Abstract: Abstract 4097 Poster Board III-1032 CD19-specific chimeric antigen receptors (CARs) based on first and second generation designs (Figure) expressed by genetically manipulated T cells are being evaluated in clinical trials and it is apparent that the long term in vivo survival of the infused T cells will be critical to achieving therapeutic successes. We have previously demonstrated that the first-generation CAR can be modified to include and provide an improved T-cell survival signal through addition of a chimeric CD28 endodomain (designated second generation CAR, CD19RCD28). However, it is recognized that optimal activation of T cells through CD28 for sustained proliferation may require co-stimulation through other signaling molecules. To address this issue we have altered the second generation CAR to include additional intracellular signal transduction domains from CD137 (4-1BB) or, CD134 (OX40) to generate (Figure) third generation CARs (CD19RCD28CD137 or, CD19RCD28CD134, respectively). The third generation CARs were electro-transferred into peripheral blood T cells using the Sleeping Beauty transposon/transposase system and propagated on K562-derived CD19+ artificial antigen presenting cells (aAPCs). We observed the selective outgrowth of CD19-specific T cells expressing similar percentages of expression and densities of CD19RCD28, CD19RCD28CD137, and CD19RCD28CD134 CARs with a subset of the propagated T cells displaying a central memory phenotype (CD62L+CD28+). The T cells expressing the third generation CARs exhibited redirected-killing and were able to produce IFN-g in response to CD19. Although the cytolytic ability and Tc1 cytokines released by both second and third generation CARs were similar, differences due to the presence of the CD137 and CD134 intracellular signaling endodomains were apparent using in vitro experiments designed to mimic in vivo T-cell persistence, by measuring long-term propagation in the absence of exogenous cytokines added to the T-cell culturing process. We observed an approximate 200% increase in the numeric expansion of CD19RCD28CD137+ and CD19RCD28CD134+ T cells as compared to the CD19RCD28+ T cells at the end of 14 days of co-culture on CD19+ aAPC. These data highlight the fact that CD19-dependent signaling through chimeric CD28 might be improved by the addition of signaling through chimeric CD137 or CD134 domains so as to fully maintain T-cell activation and sustain proliferation in vivo after adoptive transfer. Figure Schematic of first (CD19R), second (CD19RCD28), and two third (CD19RCD28CD137 and CD19RCD28CD134) generation CD19-specific CARs (shown as homodimers on the cell surface) expressing one, two, or three T-cell signaling chimeric endodomains. TM = transmembrane. Figure. Schematic of first (CD19R), second (CD19RCD28), and two third (CD19RCD28CD137 and CD19RCD28CD134) generation CD19-specific CARs (shown as homodimers on the cell surface) expressing one, two, or three T-cell signaling chimeric endodomains. TM = transmembrane. Disclosures: No relevant conflicts of interest to declare.

Abstract: Background: Allogeneic hematopoietic stem cell transplantation (allo SCT) has two potential advantages over autologous SCT: a tumor-free graft and graft-versus-myeloma (GVM) effect. Allo SCT’s potential to induce long term remission has, however, been offset by high rates of transplant-related non-relapse mortality (TRM). Reduced-intensity conditioning (RIC) regimens for allo SCT are associated with lower TRM without compromising the GVM effect. Methods: We retrospectively analyzed our experience in 69 patients (30 females and 39 males) with heavily pretreated, relapsed myeloma, who received allo SCT at our institution between1985 and 2007. Eighteen patients received myeloablative regimens (MA), while 51 received RIC regimens. MA regimens were TBI-based in 5 patients, high-dose busulfan-containing in 6 patients and high-dose melphalan containing (180–200 mg/m2) in 7 patients. RIC regimens were a combination of fludarabine (90–120 mg/m2) and melphalan (100–140 mg/m2). Median age of patients at allo SCT in both groups was 51 years. Median interval from diagnosis to allo SCT was 35.4 months in MA group, and 34.2 months in RIC group. Eight (44%) patients in MA group and 36 (70%) patients in RIC group had prior autologous SCT. Six patients (33%) in the MA group and 11 (25%) in the RIC group received allo SCT from unrelated donors (p=0.3). Median number of prior treatment regimens were 5 (range 1–10) in both groups. Stem cell source was peripheral blood in 3 patients in MA group and 41 patients in the RIC group (p=0.0001). Results: Median follow-up in surviving patients was 27 months (3–98). All patients achieved engraftment. Cumulative TRM at 1 year was 56% in the MA group and 25% in the RIC group (p=0.03). Overall response rates in evaluable patients were 69% (CR=15%, PR= 54%) in MA group, and 79% (CR=23%, PR=56%) in the RIC group (p=0.47). Disease progression at 2 years was seen in 8 patients (44%) in the MA group and 25 patients (49%) in the RIC group (p=0.78). Median progression-free survival (PFS) in MA vs. RIC groups was 4.1 and 6.8 months, respectively (p=0.003) and median overall survival (OS) ) in MA vs. RIC group was 5.3 and 13.9 months, respectively (p=0.001). Cumulative Incidence of grade II–IV acute graft-vs.