Search

Severe aplastic anemia (SAA) is a marrow failure disorder with high morbidity and mortality. It is treated with bone marrow transplantation (BMT) for those with fully matched donors or immunosuppressive therapy (IST) for those who lack such a donor, which is often the case for underrepresented minorities. We conducted a prospective phase II trial of reduced-intensity conditioning HLA-haplo BMT and post-transplantation cyclophosphamide (PTCy)-based graft-versus-host (GVHD) prophylaxis as initial therapy for patients with SAA. The median age was 25 (range 3-63) years and the median follow-up was 40.9 months (95% CI: 29.4, 55.7 mos). Over 35% of enrollment was from underrepresented racial/ethnic groups. The cumulative incidence of grade II-IV aGVHD at day 100 is 7% (95% CI: NA, 17%) and chronic GVHD at 2 years is 4% (95% CI: NA, 11%). The overall survival for 27 patients is 92% (95% CI: 83,100%) at one, two, and three years. The first 7 patients received lower dose total body irradiation (200 versus 400 cGY), but these patients were more likely to have graft failure, 3 of 7, compared to 0 out of 20 patients in the higher dose group (p=0.01, Fisher exact). HLA-haploidentical BMT with PTCy using 400cGY total body irradiation resulted in 100% overall survival with minimal GVHD in 20 consecutive patients. Not only does this approach avoid the ramifications of IST and its low failure-free survival, but also the use of haploidentical donors expands access to BMT across all populations. Clinical trial: NCT02833805

The key immunologic signatures associated with clinical outcomes after posttransplant cyclophosphamide (PTCy)-based HLA-haploidentical (haplo) and HLA-matched bone marrow transplantation (BMT) are largely unknown. To address this gap in knowledge, we used machine learning to decipher clinically relevant signatures from immunophenotypic, proteomic, and clinical data and then examined transcriptome changes in the lymphocyte subsets that predicted major posttransplant outcomes. Kinetics of immune subset reconstitution after day 28 were similar for 70 patients undergoing haplo and 75 patients undergoing HLA-matched BMT. Machine learning based on 35 candidate factors (10 clinical, 18 cellular, and 7 proteomic) revealed that combined elevations in effector CD4+ conventional T cells (Tconv) and CXCL9 at day 28 predicted acute graft-versus-host disease (aGVHD). Furthermore, higher NK cell counts predicted improved overall survival (OS) due to a reduction in both nonrelapse mortality and relapse. Transcriptional and flow-cytometric analyses of recovering lymphocytes in patients with aGVHD identified preserved hallmarks of functional CD4+ regulatory T cells (Tregs) while highlighting a Tconv-driven inflammatory and metabolic axis distinct from that seen with conventional GVHD prophylaxis. Patients developing early relapse displayed a loss of inflammatory gene signatures in NK cells and a transcriptional exhaustion phenotype in CD8+ T cells. Using a multimodality approach, we highlight the utility of systems biology in BMT biomarker discovery and offer a novel understanding of how PTCy influences alloimmune responses. Our work charts future directions for novel therapeutic interventions after these increasingly used GVHD prophylaxis platforms. Specimens collected on NCT0079656226 and NCT0080927627 https://clinicaltrials.gov/.

Purpose Allogeneic bone marrow transplantation (alloBMT) is the only cure for many primary immune deficiency disorders (PIDD), primary immune regulatory disorders (PIRD), and inherited bone marrow failure syndromes (IBMFS). Methods We report the results of 25 patients who underwent alloBMT using reduced intensity conditioning (RIC), alternative donors, and post-transplantation cyclophosphamide (PTCy). In an attempt to reduce regimen-related toxicities, we removed low-dose TBI from the prep and added mycophenolate mofetil and tacrolimus for graft-versus-host disease (GVHD) prophylaxis for all donor types in the latter 14 patients. Donors were haploidentical related (n = 14), matched unrelated (n = 9), or mismatched unrelated (n = 2). The median age was 9 years (range 5 months–21 years). Results With a median follow-up of 26 months (range 7 months–9 years), the 2-year overall survival is 92%. There were two deaths, one from infection, and one from complications after a second myeloablative BMT. Three patients developed secondary graft failure, one at 2 years and two at >3 years, successfully treated with CD34 cell boost in one or second BMT in two. The remaining 20 patients have full or stable mixed donor chimerism and are disease-free. The incidence of mixed chimerism is increased since removing TBI from the prep. The 6-month cumulative incidence of grade II acute GVHD is 17%, with no grade III–IV. The 1-year cumulative incidence of chronic GVHD is 14%, with severe of 5%. Conclusion This alloBMT platform using alternative donors, RIC, and PTCy is associated with excellent rates of engraftment and low rates of GVHD and non-relapse mortality, and offers a curative option for patients with PIDD, PIRD, and IBMFS. Trial registration ClinicalTrials.gov Identifier: NCT04232085

Allogeneic (allo-) hematopoietic stem cell transplant (HSCT) is curative for many nonmalignant pediatric disorders, including hemoglobinopathies, bone marrow failure syndromes, and immunodeficiencies. There is great success using HLA-matched related donors for these patients; however, the use of alternative donors has been associated with increased graft failure, graft versus host disease (GVHD), and transplant-related mortality (TRM). HSCT using alternative donors with post-transplantation cyclophosphamide (PT/Cy) for GVHD prophylaxis has been performed for hematologic malignancies with engraftment, GVHD, and TRM comparable to that seen with HLA-matched related donors. There are limited reports of HSCT in nonmalignant pediatric disorders other than hemoglobinopathies using alternative donors and PT/Cy. We transplanted eleven pediatric patients with life-threatening nonmalignant conditions using reduced intensity conditioning (RIC), alternative donors, and PT/Cy alone or in combination with tacrolimus and mycophenolate mofetil. We observed limited GVHD, no TRM, and successful engraftment sufficient to eliminate manifestations of disease in all patients. Allo-HSCT using alternative donors and PT/Cy shows promise for curing nonmalignant disorders; development of prospective clinical trials to confirm these observations is warranted.

There are few therapeutic options for patients with T-cell acute lymphoblastic leukemia (T-ALL) who have recurrent disease after initial matched sibling hematopoietic stem cell transplantation. While a second hematopoietic stem cell transplant (HSCT) from a haploidentical donor offers the conceptual possibility of greater graft versus leukemia effect, there is minimal literature to describe the efficacy of this approach in recurrent pediatric T-ALL. We present the case of a now 9-year-old female in whom second haploidentical HSCT, followed by successive donor lymphocyte infusions in response to minimal residual disease reemergence, has led to 3+ years of ongoing disease control without graft versus host disease and excellent quality of life.

Noninfectious fevers are common early after T cell–replete HLA haploidentical (haplo) peripheral blood transplants and have been associated with cytokine release syndrome and overall mortality. However, less is known regarding the incidence and associations of early fever after bone marrow transplantation (BMT) with post-transplant cyclophosphamide (PTCy). We hypothesized that early fever would be associated with myeloablative conditioning (MAC), because of its relative increase in tissue damage augmenting antigen presentation and class II HLA-mismatching because of recognition of antigen-presenting cells by CD4(+) T cells. In 672 recipients of MAC HLA-matched related donor (MRD) (n = 183), MAC HLA-matched unrelated donor (MUD) (n = 115), MAC haplo (n = 79), or nonmyeloablative (NMA) haplo (n = 295) T cell–replete BMT with PTCy, we retrospectively analyzed early noninfectious fever defined as temperature of ≥38.3°C once or ≥38.0°C twice or more on days 1 to 6. Fever occurred in 13% after MAC MRD, 23% after MAC MUD, 44% after NMA haplo, and 84% after MAC haplo BMT (P < .0001). Survival outcomes did not differ between patients with and without early fever. In NMA haplo BMT, mismatch in the graft-versus-host direction at HLA-DRB1 or-DPB1 (but not HLA -A, -B, -Cw, or -DQB1) was associated with early fever compared with no mismatches at these loci (P < .0001 and P =.02, respectively). In multivariable modeling, -DRB1 or -DPB1 mismatch and higher CD3(+) graft cell dose were significantly associated with early fever. Early fever is more common after haplo compared with HLA-matched BMT. Fever is associated with myeloablation, -DRB1 or -DPB1 mismatching, and higher CD3(+) graft cell dose but not survival.

Related HLA-haploidentical blood or marrow transplantation (BMT) with high-dose posttransplantation cyclophosphamide (PTCy) is being increasingly used because of its acceptable safety profile. To better define outcomes of nonmyeloablative (NMA) HLA-haploidentical BMT with PTCy, 372 consecutive adult hematologic malignancy patients who underwent this procedure were retrospectively studied. Risk-stratified outcomes were evaluated using the refined Disease Risk Index (DRI), developed to stratify disease risk across histologies and allogeneic BMT regimens. Patients received uniform conditioning, T-cell–replete allografting, then PTCy, mycophenolate mofetil, and tacrolimus. Six-month probabilities of nonrelapse mortality and severe acute graft-versus-host disease were 8% and 4%. With 4.1-year median follow-up, 3-year probabilities of relapse, progression-free survival (PFS), and overall survival (OS) were 46%, 40%, and 50%, respectively. By refined DRI group, low (n = 71), intermediate (n = 241), and high/very high (n = 60) risk groups had 3-year PFS estimates of 65%, 37%, and 22% (P < .0001), with corresponding 3-year OS estimates of 71%, 48%, and 35% (P = .0001). On multivariable analyses, the DRI was statistically significantly associated with relapse, PFS, and OS (each P < .001). This analysis demonstrates that the DRI effectively risk stratifies recipients of NMA HLA-haploidentical BMT with PTCy and also suggests that this transplantation platform yields similar survivals to those seen with HLA-matched BMT.

Recent studies have suggested that plasma-derived proteins may be potential biomarkers relevant for graft-versus-host disease and/or non-relapse mortality occurring after allogeneic blood or marrow transplantation. However, none of these putative biomarkers have been assessed in patients treated either with human leukocyte antigen-haploidentical blood or marrow transplantation or with post-transplantation cyclophosphamide, which has been repeatedly associated with low rates of severe acute graft-versus-host disease, chronic graft-versus-host disease, and non-relapse mortality. We explored whether seven of these plasma-derived proteins, as measured by enzyme-linked immunosorbent assays, were predictive of clinical outcomes in post-transplantation cyclophosphamide-treated patients using plasma samples collected at serial predetermined timepoints from patients treated on prospective clinical studies of human leukocyte antigen-haploidentical (n=58; clinicaltrials.gov Identifier: 00796562) or human leukocyte antigen-matched-related or -unrelated (n=100; clinicaltrials.gov Identifiers: 00134017 and 00809276) T-cell-replete bone marrow transplantation. Day 30 levels of interleukin-2 receptor α, tumor necrosis factor receptor 1, serum STimulation-2 (IL1RL1 gene product), and regenerating islet-derived 3-α all had high areas under the curve of 0.74-0.97 for predicting non-relapse mortality occurrence by 3 months post-transplant in both the human leukocyte antigen-matched and human leukocyte antigen-haploidentical cohorts. In both cohorts, all four of these proteins were also predictive of subsequent non-relapse mortality occurring by 6, 9, or 12 months post-transplant and were significantly associated with non-relapse mortality in univariable analyses. Furthermore, day 30 elevations of interleukin-2 receptor α were associated with grade II-IV and III-IV acute graft-versus-host disease occurring after day 30 in both cohorts. These data confirm that plasma-derived proteins previously assessed in other transplantation platforms appear to retain prognostic and predictive utility in patients treated with post-transplantation cyclophosphamide.

