Introduction Outcomes with salvage regimens in pts with MCL who progress after Bruton tyrosine kinase inhibitor (BTKi) therapy are poor. ZUMA-2 is a Phase 2, registrational, multicenter study evaluating KTE-X19, an autologous anti-CD19 CAR T cell therapy, in pts with R/R MCL (1-5 therapies, including a BTKi). We present interim efficacy and safety results. Objectives Evaluate efficacy and safety of KTE-X19 in R/R MCL. Methods Eligible pts (≥ 18 y) with R/R MCL had an ECOG of 0-1 and ≤ 5 prior therapies, including chemotherapy, an anti-CD20 antibody, and a BTKi. Pts underwent leukapheresis and conditioning chemotherapy followed by KTE-X19 infusion at 2 × 106 cells/kg. Bridging therapy with dexamethasone, ibrutinib, or acalabrutinib was permitted. The primary endpoint was objective response rate (ORR; complete response [CR] + partial response) assessed by an Independent Review Committee per the Lugano Classification (Cheson, et al. J Clin Oncol. 2014). Interim efficacy endpoints were investigator-assessed using the revised IWG Response Criteria for Malignant Lymphoma (Cheson, et al. J Clin Oncol. 2007). Key secondary endpoints were duration of response (DOR), progression-free survival (PFS), overall survival (OS), frequency of adverse events (AEs), and blood levels of CAR T cells and cytokines. Sixty pts received KTE-X19; here, we present results in pts with ≥ 1 y follow-up. Updated results in all 60 pts will be reported in the presentation. Results As of May 30, 2018, 28 pts received KTE-X19 with ≥ 1 y follow-up (median, 13.2 mo). The median age was 65 y; 43% of pts had an ECOG score of 1; 21% had blastoid morphology; 82% had stage IV disease; 50% had intermediate/high-risk MIPI; and 86% received a median of 4 prior therapies. In 20 of 28 pts with available data, the median Ki-67 index was 38%. Eight pts received bridging therapy; all had disease present post-bridging. Investigator-assessed ORR was 86% (95% CI, 67-96) with a CR rate of 57% (95% CI, 37-76). As of May 30, 2018, 75% of responders remained in response, and 64% of treated pts had ongoing responses. The 12-month rates of DOR, PFS, and OS were 83% (95% CI, 60-93), 71% (95% CI, 50-84), and 86% (95% CI, 66-94), respectively; medians were not reached. The most common Grade ≥ 3 AEs were anemia (54%), platelet count decreased (39%), and neutropenia (36%). Grade 3/4 cytokine release syndrome (CRS) assessed by Lee et al. (Blood. 2014) and Grade 3/4 neurologic events (NE) occurred in 18% and 46% of pts, respectively, with no Grade 5 events. CRS and NE were generally reversible. There was 1 Grade 5 AE of organizing pneumonia. Median CAR T cell levels measured by peak and area under the curve were 99 cells/µL (range, 0.4-2589) and 1542 cells/µL (range, 5.5-27239), respectively. Conclusion With ≥ 1 y follow-up, KTE-X19 demonstrated significant and durable clinical benefit, with a manageable safety profile in pts with R/R MCL for whom there are no curative treatment options.

Introduction LentiGlobin for SCD GT contains autologous CD34+ hematopoietic stems cells (HSCs) encoding β-globin with the anti-sickling T87Q mutation (βA-T87Q) and is being evaluated in the ongoing Phase 1/2 HGB-206 Study (NCT02140554) in patients with SCD. Levels of GT-derived hemoglobin (HbAT87Q) in 7 initial patients (Group A) were suboptimal but were maintained for ≥ 30 months of follow-up post-treatment, suggesting durable transgene expression. To increase HbAT87Q production, protocol and manufacturing changes were made (Group B; N=2). In addition, HSC collection by plerixafor mobilization and apheresis was instituted in Group C. Objective Provide an update on safety and efficacy of LentiGlobin for SCD in HGB-206 Group C. Methods Adults with severe SCD (including recurrent vaso-occlusive crisis [VOC] and acute chest syndrome [ACS]) were enrolled. CD34+ HSCs were harvested by apheresis following plerixafor mobilization and transduced with BB305 lentiviral vector (LVV). Patients received myeloablative busulfan conditioning, were infused with LentiGlobin drug product (DP) and monitored for adverse events (AEs), Hb fractions, and other parameters. LVV presence in transduced cells (%LVV+) was measured by qPCR of individual colonies from colony-forming unit assays from pre-infusion DP) and post-infusion from CD34+ bone marrow (BM) HSCs and peripheral blood mononuclear cells (PBMCs). Data are shown as median (min-max). Results As of 7 March 2019, 13 Group C patients received DP, with follow-up of 9.0 (1.0-15.2) months. All but 1 patient had neutrophil and platelet engraftment as of the data cut date. Median HbS was ≤50% of total Hb in those with ≥6 months follow-up (n=8; Figure 1). Total unsupported Hb at last visit in patients with ≥6 months of follow-up was 11.5 (10.2-15.0) g/dL, with HbAT87Q levels of 5.3 (4.5-8.8) g/dL. Six of these 8 patients had a history of VOCs or ACS; the annualized VOC+ACS rate decreased from 5.3 (3-14) pre-treatment to 0 (0-2) post-treatment (Figure 2). A decrease in hemolysis markers was also seen post-DP. Most common non-hematologic Grade ≥ 3 AEs were febrile neutropenia (n=10) and stomatitis (n=7). Serious AEs occurred in 6 patients; the most frequent were nausea and vomiting. To date, there have been no cases of DP-related AEs, graft failure, vector-mediated replication competent lentivirus, or clonal dominance. The %LVV+ colonies from PBMCs at 9 months and BM at 12 months post-DP infusion (n=5) were 79.2 (67.0-88.4) % and 81.5 (60.6-88.1) %, respectively, indicating stable engraftment of transduced cells from DP (%LVV+ was 80 [71-88] %). Conclusions Patients in HGB-206 Group C show stable LentiGlobin engraftment, with median total Hb >10 g/dL and median HbS ≤50% of total Hb in those with ≥6 months follow-up. The decrease in SCD-related complications and hemolysis in this cohort demonstrate a strong therapeutic benefit of LentiGlobin in patients with SCD.

Introduction CD19 directed chimeric antigen receptor T cell (CART19) therapy has resulted in unprecedented clinical outcomes. However, its success in chronic lymphocytic leukemia (CLL) has been modest to date. An increasing body of evidence indicates that impaired CART cell fitness is the predominant mechanism of the relative dysfunction in CLL. The immunosuppressive microenvironment in CLL is well known and may be related to the abundance of circulating extracellular vesicles (EVs) bearing immunomodulatory properties. We hypothesized that CLL derived EVs contribute to CART cell dysfunction. Objectives We aimed to investigate the interaction between EVs from CLL patients and CART19 cells. Methods We characterized circulating EVs from platelet free plasma in untreated patients with CLL and normal controls using nanoscale flow cytometry. CLL derived EVs, CART19, and CLL B cells were imaged with super resolution microscopy. We stimulated CART19 cells with CD19+ JeKo1 cells at a 1:1 ratio in the presence of increasing concentrations of CLL derived EVs and measured inhibitory receptors by flow cytometry. To interrogate the transcriptome, we stimulated CART19 cells with irradiated JeKo1 cells in the presence of CLL derived EVs at ratios of 10:1 and 1:1 EV:CART19. Finally, immunocompromised mice were engrafted with the JeKo1 and randomized to treatment with untreated, CART19 cells, or CART19 cells co-cultured ex vivo with CLL derived EVs for six hours. Results Our findings indicate that CLL derived EVs impair normal donor CART19 antigen-specific proliferation (Fig 1A). Although there wasn't a difference in total EV particle count (Fig 1B), CLL derived EVs expressed significantly more PD-L1 than normal controls (Fig 1C). On super resolution microscopy, EVs were localized at the T cell-tumor junction (Fig 1D). Furthermore, CLL derived EVs were captured by T cells (Fig 1E). Having demonstrated this interaction, we sought to establish their functional impact on CART19 cells. There was a significant upregulation of inhibitory receptors on the T cells (Fig 1F), associated with a reduction in CART effector cytokines at higher concentrations of EVs (Fig 1G), suggesting a state of exhaustion. This was further supported by RNA sequencing, which indicated a significant upregulation of AP-1 and YY1 (Figs 1H), known T cell exhaustion pathways. Finally, to confirm the impact of CLL derived EVs on CART19 functions in vivo, treatment with CART19 cultured with EVs resulted in reduced anti-tumor activity compared to treatment with CART19 alone (Fig 1I). Conclusion Our results indicate that CLL derived EVs induce significant CART19 cell dysfunction in vitro and in vivo, by direct interaction with CART cells resulting in a downstream alteration of their exhaustion pathways. These studies illuminate a novel way that circulating and potentially systemic EVs can lead to CART cell dysfunction in CLL patients.

Introduction Studies of target organs at the time of acute graft versus host disease (GVHD) are needed to identify novel inflammatory and regulatory pathways. Methods We performed a prospective transcriptome analysis of gastrointestinal (GI) biopsies at diagnosis of acute GI GVHD and compared these to bone marrow transplant (BMT) patients without GVHD. Allogeneic BMT patients ≥2 years old undergoing endoscopy for clinical indications had 2 additional rectosigmoid mucosal biopsies for transcriptome analysis. Transcriptome data from 211 ulcerative colitis (UC) patients were included to uncover genes and pathways shared between GVHD and UC. Rectal biopsies from 20 children without inflammatory bowel disease (IBD) served as controls. Results Twenty-one patients underwent endoscopy pre (n=5) or post BMT (n=16). Median age of patients was 13.5 years (range 4.5 -25 years). Four patients were excluded from analyses (pre BMT colitis n=3, no lower GI biopsy n=1). Nine patients had acute GI GVHD, while 8 patients did not (Table 1). Principal component analysis shows patient clustering of those with and without GVHD, suggesting distinct transcriptomic mechanisms driving disease (Figure 1A). We identified 1310 significant genes, 144 of which were downregulated and 1266 upregulated in acute GVHD compared to no GVHD (Figure 1B). Top upregulated biological processes in GVHD revealed an innate antimicrobial response and associated cytokine signaling, notable for interferons. Top 20 differentially expressed genes in acute GVHD (Table 2), include CXCL9, CXCL10 and CXCL11, which are chemokines involved in GVHD, and IDO1 which regulates tryptophan metabolism. Four hundred and ninety shared genes were upregulated, and 126 genes were downregulated in GVHD and UC. Notable shared downregulated pathways involved butyrate metabolism (Table 3). DUOXA2, a gene involved in regulating innate antimicrobial responses in IBD, was the top upregulated genes in pediatric IBD and was the second highest differentially expressed gene in the acute GVHD, providing additional rationale for intestinal microbiome modulation in both diseases. Genes involving the ERK1/2 cascade (Table 4) were upregulated in acute GVHD (p=1.67 × 10 −3) providing rationale for exploring ERK1/2 inhibition as novel acute GVHD treatment. Conclusions Acute GI GVHD has a distinct transcriptome signature which includes pathways of innate antimicrobial responses highlighting the role for microbiome modulation. The ERK1/2 cascade pathway is upregulated, suggesting a novel target in acute GI GVHD. Acute GI GVHD has important similarities with the UC transcriptome, including upregulation of DUOXA2 and downregulation in butyrate metabolism in both.

