, proliferation index, and pre-defined gene sets. Results In stark contrast to mRNAs, we discovered that all tested and expressed miRNAs were significantly up-regulated in high-risk disease as defined by a validated 70-gene risk score (P

Description: Abstract 3717 Human placenta has emerged as a valuable, uncontroversial source of transplantable cells for many cytotherapeutic purposes, including modulation of inflammation, bone repair, and cancer. Placenta-derived adherent cells (PDAC) are mesenchymal like adherent cells isolated from postpartum human placenta and capable of supporting bone formation in vivo. Multiple myeloma (MM) is closely associated with induction of bone disease and large lytic lesions, which are often not repaired and are usually the sites of relapses. The aim of the study was to evaluate the antimyeloma therapeutic potential, in vivo survival, and trafficking of PDAC in the SCID-rab model of MM-associated bone disease. SCID-rab system constructed by implanting a 4-weeks old rabbit bone into which primary human myeloma cells were directly injected (Yatta et al., Leukemia 2004; Yaccoby et al., Blood 2007). Bone disease was evaluated by measurements of bone mineral density (BMD) and X-rays. MM growth was determined by human immunoglobulin (hIg) ELISA and histologically. For in vivo tracking PDAC were transduced with a luciferase/GFP reporter in a lentiviral vector. SCID-rab mice engrafted with primary myeloma cells from 2 patients. Upon establishment of MM growth, PDAC (1×106 cells/bone) or vehicle were injected into the implanted myelomatous bone (Patient's 1, 5 mice/group; Patient's 2, 7 mice/group). While BMD of the implanted bones was significantly reduced in control hosts, intralesional PDAC cytotherapy significantly increased BMD of the implanted bones from pretreatment levels by 〉37% (p

Description: Abstract 2896 Background: Focal lesions (FL) are a well-recognized consequence of active multiple myeloma (MM), and distinguish it from its antecedent monoclonal gammopathy of undetermined significance (MGUS). We have recently reported on the superior performance of gene expression profiling (GEP) based on random trephine bone marrow aspirate sampling (RS) in distinguishing between 85% of patients with low-risk (LO) MM and 15% with high-risk (HI) MM compared to conventional prognostic variables. MM occasionally presents as macro-focal disease, in which cases RS may be inconclusive because of paucity of malignant cells in the sample. Here we report the comparison of GEP data from paired FL and RS samples from 106 untreated patients. Methods: We identified 106 newly diagnosed patients with paired samples, who were treated on Total Therapy (TT) protocols (3 in TT2, 19 in TT3, 75 in TT4, and 9 in TT5) in our multiple myeloma database. GEP risk scores, molecular subgroup classifications, overall survival (OS), and event-free survival (EFS) were compared and tested with the RS-derived 70-gene risk prediction and molecular subgroup classification models. Results: GEP defined molecular subgroups were correlated in 90 of 106 patients (85%). Looking at GEP-defined risk designation, we found a high degree of correlation between RS and FL samples with 95 of 106 samples showing the same designation (90%). For the 11 patients with divergent GEP designations, 8 (73%) were located at the boundary of the RS GEP risk score cutoff of +0.66 (range +0.43 to +0.84). For these risk designation-divergent patients, FL- but not RS-defined risk determined clinical outcome. Conclusion: Both the 70-gene risk prediction and molecular subgroup classification models can be used in FL-GEP samples. But, more importantly, for patients with RS-GEP risk score close to cutoff boundary (about 14% in our current data sets), FL-GEP provides better risk stratification, suggesting that the FL signal more adequately reflects to disease biology, progression and treatment response in MM. We therefore recommend that, for patients with borderline RS-based GEP risk scores, FL-GEP be used for staging and prognosis assessment in myeloma. Studies are in progress to determine, among multiple FL samples from the same patient, the variability in risk score compared to multiple RS samples. Disclosures: Shaughnessy: Myeloma Health, Celgene, Genzyme, Novartis: Consultancy, Employment, Equity Ownership, Honoraria, Patents & Royalties. Barlogie:Celgene: Consultancy, Honoraria, Research Funding; IMF: Consultancy, Honoraria; MMRF: Consultancy; Millennium: Consultancy, Honoraria, Research Funding; Genzyme: Consultancy; Novartis: Research Funding; NCI: Research Funding; Johnson & Johnson: Research Funding; Centocor: Research Funding; Onyx: Research Funding; Icon: Research Funding.