Natural killer (NK) cell alloreactivity, which may contribute to the graft-versus-leukemia effect of allogeneic hematopoietic stem cell transplantation (HSCT), is influenced by the interaction of killer-cell immunoglobulin-like receptors (KIRs) on donor NK cells and their ligands, human leukocyte antigen (HLA) Class I molecules on recipient antigen-presenting cells. Distinct models to predict NK cell alloreactivity differ in their incorporation of information from typing of recipient and donor KIR and HLA gene loci, which exist on different autosomes and are inherited independently as haplotypes. Individuals may differ in the inheritance of the two KIR haplotypes, A and B, or in the expression of individual KIR genes. Here, we examined the effect of KIR and HLA genotype, in both the recipient and donor, on the outcome of 86 patients with advanced hematologic malignancies who received nonmyeloablative, HLA-haploidentical HSCT with high-dose, post-transplantation cyclophosphamide (Cy). Compared to recipients of marrow from donors with identical KIR gene content, recipients of inhibitory KIR (iKIR) gene-mismatched marrow had an improved overall survival (HR=0.37; CI: 0.21- 0.63; p=0.0003), event-free survival (HR=0.51; CI: 0.31- 0.84; p=0.01), and relapse rate (cause specific hazard ratio, SDHR=0.53; CI: 0.31-0.93; p=0.025). Patients homozygous for the KIR “A” haplotype, which encodes only one activating KIR, had an improved overall survival (HR=0.30; CI: 0.13-10.69; p=.004), event-free survival (HR=0.47; CI: 0.22-1.00; p=.05) and non-relapse mortality (NRM; cause specific HR=0.13; CI: 0.017-0.968; p=0.046) if their donor expressed at least one KIR B haplotype, which encodes several activating KIRs. Models that incorporated information from recipient HLA typing, with or without donor HLA typing, were not predictive of outcome in this patient cohort. Thus, nonmyeloablative conditioning and T cell-replete, HLA-haploidentical HSCTs involving iKIR gene mismatches between donor and recipient, or KIR haplotype AA recipients of marrow from KIR Bx donors, were associated with lower relapse and NRM and improved overall and event-free survival. These findings suggest that selection of donors based upon inhibitory KIR gene or haplotype incompatibility may be warranted.

Hematopoietic SCT from a partially HLA-mismatched (HLA-haploidentical) first-degree relative offers the benefits of rapid and near universal donor availability but also the risks that result from traversing the HLA barrier; namely, graft failure, severe GVHD and prolonged immunodeficiency. Improvements over the last 10 years in conditioning regimens, graft engineering and pharmacological immuno-prophylaxis of GVHD have substantially reduced the morbidity and mortality of HLA-haploidentical SCT. Highly immunosuppressive but nonmyeloablative conditioning extends the availability of HLA-haploidentical SCT to elderly hematologic malignancy patients lacking HLA-matched donors and permits recovery of autologous hematopoiesis in the event of graft failure. Current regimens for HLA-haploidentical SCT are associated with a 2-year non-relapse mortality of 20 ± 5%, relapse of 35 ± 15% and overall survival of 50 ± 20%. Major developmental areas include harnessing natural killer cell alloreactivity to reduce the risk of disease relapse and improving immune reconstitution by delayed infusions of lymphocytes selectively depleted of alloreactive cells. Hematologic malignancy patients who lack suitably matched related or unrelated donors can now be treated with HLA-haploidentical related donor or unrelated umbilical cord blood SCT. Future clinical trials will assess the relative risks and benefits of these two graft sources.

We evaluated the safety and efficacy of high-dose, posttransplantation cyclophosphamide (Cy) to prevent graft rejection and graft-versus-host disease (GVHD) after outpatient nonmyeloablative conditioning and T cell-replete bone marrow transplantation from partially HLA-mismatched (haploidentical) related donors. Patients with advanced hematologic malignancies (n = 67) or paroxysmal nocturnal hemoglobinuria (n = 1) received Cy 50 mg/kg i.v. on day 3 (n = 28) or on days 3 and 4 (n = 40) after transplantation. The median times to neutrophil (>500/μL) and platelet recovery (>20,000/μL) were 15 and 24 days, respectively. Graft failure occurred in 9 of 66 (13%) evaluable patients, and was fatal in 1. The cumulative incidences of grades II-IV and grades III-IV acute (aGVHD) by day 200 were 34% and 6%, respectively. There was a trend toward a lower risk of extensive chronic GVHD (cGVHD) among recipients of 2 versus 1 dose of posttransplantation Cy (P = .05), the only difference between these groups. The cumulative incidences of nonrelapse mortality (NRM) and relapse at 1 year were 15% and 51%, respectively. Actuarial overall survival (OS) and event-free survival (EFS) at 2 years after transplantation were 36% and 26%, respectively. Patients with lymphoid malignancies had an improved EFS compared to those with myelogenous malignancies (P = .02). Nonmyeloablative HLA-haploidentical BMT with posttransplantation Cy is associated with acceptable rates of fatal graft failure and severe aGVHD or cGVHD. © 2008 American Society for Blood and Marrow Transplantation.

The "allogeneic effect" refers to the induction of host B cell antibody synthesis or host T cell cytotoxicity, including tumoricidal activity, by an infusion of allogeneic lymphocytes. We show that treatment of mice with cyclophosphamide (Cy) followed by CD8(+) T cell-depleted allogeneic donor lymphocyte infusion (Cy + CD8(-) DLI) induces regression of established tumors with minimal toxicity in models of both hematologic and solid cancers, even though the donor cells are eventually rejected by the host immune system. The optimal antitumor effect of Cy + CD8(-) DLI required the presence of donor CD4(+) T cells, host CD8(+) T cells, and alloantigen expression by normal host but not tumor tissue. The results support a model in which a donor CD4(+) T cell-mediated graft-versus-host (GVH) reaction effectively awakens antitumor immunity among Cy-resistant host CD8(+) T cells. These events provide the cellular mechanism of the "allogeneic effect" in antitumor immunity. Cy + CD8(-) DLI may be an effective and minimally toxic strategy for awakening the host immune response to advanced cancers.

Hematopoietic SCT is currently the only curative therapy for a range of benign inherited and acquired primary hematologic disorders in children, including BM failure syndromes and hemoglobinopathies. The preferred HLA-matched sibling donor is available for only about 25% of such children. However, there has been substantial progress over the last four decades in the use of alternative donors for those without a matched sibling—including HLA-matched unrelated donors, HLA-haploidentical related donors and unrelated-donor umbilical cord blood—so that it is now possible to find a donor for almost every child requiring an allograft. Below, we summarize the relative merits and limitations of the different alternative donors for benign hematologic conditions, first generally, and then in relation to specific disorders, and suggest recommendations for selecting such an alternative donor.

Inherited bone marrow failure syndromes (IBMFS) and acquired aplastic anemia (AA) are life-threatening bone marrow failure (BMF) disorders. These entities can be difficult to distinguish because they both present with pancytopenia and a hypocellular marrow (1, 2). IBMFS can be caused by germline mutations that result in impaired DNA repair (e.g., Fanconi anemia), abnormally short telomeres (e.g., dyskeratosis congenita), ribosomopathies (e.g., Schwachman– Bodian–Diamond syndrome), abnormalities in the thrombopoietin gene (amegakaryocytic thrombocytopenia), or other causes (3). However, the majority of AA is acquired and mediated by immune destruction of hematopoietic progenitors. In contrast to IBMFS, acquired AA is often responsive to immunosuppressive therapy (4). It is therefore critical to distinguish between these two etiologies, both for diagnosis and appropriate treatment. Paroxysmal nocturnal hemoglobinuria (PNH) is an acquired bone marrow failure disorder characterized by clonal populations of hematopoietic progenitors with somatic mutations in PIGA. The PIGA gene product is necessary for glycosylphosphatidylinositol anchor protein (GPI-AP) biosynthesis. Up to 70% of patients with acquired AA have a small PNH clone at diagnosis (5, 6). The deficiency of GPI-AP on the surface of these mutated stem cells renders them relatively resistant to immune attack; thus, the cells do not have an inherent proliferative advantage but rather have a conditional advantage because they escape immune destruction (6). This is believed to be the basis for the high incidence of PNH clones in AA. Based on this finding, we hypothesized that the presence of a PNH clone could distinguish immune vs. genetic causes of AA.

s / Biol Blood Marrow Transplant 21 (2015) S322eS354 S335 Christen Mumaw, Shannon R. McCurdy , Heather J. Symons , Andrea M. Towlerton , Edus H. Warren , Paul V. O’Donnell , Sophie Paczesny , Leo Luznik . 1 Department of Oncology, The Johns Hopkins University School of Medicine, Baltimore, MD; 2 School of Medicine, Indiana University, Indianapolis, IN; 3 Clinical Research Division, Fred Hutchinson Cancer Research Center

Background: With PTCy as GVHD prophylaxis, outcomes of NMA haplo and matched BMT are similar, and relapse rather than toxicity is the leading cause of treatment failure. Early discontinuation of IS may augment a graft-versus-tumor effect and permit early implementation of strategies to reduce relapse, but may increase GVHD. We present a completed, prospective single-center trial of stopping tacrolimus (tacro) 3 or 4 months earlier than our Day (D) 180 standard after NMA haplo BMT (ClinicalTrials.gov: NCT01342289). Methods: From 8/2011-11/2015,105 evaluable patients (pts) with hematologic malignancies received NMA haplo BMT on this trial. The primary objective was to evaluate the feasibility and safety of reduced-duration tacro, stopping tacro without taper before D 180.Transplant criteria included age ≤ 75, ECOG PS ≤ 2, LVEF ≥ 35%, FEV1 and FVC ≥ 40% predicted, transaminases < 5 x ULN and no prior allogeneic BMT. All received Cy (14.5 mg/kg IV D -6 and -5), fludarabine (D -6 to -2), TBI (200 cGy D -1) and T-cell replete bone marrow. GVHD prophylaxis consisted of high-dose PTCy (50 mg/kg IV D 3 and 4), mycophenolate mofetil (D 5-35) and tacro from D 5. Pretransplantation, pts were assigned to stop tacro early if eligible, contingent on having ≥ 5% donor T cells at ~D 56 onward, no relapse and no grade 2-4 acute or significant chronic GVHD. Tacro was first planned through D 90 (n=47), then through D 60 (n=55). A D 120 cohort (n=3) enrolled while D 90 safety data were maturing. For pts ineligible for planned early tacro cessation, IS was individualized and continued to at least D 180. Monitoring rules declared reduced IS feasible if ≥ 33% of pts stopped tacro early as planned. Safety stopping rules for early tacro cessation were based on ≥ 65% probability of a ≥ 20% incidence of grade 3-4 acute plus severe chronic GVHD, ≥ 10% nonrelapse mortality (NRM) or ≥ 5% graft failure, measured from the tacro stop date to ~D 180. Historical data from 212 haplo transplants at our center using the same regimen but tacro until D 180 informed safety risk calculations. Results: Of the 105 pts (median age 61, range 13-74), the most common diagnoses were acute leukemia (50%), MDS (17%), NHL (16%) and HL (8%). By refined Disease Risk Index, 11% were low risk, 70% intermediate and 19% high. Shortened IS was feasible in 63 pts (60%) overall. Ineligibility for shortened IS was due most commonly to GVHD, followed by low donor chimerism or graft failure and early relapse. Of the 47 pts in the D 90 cohort (median follow-up 44 months), 23 (49%) stopped tacro early as planned. Safety stopping criteria were not met. Of these 23 pts, 16 (70%) had no safety events before D 180, 5 (22%) developed grade 2 acute GVHD (1 complicated by severe chronic GVHD) and 2 (9%) developed grade 3-4 acute GVHD. Of the 55 pts in the D 60 cohort (median follow-up 14 months), 38 (69%) stopped tacro early as planned, and safety stopping criteria were likewise not met. Of these 38 pts, 25 (66%) had no safety events before D 180, 1 developed graft failure, 9 (24%) developed grade 2 acute GVHD and 3 (8%) developed grade 3-4 acute GVHD. GVHD outcomes by cohort relative to historical outcomes are shown in Figures A and B. In both cohorts, the D 180 CuI of grade 2-4 acute GVHD was < 40% and was < 10% for grade 3-4 acute GVHD and NRM. The 1-year CuI of any chronic GVHD was 11% for the D 90 arm and 13% for the D 60 arm (12% historically). The 1-year probabilities of PFS, OS and GVHD-free relapse-free survival (GRFS, Figure C) were 40%, 59% and 27% respectively for the D 90 arm and 63%, 77% and 53% respectively for the D 60 arm. Conclusion: These data suggest that reduced-duration tacro is feasible and carries an acceptable safety profile in pts receiving NMA haplo BMT with PTCy. Risks of acute GVHD, chronic GVHD, graft failure and NRM appear similar to historical outcomes with tacro until D 180. A larger prospective study is needed to define the optimal duration of IS that balances GVHD risk and relapse risk. However, these data show that many pts (60% in this trial) can discontinue tacro without taper well before D 180. There is even a suggestion of improved PFS and GRFS in the D 60 arm compared to the D 90 arm, although the trial was not powered for these endpoints. The favorable toxicity profile of the PTCy platform, coupled with the feasibility and safety of early tacro cessation, provides an ideal setting to incorporate novel posttransplantation approaches for relapse reduction. Disclosures No relevant conflicts of interest to declare.

