Abstract: Immune evasion is a hallmark of cancer and a central mechanism underlying acquired resistance to immune therapy. In allogeneic hematopoietic cell transplantation (alloHCT), late relapses can arise after prolonged alloreactive T-cell control, but the molecular mechanisms of immune escape remain unclear. To identify mechanisms of immune evasion, we performed a genetic analysis of serial samples from 25 patients with myeloid malignancies who relapsed ≥1 year after alloHCT. Using targeted sequencing and microarray analysis to determine HLA allele-specific copy number, we identified copy-neutral loss of heterozygosity events and focal deletions spanning class 1 HLA genes in 2 of 12 recipients of matched unrelated-donor HCT and in 1 of 4 recipients of mismatched unrelated-donor HCT. Relapsed clones, although highly related to their antecedent pretransplantation malignancies, frequently acquired additional mutations in transcription factors and mitogenic signaling genes. Previously, the study of relapse after haploidentical HCT established the paradigm of immune evasion via loss of mismatched HLA. Here, in the context of matched unrelated-donor HCT, HLA loss provides genetic evidence that allogeneic immune recognition may be mediated by minor histocompatibility antigens and suggests opportunities for novel immunologic approaches for relapse prevention.

Abstract: Background: Chronic GVHD contributes to morbidity and mortality after allogeneic transplantation. Several phase 2 and 3 open label studies suggested that ATLG (formerly referred to as ATG-F) reduces cGVHD without impacting relapse or survival. This report describes the first prospective randomized double blind phase 3 trial to assess the effect of ATLG (Neovii) when added to tacrolimus / methotrexate prophylaxis on cGVHD free survival. Patients and Methods: This study was conducted at 27 sites in the United States and Australia. 260 patients were randomized (132 placebo, 128 ATLG). 6 patients who were randomized never received ATLG/placebo (4 placebo and 2 ATLG). Data presented include the 128 placebo and 126 ATLG patients who underwent treatment and transplant. Patients were ages 18-65 years with a diagnosis of acute myeloid (AML, 64%) or lymphoblastic (ALL, 21%) leukemia in first or subsequent complete remission or myelodysplastic syndrome (MDS, 15%) with less than 10% marrow blasts. Median age was 48 years (18-65) with 55% males. All patients received 8/8 HLA (A, B, C, DR) allele matched unrelated products (80% mobilized peripheral blood, 20% bone marrow). Myeloablative conditioning was with cyclophosphamide and total body irradiation (Cy/TBI, 27%), busulfan and cyclophosphamide (Bu/Cy, 33%), or busulfan and fludarabine (Bu/Flu, 40%). All patients received tacrolimus starting on Day -1 and standard mtx on days 1, 3, 6, and 11 for GVHD prophylaxis. Patients received ATLG or placebo on Days -3, -2, -1 at an ATLG dose of 20 mg/kg per day (60 mg/kg total). The primary endpoint of the study was moderate/severe chronic GVHD free survival. cGVHD was scored by the investigator and then confirmed or overturned by an independent adjudication committee. Secondary endpoints included engraftment, acute GVHD , moderate/severe cGVHD, non-relapse mortality, GVHD and relapse free survival (GRFS), progression free (PFS) and overall survival (OS). The impact of ATLG on immune reconstitution was assessed in 91 patients. Results: Treatment arms were balanced with respect to age, diagnosis, remission status, cytogenetics, graft source (PB or BM), CMV serostatus and conditioning regimen. Median follow up of survivors was 745 days (61, 1425). Regarding the primary endpoint, there was no difference in 2 year moderate/severe cGVHD free survival between ATLG and placebo (48 vs 44%, p = 0.57), nor was there a difference in GRFS (Table 1). Neutrophil engraftment at Day 30 was less in the ATLG arm (85% vs 95%, p=0.00001). Grade 2 or higher infusion