Autologous hematopoietic stem cell transplantation (Auto-HSCT) with gene-modification represents a potential cure for multiple genetic diseases, but its broad curative potential, is limited because of morbidity/mortality from cytotoxic chemotherapy-based conditioning. To overcome these limitations, we developed antibody drug conjugates (ADC) targeting CD117 (C-KIT) to specifically deplete the hematopoietic stem and progenitor cells (HSPC). To validate CD117 ADC-mediated depletion prior to HSCT, we developed an optimized non-human primate (NHP) tool anti-CD117 ADC and evaluated it in an auto-gene modified HSCT in a rhesus model. The CD117-ADC is potent on primary human and NHP CD34+ cells in vitro (Figure 1A). Humanized NSG mice treated with a single dose had full depletion of human HSPCs in the bone marrow, while maintaining peripheral immune cells. In rhesus a single administration was fully myeloablative (>99% HSPC depletion) and comparable to HSPC depletion observed following busulfan conditioning (6 mg/kg/day x4). There was no effect of the ADC on the peripheral and bone marrow lymphocytes and the ADC was well tolerated compared to busulfan where the animals had diarrhea, loss of appetite and weight loss. To facilitate use in HSCT, the CD117-ADC was engineered to have a fast clearance and the half-life was < 10 h in NHP. We next explored whether the tool CD117-ADC could enable auto gene modified HSCT in the rhesus model. Two rhesus NHP were mobilized with GCSF and plerixafor. The selected CD34+ cells were transduced with β-globin encoded lentivirus and cryopreserved. The tool CD117-ADC was dosed on day -6 and the CD34+ cells were thawed and infused on day 0. Bone marrow aspirates analyzed on the day of infusion (day 0) demonstrated >99% depletion of the HSPCs and maintenance of the bone marrow lymphocytes (Figure 1B). The primates engrafted neutrophils (day 8 and 10) and platelets (day 10 and 11), and the peripheral lymphocytes were maintained throughout the transplant (Figure 1C-D). The gene marking in the granulocytes is comparable to busulfan conditioned animals previously reported (Tisdale, Molecular Therapy 2019). Longer follow up and data from additional animals will be presented. In summary, we have developed a tool CD117 ADC that shows potent activity on NHP CD34+ cells, is fully myeloablative, has a favorable safety profile, spares the immune system and is cleared rapidly as designed. In a rhesus model of auto-gene modified HSCT, a single dose of the ADC enables engraftment of gene modified HSC. These proof of concept studies validate the use of CD117-ADC for targeted HSPC depletion prior to transplant and support its use as a new conditioning agent for auto-gene modified HSCT. This targeted approach for safer conditioning could improve the risk benefit profile for patients undergoing HSCT and enable more patients to benefit from these potentially curative therapies.

Background Traditionally, pre-transplant conditioning regimen is given over 4-6 days before hematopoietic cell transplant (HCT). Delivering higher dose chemotherapy preparative regimen over a longer time period has not been tested previously. We hypothesized that the delivery of myeloablative dose of busulfan over a 3-week period may reduce toxicity and non-relapse mortality (NRM), without affecting relapse, and tested this in a prospective phase II study. Methods Patients between 18 and 70 years of age with hematological malignancies and adequate organ function, with 8/8-HLA matched related or unrelated donor were eligible. They received a fixed dose of busulfan 80mg/m2 as outpatient on days -20 and -13. Then, fludarabine 40mg/m2 was given on days -6 to -2 followed by busulfan dosed to achieve target area under the curve (AUC) of 20,000 mol/min for the whole course based on pharmacokinetic studies. GVHD prophylaxis was cyclophosphamide (PTCy) 50mg/kg on days 3 and 4 and tacrolimus. Mycophenolate mofetil (MMF) was added to later unrelated donor recipients. All patients received standard supportive care. The primary endpoint was day 100 NRM. Results We enrolled 52 patients with a median age of 62 (range, 39-69) years. Almost half (n=25, 48%) had AML or MDS and the other half (n=26, 50%) had had CML or MPD; 1 (2%) had multiple myeloma. Low, intermediate, high and very-high disease risk index (DRI) was present in 3 (6%), 34 (65%), 14 (27%) and 1 (2%). HCT-comorbidity index was ≥3 in 23 (44%) and 1-2 (n=23, 44%). A majority (n=32, 62%) had an unrelated donor. With a median follow up of 14 months (range, 3-23), NRM at day 100 was 1.9% (n=1) and 8% (95% CI, 0-15) at 1 year. Overall survival, progression-free survival and relapse at 1-year were 83% (95% CI, 73-95%), 78% (95% CI, 67-91%), and 14% (95% CI, 4-24%), respectively [Table]. There were no graft failures. The median time to neutrophil engraftment was 17 days (range, 13-33) and that of platelets (≥ 20K/µL, n=45) was 24 days (range, 9-266). Day 100 grade II-IV and III-IV acute GVHD rates were 37% (95% CI, 23-50%) and 6% (95% CI, 0-12%), respectively; 1-year chronic GVHD and extensive chronic GVHD rates were 9% (95% CI, 0-17%) and 7% (95% CI, 0-14%), respectively. Overall survival at 1-year differed significantly among patients with low/intermediate DRI (94%; 87-100%) and those with high/very high DRI (53%; 31-91%), P=0.001. Conclusion Myeloablative fractionated busulfan regimen with PTCy GVHD prophylaxis is feasible in older patients, has low incidence of severe acute GVHD, chronic GVHD, and NRM and results in promising overall survival.

Introduction In QuANTUM-R, the once-daily, oral, highly potent and selective FLT3 inhibitor Q improved clinical benefit vs SC (median overall survival [mOS], 6.2 vs 4.7 mo [HR, 0.76 (95% CI, 0.58-0.98); P = .02]) in R/R FLT3-ITD AML (NCT02039726). Before randomization, 25% (Q) and 23% (SC) of pts had 1 prior HSCT. Objective To describe post hoc analyses in pts who underwent subsequent HSCT during QuANTUM-R. Methods Pts with FLT3-ITD R/R AML received Q (60 mg [30-mg lead-in]) or SC. Pts in the Q arm could resume Q 30-100 d after HSCT. Decisions to proceed to HSCT and resume Q after HSCT were made per investigator discretion/institutional policy. Results Of 367 randomized pts, 85 in the Q arm underwent any subsequent HSCT (allogeneic HSCT [allo-HSCT], 84 [6 with and 78 w/o additional AML therapy]; autologous HSCT, 1), and 19 in the SC arm underwent any HSCT (5 with and 14 w/o additional AML therapy]). Q + SC pooled data showed a longer mOS (95% CI) in 104 pts with any HSCT vs 263 w/o (12.2 [10.0-24.1] vs 4.4 [4.1-4.9] mo; P < .0001; Fig 1); 1-year OS rates were 50% vs 13%. Among pts preselected for low-intensity therapy, 13/57 in the Q arm and 0/29 in the SC arm underwent any HSCT. Q + SC pooled data also showed a longer mOS (95% CI) in pts with a composite complete remission (CRc) as last recorded response before allo-HSCT vs pts w/o CRc (20.1 [11.7-NA] vs 8.8 [7.0-11.4] mo). Regardless of treatment, mOS was longer with any HSCT vs w/o (Q, 11.9 [10.2-25.1] vs 4.5 [4.1-5.4] mo; SC, 12.7 [6.1-NA] vs 4.0 [2.7-5.0] mo); respective 1-year OS rates were 50% vs 13% and 51% vs 12%. In the Q arm, mOS (95% CI) was longer in pts with a best response of CRc who resumed Q after allo-HSCT (27.1 [18.2-NA] mo) vs pts not resuming Q (5.4 [4.7-11.4] mo; Fig 2). In 48 pts (62%) in the Q arm resuming Q after allo-HSCT, median time from allo-HSCT to Q resumption was 65 d (range, 30-106 d). Four pts (5%) in the Q arm died < 30 d after allo-HSCT. As of 2/22/2018, 46/78 pts in the Q arm (59%) and 9/14 pts in the SC arm (64%) with allo-HSCT w/o additional AML therapy died, mostly due to AML progression (Q, 31 [40%]; SC, 7 [50%]). The frequency of treatment-emergent adverse events (TEAEs) and TEAEs of interest was similar in pts resuming Q after allo-HSCT and in the overall Q population. Conclusion Independent of HSCT, Q improved survival vs SC in pts with FLT3-ITD R/R AML in QuANTUM-R. Q + SC pooled analyses showed longer survival in pts with HSCT vs pts w/o and in pts with CRc before allo-HSCT. Survival in transplanted pts was similar in both arms, indicating that HSCT-eligible pts received transplants appropriately, and the higher HSCT rate in the Q arm was beneficial. In pts preselected for low-intensity SC at study entry, 13 were able to undergo HSCT after Q treatment. Q resumption after HSCT was associated with better survival outcomes and was tolerable. These data illustrate the value of using Q to target the FLT3-ITD mutation as a part of the overall treatment sequence in pts with FLT3-ITD R/R AML.