Introduction: With improvements in supportive care, including graft-versus-host disease (GVHD) control, relapse has become the major complication of allogeneic BMT. The use of haploidentical donors often allows the potential to select a different donor for the second transplant. We examined the outcomes of patients who underwent a second BMT for relapse at Johns Hopkins between 2005 and 2014. We hypothesized that if loss of heterozygosity is a mechanism for relapse after haploBMT, then utilizing a donor that recognizes the other shared haplotype could improve antitumor activity and thus overall survival (OS). In the case of a failed matched transplant, recognition of a non-shared haplotype on the tumor by the second allograft would be beneficial. Methods: We identified all patients who received a second transplant for relapse of a hematologic malignancy between 2005 and 2014. Acute GVHD (aGVHD) is reported as Glucksberg grades 1-4; chronic GVHD (cGVHD) is reported as limited or extensive. GVHD and non-relapse mortality (NRM) was calculated via cumulative incidence (CI) with competing risk analysis. OS was calculated as the time from second transplant to death or last known follow-up and estimated using the Kaplan Meier method. Differences in time to death between patient groups were estimated using Cox proportional hazards models. Non-relapse mortality was estimated using Fine and GrayÕs method. Results: Between 2005 and 2014, 40 patients received a second BMT for a relapsed hematologic malignancy (Table 1). The median age at second transplant was 43.9 (range 1 to 74); 14 (35%) were female. Fifteen (41%) received myeloablative conditioning for their first transplant, and 35 (92%) received high dose post-transplant cyclophosphamide (PTCy) as part of their GVHD prophylaxis. The incidence of grade 2-4 aGVHD after the first transplant was 13%, and limited cGVHD was 13%. There was no extensive cGVHD. Re-induction therapy and second allograft characteristics (Table 2): Refractory or progressive disease at the time of second transplant was present in 20% of patients, and 7/40 (18%) had their second transplant within 1 year. After a first failed HLA-matched transplant, 15 received a haploidentical graft, and 4 received a second HLA-matched graft. Among the 21 who relapsed after a haploBMT, 7 received an HLA-matched allograft, and 14 received a haploidentical graft; 6 out of those 14 used a donor that switched the shared haplotype. Most (81%) received PTCy with the second BMT. Median follow up is 598 days. Median OS in the cohort is 911 days (95% CI 602 Ð NR); 4 year OS is 38%. The CI of NRM by 2 years was 29% - 4 patients died of complications of GVHD (only one of whom received post-transplant Cyclophosphamide), 3 associated with prolonged cytopenias, 1 within a month of BMT, and 3 of pneumonia. Grade 2-4 aGVHD occurred in 11 pts (28%), grade 3-4 in 5 (13%). Extensive cGVHD was seen in 7 (18%). Median OS for the patients who relapsed after initial haploBMT who were then retransplanted with an HLA-haploidentical donor with a different shared haplotype has not been reached, versus a median OS of 501 [317-2950+] days among the 15 who did not receive a graft with an alternate shared haplotype (HR 0.21 [0.03, 1.61]; p=0.06). If the 19 patients who received a matched allograft at first transplant are included, median OS is 552 days in the group transplanted with a second graft with the same shared haplotype(s), and 1308 days (Figure 1) in the group whose allograft contained a different non-shared haplotype (HR 0.39 [0.16, 0.98]; p=0.04). OS was better when the second transplant was performed more than one year after the first (HR 0.30 [0.11, 0.80]; p=0.03). A trend towards improved OS was associated with disease status at time of second transplant (remission or chemo-responsiveness versus progressive or refractory disease; HR 0.41 [0.16, 1.03]; p=0.07). Source of graft (BM versus peripheral), conditioning intensity, era of transplant (2005-2010 versus 2010-2015), same versus different donor, and age group ( Conclusion: Second transplants are feasible in all age groups and provide a reasonable chance of long-term survival. GVHD rates using PTCy are low, and similar to that seen with first BMTs. An allograft sharing a different haplotype in the case of relapse after haploBMT, or a haploidentical donor after failure of a matched graft, may be beneficial. OS by Second Donor Haplotype OS by Second Donor Haplotype Figure 1 Figure 1. Disclosures Brodsky: Apellis Pharmaceuticals Inc: Membership on an entity's Board of Directors or advisory committees; Alexion Pharmaceuticals Inc: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Achillion Pharmaceuticals: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees.

Human Leukocyte Antigen-haploidentical stem cell transplantation, or HLA-haploidentical SCT, refers to the transplantation of blood or marrow from a donor into a recipient who is genotypically identical for one HLA haplotype and is variably mismatched for HLA alleles on the unshared haplotype. Typically, donors for HLA-haploidentical SCT are first-degree relatives, such as siblings, biological parents, or biological children, although aunts, uncles, cousins, and half-siblings may be HLA-haploidentical to the recipient. As compared to blood or marrow transplantation from an HLA-matched sibling, a distinguishing feature of HLA-haploidentical, or partially mismatched related donor (PMRD), SCT is intense, bidirectional T cell-mediated alloreactivity resulting in increased risks of graft failure, severe graft-versus-host disease, and nonrelapse mortality (NRM) [1-3]. The poor results of early trials of HLA-haploidentical SCT have motivated strategies to mitigate alloreactivity by depleting T cells from both the host and from the graft. While these strategies have reduced the risks of graft failure and GVHD, respectively, they have led to a compensatory increase in the risk of severe infectious complications from prolonged immunodeficiency, and possibly to an increased risk of disease relapse. More recent efforts have focused upon selective depletion of alloreactive T cells to enhance immune reconstitution without risking severe GVHD. Advances in the control of alloreactivity and in supportive care have combined to improve the results of HLA-haploidentical SCT.

We compared the outcome of nonmyeloablative allogeneic hematopoietic cell transplantation (HCT) for patients with relapsed or refractory Hodgkin lymphoma (HL) based on donor cell source. Ninety patients with HL were treated with nonmyeloablative conditioning followed by HCT from HLA-matched related, n = 38, unrelated, n = 24, or HLA-haploidentical related, n = 28 donors. Patients were heavily pretreated with a median of 5 regimens and most patients had failed autologous HCT (92%) and local radiation therapy (83%). With a median follow-up of 25 months, 2-year overall survivals, progression-free survivals (OS)/(PFS), and incidences of relapsed/progressive disease were 53%, 23%, and 56% (HLA-matched related), 58%, 29%, and 63% (unrelated), and 58%, 51%, and 40% (HLA-haploidentical related), respectively. Nonrelapse mortality (NRM) was significantly lower for HLA-haploidentical related (P = .02) recipients compared to HLA-matched related recipients. There were also significantly decreased risks of relapse for HLA-haploidentical related recipients compared to HLA-matched related (P = .01) and unrelated (P = .03) recipients. The incidences of acute grades III-IV and extensive chronic graft-versus-host disease (aGVHD, cGVHD) were 16%/50% (HLA-matched related), 8%/63% (unrelated), and 11%/35% (HLA-haploidentical related). These data suggested that salvage allogeneic HCT using nonmyeloablative conditioning provided antitumor activity in patients with advanced HL; however, disease relapse/progression continued to be major problems. Importantly, alternative donor stem cell sources are a viable option.

We evaluated the utility of allogeneic HCT using nonmyeloablative conditioning as a potential treatment option for 79 patients (pts) with HL who were ineligible for high dose conventional allogeneic HCT and compared outcome based on donor cell source (MRD n=34, URD n=24, Haplo n=21). Conditioning consisted of 2 Gy TBI alone (n=15, MRD) or combined with either 90 mg/m2 (MRD n=19, all URD) or 150 mg/m2 (all Haplo) fludarabine. All haplo recipients also received cyclophosphamide pre (29 mg/kg) and post (50 or 100 mg/kg) HCT. GVHD prophylaxis consisted of mycophenolate mofetil and a calcineurin inhibitor. Pts were heavily pretreated with a median number of 5 (range, 2–10) prior regimens. Most pts failed prior autologous HCT (91%) and radiation therapy (86%). There were no graft rejections. The incidences of acute grades III–IV and extensive chronic GVHD were 15%/47% (MRD), 8%/60% (URD), and 10%/31% (Haplo). There were no statistically significant differences in the ability to discontinue immunosuppression based on donor cell sources. With a median follow up of 20 (range, 4–91) months for living pts; the 18 month overall survivals (OS), progression free survivals (PFS), and incidences of relapse/progressive disease (PD) were 47%, 20%, and 55% (MRD), 74%, 27%, and 65% (URD), and 71%, 60%, and 35% (Haplo), respectively. Nonrelapse mortalities (NRM) were significantly lower for URD (HR 0.16; p=0.03) and Haplo (HR 0.13; p=0.02) recipients compared to MRD recipients. There were also significantly decreased risks of relapse for Haplo recipients compared to MRD (HR 0.35; p=0.03) and URD (HR 0.36; p=0.02) recipients. These data suggest that salvage allogeneic HCT using nonmyeloablative conditioning provides anti-tumor activity in pts with very advanced disease; however, Rel/PD continue to be major problems regardless of stem cell source. Importantly, alternative donor sources are a viable option particularly for those pts who may have a limited window of opportunity to proceed to HCT. Pt Characteristics MRD (n=34) URD (n=24) Haplo (n=21) Median Age (yrs) 33 28 31 Disease Status at HCT CR 11 (32%) 5 (21%) 5 (24%) PR 16 (47%) 8 (33%) 5 (24%) Rel/Ref 7 (21%) 11 (46%) 11 (52%) Disease Bulk > 5cm 3 (9%) 3 (13%) 3 (14%) HCT-Comorbidity Index >2 21 (62%) 17 (74%) 13 (62%) Regimens > 5 16 (47%) 16 (67%) 12 (57%) Failed Prior Auto HCT 30 (88%) 23 (96%) 19 (90%) Median Time Diagnosis - Allo HCT (months) 33 31 32 Median Time Auto-Allo HCT (months) 16 19 17 Outcomes by donor type MRD (n=34) URD (n=24) Haplo (n=21) Median f/u living pts months (range) 15 (4–91) 26 (8–58) 15 (4–49) Acute GVHD grade II–IV, III–IV 53%, 15% 50%, 8% 43%, 10% 18 month extensive chronic GVHD 47% 60% 31% Day 200 NRM 18% 0% 0% 18 month OS 47% 74% 71% NRM 25% 8% 5% Rel/PD 55% 65% 35% PFS 20% 27% 60%