reactions were higher in the ATLG arm (16% vs 2.3%, p=0.0001). Reactivation of CMV was similar (27% vs 31%, p=0.58). The Day 180 cumulative incidence of grades 2-4 acute GVHD was lower in the ATLG arm (23 vs 40%, p = 0.003) as was the 2-year cumulative incidence of moderate/severe cGVHD (12 vs 33%, p = 0.000006). However, both 2-year OS and PFS were lower in ATLG treated patients (58 v 76% and 47 vs 67%, respectively) due in part to a higher incidence of relapse (32 vs 19%, p=0.068). Multivariable models confirmed that ATLG was associated with inferior OS, HR 1.85 (1.1-2.9, p=.0075) and PFS, HR 1.63 (1.1-2.41, p=0.015). We then conducted an unplanned post-hoc analysis and discovered a striking influence of conditioning regimen on outcomes. There were no differences in OS or PFS between ATLG and placebo arms in patients receiving Bu/Cy or Bu/Flu but a dramatic difference in patients receiving Cy/TBI conditioning (OS, 88% vs 48% p =0.006 and PFS, 75 vs 29%, p=0.007 in placebo and ATLG arms, respectively). 91 patients (44 ATLG, 47 placebo) participated in the immune reconstitution study. ATLG treatment was associated with lower CD3, CD4, and CD8 counts at Days 30, 100, and 360, respectively. In turn lower CD3, CD4, and CD8 counts assessed as time dependent variables were all associated with inferior OS and PFS as well as higher NRM (data not shown). Conclusion: In this first ever prospective, randomized, double blind trial of ATLG added to tacrolimus /methotrexate in recipients of matched unrelated donor grafts after myeloablative conditioning, there was no significant difference in 2 year moderate/severe cGVHD free survival or GRFS. ATLG did significantly reduce grades 2-4 acute and moderate/severe chronic GVHD. However, overall survival was higher in the placebo arm, driven mainly by patients who received Cy/TBI conditioning. Further analyses are needed to understand the proper role for ATLG in HCT. Disclosures Soiffer: Juno: Membership on an entity's Board of Directors or advisory committees; Kiadis: Membership on an entity's Board of Directors or advisory committees; Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees. Porter:Genentech: Employment; Novartis: Patents & Royalties, Research Funding. Jagasia:Therakos: Consultancy. Szer:Alnylam: Honoraria, Membership on an entity's Board of Directors or advisory committees; Ra Pharma: Honoraria, Membership on an entity's Board of Directors or advisory committees; Alexion Pharmaceuticals, Inc.: Honoraria, Membership on an entity's Board of Directors or advisory committees. Ritz:Kiadis: Membership on an entity's Board of Directors or advisory committees. Glavin:Neovii Biotech: Employment. Chen:Incyte Corporation: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Research Funding.

Abstract: BACKGROUND: Pathogenesis of multiple myeloma is partly attributed to an aberrant expression of proliferation-, pro-angiogenic and bone-metabolism modifying factors by malignant plasma-cells. AIM. Given the long and variable time-span from first diagnosis of early-stage plasma-cell dyscrasias to overt myeloma and the low proliferation rate of malignant plasma-cells, we hypothesize these to concomitantly express a novel class of anti-proliferative factors of potential prognostic relevance. Here, bone morphogenic proteins (BMPs) represent possible candidates, as they inhibit proliferation, stimulate bone formation, and have an impact on the survival of cancer patients. PATIENTS AND METHODS. We assessed expression of BMPs and its receptors by Affymetrix DNA-microarrays (n=434) including CD138-purified primary myeloma-cell-samples, normal bone-marrow plasma-cell-samples, polyclonal plasmoblasts-samples, human myeloma-cell-lines (HMCL), and whole bone-marrow. Presence and differential gene expression was determined by PANP-algorithm and empirical Bayes statistics. Event-free (EFS) and overall survival (OAS) were investigated for the 168 patients undergoing high-dose