Description: Abstract 988 Background: Dexamethasone remains a key component of myeloma therapy. We previously showed that expression levels of the nuclear glucocorticoid receptor NR3C1 were associated with OS and PFS in patients treated on Total Therapy 3 (TT3). Post relapse survival (PRS) is an important component of OS following initial PFS. In this study we investigate the effect of NR3C1 expression levels at baseline (BL) and at relapse (RL) on OS, PFS and PRS for patients treated on Total Therapy 2 (TT2). Methods: TT2 was a randomized phase-III trial evaluating the impact of the up-front addition of thalidomide (T) to a multi-agent chemotherapy and high-dose melphalan program supported by tandem autotransplants. Between October 1998 and February 2004, 668 patients with newly diagnosed progressive or symptomatic multiple myeloma up to age 75 were enrolled. 351 patients had gene expression profiling (GEP) obtained at BL and 121 had GEP data obtained at RL. There was no significant difference in OS, PFS or PRS between patients with or without GEP data at BL or at RL. This analysis is based on the cohort of patients with GEP data with a cut-off date of April 22, 2011. OS and PFS were measured from the time of initiation of protocol-based therapy while PRS was measured from the initiation of salvage therapy. Results: Among patients randomized to the control arm (-T), OS improved progressively with the transition from lower to middle to upper NR3C1 tertiles. No such difference was noted in the experimental group with T (+T). Examining treatment arms within NR3C1 expression tertiles showed significant benefit from T in both OS and PFS among patients presenting with lower tertile expression; in the mid tertile, this effect was limited to PFS while such benefit was absent in the top-tertile subgroup. Multivariate analysis accounting for the interaction between NR3C1 expression and T showed that randomization to T and lower tertile NR3C1 expression reduced the hazard of death, while high risk designation by GEP, elevated B2M, elevated LDH, cytogenetic abnormalities and MS subgroup designation all conveyed a higher HR for death. With an overall median PRS of 3.0 years, there was no difference related to the initial treatment randomization or between baseline NR3C1 expression levels. In the context of treatment arms in patients randomized to -T, high- and mid-tertile NR3C1 levels at baseline dramatically improved PRS; such an effect was not observed in +T. We next investigated the implications of NR3C1 levels at relapse on PRS, revealing a significant shortening of PRS with transition from top- to mid- to low-tertile expression. Patients maintaining low NR3C1 levels also at relapse had the shortest PRS, followed by those who lost mid to high expression levels of NR3C1 at relapse, while those with higher than low-tertile levels at relapse enjoyed the longest PRS regardless of NR3C1 expression levels at baseline. Multivariate analysis including data from baseline and relapse as well as accounting for interaction between NR3C1 and T identified GEP high risk status at relapse as the only feature linked to inferior PRS whereas upper tertile expression of NR3C1 conveyed improved PRS. Conclusion: Low expression of NR3C1 at baseline is associated with poor PFS and OS in patients treated on the control arm of TT2. This confirms our previously reported findings in TT3. Addition of T overcomes the poor OS in patients with low NR3C1 expression. Low expression of NR3C1 at relapse correlated with poor PRS. This study indicates that patients with low NR3C1 expression levels at baseline benefit from the addition of T and offers a scientific rationale for NR3C1 gene expression level testing. Future trials, that include thalidomide or lenalidomide in their treatment regimen, will have to take NR3C1 expression levels into account. Disclosures: Shaughnessy: Myeloma Health, Celgene, Genzyme, Novartis: Consultancy, Employment, Equity Ownership, Honoraria, Patents & Royalties. Barlogie:Celgene: Consultancy, Honoraria, Research Funding; IMF: Consultancy, Honoraria; MMRF: Consultancy; Millennium: Consultancy, Honoraria, Research Funding; Genzyme: Consultancy; Novartis: Research Funding; NCI: Research Funding; Johnson & Johnson: Research Funding; Centocor: Research Funding; Onyx: Research Funding; Icon: Research Funding.