SAA is a life-threatening hematopoietic stem cell disorder that is treated with either bone marrow transplantation (BMT) or immunosuppressive therapy (IST). BMT is the treatment of choice for young patients with a matched sibling donor. Those with acquired SAA lacking a matched sibling donor and older adults (> 40yrs) generally receive IST because of the high toxicity of BMT in these patients. Roughly two-thirds will respond but up to 40% of patients eventually relapse or acquire a secondary clonal disease such as paroxysmal nocturnal hemoglobinuria (PNH) or myelodysplastic syndromes. Thus, a major challenge in treating SAA is the management of patients who are have relapsed after or are refractory to IST or who have acquired a secondary clonal disorder. BMT is the only curative therapy for patients with inherited SAA but problems with graft failure, graft-versus-host disease (GVHD), and lack of matched unrelated donors especially for ethnic minorities have limited the success of alternative donor BMT, especially for adult patients. PTCy after HLA-haplo BMT has been shown to facilitate engraftment and yield rates of GVHD comparable to those of matched sibling BMT in hematologic malignancies. We hypothesized that PTCy would improve outcomes for patients with refractory SAA by reducing the risk of GVHD and expanding the donor pool to include haplo related and mismatched unrelated donors (URD). Patients received an allogeneic BMT (alloBMT) for relapsed or refractory SAA. Relapsed/ refractory disease was defined as having failed previous course of IST. In the majority of the cohort, haplo donors were used due to lack of availability of suitable and timely URDs. Nine patients underwent haploBMT. A mismatched URD was used in one case of no suitably available haplo donor given the inherited nature of her disorder. The nonmyeloablative preparatory regimen included antithymocyte globulin (0.5 mg/kg day-9 and 2 mg/kg on days -8,-7), fludarabine (30mg/m2/day on days - to -2), low dose CY (14.5 mg/kg on days -4, -5 ) and total body irradiation (200cGy day -1). After the unmanipulated marrow graft was infused on day 0, cyclophosphamide 50mg/kg/day IV on days +3,+4 post-transplant was administered. All patients received tacrolimus and mycophenolate mofetil (MMF) beginning day+5 through 35 (MMF) and day +5 through 1 year (tacrolimus) for additional graft versus host disease (GVHD) prophylaxis. Neutrophil engraftment was defined as an ANC >1.0 x109/L measured for three consecutive measurements on different days. Red cell engraftment was defined as days from last red blood cell transfusion. Platelet engraftment was defined as a platelet count >50,000 for 7 days without transfusion. Ten patients were treated according to this IRB approved protocol. Eight patients had documented acquired disease and received grafts from 5/10 related donors. Six had failed ATG-containing regimens previously and two relapsed after high dose cyclophosphamide as IST. Two patients had inherited syndromes: one with Diamond Blackfan who received a graft from 5/10 related donor and one with telomeres less than the first percentile in length with a presumed familial syndrome who received a 9/10 unrelated graft. Median follow-up time is 17.5 months (range 5-49). The median age was 34 years (range 17-54) with 6 patients greater than age 30 years, and 50% were males. Median time to neutrophil engraftment was 18 days (range 16-24). Median time to red cell engraftment was 25 days (range 16-48). Median time to platelet engraftment was 27.5 days (range 22-108). At the time of BMT, six patients had PNH clones, and all were eliminated. All patients are alive and well, fully engrafted with 100% donor chimerism in blood and marrow. Two patients had grade 1-2 skin acute GVHD. These two also had mild chronic GVHD of the skin/mouth requiring systemic steroids. One patient was able to come off all immunosuppression by 15 months and the other by 17 months. Nonmyeloablative alloBMT with PTCy appears promising in patients with refractory acquired and inherited SAA with acceptable rates of engraftment, eradication of pre-existing clonal diseases, and expansion of the donor pool. Given these promising results, this approach is being utilized in the next BMTCTN national trial for acquired refractory SAA. We believe this approach could also be considered as upfront therapy for SAA, and such a trial is in development at Hopkins Disclosures Brodsky: Alexion Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.

Background: High-dose PTCy has been shown to effectively reduce acute (a) and chronic (c) GVHD after MRD and MUD BMT and to enable the safe implementation of haplo BMT. However, GVHD-related risk factors and survival outcomes, the impact of donor type, and novel composite endpoints, such as cGVHD-free relapse-free survival (cGFRFS), have not been fully characterized in BMT with PTCy. Methods: We retrospectively analyzed 582 consecutive adult pts with advanced or poor-risk hematologic malignancies who received allogeneic BMT with PTCy at Johns Hopkins from 2002-2012. All pts received a T-cell replete bone marrow graft and PTCy (50 mg/kg/d IV) on D+3 and +4. Platforms consisted of: 1) myeloablative conditioning (MAC) with Busulfan/Fludarabine or Busulfan/Cy followed by MRD (n = 193; 33%) or MUD (n=119; 20%) allografting and PTCy as sole GVHD prophylaxis; or 2) nonmyeloablative (NMA) conditioning with Fludarabine/Cy/TBI followed by related haplo (n = 270; 46%) allografting and PTCy, mycophenolate mofetil D5-35, and tacrolimus D5-180. Results: The median follow up was 4 (range 0.3-11.4) years (y). The median pt age at time of BMT was 49 (range 18-66) in the MAC cohort and 54 (range 18-73) in the NMA cohort. On competing risk analysis, the D200 probability of grade II-IV aGVHD was 27% after haplo, 37% after MRD, and 50% after MUD BMT, and grade III-IV aGVHD was 4%, 11%, and 14% respectively. The 3-y probability of cGVHD was 12% after haplo, 8% after MRD, and 19% after MUD BMT. On multivariate analysis adjusted for age and original disease risk index (DRI), MRD and MUD BMT were associated with a statistically significantly higher risk of grade II-IV and grade III-IV aGVHD compared to haplo BMT (each p ≤ 0.001). Compared to haplo BMT, MUD, but not MRD, BMT was also associated with a significantly higher risk of cGVHD (p = 0.009). On multivariate analysis, female into male allografting was independently associated with a higher risk of cGVHD (p = 0.007), whereas age and CMV serostatus of the pt and donor were not statistically significantly associated with risks of acute or chronic GVHD. Pts with grade II-IV aGVHD or cGVHD had a lower probability of relapse, but a higher probability of nonrelapse mortality (NRM) at 3 y compared to pts without GVHD. Among pts with GVHD, an HCT-CI score ≥ 5 was associated with significantly inferior overall survival compared to lower HCT-CI scores. Although MAC matched BMT was associated with less relapse risk than NMA haplo BMT, the 3-y probabilities of NRM, DFS, and overall survival were comparable between the transplant platforms (Table). On multivariate analysis adjusted for pt age and DRI, there were no statistically significant differences in DFS, OS, or cGFRFS in MAC related or unrelated BMT compared to NMA haplo BMT. Conclusions: High-dose PTCy safely modulates acute and chronic GVHD across multiple BMT platforms. The apparently lower risk of aGVHD after NMA haplo BMT with PTCy likely reflects the reduced conditioning intensity and longer duration of immunosuppression utilized. Despite the limitations inherent to a retrospective analysis, these data suggest that key outcomes after BMT with PTCy are comparable across donor types and conditioning intensities. Notably, in all these settings, 3-y estimates of cGFRFS appeared to approach those of DFS. Thus, PTCy may minimize the late morbidity and mortality of cGVHD and allow earlier implementation of novel posttransplantation strategies for relapse prevention. TableOutcomes of BMT with PTCy3-y probabilitiesNMA HaploMAC MRD*MAC MUD*NRM141717DFS384144OS495856cGFRFS313533 * p = NS for all comparisons to haplo BMT on multivariate analysis Disclosures Off Label Use: High-dose posttransplantation cyclophosphamide.

Background HaploBMT with PTCy has now been shown by multiple groups to produce results similar to that seen with matched sibling allogeneic BMT. The use of donor lymphocyte infusion (haploDLI) to treat relapse after haploBMT is not well documented due to theoretical concerns of inducing severe graft-versus-host-disease (GvHD). Here we report our experience of using DLI for relapse after haploBMT. Methods We identified all haploBMT recipients at Johns Hopkins University who received haploDLI for relapse between 6/2003 and 10/2012. Responses were evaluated by standard disease-specific criteria. GvHD was scored using the Keystone and Seattle criteria for acute (a) and chronic (c) GvHD. Duration of survival and response were calculated from date of first haploDLI administration. Patients (pts) were followed till 12/2012. Results Forty pts received 52 haploDLI doses for relapse after haploBMT (16 acute myeloid leukemia [AML], 3 acute lymphoid leukemia [ALL], 5 Hodgkin lymphoma [HL], 7 non-Hodgkin Lymphoma [NHL], 1 myelodysplastic syndrome [MDS], and 9 others). All pts received PTCy, tacrolimus and mycophenolate mofetil for GvHD prophylaxis after haploBMT. Median time from haploBMT to relapse was 183 days; 7 relapses diagnosed within 3 months (M) of haploBMT, 25 relapses within 1 year (yr). Chemotherapy was preadministered to 33 haploDLI doses in 28 pts. 35 pts received haploDLI for florid hematologic relapse, 4 for new minimal residual disease (MRD) relapse detected by cytogenetics or flow cytometry, and 1 for progressive loss of donor chimerism concerning for relapse. Median age at haploDLI administration was 48 yrs (3-70 yrs). No GvHD prophylaxis was administered after haploDLI. The first haploDLI dose received was 1X10^5 CD3+/kg of recipient ideal weight (IBW) in 5 pts, 1-5x10^6 in 31, and 1x10^7 in 4. Thirty-two pts received 1 dose of haploDLI while 8 pts received escalating doses of DLI. Median follow-up was 7M (0.5–98M). Ten pts (25%) developed aGvHD (6 pts had grade 3-4 and 8 pts had grade 2-4). Of 8 pts with grade 2-4 aGvHD, 5 pts received a dose of 1x10^6 and 2 pts received 1x10^7 CD3+/kg dose. 2 pts with GvHD (both grade 3) died within 6 weeks of DLI from sepsis. 3 pts developed cGvHD (2 extensive, 1 limited). Twelve pts responded to haploDLI (30%), all with complete responses (CR), with a median response of 13 M (1.5-96 M). 5 pts had AML, 1 ALL, 1 MDS, 1 HL, 3 NHL, and 1 prolymphocytic leukemia. 10 responders received haploDLI within 4 M of relapse. 10 responders received only 1 haploDLI. 9 pts responded to haploDLI dose of 1-5x10^6 CD3+/kg. 11 had pre-haploDLI chemotherapy or radiation. 3 of the 4 pts treated for MRD relapse responded (75%), while 9 of the 35 treated for florid hematologic relapse responded (26%). Eight CR occurred following the 30 haploDLI doses of 1x10^6 CD3+/kg (26.7%). The CR rate following haploDLI among AML pts was 31.3%. At end of follow-up, 8 of the 12 responders were still alive and in CR, including 5 pts beyond 1-yr post DLI. Five responders developed clinically significant GvHD post haploDLI (3 grade 2-4 aGvHD, 2 extensive cGvHD). Conclusions In our experience in a relatively small number of pts, haploDLI administration for relapse following T-cell replete haploBMT with PTCy is feasible, not associated with unacceptable toxicities, and can result in lasting favorable outcomes. HaploDLI dose of 1x10^6 CD3+/kg of recipient’s IBW appears to be a reasonable starting dose in this setting. Similar to DLI use in the HLA-matched setting, use of haploDLI for MRD relapse seems to result in better outcomes. The CR rate after haploDLI in AML was 31% (5 of 16 pts), similar to what is reported for DLI use in the HLA-matched setting. Small numbers preclude conclusions about disease-specific efficacy of haploDLI in other diseases. Prospective evaluation is required to confirm these findings. Disclosures: Off Label Use: entinostat for MDS. Luznik:Otsuka Pharmaceutical: Research Funding.

Background Given the life-threatening nature of severe VOD (sVOD) and associated multi-organ failure (MOF), the absence of other approved therapies for this complication in the USA, and the promising results to date with DF in this setting, DF has been made available since 2007 through a prospective T-IND. The aim of the T-IND is to gather additional data on safety and response to DF in a broader patient (pt) population, including those with sVOD/MOF and those with non-severe VOD. Thus far, this is the largest prospective evaluation of DF for the treatment of sVOD/MOF in pts undergoing HSCT, and in pts who developed VOD following chemotherapy (non-HSCT pts). Here we provide an update on the efficacy and safety of DF in HSCT and non-HSCT pts, together with analysis of other clinical features of interest, including GvHD. Methods The original T-IND protocol required pts to have a diagnosis of VOD by Baltimore criteria (total bilirubin ≥2.0 mg/dL with ≥2 of the following: hepatomegaly, ascites or 5% weight gain) with MOF (either renal and/or pulmonary failure) following HSCT, and was amended to allow inclusion of pts with non-severe VOD (defined as no MOF) occurring either post-HSCT and post-chemotherapy. Key exclusion criteria included clinically significant bleeding or the need for >1 vasopressor. Complete response (CR) was defined as bilirubin Results The current interim analysis is based on 470 VOD pts enrolled between December 2007 and December 2012 at 75 centers, 45 of which developed VOD following chemotherapy and 425 had undergone HSCT (376 allogeneic HSCT, 89%). In HSCT patients, median age was 15 yrs (range 0.1–70), 55% were male and 18% had undergone multiple HSCTs (>1 HSCT); 284 pts had severe disease at study entry. The most common diagnosis was leukemia (29% AML; 22% ALL; 6% other). Conditioning regimen included CY (66%), BU (50%) and TBI (36%). Median onset of VOD was 15 d post-HSCT. Of HSCT pts, 35% (147/425) achieved CR and 55% (Kaplan-Meier estimate) survived to D+100. In pts with sVOD, CR was 29% and D+100 survival was 48%. For pts with non-severe VOD, CR and D+100 survival was 47% and 69%, respectively. In all HSCT pts, delay of >2 d (vs ≤2 d) in the start of DF after VOD diagnosis resulted in reduced CR (25% vs 39%, p=0.0052) and survival (Kaplan-Meier estimate) (38% vs 61%, p Conclusions DF therapy in sVOD/MOF pts achieved significantly improved outcome compared to that expected based upon historical data. Given the results of this large cohort of pts, early treatment with DF (i.e. within 2 d of VOD diagnosis) is recommended, and consistently improved outcome was seen in pts who have not yet progressed to sVOD. Generally, DF was well-tolerated and as with prior studies, there was a low incidence of DF-associated toxicities. These results confirm the findings of previous trials and strongly support early intervention with DF once the diagnosis of VOD is made after HSCT, which also reinforces the validity of positive studies in prophylaxis to date and suggests further trials in prevention are warranted. Outcomes in children appear to be better than in adults, supporting the importance of this agent in the pediatric population in particular. In addition, low rates of GvHD were seen and this is an important area in which further studies are planned. Enrollment to the T-IND study continues. Disclosures: Richardson: Gentium : Membership on an entity’s Board of Directors or advisory committees. Hannah:Gentium: Consultancy. Hume:Gentium: Employment. Bandiera:Gentium: Employment. Heringa:Gentium: Employment. Study Group:Gentium: Employment.