Introduction Acute myeloid leukemia bearing a RUNX1 gene mutation (RUNX1+ AML) has been proposed as a provisional entity in the 2016 WHO classification. Clinically, it has been associated with inferior response rates and outcome after conventional chemotherapy. Accordingly, RUNX1+ AML is allocated in the unfavorable prognostic category of the 2017 European Leukemia Net classification. Following allogeneic stem cell transplantation (alloSCT), RUNX1 was an unfavorable factor in one study in MDS/secondary AML, while data in de novo AML are scarce. Objectives Here, we present a retrospective study by the EBMT Acute Leukemia Working Party, aiming to elucidate the prognostic value of RUNX1 mutation in patients undergoing alloSCT for AML in first complete remission (CR1). Methods Adults undergoing alloSCT for AML in CR1 from matched related or unrelated donors between 2013 and 2018 with complete information on conventional cytogenetics and RUNX1 mutational status were selected from the EBMT registry. Variables of interest were overall and leukemia-free survival (OS/LFS), GvHD/relapse free survival (GRFS), cumulative relapse incidence (RI), non-relapse mortality (NRM) and GvHD. Log rank test, Gray test and Cox regression models were used. Results 128 RUNX+ and 388 RUNX- patients were identified, >80% of both subgroups presenting as de novo AML. As expected, RUNX1+ patients rarely had co-mutations in NPM1 (6% vs. 26%, p=10−3), and showed a positive correlation with ASXL1 mutations (50% vs. 16%, p=10−4). Cytogenetic categories and other mutations (FLT3-ITD, CEBPA) were equally distributed between the two groups, as were age, donor and graft type, CMV, conditioning and T cell depletion (TCD). Median follow-up was 16.4 (RUNX+) and 19.8 (RUNX-) months. 2y OS/LFS of the entire cohort were 64% [59-69]/57% [52-62], with no difference between RUNX1+ and RUNX1- patients either in univariate or multivariate analysis (2y OS: 67.9% [57.3-78.5] vs. 63.1%v[57.4-68.7]p=0.15; 2y LFS: 57.6% [46.4-68.7] vs. 57% [51.4-62.6], p=0.38], figure 1). RUNX1 mutation neither had any impact among patients with normal karyotype (figure 2). Similarly, no other outcome parameter was influenced by RUNX1 mutational status. Instead, multivariate analysis revealed age and donor type as risk factors for OS, LFS and NRM. Poor cytogenetic was associated with higher RI and inferior LFS/GRFS, in vivo TCD with a lower rate of aGvHD II-IV, cGvHD, and better GRFS. Among patients with available information, FLT3-ITD was an independent risk factor for relapse, LFS and GRFS. RUNX1 did not modify the role of FLT3-ITD. Conclusion Within the limits of a retrospective registry analysis, we could not find a negative influence of RUNX1 mutation on outcome after allogeneic SCT in CR1. Hence, transplantation in CR1 might overcome the unfavorable prognostic value of RUNX1 mutation and can be recommended as consolidation treatment in this entity.

Introduction The WHO 2016 AML-MRC designation applies to AML patients (pts) with a history of myelodysplastic syndrome (MDS) or MDS/myeloproliferative neoplasm, a MDS-related cytogenetic abnormality, or multilineage dysplasia in >50% of ≥2 cell lineages in the absence of NPM1 or biallelic CEBPA mutations. AML-MRC pts typically have a poor prognosis after induction chemotherapy. CPX-351 (Vyxeos®; daunorubicin and cytarabine liposome for injection), a dual-drug liposomal encapsulation of cytarabine [C] and daunorubicin [D] at a synergistic ratio, is approved by the FDA and EMA for the treatment of adults with newly diagnosed therapy-related AML or AML-MRC. A phase 3 study (NCT01696084) in older pts (60-75 y) with newly diagnosed high-risk/secondary AML found that CPX-351 significantly improved median overall survival (OS) vs conventional 7+3, with a comparable safety profile. Objectives An exploratory subgroup analysis of the phase 3 study compared outcomes in pts with AML-MRC who achieved complete remission (CR) or CR with incomplete neutrophil or platelet recovery (CRi). Methods Pts were randomized 1:1 to receive 1-2 induction cycles with CPX-351 (100 units/m2 [C 100 mg/m2 + D 44 mg/m2] as a 90-min infusion on Days 1, 3, and 5 [2nd induction: Days 1 and 3]) or 7+3 (C 100 mg/m2/d continuously for 7 d [2nd induction: 5 d] + D 60 mg/m2 on Days 1-3 [2nd induction: Days 1-2]). Pts achieving CR or CRi could receive up to 2 consolidation cycles with CPX-351 (65 units/m2 [C 65 mg/m2 + D 29 mg/m2] on Days 1 and 3) or 5+2 (as in 2nd induction). Pts could receive hematopoietic cell transplantation (HCT) at the physician's discretion. Results AML-MRC was diagnosed in 246/309 (80%) enrolled pts (123 pts/arm). More AML-MRC pts achieved CR+CRi with CPX-351 (59/123 [48%] vs 40/123 [33%]; odds ratio = 1.83 [95% CI: 1.09-3.09]). Median OS in pts with AML-MRC who achieved CR+CRi was longer with CPX-351 vs 7+3 (Figure 1). The HCT rate in AML-MRC pts with CR+CRi was 54% with CPX-351 vs 43% with 7+3 (relative risk = 1.18 [95% CI: 0.79-1.76]), and OS landmarked from the HCT date was longer with CPX-351 (Figure 2). The safety profile was similar between arms (Table 1), except CPX-351 was associated with longer recovery of neutrophils to ≥500/μL (35 vs 29 d) and platelets to ≥50,000/μL (37 vs 28 d) vs 7+3 in pts who received 1 induction. Conclusions CPX-351 improved median OS overall and OS landmarked from the HCT date vs 7+3 chemotherapy in AML-MRC pts who achieved CR+CRi. The CPX-351 safety profile in this subgroup was consistent with the overall study population and known profile of 7+3.

Background Studies evaluating the impact of 11q23 / KMT2A / MLL translocations on post-transplant outcomes as compared to other intermediate- and adverse-risk cytogenetic characteristics are lacking. Methods We considered 8709 AML patients from the CIBMTR database for inclusion. We evaluated relapse, non-relapse mortality (NRM), disease-free survival (DFS), and overall survival (OS) for patients who were transplanted in CR1 stratified by presence of 11q23 rearrangement, intermediate- or adverse-risk disease. Assignment of cytogenetic risk was performed by hand, based on the 2010 AML-MRC Classification System. A multivariate analysis was performed using a Cox proportional hazards model. Results Characteristics of 3779 patients, selected based on disease and donor type, HSCT in CR1, and availability of data, are shown in Figure 1. Of these, 426 patients had an 11q23-rearrangement, 2384 had intermediate-risk disease, and 969 had adverse-risk disease. Of the patients with 11q23 translocations, 112 (26%) had t(9;11), 62 (15%) had t(11;19), 41 (10%) had t(6;11), 28 (7%) had t(10;11), and 47 (11%) had other partners. No partner was provided in 136 cases (32%). Figures 2 and 3 show the DFS, OS, NRM, and relapse data stratified by cytogenetic category. Based on the multivariate analysis, DFS was lowest in adverse-risk patients (HR for death 1.47, 1.26 for MLL, p < 0.001 and 0.002, respectively). OS was shorter for the MLL- and adverse-risk groups, with HR of 1.32 and 1.45, respectively, and p < 0.001 for both. The HR for NRM was significantly worse for patients with adverse-risk disease (HR 1.17, P=0.05), but not for patients with MLL disease. The HR for relapse in MLL patients was 1.27, p = 0.01 and 1.71, p < 0.001 in adverse risk. Stratification of outcomes by translocation partner did not yield any statistically significant differences between groups, however a descriptive analysis of these data will be presented. Measurable residual disease (MRD) status did not appear to play an independent role in NRM or in OS. However, the multivariate HR for relapse in patients with MRD-positivity was 1.23, p < 0.006. For DFS, the HR for death was 1.13 for MRD-positive patients (p = 0.04). Further data and figures will be presented. Conclusions Patients with MLL-rearranged AML had post-HSCT DFS and OS that were more similar to patients with adverse-risk disease than to those with intermediate-risk characteristics. MLL patients had a higher probability of relapse than intermediate-, but not adverse-risk, patients. NRM was similar among all groups. Both prognostic classification and MRD status independently contributed to the risk of relapse and DFS, but MRD was not an independent factor in predicting NRM or OS. Our study suggests that patients with MLL rearrangements do almost as poorly as those with adverse-risk disease and that novel approaches for this high-risk population represent an urgent unmet clinical need.

BMT CTN 0901 enrolled patients with MDS (N=54) or AML (N=218) (median age of 55 years) who had < 5% marrow blasts by morphology prior to HCT. The most common RIC regimen was FluBu2 (81%) and the most common MAC regimen was FluBu4 (64%). The planned enrollment was 356 patients; however, accrual was stopped early (N=272) due to a marked imbalance of relapse rates. While MAC resulted in longer relapse-free survival (RFS) at 18 months, there were no survival differences between conditioning intensities (Scott BL et al, JCO 2017). Data on all enrolled patients reported to the CIBMTR were used to supplement clinical trial outcomes beyond 18 months. Here we present an updated clinical trial analysis with a median follow up of 50 months. Overall survival (OS) at 4 years was 65% and 49% for MAC and RIC (p=0.02), respectively. In multivariate analysis of overall mortality, the hazard ratio (HR) for death was 1.54 (95% Confidence Interval [CI] 1.07-2.20, p=0.02) for RIC compared to MAC. Other risk factors for mortality were protocol-defined high risk disease (HR 1.77) and age of ≥50 ys (HR 2.20). RFS rates at 4 years was 58% and 34% for MAC and RIC (p<0.001); the corresponding HR was 2.06 (95% CI 1.48-2.85, p<0.001) for RIC vs. MAC. Cumulative incidences of transplant-related mortality and relapse are shown in the figure. Subset analysis of MDS patients (n=54) were directionally similar to the overall analysis but not statistically significant, likely due to limited power. Post-HCT relapse survival among patients with AML was 24% and 26% for MAC and RIC (p=0.87) 3 years after leukemia relapse, respectively. Long term follow up of MAvRIC showed longer survival for MAC recipients. Thus, this trial confirms the importance of intensity for HCT; for patients eligible for either intensity regimen, MAC is the superior choice.