Description: The impact of cumulative dosing and premature drug discontinuation (PMDD) of bortezomib (V), thalidomide (T), and dexamethasone (D) on overall survival (OS), event-free survival (EFS), time to next therapy, and post-relapse survival in Total Therapy 3 were examined, using time-dependent methodology, relevant to induction, peritransplantation, consolidation, and maintenance phases. Univariately, OS and EFS were longer in case higher doses were used of all agents during induction, consolidation (except T), and maintenance (except V and T). The favorable OS and EFS impact of D induction dosing provided the rationale for examining the expression of glucocorticoid receptor NR3C1, top-tertile levels of which significantly prolonged OS and EFS and rendered outcomes independent of D and T dosing, whereas T and D, but not V, dosing was critical to outcome improvement in the bottom-tertile NR3C1 setting. PMDD of V was an independent highly adverse feature for OS (hazard ratio = 6.44; P 〈 .001), whereas PMDD of both T and D independently imparted shorter time to next therapy. The absence of adverse effects on postrelapse survival of dosing of any VTD components and indeed a benefit from V supports the use up-front of all active agents in a dose-dense and dose-intense fashion, as practiced in Total Therapy 3, toward maximizing myeloma survival.

Description: Abstract 1932 Background: Most MM develops through an asymptomatic precursor stage. Current analyses understanding the risk of progression to clinical MM are based on limited set of variables. In 2002, South West Oncology Group (SWOG) began a prospective observational clinical trial of patients with asymptomatic plasmaproliferative disorders (PCD) to evaluate clinical, imaging, genetic and immunologic features predictive of disease progression. Patients and Methods: All patients with asymptomatic PCD underwent detailed staging evaluation at study entry and were prospectively followed at 3–6 month intervals until disease progression requiring therapy. Data from clinical features at baseline were evaluated to assess the risk for progression to clinical MM (P-MM). Results: Between 11/2002 and 4/2010, 230 patients have been enrolled, of whom 140 were deemed to have smoldering MM (SMM) and 90 monoclonal gammopathy of undetermined significance (MGUS), based on international myeloma working group criteria. The two groups differed in terms of abnormal serum free light chain ratio (FLCR; SMM: 79%, MGUS: 52%; p=3.5mg/L, baseline serum albumin 8 mg/dL, abnormal FLCR and low levels of uninvolved Ig were all associated with greater propensity of P-MM on univariate analysis. However upon multivariate analysis, abnormal serum FLCR and SMM disease type were the only surviving adverse variables predictive of disease progression. Conclusions: These data demonstrate in the context of a large prospective multicenter trial that abnormal FLCR is a dominant predictor of clinical malignancy in asymptomatic PCD. These findings support the need to understand the biology of abnormal FLCR and incorporate novel criteria based on imaging, genetics and host response. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4800 Background: Despite availability of novel agents, many MM patients still relapse and require salvage interventions. In the Arkansas program, we have attempted to procure initially sufficient hematopoietic precursor cells, for use in high-dose therapy salvage regimens once phase I-II trials have been exhausted. We are reporting on the efficacy in terms of response rate, EFS and OS of ARMM patients receiving S-BEAM. Patients and Methods: S-BEAM comprised standard BEAM (carmustine 300 mg/m2 on day 1, etoposide 200 mg/m2 days 1–4, cytarabine 400 mg/m2 days 1–4, melphalan 140 mg/m2 on day 5) with the addition of cisplatin (10-12.5mg/m2/d CI × 5d), bortezomib (1.3-1.5mg/m2 on days 1 + 4), thalidomide (100-200mg/d for 5 days) or lenalidomide (25-100mg/d for 5 days), DEX (40-100mg/d for 5 days) plus rapamycin (3mg d1, 1mg d2-5). Statistical methods included Cox regression modeling using significance level 0.05 and Kaplan-Meier methodology for all figures. Comparisons within figures were made using the log-rank test. Results: The characteristics of 147 patients treated included prior transplant (Tx) in 67% (2Tx, 29%; =〉3Tx, 