Abstract 4206 Background: Relapse is the primary cause of treatment failure post alloHCT. We sought to identify risk factors that predict relapse of hematologic malignancies post-alloHCT to identify those at highest risk of relapse and help guide who may benefit from novel approaches to reduce relapse risk. Design: This was a single institution, retrospective cohort study of children with acute lymphoblastic leukemia, acute myelogenous leukemia, mixed phenotypic acute leukemia or myelodysplastic syndrome who had undergone alloHCT between 1/1/2003 and 12/31/2010. Relapse was defined as any evidence of increasing disease post-alloHCT, including minimal residual disease (MRD). Relapse-free survival (RFS) and associations with various characteristics were determined by the Kaplan-Meier method and log-rank test. The Cox proportional hazards model was used to identify factors that may jointly associate with RFS. Results: 40 of 93 patients (43%) undergoing alloHCT experienced relapse at a median of 144 days (range 1 day – 58 months) post-HCT. Univariate actuarial analysis demonstrated that Black race (Figure 1), high white blood cell (WBC) count at diagnosis, central nervous system (CNS) disease at diagnosis (Figure 2), short first complete remission (CR1) duration (limited to those attaining CR1 with subsequent pre-HCT relapse), greater number of regimens given to induce remission, poorer performance status, non-myeloablative (NMA) preparative regimen, lack of remission and MRD pre-HCT were associated with higher relapse probability (Table 1). These factors were considered for inclusion in a Cox model; CNS disease at diagnosis (p=0.009), lack of morphologic remission (p=0.01) and a NMA HCT (p=0.04) were factors independently associated with worse RFS. 84 patients surviving to day 45 without relapse were included in a subset analysis to allow for incorporation of post-alloHCT factors that may modify relapse risk (i.e., graft-versus-host disease). When factors were considered jointly, only WBC at diagnosis was significantly associated with RFS. For patients in a morphologic remission who underwent a myeloablative HCT, MRD detection approximately doubled relapse risk. (Table 2). Conclusion: The association of CNS disease at diagnosis with higher post-HCT relapse risk has not previously been reported. As a sanctuary site, the CNS may be less amenable to an allogeneic effect and CNS involvement might reflect higher-risk disease. Those with CNS disease may benefit from earlier or more intensive CNS-directed therapy to reduce relapse risk. The association of black race with increased relapse risk in the univariate analysis is also notable. For these patients, increased utilization of alternative donor sources or distinct biological differences may lead to higher relapse risk. Validation of these risk factors in a larger population and development of a prognostic score to identify those at highest risk of relapse is planned with possible prospective use of this prognostic tool in the development of relapse prevention trials. Disclosures: No relevant conflicts of interest to declare.

Abstract 4205 Background: Relapse is the primary cause of treatment failure in patients who undergo alloHCT for hematologic malignancies. We describe the presentation, management, and outcomes of children who relapse post-HCT to improve monitoring and intervention strategies. Design: This was a single institution, retrospective cohort study of children with relapse or progression of acute lymphoblastic leukemia (ALL), acute myelogenous leukemia (AML), mixed phenotypic acute leukemia (MPAL) or myelodysplastic syndrome (MDS) post-alloHCT between January 1, 2003 and December 31, 2010. Relapse was defined as any evidence for disease detected by any modality (including morphologic, cytogenetic, flow cytometric, molecular and radiographic methods) after previously negative results. Progressive disease was defined as an increase in any measure from baseline results. Minimal residual disease (MRD) was defined as disease detectable by immunophenotypic, cytogenetic or molecular methods that did not meet classic morphologic criteria for relapse (defined as > 5% disease). Results: 40 patients with AML (n=18), ALL (n=16), MPAL (n=4) and MDS (n=2) were identified who relapsed from among 93 patients who underwent a first alloHCT. The median time from alloHCT to relapse was 144 days (range 1 day– 58 months). Presentation of relapse included: detection of asymptomatic disease found on scheduled restaging (n=17); presentation with peripheral blasts or overt extramedullary disease (n=17); and presentation with symptoms prompting evaluation for recurrent disease, including hematologic abnormalities and pain (n=5). Two patients had isolated, asymptomatic central nervous system relapse. Salvage therapies included: supportive care, including hospice, palliative or alternative care only, (n=6); withdrawal of immunosuppression (WIS) (n=3); cytoreductive and/or radiation therapy with or without additional therapy (n=24), donor lymphocyte infusion (DLI) with or without prior chemotherapy (n=13) and second HCT (n=12). Nine patients with only MRD at the time of relapse presented at a median time of 35 days post-HCT (range 28–182 days). Although these patients were immediately considered candidates for intervention and despite 4 patients being treated with WIS, most had very rapid progression of disease with evidence for frank relapse at a median time of 23 days post-initial MRD detection. 2 patients died from transplant-related-mortality shortly after detection of MRD. Median survival after relapse was 123 days (range 4 days–5 years). Estimated 6-month and 1-year post-relapse survival were 30% and 17.5%, respectively. Five of 40 (12.5%) patients are currently alive with a median follow-up of 39 months (range 1 day–58 months), including 1 patient with active disease. 1 survivor had MDS and presented with MRD alone. The remaining 4 presented with overt disease and included 3 patients with ALL and 1 with MPAL. 3 of 5 survivors underwent a second HCT. Overall survival (OS) was superior in patients who underwent a second HCT (p=0.0001), with a 25% OS after second HCT (Figure 1). No patients with AML survived post-relapse even though 6/18 underwent second HCT. There was no survival advantage for those whose relapse was detected as MRD as compared to overt disease. Long-term survivors include 3 patients with overt relapse between 146 and 411 days post-HCT. Conclusion: 40 of 93 (43%) patients with hematologic malignancies relapsed after alloHCT. After relapse, 1-year OS was 17.5% with a median survival of 124 days post-relapse. Although pre-emptive treatment of relapse in the setting of MRD is felt to be ideal, it may not be feasible. In our study, patients presenting with MRD only presented very early post-HCT (median time 35 days) at a time when post-HCT complications can be high and limit therapeutic options. Additionally, once MRD was detected, disease progression to frank relapse was rapid (median time 23 days) limiting the chance to respond to many frontline immunotherapeutic options such as WIS or DLI. Given the poor outcomes of relapse post-HCT and limited ability to treat relapse even when detected at the stage of MRD, efforts should focus on developing more effective therapies and preventing relapse by identifying those at highest risk of relapse as candidates for novel methods to enhance efficacy of alloHCT. Disclosures: No relevant conflicts of interest to declare.

Abstract 4151 Historically, myeloablative BMT without T cell depletion from ≥ 2 HLA-antigen-mismatched related (HLA-haploidentical) donors has been associated with excessive rates of severe graft-versus-host disease (GVHD) and non-relapse mortality (NRM). Using the same GVHD prophylaxis regimen that has produced outcomes in non-myeloablative haploidentical BMT similar to those seen in non-myeloablative matched related BMT, we have completed a single-institution phase II clinical trial utilizing high-dose, post-transplantation Cy after myeloablative conditioning and T-cell-replete, HLA-haploidentical BMT. The trial initially enrolled subjects with refractory hematologic malignancies only, with the later addition of high risk leukemias in remission and chemosensitive lymphomas. The median age of the 30 patients enrolled on trial was 43 years (range, 2–64); 13 were male and 19 Caucasian. Diagnoses included acute myeloid (10 refractory, 3 CR1, and 3 CR2), acute lymphocytic (2 CR1), and 1 refractory biphenotypic leukemia; 2 chronic myelogenous leukemia in blast crisis; and non-Hodgkin lymphoma (6 refractory and 2 chemosensitive); and 1 chemosensitive gray zone lymphoma. The majority (67%) of patients were not in remission at the time of transplant. Conditioning consisted of IV Busulfan (pharmacokinetically adjusted) on days −6 to −3, Cy (50 mg/kg/day) on days −2 and −1 in twenty-seven patients or Cy (50 mg/kg/day) on days −5 and −4 and total body irradiation (300 cGy /day) on days −3 to 0 in three patients, followed by T-cell-replete bone marrow in all patients. Postgrafting immunosuppression in all patients consisted of Cy (50 mg/kg/day) on days 3 and 4, followed by mycophenolate mofetil for 30 days, and tacrolimus for 6 months. Donor engraftment at Day 60 occurred in all but one evaluable patient (96%, 24/25). The median times to neutrophil (>500/μL) and platelet recovery (>20,000/μL) were 25 and 32 days, respectively. The cumulative incidences of grades II-IV and grades III-IV acute GVHD at day 100 were 14% (95% CI: 4.3%, 29.4%) and 7.3% (95% CI: 1.2, 21.2%), respectively (Figure a). The cumulative incidence of chronic GVHD at one year was 13% (95% CI: 1.5, 37.6%). The cumulative incidence of NRM at 100 days was 12% (95% CI: 3%, 28%) (Figure b). Non-relapse causes of death were multi-organ system failure in two patients with bulky mediastinal lymphoma and veno-occlusive disease in one patient who had received gemtuzumab ozogamicin 3 weeks before BMT. There were no deaths from infection. The cumulative incidence of relapse at 1 year was 66% (95% CI: 38%, 83.5%), in this poor-risk cohort. The cumulative incidence of relapse among patients in CR prior to transplant was 13% (95% CI: 1.5%, 37.6%) at 1 year. With a median follow-up of surviving patients of 168 days (range 60–729), actuarial overall survival (OS) was 40% at one year. With a median follow-up of event-free patients of 144 days (range 60–729), actuarial event-free survival (EFS) was 23.5% at one year. Myeloablative HLA-haploidentical BMT with T cell replete bone marrow and post-transplantation Cy is associated with promising rates of engraftment, GVHD, and NRM, similar to those seen with non-myeloablative haploidentical BMT with post-transplantation Cy and with conventional myeloablative matched sibling BMT. Based on these data, this approach to myeloablative haploidentical BMT is now being performed in better risk patients. Disease progression remains a problem in patients with refractory leukemia; low rates of GVHD and NRM allow for combining other novel immunotherapies with myeloablative haploidentical BMT in the future to improve disease free survival. Disclosures: Symons: Otsuka Pharmaceuticals: Research Funding. Chen:Genentech: Membership on an entity's Board of Directors or advisory committees. Luznik:Otsuka Pharmaceuticals: Research Funding.