-host (GVHD) in MA vs. RIC groups disease was 33 vs. 27% (p=0.76); cumulative incidence of chronic GVHD in MA vs. RIC group was 54% vs. 47% (p=0.41) in evaluable patients. At the time of this analysis, 13 patients (25%) were still alive in RIC group, 7 of whom (14%) were in remission for up to 6 years post allo SCT. The most common causes of death were recurrent disease (30 patients; 43%), acute or chronic GVHD (16 patients; 23%) and opportunistic infections (5 patients: 7%). Conclusions: Allo SCT after RIC regimens is associated with longer PFS and OS and lower TRM. There was no increase in the risk of relapse, or acute or chronic GVHD. These regimens can safely replace MA regimens and may offer greater benefit if utilized earlier in the course of disease. Figure Figure

Abstract: T-cell replete haploidentical stem cell transplantation (HaploSCT) using post-transplant cyclophosphamide (PTCY) has been performed primarily with using non-myeloablative conditioning. This approach has been associated with very low non-relapse mortality (NRM); however, a high relapse rate was noted for pts in remission, suggesting that a more intense conditioning could be used to decrease rate of relapse post-transplant. We explored a myeloablative yet reduced-intensity melphalan-based conditioning regimen in an ongoing phase II clinical trial. Patients and Methods Results from the first 57 patients (pts) with advanced hematologic malignancies between 03/2009 and 4/2013 are reported. All pts received a melphalan-based conditioning regimen using fludarabine 160mg/m2, melphalan 140mg/m2 (N=35) or 100mg/m2 (N=22) +/- thiotepa 5-10mg/kg. GVHD prophylaxis was with post transplant cyclophosphamide 50 mg/kg on days +3 and +4 followed by tacrolimus and mycophenolate, for at least 6 and 3 mo post-transplant, respectively. Overall survival (OS) and progression-free survival (PFS) were assessed using the Kaplan-Meier method and group differences were determined using the log-rank test. Cumulative incidence (CI) of relapse, GVHD, and non-relapse mortality (NRM) were assessed using the competing risks method, where the competing risk for CIR was death, for NRM was relapse, and for GVHD was death and relapse. Group differences were determined using Gray's test. Results 34 (60%) pts were males and the median age was 47 years (range 21-66). Diagnoses were AML/MDS 28 (49.1%), CML 7 (12.3%), ALL 7 (12.3%), and lymphoma/CLL 10 (3 Hodgkin's, 3 NHL, 4 CLL) (17.5%), myelofibrosis/AML 4 (7%) and myeloma 1 pts. Racial distribution was: 18 (31.6%) African-Americans, 16 (28%) Caucasians, 12 (21%) Hispanics, 4 (7%) Asians, 7 (12.3%) Middle-Eastern. All patients except 2 (96.5%) had a bone marrow graft. 20/28 (71.4%) pts with AML/MDS were in complete remission (CR) at transplant (CR1+CR2) and 15/28 pts with AML had poor-risk by cytogenetics (N=12) or FLT3 mutation (N=3). All ALL pts had disease beyond CR1 except 1; 7/10 (70%) lymphoma/CLL pts had advanced disease. All patients with myelofibrosis have progressed to AML. Donors were children (N=26, 46.5%), siblings (N=23, 40.3%), parents (N=7, 12.3%), cousin (N=1), with 3-5 HLA antigen mismatches. The median follow-up of survivors was 14 mo (range: 1.0-48.4 months). One patient had early death due to RSV infection. Primary engraftment was achieved in 54/56 (96.5%) of evaluable patients, with majority (N=51, 90%) achieving full donor chimerism. Of the 6 who had mixed chimerism, 2 relapsed and 4 pts converted to full chimerism. Median time to neutrophil engraftment was 18 days (range 11-43). The CI of gr 2-4 aGVHD and 3-4 aGVHD were 29.7% and 3.7%, respectively, while CI of cGVHD limited+extensive and extensive only were 23.7% and 11.5%. Overall, the NRM was 21.4%, CI of relapse was 25.8% and PFS was 52.8% at the last follow-up (Figure 1A). NRM for myeloid pts transplanted in remission was 9% and PFS 66.8% at 50 mo median follow-up (Table 1, Figure 1B). Conclusions Melphalan-based conditioning for HaploSCT offers good disease control with acceptable NRM for patients in remission at transplant. A low relapse rate was observed in lymphoma/CLL patients despite the fact that most patients were not in remission at transplant. Overall, patients with advanced disease remain a challenge due to higher relapse rate and novel strategies are needed to significantly improve outcomes of these patients. Disclosures: No relevant conflicts of interest to declare.