11%), and prior exposure and resistance in virtually all patients (92%) to bortezomib, thalidomide, lenalidomide applied in VTD, VRD or with chemotherapy VTD-PACE. Pre-S-BEAM high-risk features included low albumin (=3.5mg/L; 32%), high LDH (〉=ULN; 44%), and presence of cytogenetic abnormalities (CA) in 70%. Clinical outcomes included at least PR in 62% including 48% with n-CR and 29% with CR. Two-year estimates of EFS and OS were 29% and 33%; TRM within 60 days was 3%. At 4 years, 23% remain alive and 15% event-free. Independently significant variables affecting both OS and EFS adversely included, in a model without GEP, high B2M (〉5.5mg/L), high LDH (〉=ULN), low hemoglobin (=10 in 42%. When GEP data were included in a subset of 103 patients, high-risk designation, high LDH and age 〉=65 were identified on the basis of highest R2 values (49% for OS, 41% for EFS). Among 28 patients lacking any of these 3 features, 1-year OS/EFS was 83%/67%, with 1 variable (n=36) 53%/38%, with 2 (n=31) 22%/6% and with 3 (n=8) 0%/0% (both P

Description: Abstract 303 Background: Gene expression profiling (GEP) of purified plasma cells identifies 15% of newly diagnosed MM as high-risk with a median survival of 2yr compared to 10+yr for the remainder. A validated 70-gene GEP risk model (GEP70) making such determinations is related to copy number increases in chromosome 1q21. Moreover, FISH-defined gains of 1q21 at diagnosis are associated with poor outcome and serial studies have shown that both the percentage of cells with 1q21 gains and 1q21 copies in these cells invariably increase at relapse. Combined with the fact that 1q21 is the only recurrent high-level amplicon in MM, these data suggests that 1q21 harbors a copy number sensitive gene or genes that confer resistance to apoptosis. PSMD4 and CKS1B are the only genes in the GEP70 model that map to the 1q21 amplicon. PSMD4 is the polyubiquitin receptor for the proteasome and the only component of the proteasome that exists free of the proteasome complex. High levels of free cytoplasmic PSMD4 and a small proteolytic fragment of PSMD4, known as anti-anti-secretoy factor, may be able to reduce proteasome load thereby reducing sensitivity of MM cells to proteasome inhibition-induced apoptosis. Patients and Method: In TT3, we added BOR to TT2 and performed GEP at baseline and 48hr after BOR test-dosing (1.0mg/m2). We correlated post BOR GEP (TT3), baseline GEP (TT2 and TT3), and baseline 1q21 FISH (TT2 and TT3) with outcomes in over 600 cases. Result: PSMD4 and 14 other proteasome genes were among 80 genes in a post-BOR GEP model (GEP80) created in TT3 and validated in TT3B, whose post-BOR elevated expression was related to poor outcome. The absence of hyper-activation of PSMD4 and proteasome genes after in-vivo thalidomide, dexamethasone or lenalidomide test dosing suggested that this effect was BOR-specific. There was strong but not complete overlap between risk designations by the GEP70 and GEP80 models in TT2 and TT3. We combined the risk predictions of the two models in baseline samples creating four risk combinations. Kaplan Meier analysis revealed a dramatic improvement in outcomes of GEP70 high-risk/GEP80 low-risk cases in TT3 relative to TT2. Similarly, while a significant improvement in outcomes were observed in cases with 3 copies of 1q21, there was no difference for cases with 4+ copies of 1q21. To determine if 1q21 copy number-driven expression changes could account for these differences, we correlated GEP of candidate genes with the presence of 2, 3 or 4+ copies of 1q21. Using FISH-defined tertiles we discovered that intermediate levels of PSMD4, corresponding to 3 copies of 1q21, was associated with significant improvement in outcome in TT3. Conclusion: BOR incorporated into TT3 overcomes GEP70 high-risk disease with 3, but not 4+ copies of 1q21. PSMD4, is a copy number dependent gene at 1q21 and appears to be a strong prognostic biomarker for BOR-containing therapies. We propose that TT3-like therapies can overcome the anti-apoptotic effects of modest increases in PSMD4 levels in MM, but that novel therapeutic strategies specifically targeting PSMD4 function might be needed to improve the currently dismal outcomes associated with high-level expression of PSMD4. Disclosures: No relevant conflicts of interest to declare.