Abstract 487FN2 Background: Given the life-threatening nature of severe VOD (sVOD) and multi-organ failure (MOF), DF was made available in the US through a prospective treatment IND protocol (T-IND) to gather additional safety and response data from SCT patients (pts) with sVOD who were not eligible for a pivotal Phase 3 trial. Following completion of this Phase 3 trial, the T-IND protocol has continued and been amended to include pts who would have met eligibility criteria for the completed phase 3 trial, as well as pts with non-severe VOD and non-SCT pts who develop VOD after chemotherapy (chemo). This is the largest prospective evaluation of DF for the treatment of sVOD/MOF post SCT and non-SCT pts to date. Methods: Pts were initially required to have a diagnosis of VOD by Baltimore criteria (total bilirubin > 2.0 mg/dL with > 2 of the following: hepatomegaly, ascites or 5% weight gain) with MOF (either renal and/or pulmonary failure) that followed SCT. Following treatment of 104 pts, an amendment expanded eligibility criteria to include pts with VOD after chemo and pts with non-severe VOD (defined as no MOF). Key exclusion criteria included clinically significant bleeding or >1 pressor to maintain BP. CR was defined as bilirubin < 2 mg/dL + resolution of MOF (if applicable). Mortality was assessed at Day +100 (D+100) in all pts. DF was given at 6.25 mg/kg IV q6h (25 mg/kg/d) with treatment duration recommended for at least 21d. Results: This interim analysis is based on 269 pts enrolled between December 2007 and March 2011 at 67 centers. Nearly all pts (n=251) had undergone SCT (with allogeneic SCT in 225, 90%); 18 developed VOD after chemo alone. Of the 269 cases of VOD, 200 were severe at study entry, with 25% (66/269) of all pts dialysis dependent and 31% (83/269) ventilator dependent. Median age was 16 years (range 0.1 – 70); 55% were male. In the SCT pts, the most common diagnosis was leukemia (29% AML; 22% ALL; CML 3%; 4% other), with conditioning of CY (71%), BU (46%) and TBI (38%) respectively; 18% had undergone multiple SCTs (>1 SCT). Median onset of VOD was 15 d post-SCT. When presenting after chemo alone (n=18), median onset of VOD was 16 d after the first dose, most frequently after CY (61%), cytarabine (44%) and vincristine (39%), and for treatment of leukemia (AML 33%, ALL 33%, other 6%). Overall mean number of days of DF administration in all pts was 22 (range 1–88).Of 269 pts, 32% (85/269) achieved a CR and 50% [ by Kaplan-Meier estimate] survived to D+100. In the subgroup of SCT pts, 31% (78/251) achieved CR and 50% survived to D+100; 134 pts met entry criteria for the original Phase 3 trial and comparison to the Phase 3 historical control showed a statistically improved outcome in CR (30% vs 9%, p=0.0006) and D+100 survival (46% vs 25%; p=0.006). Of 200 pts with sVOD, CR was 28% and D+100 survival was 44%. In the 18 chemo only pts, CR was 39% and D+100 survival was 50%. CR rate and D+100 survival for the 69 pts with non-severe VOD were 42% and 62%, respectively. Delay of >2 d (vs < 2 d) in the start of DF after VOD diagnosis resulted in reduced CR (23% vs 35%, p=0.0339) and survival (37% vs 56%, p=0.01). Children as compared to adults had higher rates in CR (35% vs 29%) and survival (55% vs 43%). Toxicity proved generally manageable: 22% of pts experienced a total of 81 related AEs, primarily consisting of hemorrhage (19%) and hypotension (4%). Hemorrhage included pulmonary bleeding (6%), GI hemorrhage (3%), epistaxis (3%) and hematuria (3%). Similar to the observation of decreased GvHD in other studies, the incidence of all grade GvHD in the allogeneic SCT pts was 8%. Conclusions: In 269 pts with mainly sVOD/MOF, DF therapy achieved significantly improved outcome compared to an untreated historical control. Importantly, CR and survival were improved in pts who were treated within 2d of VOD diagnosis (vs. later) and in pts who had not yet progressed to sVOD. As with prior studies, there was a low incidence of DF-associated toxicities. Interestingly, the incidence of GvHD in allo-SCT pts was low, consistent with the reduction of GvHD seen in the large randomized EBMT pediatric prevention study. These results confirm the findings of previous trials and strongly support early intervention with DF once the diagnosis of VOD is made after SCT, as well as its use in pts who have not progressed to advanced MOF. The use of DF for VOD following intensive chemotherapy without SCT also appears promising, but more research in this patient population is needed. Enrollment to the T-IND study continues. Disclosures: Richardson: Gentium: Membership on an entity's Board of Directors or advisory committees. Arai:Gentium: Research Funding. Symons:Otsuka Pharmaceuticals: Research Funding. Martin:Gentium: Research Funding. Massaro:Gentium: Consultancy. D'Agostino:Gentium: Membership on an entity's Board of Directors or advisory committees. Hannah:Gentium: Consultancy. Tudone:Gentium: Employment. Hume:Gentium: Employment. Iacobelli:Gentium SpA: Employment. Soiffer:Gentium: Honoraria.

Abstract 840 Significance of Missing Inhibitory KIR Ligands in Nonmyeloablative, HLA-Haploidentical (Haplo) BMT with Posttransplantation High-Dose Cyclophosphamide (PT/Cy). Yvette L. Kasamon 1, Leo Luznik1, Mary S. Leffell1, Hua-Ling Tsai1, Heather J. Symons1, Javier Bolaños-Meade1, Gary Rosner1, Lawrence E. Morris2, Pamela A. Crilley3, Richard J. Jones1 and Ephraim J. Fuchs1, (1)Johns Hopkins University, Baltimore, MD, (2)Northside Blood and Marrow Transplant Program, Atlanta, GA, (3)Hahnemann University Hospital, Philadelphia, PA Introduction: NK cells can influence haplo BMT outcomes including risks of relapse and GVHD. Our group previously reported that, in allogeneic BMT for hematologic malignancies that incorporates PT/Cy, donor-recipient iKIR (inhibitory killer-cell immunoglobulin-like receptor) gene mismatches and having a KIR haplotype B donor were associated with improved outcomes (BBMT 2010;16:533). Because the reported impact of NK cell alloreactivity models in haplo BMT has been variable and some models are relevant to donor selection, we expanded our analysis of iKIR ligand status in this transplantation platform. Patients and methods: Outcomes of 212 uniformly treated patients (pts) enrolled on two similar clinical trials of related-donor, haplo BMT were retrospectively analyzed. Planned treatment consisted of fludarabine (30 mg/m2 IV on days −6 to −2), Cy (14.5 mg/kg IV on days −6 and −5), total body irradiation (200 cGy on day −1), and non-T-cell depleted bone marrow infusion. GVHD prophylaxis consisted of high-dose Cy (50 mg/kg IV on days 3 and 4), mycophenolate mofetil on days 5–35, and tacrolimus on days 5–180 without taper, with filgrastim begun on day 5. All pts (median age 51, range 1–73) had poor-risk hematologic malignancies; 60 (28%) had prior BMT. Diagnoses were Hodgkin lymphoma (31 pts), NHL (69), CLL (21), multiple myeloma (6), acute leukemia or lymphoblastic lymphoma (62), MDS (9), CML (9), CMML (4), PV (1). Missing ligands (ML; defined as absence in the recipient of one or more HLA ligands for iKIRs) and donor/recipient iKIR ligand incompatibility (LI; defined as presence of an iKIR ligand in the donor that is absent in the recipient) were determined using high-resolution HLA typing of class I alleles. For study purposes, HLA-A*2301, A*2402, and A*3201 were included in the Bw4 serologic group and effects attributable to HLA-A3 and A11 ligand groups were excluded. Results: With a median 2.9 year follow-up (range, 0.3–7 years) in pts without events, the actuarial 2-year progression-free survival (PFS) was 34%. On competing-risk analysis, cumulative incidences of grade II–IV acute GVHD and chronic GVHD were 28% and 14%, respectively; 1-year cumulative incidences of relapse and nonrelapse mortality (NRM) were 42% and 14%. Baseline characteristics in pts with ML (157 pts, of whom 76 had LI) and without ML were similar. On univariate analysis, pts with LI had no significant difference in PFS (figure A), grade II–IV acute GVHD, relapse or NRM compared to those without LI. In contrast, as compared to no ML, the presence of ML was associated with a statistically significantly improved PFS on univariate analysis (hazard ratio [HR] = 0.68, p = 0.03; figure B). This association was not identified in our previous analysis, potentially due to underpowering or inclusion of pts receiving one dose of PT/Cy. No statistically significant difference was detected according to the presence or absence of ML in the cumulative incidences of acute GVHD, relapse, or NRM, although presence of ML was associated with a tendency toward lower NRM risk (HR = 0.61, p = 0.17). On multivariate analysis, the presence of ML was found to be independently associated with a significant improvement in PFS (HR = 0.66, p = 0.03). This multivariate model also confirmed our previous observation (BBMT 2010;16:482) that greater donor-recipient HLA disparity was not detrimental to PFS (HR = 0.57, p = 0.04 for 3–4 antigen mismatches versus fewer at HLA-A, B, C, and DRB1 combined). We are currently investigating the impact of KIR haplotypes in this setting with the goal of optimizing donor selection. Conclusion: The presence of ML may be beneficial in haplo BMT with postgrafting immunosuppression that includes high-dose Cy. Further studies are needed to confirm and define the mechanisms of this effect. Our findings do not support selection of donors on the basis of LI in this transplantation platform. Disclosures: No relevant conflicts of interest to declare.

Abstract 1009 Delayed immune reconstitution with increased risk of opportunistic infection is a major complication of HLA-haploidentical stem cell transplantation, especially in protocols employing extensive T cell depletion of the graft. Previous studies at our institution with high-dose, post-transplantation Cy (PT/Cy) have reported low rates of non-relapse mortality and serious opportunistic infections. Here we characterize immune reconstitution in fifty-three consecutive hematologic malignancies patients receiving nonmyeloablative conditioning, T cell-replete, HLA-haploidentical bone marrow transplantation (BMT), and graft versus host disease prophylaxis including PT/Cy. Patients with advanced hematologic malignancies (median age 51, range 14–71; 5 AML, 2 ALL, 4 MDS, 2 CML, 4 CLL, 1 CMML, 25 NHL, 7 Hodgkins, 3 mantle cell) received Cy 14.5 mg/kg/day IV on days −6 and −5, fludarabine 30 mg/m2/day IV on days −6 to −2, 200 cGy of TBI on day -1 and T cell replete bone marrow from donors with a median age of 44 (range 14–68). GVHD prophylaxis consisted of Cy (50 mg/kg/day) on days 3 and 4, mycophenolate mofetil for 30 days, and tacrolimus for 6 months. Grafts contained an infused median TNC/kg of 4.1 e8 (range 2.6–6.6 e8), CD3+/kg 3.6 e7 (range 1.7–6.7e7) and a CD34+/kg of 3.5e6 (range 1.4–7.0e6). Sustained engraftment of donor cells occurred in 86% of evaluable patients (44/51).The median times to neutrophil (>500/μL) and platelet recovery (>20,000/μL) were 17 days (range, 13–92 days) and 28 days (range, 13–580 days), respectively. Post-transplantation recovery of lymphocyte subsets is shown in Table 1 and Figure 1 and is notable for the following: 1) The median lymphocyte count at day 30 after transplantation is >180/ml and recovers to over 800/ml by day 60; 2) CD4+ T cell counts recover to a median >120/ml by day 60 and >220/ml by day 180 after transplantation; and 3) recovery of CD31+ recent thymic emigrants and CD45RA+ naïve T cells is delayed compared to recovery of memory T cells. T cell receptor spectratyping analysis on a subset of 10 patient/donor pairs chosen specifically for having no relapse/no GVHD (n=4), GVHD and no relapse (n=3), or late relapse (n=3) revealed that patients without relapse, GVHD, or recent viral infection had excellent reconstitution of the T cell repertoire to the level of the pre-transplant donor, as early as 6 months post-transplant (Figure 2). CMV specific T cell response using ELISPOT measured on a subset of 17 patients whose donors were reactive to CMV, revealed that donor-derived immunity to CMV returns by Day 60 in about 70% of patients (12/17) (Figure 3). In conclusion, immune reconstitution after non-myeloablative haploidentical T cell replete BMT with PT/Cy compares favorably with other reduced intensity conditioning alternative donor regimens and suggests that PT/Cy selectively preserves pathogen-specific memory T cells necessary to protect against infection. Further correlations of immune reconstitution with specific infectious and overall outcomes are being analyzed.Figure 1T-, B-, and NK-cell ReconstitutionFigure 1. T-, B-, and NK-cell ReconstitutionFigure 2T cell receptor spectratypingFigure 2. T cell receptor spectratypingFigure 3CMV-specific T cell frequencyFigure 3. CMV-specific T cell frequencyTable 1.Post-transplantation Lymphocyte Subset RecoveryMedian (cells/μL) (N)Interquartile range (cells/μL)ALCDonor1765 (46)1480–2100Recipient pre-BMT840 (45)425–1295Day 30184 (49)54–402Day 60820 (38)470–1260Day 180915 (34)670–1560Day 3653060 (22)820–2030CD3+CD4+CD45RA+ (naïve)Donor119 (33)82–189Recipient pre-BMT22 (34)4–38Day 300.33 (35)0.07–1Day 603 (29)1–9Day 18011 (23)5–31Day 36523 (13)13–92CD3+CD4+CD45RA−CCR7+ (central memory)Donor135 (33)95–158Recipient pre-BMT54 (34)11–79Day 302 (35)0.5–11Day 6034 (29)10–79Day 18061 (23)35–117Day 36589 (13)60–122CD3+CD4+CD45RA−CCR7− (effector memory)Donor187 (33)130–245Recipient pre-BMT87 (34)14–134Day 303 (35)1–18Day 6059 (29)14–122Day 180102 (23)43–179Day 365142 (12)64–204CD3+CD4+CD45RA+CD31+ (recent thymic emigrants)Donor61 (33)30–97Recipient pre-BMT6 (34)1–16Day 300.9 (35)0.03–0.4Day 601 (29)0.5–2Day 1804 (23)1–11Day 3657 (13)2–12CD3+CD4+Foxp3+Donor28 (33)23–35Recipient pre-BMT13 (34)6–24Day 301 (35)0.1–5Day 608 (29)4–16Day 18013 (23)7–25Day 36514 (13)7–17ALC, absolute lymphocyte count; WBC, white blood cell count; Treg, regulatory T cell Disclosures: Jones: Aldagen: Patents & Royalties.