Abstract: Background: The combination of low intensity therapy with inotuzumab ozogamicin improved survival compared to intensive chemotherapy and to single agent inotuzumab ozogamicin in first salvage (Jabbour et al. Cancer. 2018 (in press)). The incidence of veno-occlusive disease (VOD) is minimized with weekly divided dosage and reduced dose of inotuzumab ozogamicin per cycle. Blinatumomab single agent improves survival in relapsed / refractory ALL compared to that of standard chemotherapy. The sequential addition of blinatumomab to mini-hyper-CVD + inotuzumab ozogamicin might further improve survival and minimize the risk of veno-occlusive disease (VOD) by allowing a reduction of inotuzumab dose and spacing allogeneic stem cell transplant (ASCT) from the last dose of inotuzumab. Methods: Patients with relapsed / refractory Philadelphia chromosome negative ALL were eligible. The mini-hyper-CVD (cycles 1, 3, 5, 7) comprised cyclophosphamide (150 mg/m2 every 12 h on days 1-3), vincristine (2 mg flat dose on days 1 and 8), and dexamethasone (20 mg on days 1-4 and days 11-14) without anthracycline. Even cycles (cycles 2, 4, 6, 8) comprised methotrexate (250 mg/m2 on day 1) and cytarabine (0.5 g/m2 given every 12 h on days 2 and 3). Rituximab and intrathecal chemotherapy were given for first 4 courses. Inotuzumab ozogamicin was originally given on day 3 of the first four cycles at the dose of 1.3-1.8 mg/m2 at cycle 1, followed by 1.0-1.3 mg/m2 in subsequent cycles. After 67 pts were treated, an amendment was made to incorporate 4 cycles of blinatumomab after 4 cycles of mini-hyper-CVD + inotuzumab ozogamicin. Inotuzumab ozogamicin was given on days 2 and 8 at the dose of 0.6 and 0.3 mg/m2 at cycle 1, respectively, followed by days 2 and 8 at the dose of 0.3 and 0.3 mg/m2 at subsequent cycles; blinatumomab was continuously infused over 28 days every 42-day cycle for 4 cycles. The decision to proceed with ASCT was based on the discretion of the treating physician after discussion with the patient. Results: From 2/2013 to 5/2018, 84 patients were enrolled and treated including 17 patients with mini-hyper-CVD + inotuzumab + blinatumomab. The median follow-up is 31 months (range, 0.1-64.1). Patient characteristics and outcome are summarized in Table 1. The median age was 35 years (range, 9-87), and 23% of patients had received prior ASCT. The overall response rate was 80% (CR, 58%, CRp/CRi, 21%). These rates were 92% in S1 (primary refractory, 100%; CR1 duration 〈 12 months, 82%; CR1 duration 〉 12 months, 100%) and 56% in S2, and 60% in S3 or higher. Among 64 evaluable patients for minimal residual disease (MRD) assessment, 51 patients (80%) achieved negative MRD by 6-color flow cytometry with higher rates of negative MRD at 85% in salvage 1. Thirty four patients (40%) received ASCT. Three-year CR duration and overall survival (OS) rates were 49% and 33%, respectively (Figure 1). The median OS was 25 months, 6 months, and 7 months in salvage 1, salvage 2, and salvage 3 or more, respectively (p=0.001). Historical comparison showed median OS of 14 months and 6 months in hyper-CVD + inotuzumab ozogamicin +/- blinatumomab and inotuzumab ozogamicin single agent, respectively (p=0.001) (Figure 2). Among the 79 evaluable patients, VOD was observed in 9 (11%). The incidence of VOD was reduced from 9/61 (15%) with single dose of inotuzumab ozogamicin to 0/18 (0%) with weekly divided dose schedule. Of the 17 patients treated with mini-hyper-CVD + inotuzumab ozogamicin + blinatumomab, 3 patients underwent ASCT (2, haploidentical transplant; 1, cord blood transplant). Conclusion: The combination of inotuzumab ozogamicin plus/minus blinatumomab with low-intensity mini-hyper-CVD chemotherapy is effective and shows encouraging results in patients with relapsed/refractory ALL. The risk of VOD can be minimized with fractionated inotuzumab ozogamicin dosing. Disclosures Sasaki: Otsuka Pharmaceutical: Honoraria. Ravandi:Xencor: Research Funding; Amgen: Honoraria, Research