Purpose We previously demonstrated poor outcomes of patients with acute myeloid leukemia (AML) with antecedent myeloproliferative neoplasms (MPN) undergoing allogeneic hematopoietic stem cell transplantation (HCT) (Gupta et al, BBMT, 2019; abstract 140). In particular, we did not find any difference in outcomes of patients who received transplant “in remission” defined as blood and bone marrow blasts < 5% compared to those with “active leukemia” defined as blood and or bone marrow blasts ≥ 5%. We hypothesized that genetic mutations may help in predicting the efficacy of HCT in molecularly defined sub-groups of post MPN AML. In addition, we hypothesized that patients considered “in remission” at the time of HCT may have a molecular profile similar to those who have “active leukemia” at the time of HCT. Patients and Methods Of the 177 patients with post MPN AML identified in the Center for International Blood and Marrow Transplant Research (CIBMTR) database, 95 (54%) had sufficient DNA to perform molecular analysis that consisted of targeted Next-Generation sequencing on 49 genes clinically relevant in hematologic malignancies. Results Of the 95 patients analyzed, 54 were “in remission”, and 41 had “active leukemia”. Adverse risk cytogenetics was seen in 39 patients. The most frequently mutated genes (≥10% patients) in study cohort were: JAK2 (55%); TP53 (23%); ASXL1 (22%); TET2 (19%), SRSF2 (16%); DNMT3A (14%), RUNX1 (14%); CALR (13%); SF3B1 (10%). Cumulative incidence of non-relapse mortality (NRM), relapse, progression-free survival (PFS), and survival in the study cohort at 5-years were 23% (95% CI 15-32), 66% (95% CI 56-75), 11% (95% CI 6-19) and 16% (95% CI 9-25). In a multivariate model, TP53 mutation status was the only factor associated with outcomes of HCT; patients with mutated TP53 mutations had inferior overall survival [RR 1.99 (95% CI 1.14-3.49)] and increased relapse [RR 2.59 (95% CI 1.41-4.74)], and inferior PFS (RR 2.18 (95% CI 1.25-3.82)]. Disease status at HCT was not associated with outcomes. Moreover, there were no differences in the mutational landscape, number of mutations and variant allele frequency (VAF) between patients who underwent HCT “in remission” (n=54) versus those with “active leukemia” (n=41). Conclusions We conclude that there is minimal benefit of HCT in patients with post MPN AML with mutated TP53, and novel strategies are required for these patients. We did not observe any difference in clinical outcomes or mutational profile of patients undergoing transplant “in remission” compared to those with “active leukemia” questioning the clinical benefit of blast reduction strategies in post MPN AML.

Background High-resolution genome-wide SNP-arrays detect large chromosomal aberrations including copy-neutral loss of heterozygosity (CNLOH), which is not captured in conventional cytogenetics. Methods We used SNP-array genotyping data generated by the DISCOVeRY-BMT study to detect chromosomal aberrations in pre-HCT blood samples from 1,974 acute myeloid leukemia (AML) patients. Clinical data and blood samples were available through the Center for International Blood and Marrow Transplant Research. We used Cox proportional hazard model for statistical analyses. Results AML patients in this study received unrelated donor HCT between 2000-2011 at a median age of 48.2 (range=0.6-78.0) years. About 53% of the patients were males, 7.4% had therapy-related disease, and 73% were in complete remission. Chromosomal aberrations were detected in 14.4% of the patients (Figure 1). Most common aberrations in patients with advanced disease at HCT (n=536) were copy losses in chr7 (5.8%), chr5 (4.5%), CNLOH in the following: chr13 (4.7%), chr11 (3.9%), or chr17 (3.5%), and copy gain in chr8 (3.5%). Common aberrations in patients in complete remission (n=1438) were copy-loss in chr7 (1.0%), chr5 (0.7%), and CNLOH in chr9 (0.9%) or in chr17 (0.8%) (Figure 2). Among patients with normal cytogenetics at diagnosis (n=572), 27.5% received HCT in advanced disease (of them 8.3% had chr13-CNLOH). In multivariable models including commonly detected aberrations and adjusted for important factors (footnote Figure 3), chr13-CNLOH was associated with an approximately 3-fold increased risk of leukemia relapse and post-HCT mortality in patients with advanced disease (relapse: HR=2.67, 95% CI=1.67-4.26; mortality: HR=2.79, 95% CI=1.81-4.28, p<0.0001). In contrast, chr17-CNLOH was associated with a statistically significant risk of post-HCT mortality in patients with remission (HR=2.85, 95% CI=1.51-5.37, p=0.001); the association with relapse was not statistically significant (HR=1.84, p=0.21). Having aberrations in ≥3 chromosomes was associated with a statistically significant excess risk of mortality in patients with advanced disease (HR=1.59, 95% CI=1.07-2.36, p=0.02) (Figure 3). In patients with normal cytogenetics at diagnosis, chr13-CNLOH pre-HCT refined risk stratification of patients with advanced disease; compared to patients in remission, the HR for OS in patients with advanced disease in the presence or absence of chr13-CNLOH= 5.7 vs. 1.7, respectively, p<0.0001 for both; HR=3.6, p=0.0004 when comparing presence or absence of chr13-CNLOH in patients with advanced disease. Conclusion Pre-HCT CNLOH in chr13 or chr17 are associated with inferior post-HCT survival in AML; possibly as a consequence of high risk of post-HCT leukemia relapse. CNLOH in chr13 or chr17 may provide new genomic prognostic markers that guide patient risk stratification and possibly therapeutic options.

Introduction In a clinical trial of patients vaccinated after chemotherapy-induced remission with patient-derived AML cells fused with autologous dendritic cells (DC/AML fusions) we demonstrated durable AML-specific immunity. 71% of patients sustained long-term remission from AML at 55-91 months (responders) while those who experienced relapse had recurrence at 2-26 months (non-responders). Objectives We performed genomic analysis of the marrow microenvironment to identify factors associated with durable remission after vaccination. Methods RNA sequencing was done using banked formalin-fixed paraffin embedded (FFPE) marrow at serial time points. Ingenuity Pathways IPA 9.0 was used to define pathways and upstream regulators. To characterize the cellular components, single cell RNA-seq was performed on fresh frozen aspirates with cell cluster annotation performed by Single Cell Wizard software and peripheral blood was used for TCR clonal diversity by NGS. Results Heatmaps depict significant (p-value ≤0.01 and fold change ≥2) differential gene expression both pre- and post-vaccination in responders compared to non-responders. Prior to vaccination there was inhibition of TGF-β in responders. Notable upregulated pathways in responders after vaccination (p value <0.01) were related to immune activation including NO and ROS in macrophages, IL-2, IL-15, IL-6, IL-7, IL17A, and B cell activating factor. TGF-β was also downregulated in responders post-vaccination. Key upstream transcription regulators of response pre-vaccination included inhibition of IL-4, IGF-1, VEGF, PDGFBB and TGF-β1 while post-vaccination included activation of CD3, CD28, tumor-necrosis factor and IL-1B. We also identified a highly connected cohesive network of immune system activation including these key regulators as well as GBP2, CD226 and toll-like receptor genes. In responders, increased cellular heterogeneity was present pre-vaccine, and increases in T and NK populations post-vaccine. Also, temporal changes in expression of TCR clonotypes showed a sustained increase in TCR diversity after vaccination. Conclusion In a cohort of AML patients treated with DC/AML fusions, we found distinct gene signatures amongst patients with long term remisison vs those with early relapse. The transcriptomes indicate a microenvironment in which TGF-β-mediated immunosuppression is inhibited after chemotherapy and in which interleukins, B cell and macrophage signaling are activated are associated with vaccine response, and have potential for combinatorial therapeutic strategies. T cell costimulation and activation are key components of response with NK cells and DC activation playing a supporting role. A durable oligoclonal T cell expansion suggests vaccine is capable of enhancing the T cell repertroie which could provide a unique opportunity to identify target antigens by TCR-epitope pairing.

Mutations resulting in increased PI3K signaling confer increased risk of B cell lymphoma, recurrent infections, poor viral control, and autoimmunity. Allogeneic HCT is curative, but the optimal approach to HCT for these patients is still evolving as experience grows. Herein, we present the clinical outcomes of 27 patients transplanted for activating PI3K mutations. Required approvals were obtained by contributing centers. Nineteen PIK3CD patients and 8 PIK3R1 patients received 22 and 10 HCTs respectively, at median age 12 years (range 2-66) at time of first HCT. Significant pre-HCT comorbidities were common, Fig 1. Conditioning platforms varied in intensity; most included serotherapy (84%). With median survivor follow up of 26.3 months from first HCT (range 2.4-71.8), overall survival was 85% (3 infectious deaths and 1 due to organ toxicity); graft failure (GF)-free survival (GFFS) was estimated at 80% at 1 year but declined to 68% by 2 years, Fig 2. Neutrophil engraftment occurred at median 15.5 days (range 11-33) with autologous recovery in 1 patient at day +14 and another deceased prior to engraftment. Four engrafted patients are alive and well despite mixed donor chimerism (<95%). Nine (33%) patients required 24 total subsequent unplanned donor cell infusions (DCI) for mixed chimerism (n=15), poor graft function (n=2), infection (n=1), GF (n=3), malignancy relapse (n=1), or to improve immune reconstitution (n=2). Grade 2-4 acute graft versus host disease (GVHD) occurred in 27% patients, with Grade 3-4 acute GVHD and limited self-resolved chronic GVHD in 1 (4%). Organ toxicities and infectious complications are summarized in Fig 3; CMV infection requiring treatment occurred in 11 (41%) patients with disease in 3 (11%). Serotherapy use, conditioning intensity, and degree of donor human leukocyte antigen (HLA) match were not significantly associated with cumulative incidence of GF or subsequent unplanned DCI (with death as a competing risk), while mTOR inhibitor use post-HCT was associated with subsequent unplanned DCI for graft augmentation or GF (p=0.006). An mTOR inhibitor was used within the year following 12 HCTs, for primary GVHD prophylaxis (n=7) or to treat disease manifestations or GVHD; loss of chimerism and/or need for subsequent DCI occurred in 9 of these. Patients with activating PI3K mutations are at overall significant risk of GF or need for subsequent DCI. While we found no association with GF or DCI for discrete elements such as serotherapy use, donor source, or conditioning intensity, small numbers preclude conclusions regarding the optimal HCT approach, and more HCTs with longer follow up are needed to refine these preliminary findings. Post-HCT mTOR inhibitor exposure was associated with need for subsequent unplanned DCI, perhaps by providing a survival advantage to host lymphocytes, and may thus be detrimental in these patients during the early post-HCT timeframe.