Abstract 2362 HLA-Haploidentical Bone Marrow Transplantation (BMT) for High Risk Hematologic Malignancies Using Myeloablative Conditioning and High-Dose, Posttransplantation Cyclophosphamide Heather J Symons, Allen R. Chen, M. Sue Leffell, Marianna Zahurak, Kathy Schultz, Heather Karaszkiewicz, Leo Luznik, Javier Bolanos-Meade, Yvette Kasamon, Lode Swinnen, Douglas Gladstone, Richard J. Jones, Ephraim J. Fuchs Historically, myeloablative BMT using T cell-replete bone marrow from partially HLA-mismatched (HLA-haploidentical) related donors has been associated with excessive rates of severe graft-versus-host disease (GVHD) and non-relapse mortality (NRM). Based upon promising clinical outcomes using high dose, post-transplantation cyclophosphamide (Cy) as GVHD prophylaxis after non-myeloablative, HLA-haploidentical BMT(Luznik, BBMT 2008;14:641) or after myeloablative, HLA-matched BMT (Luznik, Blood 2010;115(16):3224), we evaluated the safety and efficacy of high-dose, posttransplantation Cy after myeloablative conditioning and T cell-replete, HLA-haploidentical BMT for advanced and refractory hematologic malignancies. Seventeen patients (median age 40, range 13–64; 7 AML, 1 bilineage leukemia, 2 CML, 6 NHL, 1 grey zone lymphoma) have been enrolled, the majority (88%) of whom were not in remission at the time of transplant. Conditioning consisted of IV Busulfan (pharmacokinetically adjusted) on days -6 to -3, Cy (50mg/kg/day) on days -2 and -1 in sixteen patients or Cy (50mg/kg/day) on days -5 and -4 and total body irradiation (300cGy /day) on days -3 to 0 in one patient, followed by T-cell replete bone marrow in all patients. Postgrafting immunosuppression consisted of Cy (50 mg/kg/day) on days 3 and 4, mycophenolate mofetil for 35 days, and tacrolimus for 6 months in all patients. Donor T-cell engraftment occurred in all evaluable patients (n=13). The median times to neutrophil (>500/μ L) and platelet recovery (>20,000/μ L) are 24 days (range, 17–44 days) and 22 days (range, 13–91 days), respectively. On competing risk analysis, the cumulative incidences of grades II-IV and grades III-IV acute graft versus host disease (aGVHD) at day 100 are 24% (95% CI: 7%, 46%) and 12.5% (95% CI: 1.8, 33.9%), respectively (Figure a). On competing risk analysis, the cumulative incidence of chronic GVHD (cGVHD) at one year is 14.7% (95% CI: 0.76, 47.1%). On competing risk analysis, the cumulative incidence of NRM at 100 days is 18% (95% CI: 4%, 39%) (Figure b). NRMs were due to multi-organ system failure in two patients with bulky mediastinal lymphoma and veno-occlusive disease in one patient who had received gemtuzumab ozogamicin 3 weeks prior to BMT. There have been no infectious deaths to date. On competing risk analysis, the cumulative incidence of relapse at 100 days is 29% (95% CI: 10%, 52%) in this poor-risk cohort. With a median follow-up of 10.1 months (range 5.3–12.8 months) in those without events, actuarial overall survival (OS) is 58% at 6 months and 37% at one year and actuarial event-free survival (EFS) is 22% at 6 months and 14% at one year. Myeloablative HLA-haploidentical BMT with T cell replete bone marrow and posttransplantation Cy is associated with promising rates of engraftment, GVHD, and NRM that do not appear substantially different than that seen with non-myeloablative haploidentical BMT with posttransplantation Cy or with myeloablative matched BMT. Based on these data, this approach is being moved into better risk patients. Disclosures: Kasamon: Genentech: Research Funding. Swinnen: Genentech: Membership on an entity's Board of Directors or advisory committees, Research Funding.

NM-HCT from related donors who share at least one HLA haplotype with the recipient is a treatment option for patients with hematologic malignancies who do not have suitable HLA-matched related or unrelated donors. We have shown that addition of cyclophosphamide pre-transplant (29 mg/kg) and post-transplant (50 mg/kg on day+3 or days+3 and +4) to a standard nonmyeloablative conditioning regimen of fludarabine and TBI combined with tacrolimus and MMF prophylaxis of GvHD is an effective means of achieving complete engraftment of donor T-cells and granulocytes with low non-relapse mortality and an incidence and severity of acute and chronic GvHD which does not differ from studies of NM-HCT using HLA-matched donors. In a study of 89 patients (FHCRC,N=30; JH,N=59) with advanced myeloid and lymphoid malignancies, patients with relapsed HL had significantly better survival than patients with other diagnoses. Fifteen patients with HL were studied (median follow-up of 17 mo). Patients were heavily pre-treated with a median number of prior cytotoxic therapy regimens of 5 (range: 3–10); 14 patients had failed prior autologous HCT with a median time of 18 mo from auto-HCT to NM-HCT. Donors for 9 patients were HLA-mismatched at ≥4/10 loci. All evaluable patients were complete donor CD3 and CD33 chimeras by day +28 (one patient died on day +28 and was not evaluable). Clinically significant acute GvHD occurred in 9/14 (64%) patients and Grades III/IV GvHD in 3/14 (21%) patients. Extensive chronic GvHD occurred in 5/14 (36%) patients. Two patients (13%) died of non-relapse causes at days +236 and +633 secondary to chronic GvHD. Median failure-free survival (FFS) was 21 mo compared to 6 mo for patients with other lymphoid malignancies (N=23) or myeloid malignancies [(N=51), see Figure]. The hazard of mortality was less among patients with HL compared to those with other lymphoid malignancies [hazard ratio (HR)=0.36 (p=0.05)] yet patients with myeloid malignancies had a similar hazard of mortality compared to those with lymphoid malignancies other than HL [HR=0.88 (p=0.66)]. Difference in FFS between HL and other lymphoid malignancies was not statistically significant [HR=0.54 (p=0.16)], nor was the difference between myeloid and other lymphoid malignancies [HR=1.26 (p=0.44)]. More patients will need to be studied to better demonstrate a graft-versus-lymphoma effect of haploidentical transplantation in relapsed HL. NM-HCT from haploidentical donors may be a good option for such patients who have a limited window of opportunity to proceed to transplant if responsive to salvage chemotherapy. Download : Download high-res image (95KB) Download : Download full-size image Download : Download high-res image (79KB) Download : Download full-size image

Cytomegalovirus (CMV) disease remains an important complication of allogeneic stem cell transplantation (SCT). A major goal has been to develop bone marrow transplant (BMT) preparative regimens that minimize the duration of post-transplantation immunosuppression, and thus the risk of severe infectious complications. Previously, we reported a low rate of CMV reactivation (17.6%, 6/34 eligible patient-donor pairs) among hematologic malignancies patients receiving partially HLA-mismatched (haploidentical) BMT after non-myeloablative conditioning including high dose, post-transplantation Cy (50 mg/kg IV on days 3 and 4). Here, we have retrospectively analyzed an additional 99 recipients of post-transplantation Cy after fludarabine/200 cGy TBI-based nonmyeloablative conditioning and haploidentical BMT (n=44) or busulfan/Cy-based myeloablative conditioning and HLA-matched BMT (n=55). After high-dose Cy, patients undergoing haploidentical BMT received mycophenolate mofetil and tacrolimus, while recipients of HLA-matched grafts received no further graft-versus-host disease (GVHD) prophylaxis. Patients did not receive CMV prophylaxis. CMV reactivation, measured weekly by quantitative PCR, occurred in 38.7% (12/31 eligible donor-patient pairs) in the mini-haploBMT trial and 31% (13/42 eligible patient donor pairs) in the myeloablative trial. CMV reactivation occurred twice in two patients in each trial. CMV disease occurred in two patients in the haploidentical BMT trial (1 retinitis, 1 pneumonia) but in no recipients of HLA-matched grafts. No deaths have been attributed to CMV infection on either trial. In the mini-haplo BMT patients, 58.3% (7/12) of those with CMV reactivation were considered high-risk (recipient CMV IgG positive, donor CMV IgG positive or negative) and 25% (3/12) were considered to be at intermediate risk (recipient CMV IgG negative and donor CMV IgG positive). Of the myeloablative BMT patients, 100% (13/13) of patients with CMV reactivation were part of the high-risk group. In addition to recipient CMV positivity, the presence of graft-versus-host disease (GVHD) was a risk factor for CMV reactivation on the myeloablative trial as 77% (10/13) of patients who reactivated their CMV had GVHD, but not on the “mini-haploBMT” trial where only 41.7% (5/12) had GVHD. The median number of days post-transplant to CMV reactivation was 49 (range 19–197) in the “mini-haploBMT” patients and 56 (range 19–140) in the myeloablative group. The median number of weeks to clear detectable CMV was 1 (range 1–11 in the mini-haploBMT group and 1–4 in the myeloablative group). These rates of CMV reactivation and infection compare favorably to the rates seen in patients receiving haploidentical and HLA-matched grafts following myeloablative conditioning without post-transplantation Cy. The low rates of CMV reactivation may be secondary to low rates of Grade 3–4 GVHD (9.1%, 4/44 “mini-haploBMT patients”; 7.3%, 4/55 myeloablative patients) and rapid immunologic recovery facilitated by the absence of prolonged pharmacologic immunosuppression, especially among recipients of HLA-matched grafts.

Donor vs. recipient natural killer (NK) cell alloreactivity in mismatched (haploidentical) bone marrow transplantation (BMT) has been defined as a mismatch between killer inhibitory receptors (KIR) on donor NK cells and their inhibitory HLA class I ligands on recipient cells. Data reporting the impact of KIR incompatibility in the haploidentical BMT setting have been controversial. Here, we examine the effect of KIR ligand mismatches in 60 patients with advanced hematologic malignancies who were enrolled in a clinical trial of nonmyeloablative haploidentical BMT (“mini-haploBMT”). All patients were conditioned prior to transplantation with fludarabine, low dose cyclophosphamide (Cy), and 200 cGy TBI, and after transplantation received high dose Cy, mycophenolate mofetil (MMF) and tacrolimus. The median age of transplant recipients was 48 years (range, 1–71years). Patients received a median of 1.27x 108/kg mononuclear cells/kg and a median of 3.9 x 106 CD34+ cells/kg. Donor-recipient pairs had an average of 3 out of 8 HLA antigen mismatches in both the host-vs.-graft (HVG) and graft-vs.-host (GVH) directions. Graft failure occurred in 12/58 (21%) evaluable patients, and grade 3–4 GVHD occurred in 7/60 patients (11.7%). Median time to disease progression was 222 days and median overall survival was 255 days. The median follow-up at the time of this analysis was 251 days (range 41–1715 days). KIR mismatches were identified via KIR genotyping (KIR genes determined by SSP from genomic DNA extracts) as well as via the previously described receptor:ligand model which identifies the presence of inhibitory KIR on donor cells, with absence of HLA ligand on host cells. Utilizing the receptor:ligand model, our results demonstrate an improvement in time to relapse (TTR) in patients with ≥1 inhibitory KIR mismatches (32 patients, median TTR 359 days) versus no mismatches (27 patients, median TTR 185 days) (p=0.036) and an improved overall survival (OS) in patients with ≥1 inhibitory KIR mismatches (32 patients, median OS 622 days) versus no mismatches (27 patients, median OS 208 days) (p=0.008). No significant difference in OS or TTR was found when patients were separated by disease type (myeloid versus lymphoid; OS p= 0.095 vs. p=0.1 and TTR p=0.2 vs. 0.1). There were also no significant differences in TTR or OS when using the KIR genotyping method of identifying mismatches. No differences in Grades 2–4 graft-versus-host disease (GVHD) or engraftment were found based on the presence of inhibitory or stimulatory KIR mismatches, although the small numbers of patients with GVHD or engraftment failure prevent a definitive analysis. KIR ligand incompatibility in the GVH direction is associated with a prolonged time to relapse and improved overall survival after nonmyeloablative haploidentical BMT. These findings may warrant the selection of donors based upon KIR ligand incompatibility.