Funding, Speakers Bureau; Jazz: Honoraria; Seattle Genetics: Research Funding; Seattle Genetics: Research Funding; Abbvie: Research Funding; Sunesis: Honoraria; Astellas Pharmaceuticals: Consultancy, Honoraria; Sunesis: Honoraria; Bristol-Myers Squibb: Research Funding; Jazz: Honoraria; Abbvie: Research Funding; Orsenix: Honoraria; Xencor: Research Funding; Astellas Pharmaceuticals: Consultancy, Honoraria; Macrogenix: Honoraria, Research Funding; Macrogenix: Honoraria, Research Funding; Amgen: Honoraria, Research Funding, Speakers Bureau; Orsenix: Honoraria; Bristol-Myers Squibb: Research Funding. Short:Takeda Oncology: Consultancy. Jain:Verastem: Research Funding; Abbvie: Research Funding; Abbvie: Research Funding; BMS: Research Funding; Seattle Genetics: Research Funding; Genentech: Research Funding; Cellectis: Research Funding; ADC Therapeutics: Honoraria, Membership on an entity's Board of Directors or advisory committees; BMS: Research Funding; Pfizer: Research Funding; Novimmune: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees; Infinity: Research Funding; Genentech: Research Funding; Infinity: Research Funding; Astra Zeneca: Research Funding; Celgene: Research Funding; Servier: Honoraria, Membership on an entity's Board of Directors or advisory committees; Abbvie: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; Astra Zeneca: Honoraria, Membership on an entity's Board of Directors or advisory committees; Incyte: Research Funding; Servier: Research Funding; Pharmacyclics: Research Funding; Abbvie: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pharmacyclics: Research Funding; Incyte: Research Funding; Verastem: Honoraria, Membership on an entity's Board of Directors or advisory committees; Adaptive Biotechnologies: Honoraria, Membership on an entity's Board of Directors or advisory committees; ADC Therapeutics: Research Funding; Astra Zeneca: Honoraria, Membership on an entity's Board of Directors or advisory committees; Cellectis: Research Funding; Adaptive Biotechnologioes: Research Funding; Pharmacyclics: Honoraria, Membership on an entity's Board of Directors or advisory committees; Verastem: Research Funding; Servier: Research Funding; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Astra Zeneca: Research Funding; Adaptive Biotechnologies: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novimmune: Honoraria, Membership on an entity's Board of Directors or advisory committees; Celgene: Research Funding; Servier: Honoraria, Membership on an entity's Board of Directors or advisory committees; ADC Therapeutics: Honoraria, Membership on an entity's Board of Directors or advisory committees; Seattle Genetics: Research Funding; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; ADC Therapeutics: Research Funding; Verastem: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pfizer: Research Funding; Pharmacyclics: Honoraria, Membership on an entity's Board of Directors or advisory committees; Adaptive Biotechnologioes: Research Funding; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees. Konopleva:Stemline Therapeutics: Research Funding. Champlin:Otsuka: Research Funding; Sanofi: Research Funding. Kadia:Pfizer: Consultancy, Research Funding; BMS: Research Funding; Jazz: Consultancy, Research Funding; Jazz: Consultancy, Research Funding; Pfizer: Consultancy, Research Funding; Amgen: Consultancy, Research Funding; Amgen: Consultancy, Research Funding; Takeda: Consultancy; Celgene: Research Funding; Abbvie: Consultancy; Abbvie: Consultancy; Takeda: Consultancy; Novartis: Consultancy; Celgene: Research Funding; BMS: Research Funding; Novartis: Consultancy. Cortes:Novartis: Consultancy, Research Funding; Daiichi Sankyo: Consultancy, Research Funding; Pfizer: Consultancy, Research Funding; Astellas Pharma: Consultancy, Research Funding; Arog: Research Funding. Jabbour:Takeda: Consultancy, Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Novartis: Research Funding; Abbvie: Research Funding; Pfizer: Consultancy, Research Funding.