chemotherapy (HM-group) using Cox’s proportional hazard model. Findings were validated using the same strategy on an independent group of 345 patients from the Arkansas-group. For validation, quantitative real-time PCR and flow cytometry were performed. In vitro induction of angiogenesis was assessed using the AngioKit-assay. Effect of BMP6 on proliferation of HMCL was assessed by 3H-thymidine uptake. RESULTS. BMP6 is the only BMP expressed by normal- (13/14 samples) and malignant plasma-cells (228/233 samples). It is significantly lower expressed in proliferating non-malignant plasmablastic cells and human myeloma cell-lines. In vitro, BMP6 significantly inhibits proliferation of myeloma-cell-lines with an IC50 ranged from 0.08–2.15μg/ml, survival of primary myeloma-cells, and in vitro tubule formation down to the level of the negative control. High BMP6-expression in malignant plasma cells delineates significantly superior overall-survival for patients undergoing high-dose chemotherapy in both independent series of patients (n=168, P=.02 and n=345, P=.03, respectively, see below). CONCLUSION. With BMP6 we report for the first time the autocrine expression of a prognostically relevant anti-angiogenic and anti-proliferative factor and its receptors by normal and malignant plasma-cells. Figure Figure

Abstract: Background: Myelodysplastic syndromes (MDS) are hematological disorders characterized by chronic anemia, and are often refractory to cytokine therapy. Two phase 2 studies evaluating lenalidomide in MDS (MDS-001 and MDS-003), demonstrated a high frequency of erythroid response in patients with a chromosome 5q31 deletion (List AF et al., N Eng J Med2005; 352:549–557 and List AF et al., in press). Herein we report on the durability of response with long term follow-up approaching 4 years. Methods: All enrolled subjects were evaluated for frequency and duration of red blood cell transfusion independence (RBC-TI), change in hemoglobin (Hgb), pathologic and cytogenetic response, and safety. Results: Ten 5q31- patients achieved a cytogenetic and/or major erythroid response in MDS-001. Four of these patients have maintained a major response for at least 2 years (range 2.2 to 3.7 years). In MDS-003, 99 (67%) of 148 enrolled patients achieved RBC-TI. Of the 99 responders 83 were Low/Int-1, 3 were Int-2/High and 13 were not classified. Median increase in Hgb from baseline to maximum achieved during RBC-TI was 5.4 g/dL (range 1.1, 11.4). The estimated median duration of response was 115.9 weeks, and 52 (53%) of 99 responders remain on study with ongoing RBC-TI response as of 30 June 2006. Duration of response was at least 52 weeks for 63 (64%), 78 weeks for 52 (53%), and 104 weeks for 36 (36%) of 99 responding patients. Variables (or co-variates) associated with longer duration of TI in multivariate analysis included low baseline transfusion requirement ( 〈 4 units per 8 wks, p=0.002), 5q- syndrome (p=0.01) and low IPSS (p=0.02) classifications. Among 85 evaluable patients, major cytogenetic responses were achieved in 37 (44%) and minor responses in 24 (28%). Cytogenetic responses occurred in patients with complex karyotypes as well as those with an isolated 5q31 deletion. The frequencies of major and minor cytogenetic responses were similar among patients with and without 5q- syndrome. The most common drug-related adverse events (AE) were Grade 3/4 neutropenia and thrombocytopenia, reported in 87 (59%) and 74 (50%) of 148 patients, respectively although hematologic AEs led to study discontinuation in only 13 (9%) of patients. Pneumonia was the only serious adverse event that occurred in 〉 10% of patients (15/148; 10%). Adverse events were manageable through supportive treatment and dose reduction. Conclusion: Lenalidomide is an effective treatment in MDS patients with an associated chromosome 5q31 deletion, with RBC-TI and major erythroid response maintained for median duration 〉 2 years.