The goal of conditioning for nonmyeloablative allogeneic stem cell transplantation, or NST, is to permit the stable engraftment of donor cells with a minimum of toxicity to the patient. We have previously developed a minimal conditioning regimen, consisting of low dose total body irradiation (TBI; 100–200 cGy) and post-transplantation cyclophosphamide (Cy), that permits the induction of stable mixed hematopoietic chimerism following the transplantation of modest doses of MHC-identical allogeneic stem cells in mice. The goal of the current study was to determine whether intravenous (IV) busulfan (Busulfex; ESP Pharma, Edison, NJ), an agent with myelotoxic but minimal immunosuppressive activity, could be used in place of TBI in this regimen and still promote allograft tolerance. TBI was more immunosuppressive than Busulfex, because engraftment of allogeneic stem cells could be achieved in mice conditioned with 700 cGy TBI alone but not in mice conditioned with an equivalently myelotoxic dose of 60 mg/kg Busulfex IV. However, when post-transplantation Cy was added, stable mixed hematopoietic chimerism was achieved in mice conditioned with doses as low as 100 cGy TBI or 20 mg/kg Busulfex. Busulfex and post-transplantation Cy were both required for the induction of allograft tolerance, because graft rejection occurred in transplanted mice treated with either agent alone. The capacity of Busulfex to promote allograft tolerance in this model could not be explained by effects of the drug on the host immune system, because the graft rejection response was not restored in Busulfex-conditioned mice reconstituted with mature host lymphocytes at the time of transplantation. These results demonstrate that Busulfex effectively substitutes for TBI in promoting allograft tolerance, and suggests that the creation of marrow “space” promotes alloengraftment in conditions of suboptimal host immunosuppression.

The goals of administering transplantation conditioning to leukemia patients undergoing alloSCT are to kill tumor cells, prevent rejection of the donor graft, create “space” for donor hematopoiesis, and provide an overall milieu that favors a donor T cell-mediated “graft-versus-leukemia” effect. However, transplantation conditioning also increases the incidence and severity of acute GVHD by damaging non-hematopoietic tissues and by inducing lymphopenia, which augments the antigen-driven expansion of alloreactive T cells. The purpose of this study was to compare the relative safety and efficacy of transplantation conditioning with intravenous busulfan (Busulfex; ESP Pharma, Edison, NJ) versus total body irradiation (TBI) by administering titrated doses of either agent to mice and comparing their effects on: 1) donor chimerism after syngeneic and semi-allogeneic (parent into F1) SCT; 2) graft rejection after MHC-matched allogeneic SCT; and 3) acute GVHD after semi-allogeneic SCT. Both agents induced titratable donor chimerism after syngeneic BMT, but TBI was more immunosuppressive, since allogeneic marrow engrafted after 700 cGy TBI but not after an equivalently myelotoxic dose of Busulfex (60 mg/kg). When given in equivalently myelotoxic doses, TBI always induced more GVHD than did IV busulfan, as measured by weight loss and the augmentation of donor T cell and myeloid chimerism. As a result, stable mixed hematopoietic chimerism was generated more readily in mice conditioned with nonmyeloablative doses of busulfan than in mice conditioned with equitoxic doses of TBI. These results demonstrate that IV busulfan promotes GVHD to a lesser extent than does an equivalently myelotoxic dose of TBI. IV busulfan may be the preferred agent if stable mixed hematopoietic chimerism is the desired outcome. The relative roles of lymphopenia versus tissue damage in the differential outcomes of TBI versus Busulfex conditioning are being investigated.

A variety of hematologic malignancies that are resistant to standard chemotherapy can be treated by nonmyelaoblative allogeneic stem cell transplantation, or NST. At our institution, 84 patients have enrolled onto sequential Phase I and II protocols of NST for hematologic malignancies utilizing haploidentical donors. Patients received non-T cell-depleted marrow after conditioning with fludarabine 30 mg/m2 (days −6 to −2), cyclophosphamide (Cy; 14.5 mg/kg/d, days −6 and −5) and 200 cGy total body irradiation on day −1. Novel post-transplantation immunosuppression included high dose Cy (50 mg/kg/d on day 3 only or (n=23) days 3 and 4 (n=40), mycophenolate mofetil from days 5–35 and FK-506 from day 5–180. Sixteen patients (19%) experienced graft rejection, defined as absence of donor chimerism on day 60, including 6/10 patients with myelodysplastic syndrome (MDS), 3/7 patients with chronic myeloid leukemia (CML), 2/3 patients with chronic myelomonocytic leukemia (CMML), 2/21 patients with acute myeloid leukemia (AML), and three patients with lymphoid malignancies. Fourteen patients recovered autologous hematopoiesis, whereas two (1 MDS, 1 NHL) died in aplasia. Interestingly, objective clinical responses were observed in 3/6 of the rejectors with MDS (1 CR, 2 PR), 2/3 rejectors with CML (2 molecular CR), and in both rejectors with CMML (1 CR, 1 PR). Two of the five patients with MDS and blast counts of greater than 20% survived for greater than 2 years with a decrease in blast counts and need for transfusions. Both patients with CML had disease refractory to imatinib mesylate, and both achieved molecular complete responses following transplantation. One of the patients has relapsed, and the other remains in hematologic and cytogenetic complete remission, though in minimal molecular relapse, at 6+ months after transplantation. Of the patients with CMML, one achieved a morphologic and hematologic complete remission at six months after transplantation, and the other achieved improved blood counts and disappearance of marrow blasts before relapsing at six months after transplant. In summary, fatal graft rejection is uncommon following nonmyeloablative BMT with post-transplantation cyclophosphamide. Clinical responses occurred despite graft rejection, particularly among patients with chronic myeloid malignancies. Potential mechanisms of the anti-tumor effect include an effect of conditioning, a transient graft-versus-host reaction by donor T or NK cells, or an abrogation of host tolerance to tumor by either the graft-versus-host or host-versus-graft reaction.

Graft rejection, severe graft-versus-host disease (GVHD), and prolonged immunosuppression are major complications of partially HLA-mismatched, or haploidentical, stem cell transplantation in humans. Intensive conditioning of the recipient and rigorous T cell depletion of the donor graft reduces the risk of the first two complications, but may prolong or intensify post-transplantation immunodeficiency, increasing the patient’s susceptibility to opportunistic infection. In contrast, alloreactive T cells may be selectively eliminated in vivo by the early post-transplantation administration of high dose cyclophosphamide (Cy), leaving virus or other pathogen specific resting memory T cells relatively unharmed. To test this hypothesis, we have analyzed the infectious complications in 60 patients who were enrolled in a clinical trial of nonmyeloablative haploidentical bone marrow transplantation (“mini-haploBMT”) incorporating post-transplantation Cy. All patients received fludarabine 30mg/m2/dose on days -6 to -2, Cy on day -2, and 200 cGy total body irradiation (TBI) on day -1. The first 20 patients received a single dose of 50 mg/kg Cy IV on day 3, whereas the final 40 patients received the same dose of Cy on day 3 and 4. The median age of transplant recipients was 48years (1-71years), 12 patients had 13 prior autologous transplants prior to enrollment, and patients had an average of 2.7 treatment regimens prior to transplant (0–8). Patients received a median of 1.27x 108/kg mononuclear cells/kg with 3.9 x 106 CD34+ cells/kg. The median follow-up at the time of this analysis was 245.5 days (41–1550 days). An ANC >500/mm3 was achieved a median of day 16 (12–208). Only 5 patients (8.3%) died from infectious related causes. Of these, 3 were known to have relapsed disease. There were 34 patient/donor pairs (56.6%) at risk for CMV reactivation. In these pairs, only 6 patients developed CMV reactivation (17.6%), and 2 patients developed CMV infection. 3 patients with CMV reactivation received one dose of post-transplant Cy, and 3 patients received two. Both patients who developed CMV infections received only dose of post transplant Cy. 23% (14/60) of patients developed an invasive fungal infection, of which 21.4% (3/14) developed in the first 30 days post transplant, 14.3% (2/14) developed between days 31–60, 14.3% (2/60) developed between days 61–100, and 50% (7/14) occurred more then 100 days post transplant. 15% (9/60) had at least one systemic viral infection (other then CMV), of which 22.2% (2/9) developed in the first 30 days post-transplant and 77.8% (7/9) developed more then 90 days post transplant. No patients developed viral infections between days 31–100. These rates of infection compare favorably to the rates of infection seen in patients receiving haploidentical grafts following myeloablative conditioning without post-transplantation Cy. We propose that the very low numbers of patients with CMV reactivation, fungal or viral infections, and infectious related mortality is secondary to early engraftment, low rates of Grade 3–4 GVHD (7/60 patients, 11.7%; 5/7 patients received one dose of post-transplant Cy, 2 received two) or prolonged immunosuppression, and an intact host B cell and memory T cell function preserved with this novel prep regimen and T cell replete graft.

The “allogeneic effect” refers to the induction of host B cell antibody synthesis or host T cell cytotoxicity, including tumoricidal activity, by an infusion of allogeneic lymphocytes. We have previously shown that treatment of mice with cyclophosphamide (Cy) followed by infusion of CD8+ T cell-depleted allogeneic spleen cells (Cy + CD8− DLI) induces anti-tumor activity in a model of minimal residual leukemia, even though the donor cells are eventually rejected by the host immune system. The purpose of the current investigation was to test the activity of Cy + CD8− DLI in the treatment of well-established cancer, and to characterize the mechanisms of the anti-tumor effect. BALB/c mice were inoculated intravenously (IV) with the syngeneic A20 lymphoma/leukemia or the RENCA renal cell carcinoma on day 0 and were then treated with nothing, Cy alone on day 14, or Cy + CD8− DLI from MHC-mismatched C57BL/6 donors on day 15. In both tumor models, the combination of Cy + CD8− DLI significantly prolonged survival compared to mice treated with nothing or with Cy alone. While depletion of CD4+ T cells from the DLI significantly diminished the beneficial effect of CD8− DLI, purified CD4+ T cells alone were inactive, demonstrating that donor CD4+ T cells and another population of cells were required for optimal anti-tumor activity. Several observations pointed to an active role for the host immune system in the anti-tumor activity of Cy + CD8− DLI. First, host T cells participated in the anti-tumor effect of treatment with Cy alone, since the drug’s activity was diminished in tumor-bearing scid mice or in normal BALB/c mice depleted of T cells. Second, while Cy + CD8− DLI caused no GVHD in tumor-bearing but immunocompetent BALB/c recipients, it caused fatal acute GVHD in either tumor-bearing scid or T-cell depleted BALB/c mice. Finally, the anti-tumor effect of Cy + CD8- DLI was also significantly inhibited in BALB/c mice that were depleted of CD8+ T cells. These results demonstrate that transiently engrafting T cells administered after Cy can induce significant anti-tumor effects against both solid and liquid tumors. We propose that upon recognition of alloantigen on host antigen-presenting cells (APCs), allogeneic donor CD4+ T cells deliver activating ligands to the APCs, thereby generating effective “help” to break tolerance in tumor-specific host CD8+ T cells. This mechanism may correspond to the “allogeneic effect” in the anti-tumor response described over three decades ago.