Abstract: BACKGROUND: MCL remains incurable with cytotoxic therapy, but may be susceptible to immune-based modalities. Since the availability of NMT(1997) and rituximab(1999), we have employed the following risk-adapted strategy: patients (pts) in first remission after one chemotherapy (REM1) were offered autologous SCT (ASCT) with high-dose rituximab (R), and pts beyond REM1 were offered NMT if a donor was available, and ASCT+R if not. In order to assess the success of this strategy, we analyzed the results of 17 years of MCL transplantation at our center, including ASCT pts transplanted without R as historical controls. METHODS: 52 pts in REM1 (ASCT1) and 36 pts beyond REM1 (ASCT2) received ASCT with (43%) or without (57%) R, and 35 pts beyond REM1 received NMT with fludarabine, cyclophosphamide & R. ASCT1, ASCT2 and NMT pts were balanced for age, Ann-Arbor stage, B-symptoms, LDH and B2m, but NMT pts were more heavily pretreated (p=0.01) and were further from initial diagnosis (p=0.02) than ASCT2 pts. RESULTS {ASCT}: 96% & 91% of ASCT1 & ASCT2 pts were in CR following transplantation. Chemoresistance was an important determinant of ASCT outcome, with ASCT1 pts experiencing superior median PFS (42 v 27 months, p=0.009) and OS (94 v 52 months p=0.01) than ASCT2. B-symptoms and B2m≥3.0mg/L adversely affected OS in both ASCT groups. There was a trend to improved PFS for the 20 ASCT1 pts who received R, with no events in 10 pts followed between 24 to 94 months, compared with a continuous pattern of progression in non-R pts (p=0.06). {NMT}: 94% of pts were in CR after transplantation, and the 42 month current-PFS was 57%. In contrast to the ASCT pts, OS following NMT was not affected by high B2m, although B-symptoms remained significant (p=0.03). Despite all pts being transplanted beyond REM1, a plateau in survival was evident with no deaths occurring among 11 pts followed between 46 to 98 months (figure). Actuarial non-relapse mortality (NRM) at 90 days, 12 months and 2 years were 0%, 13% and 20% respectively. Neither the stem cell source (related 69% v unrelated 31%) nor a history of previous autologous transplantation (n=6) had a significant impact on NMT outcome (p 〉 0.30 for PFS, OS and NRM). CONCLUSIONS: These results suggest that ASCT is most effective in pts transplanted in REM1, in whom the addition of R may have improved PFS. NMT is effective in overcoming the adverse impact of prior chemotherapy and high B2m, and remains the modality of choice for pts beyond first remission. Figure Figure

Abstract: Background: Allogeneic stem cell transplantation is an established, potentially curative therapy for patients(pts) with CML. Its use has declined related to the favorable results of tyrosine kinase inhibitors (TKI) and related to concerns regarding toxicities and GVHD. Minimal residual disease is detectable in most pts treated with TKIs and some have overt progression of their leukemia. The efficacy of allotransplantation in CML is largely derived from the immune graft-vs-leukemia effect mediated by alloreactive donor T cells. Nonmyeloablative conditioning has reduced the toxicity of allotransplantation, and post transplant treatment with imatinib and donor lymphocyte infusions can produce durable molecular complete remissions in substantial fractions of patients who have persistent or recurrent disease. Methods: We conducted a prospective clinical trial designed to induce molecular CR (mCR) in pts with CML with residual disease despite imatinib treatment. Pts received a reduced intensity preparative regimen involving imatinib 800 mg daily, fludarabine 40 mg/m2 x 4 days, busulfan 130 mg/m2 x 2 days, and Thymoglobulin 2.5 mg/kg daily x 3 days followed by allogeneic stem cell transplant from an HLA-identical or one antigen mismatched related or unrelated donor. The regimen was designed to achieve engraftment with a low rate of GVHD and treatment related mortality. Pts with residual disease after 3 mo by quantitative PCR analysis for bcr-abl received imatinib and those who did not achieve mCR within 3 mo received escalating doses of donor lymphocyte infusion (DLI). Results: 33 pts were entered, median age 41 (range 22–69) yrs. All were previously treated with imatinib, 11 were in cytogenetic CR (CyCR), 32 had detectable disease by PCR. 16 pts had early disease (in chronic phase or with isolated clonal evolution) and 17 had advanced disease (with prior accelerated phase or in second chronic phase after blast crisis). Median follow up time is 2.4 (range 0.3 – 4.2) yrs. The regimen was well tolerated. One pt required a second transplant for rejection. 7 pts (21%) developed reversible grade 2–3 acute GVHD. None died within 100 days post transplant and only one patient died later from causes other than relapse. At 3 months, all 16 early pts and 12 of the 17 advanced pts achieved CyCR and 4 early and 2 advanced pts had mCR respectively. 18 pts with residual or recurrent molecular residual disease after 3 mo and received treatment with imatinib; mCR was achieved in 7 of 9 early pts and 2 of 9 with advanced CML. 8 pts subsequently received DLI; one early pt achieved mCR another pt is too early to evaluate (TE). Of 6 advanced pts receiving DLI, 2 have a mCR, and one is TE. 25 patients are alive, including 15 of 16 pts with early disease (94%), all currently in CyCR, 9 in mCR. Of the 17 pts with advanced disease, 10 are alive (58%), 8 in CyCR and 5 in mCR. 6 advanced pts succumbed to recurrent blast crisis. Conclusion: Sequential therapy including nonmyeloablative allotransplantation, post transplant imatinib, and DLI deserves further study as a well tolerated approach to induce durable mCR in imatinib treated CML patients. Best results occurred in patients transplanted before overt transformation of their leukemia.