Introduction Hemophagocytic lymphohistiocytosis (HLH) is a syndrome of excessive immune system activation largely driven by high levels of interferon gamma. The clinical HLH presentation can have considerable overlap with other inflammatory conditions. We present a cohort of patients with therapy refractory HLH referred to our center who were found to have simultaneous presentation of complement mediated thrombotic microangiopathy (TMA). Figure 1 contains the diagnostic criteria for both HLH and TMA. Objectives • Recognize presenting symptoms of HLH • Recognize presenting symptoms of TMA • Recognize that HLH and TMA can present concurrently Methods Patients of any age with frontline therapy refractory hemophagocytic lymphohistiocytosis (HLH) who were treated at our institution from January 2012 through December 2018 were identified by retrospective chart review after Institutional Review Board (IRB) approval. All patients were prospectively screened for TMA as part of clinical care as previously described. Patient demographics, disease characteristics, HLH and TMA targeted therapy, complications, intervention and laboratory assessment were captured from the electronic medical record. Results Sixteen of 23 patients with HLH met high risk TMA criteria (Figure 2). Renal failure requiring renal replacement therapy, severe hypertension, serositis and gastrointestinal bleeding were documented only in patients with HLH who had concomitant complement mediated TMA. Patients with HLH and without TMA required ventilatory support mainly due to CNS symptoms, such as status epilepticus, while those with HLH and TMA had respiratory failure predominantly associated with pulmonary hypertension, a known presentation of pulmonary TMA (Figure 2). Ten patients received eculizumab for complement mediated TMA management while being treated for HLH. All patients who received the complement blocker eculizumab, in addition to the interferon gamma blocker emapalumab, had complete resolution of their TMA and survived (Figure 2). Conclusion Our observations suggest co-activation of both interferon and complement pathways as a potential culprit in the evolution of thrombotic microangiopathy in patients with inflammatory disorders like HLH and may offer novel therapeutic approaches for these critically ill patients. TMA should be considered in children with HLH and multi-organ failure, as early institution of a brief course of complement blocking therapy in addition to HLH targeted therapy may improve clinical outcomes in these patients (Figure 3).

Introduction Allogeneic hematopoietic stem cell transplantation (HSCT) is curative therapy for a variety of primary immunodeficiency disorders (PIDs). Patients with PID have high rates of pulmonary disease from infections and immune-mediated lung damage, and post-HSCT pulmonary complications account for considerable morbidity and mortality. It is unknown whether pre-HSCT pulmonary disease places PID patients at higher risk for post-HSCT complications. We hypothesize that PID patients with pre-HSCT pulmonary disease have higher risk of transplant-related mortality (TRM), more pulmonary complications, and lower overall survival (OS) compared to PID patients without pre-HSCT pulmonary disease. Objectives The primary aim of this study is to compare TRM in PID patients with and without pre-HSCT pulmonary disease. Secondary aims of the study are to compare OS, incidence of non-infectious and infectious pulmonary disease post-HSCT, ICU transfer for any cause, incidence of acute or chronic GVHD, and immune reconstitution. Methods This is a single center, retrospective, chart review. All pediatric patients (ages 0-18 years) who underwent allogeneic HSCT for a diagnosis of a PID at Boston Children's Hospital from January 2007-June 2018 (n=115) were included in the analysis. We defined non-infectious pulmonary disease as either functional lung impairment based on pulmonary function tests (PFTs), radiographic evidence of pulmonary disease, or documentation of a specific pulmonary diagnosis. We defined infectious pulmonary disease as either viral, bacterial, fungal or other infection in conjunction with positive radiographic and/or bronch/BAL findings. Results Within our cohort, the most common PIDs were SCID (n=28), DOCK8 deficiency (n=16), Wiskott-Aldrich syndrome (n=15), and chronic granulomatous disease (n=9). Of the 115 patients, 54 (47%) had a pre-HSCT diagnosis of pulmonary disease. Pulmonary infection (n=30) and bronchiectasis (n=11) were the most common diagnoses (Fig. 1A). Compared to patients without pre-HSCT lung disease, patients with pre-HSCT lung disease had higher numbers of pulmonary complications after HSCT, such as the need for non-invasive or mechanical ventilation (23 of 54 patients [43%] vs. 12 of 61 patients [20%]) (Fig. 1B). Amongst patients with pre-HSCT lung disease, 1-year survival was 81% compared with 92% in patients without pre-HSCT lung disease. Overall survival was 76% at a median follow up time of 1.6 years versus 89% at a median follow up of 2.7 years in patients with and without pre-HSCT lung disease, respectively (censored logrank p = 0.05) (Fig. 2). Conclusion Within our single center study, PID patients with pre-HSCT lung disease had a trend towards more post-HSCT lung complications and lower OS compared to PID patients without pre-HSCT lung disease. Pulmonary risk factors should be carefully considered in PID patients prior to HSCT.

Background Our centre performed the first ex-vivo T-lymphocyte depleted (TCD) haploidentical allograft in 1987 in a child with PID. Different TCD methods have been used including CAMPATH-1M TCD marrow (n=34), CD34+ selected marrow (CD34-S, n=34), CD3/CD19 TCD PBSC (n=7), and CD3-TCRαβ/CD19 TCD PBSC (αβ-TCD, n=30)). This study aimed to examine transplant outcomes according to different methods in children with PID. Methods Between Jan 1987 and Mar 2019, 105 PID patients underwent first TCD allograft and were included in the study. Thirteen patients who received gene-modified add-back T cells were excluded. The main outcomes of interest were OS, EFS and graft failure. An event was defined as death, graft failure or second procedures for slipping chimerism. Other endpoints were neutrophil engraftment, GvHD, immune reconstitution and donor chimerism. Log-rank test was used to compare OS and EFS according to TCD methods. Cumulative incidences (CI) graft failure was calculated using a competing risk analysis, considering death as a competing event. Multilevel mixed effects modelling was performed for the longitudinal analysis of immune reconstitution and CAMPATH-1M was used as a reference group. Results Donors were haploidentical donor (n=84, 83%), MUD/MFD (n=9, 9%) and MMFD/MMUD (n=8, 8%). 94% received conditioning. Thirteen (12%) patients had aGvHD, of whom 3 had grade III-IV. None had cGvHD. The cause of death changed from predominantly infection (18/25, 72%) in CAMPATH-IM and CD34-S to non-infectious causes in all 6 deaths in CD3/CD19 and αβ-TCD. Analysis by TCD methods revealed a 5-year OS of 58% (95% CI, 40-73%) for CAMPATH-1M, 68% (49-81%) for CD34-S, 69% (22-91%) for CD3/CD19 TCD and 83% (61-93%) for (αβ-TCD (p=0.24). Age was a significant predictor of OS for non-SCID PID (p=0.02). The corresponding EFS was 46% (29-62%) for CAMPATH-1M, 47% (30-62%) for CD34-S, 69% (95% CI, 22-91%) for CD3/CD19 TCD and 83% (61-93%) for αβ-TCD (p=0.04) (Fig 1B) The CI of graft failure reduced significantly, from 29% (14-61%) for CAMPATH-1M, to 19% (95%CI, 8-45%) for CD34-S, 17% (2-18%) for CD3/CD19 TCD and none had graft failure in αβ-TCD (p=0.002) (Figure 1C). CD4+ lymphocytes were significantly higher in CD34-S at months 4 (p=0.02), 5 (p=0.01), and 6 (p=0.006) post-HCT and at month 4 (p=0.04) post-HCT in αβ-TCD when compared to CAMPATH-1M (Fig 1D). The median donor myeloid chimerism at last follow-up was higher in newer TCD; 100% (range, 0-100%) for αβ-TCD, 93% (0-100%) for CD3/CD 19 depletion, 6% (0 -49%) for CD34-S, 20% (0-100%) for CAMPATH-1M (p<0.001). There was no significant difference in donor T-lymphocyte chimerism between TCD. Conclusions Outcomes after CD3+TCR αβ/CD19+ depletion are superior to previously used TCD methods. In an experienced centre it is a safe alternative procedure and enables a wide spectrum of PID to be transplanted. The result has led to evolution of donor hierarchy in our centre.

Background With >36 years of data collected, PIDTC prospective (6901) and retrospective (6902) natural history studies provide an unprecedented opportunity to study hematopoietic cell transplantation (HCT) outcomes for SCID over time. Methods Patients in this 6901/6902 analysis met PIDTC diagnostic criteria for SCID and underwent HCT. Categorial variables were analyzed between decades (a) 1982-89 (b) 1990-99 (c) 2000-09 (d) 2010-18 using the chi-square test. Continuous outcomes were compared using the Kruskal-Wallis test. Kaplan-Meier method was used for estimates of overall survival (OS). Results 896 children with typical (n=742) and atypical (n=154) SCID requiring HCT between 1982-2018 were enrolled. Diagnosis of SCID for reasons other than family history or newborn screening was common (60%) in early cohorts (a-c) dropping to 30% in cohort (d). Distribution of SCID genotypes changed over time (Fig 1), and novel/unknown genotypes also decreased from (a) 53% to (d) 13%, p<0.001. Due to the rise of gene therapy, there were fewer patients with adenosine deaminase deficient SCID receiving HCT in cohort (d) compared to prior decades. Median age at HCT has decreased, from 193 to 112 days of life, p<0.001. While marrow was the most common stem cell source in the 1980’s (95%), currently peripheral blood stem cells and cord blood are given to 46% of patients (p<0.001). The increased use of unrelated donors parallels a decrease in mismatched related donors (Fig 2a; p<0.001). In univariate analysis, use of conditioning has increased from 12% in the 1980’s to 65% currently (Fig 2b, p<0.001); contributing factors that will be evaluated in upcoming multivariate analyses include genotype, typical vs atypical SCID, and donor type. Day +100 cumulative incidence of grades 3-4 acute graft-versus-host disease (GVHD) has dropped, from 9.6% to 5.9% (p=0.02). OS has markedly improved in the last decade (Fig 3a), with 1- and 5-year OS being 90 and 87%, respectively (p=0.02). In particular, gains have been made in OS when using alternative donors (p=0.02) (Fig 3c,d) while OS for matched sibling HCT remains high (Fig 3b). Conclusion This longitudinal data set, spanning the implementation of NBS and advancements in diagnostics and supportive care, highlights improved OS after HCT for SCID, including alternative donors. Exploration of decreased toxicity approaches that maintain high OS and engraftment are warranted.