Abstract: Abstract 4395 G-CSF-mobilized peripheral blood stem cells (PBSC) is the most commonly used graft source for autologous hematopoietic stem cell transplantation. Recent improvements in optimizing PBSC collections include the introduction of plerixafor, real-time monitoring of blood CD34+ cells prior to initiation of apheresis, and increasing use of large volume apheresis. The Spectra Optia (SPO; CaridianBCT) is a new cell separation device. This study was designed to evaluate the CD34+ cells collection efficiency of the SPO device in comparison to a consecutive series of patients recently collected using the COBE Spectra (CSP) device. We performed an IRB-approved analysis of apheresis procedures performed at two large transplant centers of the collection outcomes of 15 apheresis procedures performed on the SPO platform with 32 apheresis procedures performed on the CSP platform. SPO and CSP cases were matched in a ratio of approximately 1:2 by the percentage of CD34+ cells in the blood at the time of starting apheresis. Primary outcomes measurements were the total numbers of CD34+ cells/kg collected and the efficiency of collection of CD34+ cells by apheresis [efficiency = total CD34+ cells collected(net of sampling)/(CD34+ cells/uL blood×blood volume processed)]. A comparison of apheresis on each device is summarized in the Table. The average flow rate and whole blood volume processed were significantly less on the SPO versus on the CSP. There were no statistical differences between comparing the SPO and CSP platforms with respect to the efficiency of CD34+ cell collection (mean values of 0.53 ± 0.11 and 0.48 ± 0.17, respectively), or the total numbers of CD34+ cells collected/kg in a single apheresis 10.6 ± 8.9 × 10E6/kg and 9.2 ± 5.9 × 10E6/kg, respectively. Apheresis performed on the SPO device did result in less depletion of platelets in the blood, when platelet were measured immediately before and the day following apheresis with decrements in blood platelet counts of 36% ± 10% and 50% ± 15%, respectively, in a subset of 14 SPO and 23 CSP patients analyzed for this end-point (P=0.009). Apheresis using the SPO platform resulted an equivalent efficiency of CD34+ cell collection as the CSP platform, with somewhat less depletion of platelets. The smaller “foot-print” of the SPO device and quieter operation make it attractive as an alternative to the CSP apheresis platform. A registration clinical trial is in progress to obtain 510k approval.Table:Comparison of CD34+ Cell and Platelet Collection Efficiency on the COBE Spectra and OPTIA. Mean values ± SD are shown.OPTIA (N=15)COBE (N=32)PAge (years)59.33 ± 11.1661.63 ± 8.350.436Weight (kg)84.78 ± 18.7481.19 ± 17.710.52Blood volume (ml)5238 ± 10104856 ± 10580.249WBC (×109/L)34 ± 1545 ± 290.160CD34+ cell in the blood before apheresis (%)0.48 ± 0.480.32 ± 0.310.160CD34+ cells before apheresis (/ul)140 ± 18684 ± 720.138Flow rate (ml/min)51 ± 1074 ± 160.001*Time (minutes)300 ± 54293 ± 500.637Whole Blood volume processed (ml)15440 ± 477721362 ± 49420.001*CD34+ cells in product (×106/kg)10.6 ± 8.99.2 ± 5.90.530CD34+ cell collection efficiency0.53 ± 0.110.48 ± 0.170.369Depletion of Platelets (%)36% ± 10% (N=14)50% ± 16% (N=23)0.009**p 〈 0.05 Disclosures: No relevant conflicts of interest to declare.