Abstract: Abstract 3471 AML with the internal tandem duplication of the Fms-Like Tyrosine kinase-3 gene (FLT3-ITD) has a poor prognosis, with high relapse rates. Sorafenib is an oral multikinase inhibitor of tyrosine kinases including Flt-3 with activity in AML. We used this drug to treat patients relapsing after allogeneic HSCT and herein present the results. Patients and Methods. We retrospectively evaluated 16 patients with FLT3-ITD AML that received sorafenib for at least a 7-day course either alone (n=8; 50%) or in combination (n=8; 50%) to treat AML relapsing at a median of 3 months (range, 1–7) after HSCT. Median age was 34 years (range, 20–603). Cytogenetics was diploid in 69%, high-risk in 19% and unknown in 12%. Treatment history prior to HSCT included sorafenib in 38% of the cases, and 30% were status post a 2nd HSCT. Donors were related (50%) or unrelated (50%). Seven patients (44%) received and failed other salvage chemotherapy prior to starting sorafenib (median=2 treatments; range, 1–5). At start of sorafenib salvage treatment, median WBC was 22.5 K/mm3 (range, 0.6–119), and median bone marrow blast was 58% (range, 12–88%), while 75% of the patients had circulating blasts. Sorafenib doses were as follows: 400 mg (n=3) or 800 mg daily (n=4) in combination to idarubicin/Ara-C, or 800 mg daily with 5-azacitidine (n=1); single agent: 800 mg (n=6) or 1200 mg (n=2) daily. Median duration of sorafenib single agent therapy was 39.5 days (range, 10–100), and in combination it was 7 days (range, 7–32). Complete remission (CR) rate was zero; 2 patients achieved a PR. Median reduction in bone marrow (n=9) and peripheral blood (n=12) blasts was 0% (range, 0–46), and 50% (range, 0–88), respectively. Two patients were ‘bridged’ to a 2nd HSCT, which led to CR durations of 53 and 106 days. Median survival after sorafenib initiation was 81 days and it was similar with single agent versus combination therapy (Figure 1 and 2). All patients have died. Conclusions. Treatment of overt AML relapse with sorafenib was unsuccessful as described here. It is possible that ‘pre-emptive’ treatment in the presence of minimal residual disease will be more effective, but this remains to be proven. Disclosures: No relevant conflicts of interest to declare.

Abstract: Abstract 489 Relapse remains the major cause of treatment failure after allogeneic hematopoietic transplantation for AML and MDS. Alloreactive NK cells mediate a potent antileukemic effect and may also enhance engraftment and reduce GVHD. We performed a phase I study infusing “third party” alloreactive NK cells from a haploidentical related donor as a component of the preparative regimen for allotransplantation from a separate HLA identical donor. The goal was to augment the antileukemia cytotoxicity of the preparative regimen by infusion of alloreactive NK cells and improve the overall outcome of hematopoietic transplantation. Patients with advanced AML or high risk MDS in relapse or beyond first remission were eligible. They received the busulfan-fludarabine preparative regimen, followed by infusion of NK cells predicted to be alloreactive by KIR:KIR ligand incompatibility. The NK cell enriched product was produced from a steady state apheresis product by depleting CD3+ cells using the CliniMACS device (Miltenyi Corp). A second step positively selecting CD56+ cells was performed for the first dose level, but discontinued thereafter in order to increase the cell yield. The NK cell product was then cultured in complete media containing rIL-2 for 16 hours and infused intravenously following completion of the busulfan-fludarabine chemotherapy. After 5 days, ATG was administered followed by PBSC infusion from the HLA identical sibling or unrelated donor. Patients received tacrolimus and methotrexate for GVHD prophylaxis. Patients were treated in 4 dose levels of the NK cell enriched product; 1) 106 cells/kg, 2) 5 × 106/kg, 3) 3 × 107/kg, and 4) 3 × 107/kg followed by systemic interleukin-2 0.5 million units/m2 SQ daily for 5 days. CD3+ cells in the NK cell product were required to be 〈 105/kg (median infused 1.1 x104/kg). Median CD56+ cells infused (x106/kg) were from 0.9 (level 1), 1.5 (level 2), and 4.9 (levels 3 and 4). 13 patients were entered. Median age was 51 years (range 2–60). 