Background Hematopoietic stem cell transplant (HSCT) in IMDs prevents progression of fatal neurodegeneration. Cord blood as a preferred donor source is disadvantaged by delayed engraftment and graft failure due to low CD34+ cell numbers. Higher cell doses are correlated with improved outcomes and cross-correction of disease. MGTA-456 is a novel cell therapy which provides a high dose of CD34+ cells and is being investigated in an ongoing Phase 2 clinical study to assess the safety and long-term outcomes in IMD patients. Methods A Phase 2 trial (NCT03406962) is enrolling 12 patients <16 yo with a diagnosis of cerebral onset adrenoleukodystrophy (cALD), mucopolysaccharidosis type IH (MPS I), metachromatic leukodystrophy (MLD) or globoid cell leukodystrophy (GLD). Patients receive a myeloablative conditioning regimen. Patients are followed for 1 year after transplant and eligible patients may enroll in a long-term follow-up study (NCT04008849). Results 2 patients with cALD and 3 with MPS I, have been treated per protocol and with follow-up up to 1 year. In patients with cALD, resolution of gadolinium enhancement on MRI occurred by one month and was sustained throughout the follow-up period (Fig 1). Loes scores, a measure of disease severity, and Neurologic Function Scores remained stable over this time period consistent with a halt in disease progression. Neurocognitive function remains stable to date with no deterioration in Wechsler IQ Scale and Vineland Adaptive Behavior Scales. MPS I patients showed normalization of blood a-L-iduronidase and excretion of MPS I-specific urinary glycosoaminoglycans declined from baseline over the period of observation (Fig 2). Neurologic development and function scores as measured by Bayley Infant Development and Vineland Scales are being collected. Patients received a median CD34+ dose of 110 × 106 cells/kg and TNC dose of 26 × 107 cells/kg with a median duration of neutropenia of 1 day (range 0-9) compared to a median of 8 days for historical controls. Myeloid chimerism was ≥98% donor in evaluable patients by day +14 and immune reconstitution of CD4 and CD8 T cell subsets were comparable or better than historical controls. Two steroid-responsive episodes of skin-only aGVHD were observed and no cGVHD developed. Data on additional patients receiving cryopreserved MGTA-456 will be reported. Conclusions Treatment with MGTA-456 in patients with IMDs shows early, robust engraftment and immune reconstitution, features that are highly associated with improved clinical outcomes. Disease-specific outcomes up to one-year post-transplant in MPS I and cALD patients show crucial, durable stabilization in biochemical, MRI and clinical/neurodevelopmental assessments. In summary, the high CD34+ cell doses delivered by MGTA-456 shows compelling potential to rapidly and durably improve outcomes in IMD patients.

Patients with severe sickle cell disease (SCD) experience organ damage, poor quality of life and premature mortality. Hematopoietic cell transplantation (HCT) is curative in SCD. Excellent outcomes have been achieved in young children with HLA matched related donors (MRD), with low rates of acute and chronic graft versus host disease (GVHD). Adolescent patients with SCD receiving MRD HCT and patients with SCD receiving matched unrelated donor (MUD) HCT, remain at highest risk for GVHD and resulting morbidity and mortality. Building on the encouraging experience with abatacept (CTLA-4Ig) for GVHD prophylaxis in URD HCT for hematologic malignancies, we launched an early phase, multicenter trial (n=25) through the Sickle Cell Transplant Advocacy and Research Alliance (STAR) of abatacept in pediatric SCD patients at high risk for GVHD. Patients with severe sickle cell disease were eligible if they were receiving a MRD HCT and either they or their donor were at least 10 years or if they were receiving a MUD HCT. Twenty-five patients have been enrolled to date with 23 patients completing conditioning and HCT. All patients received reduced intensity conditioning (RIC) with distal alemtuzumab, fludarabine, thiotepa and melphalan and marrow grafts. Four doses of abatacept (10 mg/kg/dose IV on days -1,+5,+14, +28) were added to standard GVHD prophylaxis using tacrolimus and methotrexate. Abatacept dosing was extended to eight doses for the last 11 patients (+60, +90, +120, +150). Tacrolimus was tapered at day 180 if no GHVD was present. The median age at HCT was 11 (3-21 years); 13 received a MRD HCT and 10 a MUD HCT. All patients engrafted with neutrophils at a median of 21 days (range 13-26) and platelets at a median of 23 days (13-55). As of 10/1/2019, median follow up is 804 days (range 46 – 1139). The mean myeloid and T-cell chimerisms at one (n=17) and two years (n=10) were 97.9 +/- 4.9 and 94.2 +/- 10.0 and 96.6 +/- 8.9 and 97.5 +/-4.3, respectively. Three evaluable patients at 2 years did not have chimerism studies sent. There have been 4 cases of EBV viremia (1 PTLD) and 6 cases of CMV viremia (1 disease). Grades 3-4 acute GVHD was seen in one patient and chronic/overlap in 5 patients with two patients currently receiving sytemic immune suppression. There has been no cerebral hemorrhages but one patient developed non-infectious encephalopathy of unclear etiology. The overall and sickle cell free survival at 2 years post HCT in evaluable patients (N=13) is 100% and 100%, respectively. The preliminary results in first 23 patients enrolled to date suggest that incorporating abatacept into GVHD prophylaxis is a safe and efficacious approach to improving outcomes in HCT for pediatric SCD.

Background Persons with multiple sclerosis (MS) are sometimes treated with high-dose immune suppressive or cytotoxic drugs and an autotransplant. Therapy-related mortality (TRM) has decreased to <2% because of less intensive pretransplant regimens and better subject selection. Trials have reported a greater proportion of subjects with no evidence of active disease in subjects receiving high-dose therapy and an autotransplant compared with controls. Recent autotransplants use less intensive pretransplant therapy with seemingly similar efficacy and fewer adverse events. We previously reported autotransplants for several haematologic neoplasms can be done in an outpatient setting using refrigerated blood cells. We now report data from 739 subjects with MS receiving autotransplants in an outpatient setting using refrigerated blood cells. Methods 739 consecutive patients with MS were autografted in a single center using non-frozen peripheral blood stem cells, on an outpatient basis and conditioning with cyclophosphamide and rituximab (Figure 1). The protocol was registered in ClinicalTrials.gov identifier NCT02674217. Eligibility criteria included all the following: Karnofsky performance score >70%, extended Disability Status Scale (EDSS ≤8 in the 2 w pretransplant), CNS magnetic resonance image (MRI) ≤3 mo pretransplant, no prior bone marrow toxic drugs, normal heart, liver, lung and kidney function, ≥6 mo since exposure to immune suppressive drugs. Results 495 females and 244 males were included; median age was 47 years. 310 patients presented with relapsing remitting MS (RRMS), 273 with secondary progressive (SPMS) and 156 with primary progressive (PPMS). All procedures were started on an outpatient basis and only 31 persons needed to be admitted to the hospital during the procedure. In order to obtain at least 1 × 106/Kg viable CD34 cells, one to three apheresis were performed. Total number of viable CD34+ cells infused ranged between 1 and 37.83 × 106/Kg. Patients recovered above 0.5 × 109/L absolute granulocytes on day 8 whereas platelet recovery above 20 × 109/L on day 4. Seven individuals required red blood cells and eight needed platelet transfusions. There were two transplant related deaths and the 30-month overall survival of the patients is 99.7%. All patients had a significant drop in the EDSS 12 and 18-mo after the transplant. The response rate (either drop or stabilization of the EDSS score) were at 12 and 18 months respectively: 85 and 88% for RRMS; 75 and 84% for PPMS and 79 and 85% for SPMS (Figure 2). Conclusions Changes in the EDSS score consonant with neurological improvement were observed in persons with all types of MS after HSCT employing the “Mexican method”.

Background Deep and durable hematologic remissions following RA-ASCT are associated with improved organ function and extended overall survival in AL amyloidosis. While depth of hematologic response by standard criteria are important, this study assessed additional factors that influence RR and time to RR. Methods All patients (pts) with AL and renal involvement undergoing RA-ASCT at Memorial Sloan Kettering Cancer Center between January 1, 2007 to December 31, 2016 were included. Pts with follow up <12 months (mos) post RA-ASCT and hemodialysis prior to RA-ASCT were excluded. Melphalan dose was assigned based on age, cardiac involvement and renal compromise. Hematologic and organ response was assessed at 3 (early) and 12 mos post RA-ASCT. Consolidation with bortezomib and dexamethasone was offered to pts with less than complete response (CR) at 3 mos. Logistic regression models were used to assess association with RR (Palladini Blood 2014) by 12 mos. Covariates for adjustment in multivariate models were chosen based on univariate analyses and clinical relevance. Results Sixty-four patients with renal AL meeting the inclusion criteria were identified; 3 died within 12 mos post RA-ASCT were excluded. Median age was 61 years (44-73), 51% female and 34% had cardiac involvement. Median 24 hr proteinuria pre ASCT was 5014 mg/day (2632-7514) and eGFR 68 ml/min/1.73 m2 (44-91). Renal amyloid stage I:II:III was 33%:52%:15%. Mayo cardiac stage (2004) I:II:III was 28%:61%:11% and revised Mayo stage (2012) I:II:III:IV was 13%:57%:21%:8%. Median BM plasma cells pre ASCT was 9% (2-14%). 46% received treatment prior to ASCT. Melphalan dose (mg/m2) 200:140:100 was 44%:43%:11% and 46% received consolidation. Hematologic response early post ASCT was CR 44%, very good partial (VGPR) 29% and partial response (PR) 20%. dFLC <1mg/dL was achieved in 63%. RR by 12 mos was achieved in 32 pts (53%); median time to RR was 5.8 mos (5.1 – 11.3). Amongst renal responders, 50% achieved CR, 53% MRD negative, and 78% dFLC <1mg/dL early post RA-ASCT. In pts with dFLC <1mg/dL early post RA-ASCT, 66% achieved RR. By univariate analysis, early post ASCT dFLC <1mg/dL (OR 3.00, 95% CI 1.01-8.93, p = 0.048) and VGPR (7.80, 1.69-36.06, p = 0.009), at 12 mos dFLC <1mg/dL (7.20, 2.14-24.21, p = 0.001) and CR (10.27, 1.14-92.26, p = 0.038) were associated with RR. By multivariate analysis, dFLC <1mg/dL early post ASCT (OR 4.52, 95% CI 1.26-16.24, p = 0.021) was associated with RR when adjusted for renal amyloid stage and Mayo cardiac stage (2004). Conclusion In this single center study, we found that RA-ASCT results in RR in more than half (53%) of pts at 1 year. Achieving dFLC <1mg/dL early post ASCT is prognostic of RR independent of renal stage. Early MRD-negativity did not predict RR. Our study suggests maximal suppression of the pathologic light chains is important for organ recovery. Larger studies using dFLC <1mg/dL as an end point are warranted.