Abstract: Introduction: Leukemic escape from the graft versus leukemia (GVL) effect of allogeneic hematopoietic cell transplantation (HCT) is poorly understood. During prolonged periods of post-transplant remission, immunologic selection pressure can favor relapse mediated by acquired resistance to GVL-mediated clearance, such as loss of mismatched HLA in haploidentical transplants. We hypothesized that genetic mechanisms of immune evasion can cause late relapses after matched unrelated donor transplants for myeloid malignancies. Methods: We evaluated 580 adult patients with myeloid malignancies who underwent allogeneic HCT at our institution between 2001 and 2014 and experienced subsequent disease relapse. In 19% of these patients (n=112) relapse occurred at least one year after transplantation. The 25 patients included in the study had specimens banked at each of three timepoints: 1) prior to transplantation, with AML or MDS involvement, 2) approximately 100 days after transplantation, during the period of remission 3) at the time of disease relapse. The indications for transplantation were AML (n = 20) and MDS (n = 5). Transplants were from matched unrelated (n=14), matched related (n=9), and mismatched unrelated (n=2) donors. 12 patients received myeloablative conditioning and 13 received reduced-intensity conditioning regimens. Targeted sequencing of 187 genes, selected based on pathogenic involvement in myeloid malignancies or suspected involvement in immune evasion, was performed on all three timepoints from each patient. Genome-wide microarray-based copy number assessment was performed onthe pre-transplant specimen and post-transplant relapse specimens. Results: We identified three recipients with relapse-specific HLA loss via 1-8 Mb deletions or chromosome 6p arm-level copy neutral uniparental disomy (UPD). One of the HLA alterations was a deletion spanning HLA-B and HLA-C in a donor/recipient pair mismatched at HLA-C. However, in two other cases, relapse-specific HLA losses were identified in donor/recipient pairs that were fully matched at A, B, C, and DRB1. These findings suggest that HLA loss may allow leukemic cells to escape allogeneic immune recognition of minor HLA discrepancies and/or the presentation of minor histocompatibility antigens in the context of matched MHC presentation. These three HLA losses were identified among 14 recipients of matched unrelated donor HCT. HLA losses were not identified among recipients of matched related (n = 9) or mismatched unrelated (n = 2) allogeneic HCT. To evaluate potential interactions between canonical myeloid driver mutations and immunologic alterations, we defined the genetic characteristics of paired pre-transplant MDS/AML samples and post-transplant relapsed samples. In 22 out of 25 cases, at least one driver mutation that was present in the pre-transplant sample was also detected in the relapse sample. Clonal genetic evolution was common at the time of relapse and predominantly involved the acquisition of new subclonal mutations affecting mitogenic signaling (n = 9) and myeloid transcription factors (n = 11). TP53 alterations, including point mutations and 17p deletions were identified in 6 out of 25 patients, including two that remained stable before and after transplantation, and four that were newly detected at the time of relapse. Conclusions: We identified recurrent HLA loss via 6p UPD and segmental deletions in 3 out of 14 patients with late relapse after matched unrelated HCT. Although HLA loss has been observed at relapse after haploidentical HCT, the role of HLA loss as a mechanism of relapse after non-haploidentical HCT has remained unclear. HLA loss in cases with late relapse indicates a prolonged period of immunologic equilibrium, where effective GVL serves as a selective pressure for clonal genetic mechanisms of alloimmune evasion. In the context of MURD HCT, HLA loss provides genetic evidence that allogeneic immune recognition may be mediated by minor histocompatibility antigens and suggests opportunities for novel immunologic approaches for relapse prevention. Disclosures Ho: Jazz Pharmaceuticals: Consultancy. Nikiforow:Kite Pharma: Consultancy. Antin:Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees. Soiffer:Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.