11 had active disease and 2 were in a second remission. Only mild infusion toxicity occurred with the NK cell infusion. Other toxicities were similar to that experienced with the preparative regimen without NK cells. The 2 patients at dose level 4 tolerated interleukin-2 systemic treatment without fever or increased toxicity. The haploidentical NK cells were transiently detected in the blood by chimerism studies in one patient. Rapid engraftment and hematologic recovery uniformly occurred from the HLA matched PBPC donor in all patients. None had graft failure. Grade 2 acute GVHD developed in 3 patients; all responded to corticosteroid treatment. None developed grade 3 or 4 acute GVHD. 9 of 11 patients with active disease achieved a complete remission. Only one patient died of nonrelapse mortality; she died in remission from infection 2 months post transplant. This trial confirms the feasibility of producing the haploidentical NK cell product, and the lack of major toxicity attributable to the NK cell infusion in combination with an HLA compatible allogeneic transplantation. Infusion of haploidentical alloreactive NK cells was well tolerated and did not interfere with engraftment or increase the rate of GVHD after allogeneic hematopoietic transplantation. This approach merits further phase II and III study designed to reduce relapse and improve the outcome of allogeneic hematopoietic transplantation for AML/MDS. N NK cell doseCells/kg grade 2 GVHD Relapse NRM Alive in CR 4 106 1 4 0 0 4 5 × 106 0 3 0 1 (27+ m) 3 3 × 107 2 1 1 1 (11+ m) 2 3 × 107 + IL-2 0 0 0 2 (9+ m, 4+ m) Disclosures: No relevant conflicts of interest to declare.

Abstract: Abstract 519 We have previously identified a high risk of primary graft failure (PGF) in patients with DSA in T-cell depleted haploidentical transplantation (HaploSCT); 3/4 patients with DSA had PGF compared with 1/20 patients without DSA. All patients with DSA against anti-HLA A, B and DRB1 developed PGF, while 1 patient with anti-DP DSA did not. We now hypothesize that anti-DP antibodies may have a significant but less deleterious impact on engraftment. Methods: We evaluated the occurrence of PGF in 592 patients who received matched unrelated donor transplants at our institution after 1/2005. 88 % of the transplants were matched in 8/8 alleles of HLA-A,B,C,DRB1 in the HvG vector; approximately 75% of the transplants were mismatched in either DRB 3/4/5, DQB1 or DPB1. The presence of DSA was determined by testing the patients' sera with a panel of fluorescent beads coated with single HLA antigen preparations using a Luminex™ platform; results were interpreted as fluorescence intensity (FI) against DSA mismatch. All HLA loci were typed by high resolution methods. Results: 19/592 patients (3%) had either PGF (N=9) or died early without engraftment (ED) (N=10). The only DSA identified were against the DP molecule in 8 patients matched in 8/8 HLA alleles with no apparent specificity. 3/8 (37.5%) patients had PGF/ED compared with 16/584 (2.7%) who did not have DSA (p=0.001, RR=23.3). One patient with PGF in the presence of anti-DP DSA had a second transplant in the absence of antibodies and engrafted cells from the same donor. In spite of common HLA sensitization (116/592, 19.6%) only 8 patients (1.4%) presented DSA in their pre-transplant specimens, mostly females (N=7) with a median age 49 years. DSA ranged from 1558 to 9845 FI, similar in patients with and without engraftment. Overall there was a 3% risk of PGF/ED without HLA antibodies, 6% risk in the presence of antibodies but not DSA, and 37.5% risk in the presence of DSA. Multivariate analysis revealed that DSA (p=0.0001) and ABO mismatch (p=0.04) were the only variables associated with graft failure. There was a significant association between female gender and allosensitization, 30.8% of females had anti HLA antibodies vs. 12.1% males (p 〈 0.0001). While no difference in the incidence of HLA antibodies was observed between females with no prior pregnancies and males (p=0.24), this became apparent when allosensitization was evaluated in males vs. females with one pregnancy (p=0.008) and females with 2 or more pregnancies (p=0.0003). Conclusions: These results, combined with our previous findings in HaploSCT, suggest that DSA are associated with graft rejection in hematopoietic stem cell transplantation. Graft failure occurs less often with anti-DP DSA (3/9) compared with DSA against high expression HLA molecules (3/3, HLA-A, B or DRB1) and may confer a lower risk for graft rejection. The differences may reside in the lower levels of DP molecules expressed on cell surface. DSA screening is warranted when considering donors with HLA mismatches as strategies for donor selection and/or antibody level reduction may be needed to decrease the risk of PGF in allogeneic HSCT from partially HLA-matched donors. Disclosures: No relevant conflicts of interest to declare.