Introduction Disease relapse is the most common cause of therapy failure in NHL patients undergoing RIC alloHCT. Whether increasing TBI dose from 2Gy to 4Gy in RIC-platform can provide improved disease control without increasing non-relapse mortality (NRM) is not known. Using the CIBMTR database we evaluated the outcomes of NHL patients receiving RIC alloHCT with either Flu/2Gy TBI vs. Flu/4Gy TBI. Methods In the CIBMTR registry, 413 adult NHL patients underwent a first alloHCT using either a matched related or unrelated donor between 2008-2017, utilizing a RIC regimen with either Flu/2Gy TBI (n=349) or Flu/4Gy TBI (n=64). Graft-versus-host disease (GVHD) prophylaxis was calcineurin inhibitor-based. The primary endpoint was overall survival (OS). Secondary endpoints included acute (a) and chronic (c) GVHD, NRM, relapse/progression and progression-free survival (PFS). Probabilities of OS and PFS were calculated using the Kaplan-Meier estimator. Multivariable regression analysis was performed for GVHD, relapse/progression, NRM, PFS, and OS. Covariates with a p<0.05 were considered significant. Results Baseline characteristics are shown in Figure 1. The Flu/2Gy TBI cohort had significantly fewer patients with KPS ≥90 and significantly more patients with HCT-CI ≥3. On multivariate analysis (Figure 2), the two conditioning cohorts were not significantly different in terms of risk of grade 3-4 aGVHD or cGVHD. Compared to Flu/2Gy TBI, the Flu/4Gy TBI conditioning was associated with a significantly higher risk of NRM (HR 1.79, 95%CI=1.11-2.89, p = 0.02), and inferior OS (HR 1.51, 95%CI=1.03-2.23, p = 0.03). No significant differences were seen in the risk of relapse/progression (HR 0.78, 95%CI=0.47-1.29, p = 0.33) or PFS (HR 1.09, 95%CI=0.78-1.54, p=0.61) between the two regimens. Comparing Flu/2Gy TBI vs. Flu/4Gy TBI cohorts the 5-year adjusted outcomes were; NRM (28% vs. 47%; p = 0.005), relapse/progression (35% vs. 29%; p = 0.28), PFS (37% vs. 24%; p = 0.03) and OS (51% vs. 31%; p = 0.001), respectively (Figure 3). Rate of graft failure at day100 in similar order was 0.6% vs. 1.6% (p = 0.54), respectively. Relapse was the most common cause of death in both cohorts. Conclusions In NHL patients undergoing Flu/TBI-based conditioning, augmenting TBI dose from 2Gy to 4Gy is associated with higher NRM and inferior OS, without any significant benefit in terms of disease control. 2Gy is optimal dose in the RIC Flu/TBI platform for lymphomas.

Allogeneic hematopoietic stem cell transplantation (HCT), defined by its underlying graft-versus-multiple myeloma (MM) effect, is a potentially curative approach for high-risk pts. However, relapse remains the main cause of treatment failure. Predictors for post-relapse survival after HCT are not well characterized, and we hypothesized that the time to relapse and the patterns of relapse are prognostic for post-relapse survival. Methods: All pts with MM who underwent a CD34+-selected alloHCT at Memorial Sloan Kettering Cancer Center on protocol (NCT 01131169/01119066) or off study from 01/2010 to 12/2017 and progressed after alloHCT were included. Pts were conditioned with busulfan (0.8 mg/kg x 10), melphalan (70 mg/m2 × 2), and fludarabine (25mg/m2 × 5). The K-M method was used to estimate OS post-HCT relapse. Significant predictors were considered in a Cox model. Results: Of the 115 transplanted pts, 60 (52%) relapsed and were analyzed. The median age was 56 yrs (35- 66), 62% male and 82% with high-risk cytogenetics (CG). KPS was ≥ 80 in 100% and ≥ 90 in 52%. Pts received a median of 4 lines of therapy pre-HCT (1-10), with 88% achieving at least a partial response (PR) prior to alloHCT. All pts had received at least one autoHCT. Of the 38% who received a planned or preemptive post-HCT therapy, 13 received DLI and 10 other strategies. Relapse was characterized as very early (<6 mo, 28%), early (6-24 mo, 50%), or late (>24 mo, 22%). At relapse, 27% presented with extramedullary disease (EMD). Median post-relapse OS was significantly different for pts who relapsed very early, early, or late; 4 mo, 17 mo, and 72 mo, respectively (p = 0.002). Age ≥55, relapse with EMD, lower KPS,

Introduction AutoHCT recipients age 70+ who attended a GA-guided MDC had favorable outcomes with 0% 1-yr non-relapse mortality (NRM). We now examine vulnerabilities and outcomes in all older autoHCT candidates undergoing MDC, including HCT utilization rates. Methods An HCT MD, advanced practice provider, dietician, PT/OT, social worker, ID MD, and geriatric MD comprised the MDC. All pts received GA-informed optimization. Pts were classified by MDC recommendation: ‘Decline’ if unlikely to realize benefits of HCT, ‘Defer’ if significant optimization necessary before HCT, and ‘Proceed’ if autoHCT could proceed without delay. We compared autoHCT recipients (AR, n=62) to non-autoHCT recipients (non-AR, n=29). Results Differences by ‘Defer’ (n=25) vs ‘Proceed’ (n=61) included more impairments (median HCT-CI 3 vs 2, p = 0.01; more frail 4-meter walk times, 48% v 14%, p = 0.001) and more lymphoma (52% vs 30%, p = 0.01). AutoHCT occurred in 49/61 (80%) ‘Proceed’, 13/25 (52%) ‘Defer’, and 0/5 ‘Decline’. AutoHCT recipients (AR) vs non-AR did not differ by age but had fewer impairments as shown in table 1. AR vs non-AR was enriched for plasma cell dyscrasia (PCD, 77% vs 48%, p = 0.008). AR had longer OS compared to non-AR (HR 0.26, 95% CI 0.11-0.57, p = 0.0012). No AR pts experienced NRM. Within AR, MDC ‘Defer’ pts had a longer time from MDC to HCT compared to ‘Proceed’ pts (median 41 vs 20 days, p = 0.001). Although ‘Defer’ AR pts at DC had more frail 4-meter walk times (38% vs 10%, p = 0.03) and worse Mental Health Index-5 (p = 0.03), length of hospitalization, readmission rate, skilled nursing facility admission, and mortality did not differ relative to the ‘Proceed’ AR group (see figure). 1-yr OS for ‘Defer’ AR (n=13) vs ‘Defer’ Non-AR (n=12) was 92% vs 60% (p = 0.07); 3-yr OS 55% vs 48% (p = 0.8). Conclusion A GA guided MDC deferred around 27% of older autoHCT candidates likely due to significant impairments, and half went on to receive HCT. Pts pursuing autoHCT after MDC achieved excellent outcomes, even among those initially deferred. Employing GA earlier in the disease trajectory to inform early referral to a GA-guided MDC should be tested to increase utilization and improve outcomes in older autoHCT candidates.

Introduction Unprecedented depth of response is obtained with regimens incorporating monoclonal antibodies in the management of newly diagnosed multiple myeloma (NDMM). The incremental benefit of AHCT in this setting is being reconsidered but can be appraised by systematic assessment of disease burden pre- and post-AHCT utilizing NGS MRD. Methods We performed a multi-center clinical trial with Dara-KRd induction, AHCT (Melphalan conditioning), and post-AHCT, MRD response-adapted Dara-KRd consolidation. Primary endpoint was MRD <10−5 using the NGS platform clonoSEQ. Secondary endpoints included the achievement of MRD <10−6. MRD assessment happened in parallel with IWMG response assessment after completion of 4 cycles of Dara-KRd, 60-80 days after AHCT, and during each of up to two additional 4-cycle blocks of Dara-KRd consolidation. Patients with MRD <10−5 in two consecutive phases of therapy discontinue treatment without maintenance and were actively monitored for the resurgence of MRD or clinical relapse. Results To date, 82 patients have been enrolled, and 77 (94%) have disease trackable by NGS. Of those, 54 have completed post-induction and 29 post-AHCT assessment. The median age of the cohort is 61 years (range 36-79), with 46% females, 31% with high-risk FISH abnormality [t(4;14) or t(14;16) or del17p] and 20.3% ISS III disease. All patients obtained objective responses with the rate of best responses ≥VGPR of 93% after induction and 100% after AHCT and rate of sCR of 46% and 79%, respectively. MRD responses are displayed in Figure 1a. Patients with MRD <10−5 were 31% post-induction and 76% post-AHCT. Patients with MRD<10−6 were 24% post-induction and 48% post-AHCT. All 7 patients with MRD <10−6 post-induction had confirmed MRD <10−6 after AHCT. Of the remaining 22 patients, 20 (90%) had a reduction in MRD burden with AHCT (figure 1b). The median reduction in disease burden was 0.94 Log10 (range -1.3 to 3.3), and 10/22 patients (45%) had a reduction in more than one Log10. Of the two patients without a reduction in MRD with AHCT, one had minimal change from 1.3 to 1.8 × 10−5 and one patient had increased from 2.3 × 10−2 post-induction (while in sCR) to 4.6 × 10−1 post-AHCT corresponding to morphologically positive marrow and clinical disease progression. Conclusion While a subset of patients with NDMM receiving Dara-KRD achieved MRD negativity after four cycles of induction, subsequent AHCT significantly increased the number of patients obtaining this status. Despite a quadruple induction regimen utilizing what are considered the most effective anti-myeloma drugs, AHCT resulted in a substantial reduction of disease burden in a majority of patients and should remain part of current therapy for fit, NDMM patients. Future trial designs with the intent to defer AHCT for patients may need to demonstrate high rates of MRD negativity to show benefit ultimately.