Abstract: To examine the safety and efficacy of recombinant-methionyl human stem cell factor (r-metHuSCF), 38 patients with intermediate-grade or immunoblastic high-grade non-Hodgkin's lymphoma who were eligible for autologous transplantation were randomized to receive r-metHuSCF (5, 10, 15, or 20 μg/kg/d) plus Filgrastim (10 μg/kg/d) or Filgrastim (10 μg/kg/d) alone to mobilize peripheral blood progenitor cells. Subcutaneous administration of r-metHuSCF was well tolerated in conjunction with a multi-agent pre-medication regimen; local injection site reactions were the most commonly seen adverse event. The total mononuclear cell count, CD34+ cell content, granulocyte-macrophage colony-forming cells (GM-CFC), and burst-forming units-erythroid (BFU-E) per kilogram in the apheresis product was similar when all patients were analyzed by treatment cohort and mobilization regimen (Filgrastim or r-metHuSCF in combination with Filgrastim); however, when prior chemotherapy was taken into account in a supplementary analysis, clinically important differences were observed. Extensive prior therapy was defined as the amount of exposure to specific stem cell toxic chemotherapeutic agents that patients received. These agents include procarbazine, nitrogen mustard, melphalan, nitrosoureas (≥2 cycles of any of these drugs) or greater than 7.5 g of cytosine arabinoside. In these patients, there was an increased number of CD34+ cells (1.76 v 0.28 × 106/kg), GM-CFC (20.5 v 5.0 × 104/kg), and BFU-E (36.9 v 8.9 × 104/kg) in patients receiving r-metHuSCF and Filgrastim (N = 18) compared with Filgrastim alone (N = 5). These patients also had a decreased time to an untransfused platelet count of 20 × 109/L that was 10.5 days shorter in the patients who received r-metHuSCF and Filgrastim (12.5 v 23 days). These differences were not found to be statistically significant, possibly because of small size, but are clinically important.

Abstract: Disease relapse remains the leading cause of failure after autologous stem cell transplantation (ASCT) for patients with relapsed/refractory diffuse large B-cell lymphoma (R/R DLBCL). We conducted a phase 2, multicenter, single-arm study of the anti–PD-1 monoclonal antibody pembrolizumab given after ASCT in patients with chemosensitive DLBCL, hypothesizing that it would improve the progression-free survival (PFS) at 18 months after ASCT (primary endpoint) from 60% to 80%. Pembrolizumab was administered at 200 mg IV every 3 weeks for up to 8 cycles, starting within 21 days of post-ASCT discharge. Twenty-nine patients were treated on this study; 62% completed all 8 cycles. Seventy-nine percent of patients experienced at least one grade 3 or higher adverse event, and 34% experienced at least one grade 2 or higher immune-related adverse event. Overall, 59% of patients were alive and progression free at 18 months, which did not meet the primary endpoint. The 18-month overall survival was 93%. In conclusion, pembrolizumab was successfully administered as post-ASCT consolidation in patients with R/R DLBCL, but the PFS did not meet the protocol-specific primary objective and therefore does not support a larger confirmatory study. This trial was registered at www.clinicaltrials.gov as #NCT02362997.

Abstract: Helios, encoded by IKZF2, is a member of the Ikaros family of transcription factors with pivotal roles in T-follicular helper, NK- and T-regulatory cell physiology. Somatic IKZF2 mutations are frequently found in lymphoid malignancies. Although germline mutations in IKZF1 and IKZF3 encoding Ikaros and Aiolos have recently been identified in patients with phenotypically similar immunodeficiency syndromes, the effect of germline mutations in IKZF2 on human hematopoiesis and immunity remains enigmatic. We identified germline IKZF2 mutations (one nonsense (p.R291X)- and 4 distinct missense variants) in six patients with systemic lupus erythematosus, immune thrombocytopenia or EBV-associated hemophagocytic lymphohistiocytosis. Patients exhibited hypogammaglobulinemia, decreased number of T-follicular helper and NK cells. Single-cell RNA sequencing of PBMCs from the patient carrying the R291X variant revealed upregulation of proinflammatory genes associated with T-cell receptor activation and T-cell exhaustion. Functional assays revealed the inability of HeliosR291X to homodimerize and bind target DNA as dimers. Moreover, proteomic analysis by proximity-dependent Biotin Identification revealed aberrant interaction of 3/5 Helios mutants with core components of the NuRD complex conveying HELIOS-mediated epigenetic and transcriptional dysregulation.

Abstract: Graft-versus-host effects may lead to HIV-1 reactivation and cell death of infected pre-HCT CD4+ T cells. Natural killer cell activation correlates with in vitro HIV-1 transcriptional activity in the setting of HCT.