Abstract: Introduction: T-cell acute lymphoblastic leukemia (T-ALL) represents approximately 20-25% of all cases of ALL diagnoses per year. Given the rarity of T-ALL, there are limited data to guide the treatment of these patients, particularly as it relates to allogeneic stem cell transplantation (allo-SCT). Early thymic-precursor (ETP) ALL is a rare variant of T-ALL which expresses both myeloid and lymphoid markers, and has been associated with increased relapse and worse overall survival in some series. Minimal residual disease (MRD) is established as a prognostic marker for increased relapse in B-ALL, but its effect on T-ALL, particularly as it relates to allo-SCT, is less clear. We conducted a multi-center analysis of patients with T-ALL to determine the effect of T-ALL subtype and MRD on transplant outcomes to better guide decisions regarding therapy for this aggressive disease. Methods: Data from MD Anderson Cancer Center (MDACC), the National University Cancer Institute of Singapore, and the Oregon Health & Science University were reviewed. All patients with a diagnosis of T-ALL who received a first allo-SCT after January 1, 2000 were included in the analysis. Flow cytometry data were reviewed centrally at MDACC to determine the T-ALL subtype. ETP was defined according to Coustan-Smith et al. (Lancet Oncol 2009;10:147-56). T-ALL subtype was determined using a similar system based upon Ludwig (Leuk Lymph 1994: 13:Suppl 1: 71-76) and Gassmann et al (Br J Haem 1997:97:372-82). MRD was evaluated by flow cytometry within one month prior to allo-SCT on bone marrow biopsy, and was detected with a sensitivity of 10-3. Overall survival (OS) and progression free survival (PFS) were estimated using the Kaplan-Meier method. Disease progression, acute GVHD (aGVHD), and chronic GVHD (cGVHD) were determined utilizing the cumulative incidence (CI) method to account for competing risks. Prognostic factors for OS were assessed using Cox Proportional Hazards regression analysis on univariate and multivariate analysis. Results: 103 patients received first allo-SCT, with a median age of 31 years (range 2-72 years), and median follow-up time of 2.2 years (3.4-7.3 years). Of these, 91 had an identifiable subtype. The 3-year OS, PFS, CI NRM and CI progression was 34%, 31%, 11%, and 58% for the whole cohort, respectively. The CI 100-day NRM was 5%, while the incidence of grade II-IV aGVHD was 48% and CI cGVHD was 29% at 3 years. There was no difference in OS or NRM amongst T-ALL subtypes, including ETP (p=0.8), (Table 1). Furthermore, there was no difference in the rate progression of ETP (HR 1.5 (0.6-3.3, p=0.4). Amongst patients transplanted after 2007, when flow cytometry for MRD was widely utilized, 66 patients had available data: 41 MRD negative, 17 MRD positive, and 8 with no CR defined by 〉 5% marrow blasts. On univariate analysis for OS, there was no difference according to institution, age, disease status at transplant (CR1 vs CR2+), white blood count, presence of extra-medullary disease at diagnosis, presence of CNS involvement, stem cell source, TBI vs no TBI, cytogenetic risk (SWOG classification), or complex cytogenetics. No CR at transplant (HR 3.9, p=0.003) and MRD positivity (HR 2.5, p=0.015) were associated with worse OS. With a median follow-up time of 2 years, the OS rates for patients who were MRD negative, MRD positive, and no CR were 62%, 24%, and 12%, respectively (Figure 1), with CI of progression of 30%, 69%, and 87% respectively (Figure 2). On multivariate analysis, MRD positivity was strongly predictive of OS (HR 2.3, p=0.03) and PFS (HR 2.1, p=0.04). Conclusions: There is no difference in outcomes of patients with T-ALL according to disease subtype. In particular, patients with ETP-ALL who underwent allo-SCT had similar outcomes to other T-ALL subtypes, suggesting that allo-SCT may overcome this poor prognostic primary disease feature. MRD at transplant is the strongest predictor of disease progression and lower survival in patients with T-ALL, with outcomes comparable to those transplanted with active disease. Novel approaches pre-transplant and post-transplant are necessary to mitigate the poor prognostic impact of MRD in T-ALL. Table 1. Outcomes of Allogeneic Stem Cell Transplantation According to T-ALL Subtype Subtype n (%) OS HR p-value PFS HR Progression p-value Early 30 (33) 41% Ref 41% Ref Ref Cortical 28 (31) 26% 1.3 0.4 24% 0.8 1.8 0.1 Mature 17 (19) 31% 1.1 0.8 23% NE 2.4 0.04 ETP 16 (17) 29% 0.9 0.8 22% 0.9 1.5 0.4 Figure 1. Figure 1. Figure 2. Figure 2. Disclosures Brammer: Celgene: Research Funding.