Introduction Salvage autologous stem cell transplant (sASCT) may be associated with improved survival in multiple myeloma (MM) and is an option for patients with long time to relapse (TTR) after initial Autologous transplant (ASCT1). Due to potential mobilization failure after maintenance therapy (MT), adequate peripheral blood progenitor cells (PBPC) are frequently collected for later sASCT. Objectives Evaluate utilization of stored PBPC in MM. Provide a cost analysis of collecting and storing PBPC for sASCT. Determine utilization of sASCT in those eligible who had a prolonged TTR. Methods We included MM patients who received ASCT1 at our institution between 2009 and 2016. Extra day of collection (EDC) was defined as additional collection of ≥2 × 106 cells/kg PBPC for a second transplant. TTR was calculated as months from ASCT1 to relapse. Eligibility for sASCT was defined as TTR ≥18 months without MT or ≥36 months on MT. Patients who underwent tandem transplant (tASCT) (defined as two transplants within 6 months) were excluded from relapse analysis. Patients with inadequate relapse data defined as death or lost to follow up (LTFU) at ≤36 months prior to relapse, or ≤6 months follow up after relapse were excluded. Costs were obtained from the institution charge master. Descriptive statistics were calculated for the final cost benefit analysis. Results 154 MM patients received ASCT1 from 2009-2016. Median age was 59 years. Median follow up was 151 months. Median TTR was 20 months. 58% were male. Ethnicity was 76% Caucasian and 24% Black. 62% had IgG MM subtype, 23% IgA MM and 15% other. Majority (95%) were European Cooperative Oncology Group (ECOG) 0 or 1. 13% had high risk cytogenetics. 53% received MT. 137(89%) had PBSC stored for sASCT. 3 were used for sASCT and 10 for tASCT. 124 (90.5%) did not use stored cells. 54 (39%) met collection targets for two transplants in single collection while 83 (%) required ≥1 EDC with cumulative 108 EDC. (Figure 1) Cumulative cost savings for plerixafor use, cell collection, processing, analysis and storage was estimated at $2,252,244. (Figure 2). Total 60 doses of plerixafor were used for EDC. (31 patients with 1EDC, 10 with 2EDC and 3 with 3 EDC). 97 patients were studied for relapses. 54 (77.5%) relapsed. 5 deaths and 8 LTFU occurred within ≤6 months of relapse. 42 patients were further analyzed. 25 of them met criteria for sASCT based on TTR and only 2 received sASCT.1 patient with sASCT did not have MT data and was excluded. (Figure 3) Conclusion Only 25% (24/98) patients who underwent ASCT1 were eligible for sASCT. Only 2% (2/98) received sASCT. Collecting and storing PBPC for sASCT is associated with high cost, but overall utilization at our institution is low. Our findings should be compared against practices at other institutions to help develop guidelines for selection of subset of MM where this practice may be of higher value.

Objectives CAR T cell therapy has generally been limited to inpatient treatment. Infusion and management of CAR T cell therapy in the outpatient setting may lead to wider use in nonuniversity centers and improved access. We report on patients (pts) with R/R large B cell NHL treated with liso-cel in the outpatient setting in TRANSCEND NHL 001 (NCT02631044) and two phase 2 studies (≥3rd-line therapy: OUTREACH, NCT03744676; 2nd-line TNE: PILOT, NCT03483103). Methods Eligible pts had R/R large B cell NHL (TRANSCEND/OUTREACH: ≥2 lines of prior therapy and ECOG PS ≤1; PILOT: 1 line of prior therapy and deemed TNE for autologous hematopoietic stem cell transplant based on ECOG PS, organ function, or age). After lymphodepletion with fludarabine/cyclophosphamide, liso-cel was administered. All studies allowed outpatient treatment at nonuniversity (OUTREACH) or university and nonuniversity medical centers (TRANSCEND/PILOT), with hospitalization at the first sign of fever or neurological events (NEs) per management guidelines. Results At data cutoff, 37 pts across studies received liso-cel on study Day 1 and were monitored as outpatients, including pts aged ≥65 years (n = 15) and those with SPD ≥50 cm2 (n = 10) or LDH ≥500 UL (n = 2). Results are shown in the Table. Sixteen pts had any grade cytokine release syndrome (CRS) and 12 had any grade NEs (19 pts had CRS and/or NEs). Only 2 pts had had grade 3 or 4 CRS or NEs, which were reversible. Three pts received tocilizumab and corticosteroids for CRS and/or NEs; none received tocilizumab alone. Overall, 59% of pts (n = 22/37) required hospitalization at any time; all were from TRANSCEND or OUTREACH. Of 37 pts, 3 (8%) were admitted on study Day 3 or earlier (all for CRS) and 1 (3%) required ICU-level care (length of stay, 3 days). Median time to hospitalization post treatment was 5 (range, 2‒22) days; median length of stay was 6 (range, 2‒23) days. Overall, 41% of pts (15/37), including all 5 pts from PILOT, did not require hospitalization in the first 29 days post liso-cel infusion. Across all studies, most pts achieved an objective response (76%), including complete responses. Conclusions A subset of pts with R/R large B cell NHL were successfully treated with liso-cel and monitored for CAR T cell–related toxicity in the outpatient setting, including elderly pts and pts with high tumor burden. Incidences of severe CRS, NEs, and early hospitalization were low; 41% of pts did not require hospitalization in the first month post treatment. Most pts achieved an objective response.

Background Patients(pts) who undergo autologous stem cell transplant (ASCT) for multiple myeloma (MM) are routinely assessed at Day-100 using serum monoclonal protein (M-spike), free light chains(FLC), and urine studies. Limited data are available to assess the prognostic value of an increasing M-spike or FLC from immediately before ASCT to Day-100 post-ASCT. We evaluated this in our cohort of patients. Methods We retrospectively reviewed 1070 pts with MM at Mayo Clinic, Rochester, who had their first ASCT between 2000 and 2016. We stratified pts according to a rise in M-spike by at least >0.1g/dl from immediately before ASCT to Day-100 post-ASCT (cohort 1) compared to those who did not (cohort 2). We also stratified pts according to a rise in involved FLC by at least >5mg/dl (cohort 3) compared to those who did not (cohort 4). Survival analysis for progression free survival (PFS) and overall survival (OS) was performed using the Kaplan Meier method. Results 46 pts were found to have a rise in M-spike by at least >0.1g/dl between pre-ASCT and Day-100 (cohort 1) and 825 pts were found to have a declining or <=0.1g/dl increase in M-spike (cohort 2). 25 pts were found to have an increase in involved FLC by at least >5mg/dl (cohort 3) and 1045 pts were found to have a decrease or a < = 5mg/dl increase in involved FLC (cohort 4). The median PFS was significantly shorter in cohort 1 compared to cohort 2 (median PFS of 10 months vs. 27 months, respectively P<0.0001). Cohort 1 patients also had an inferior OS, compared to cohort 2 (median OS of 34 months vs. 79 months, respectively, P<0.0001). Similarly the median PFS and OS were found to be inferior in cohort 3 compared to cohort 4 (median PFS of 4 months and 28 months, respectively p<0.0001 and median OS of 11 months and 82 months, respectively, p<0.0001) Conclusions An increase of M-spike by at least >0.1 g/dl and involved FLC by at least >5mg/dl from pre-ASCT to Day-100 is predictive of inferior PFS and OS in pts with MM. These patients would benefit from closer surveillance post Day-100.

Introduction Haploidentical (HAPLO) stem cell transplantation (SCT) has been shown to be associated with lower incidence of severe acute graft-versus-host (aGVHD) compared with 8/8 HLA-matched unrelated donor (MUD) SCT. However, no information is available regarding aGVHD characteristics comparing these 2 donor sources. Methods We aimed to compare aGVHD characteristics and non-relapse mortality (NRM) in adult patients with grade 2-4 aGVHD after HAPLO (N=758) or MUD (N=2586) SCT that were reported to the Center for International Blood and Marrow Transplant Research (Table 1). GVHD prophylaxis included a calcineurin inhibitor (CNI), mycophenolate mofetil (MMF), and post-transplant cyclophosphamide (PTCy) in HAPLOs, and CNI with MMF or methotrexate in MUDs. Results The 6-m cumulative incidence (CumInc) of grade 2-4 aGVHD (35 vs 45%, <0.001) was significantly lower in HAPLOs, with a lower prevalence (28% vs 39%, p=0.001) of grade 3-4. Median (range) time from SCT to aGVHD was 35 (5-218) and 33 days (7-374) in the 2 groups, with comparable proportions (4 and 6%, p=0.2) diagnosed after day 100. Skin (66 vs 67%, p=0.8) and lower (L) gastrointestinal (GI) (76 vs 78%, p=0.4) involvement did not differ between the 2 groups; however, upper GI (46 vs 53%, p=0.04) and liver (11 vs 15%, p=0.06) involvement were less frequent in HAPLOs (Table 2). The total number and organ severity were reduced in HAPLOs with lower prevalence (26 vs 35%, p=0.008) of >2 organs involved, stage 4 skin (1 vs 5%, p=0.01), and stage 3-4 liver (4 vs 7%, p=0.04) or LGI (14% vs 21%, p=0.01). The 2-yrs and overall CumInc of NRM since aGVHD was 18% (95% CI 14-23) and 19% (95% CI 14-24) in HAPLOs; and 30% (95% CI 27-33) and 51% (95% CI 30-68) in MUDs. In multivariate analysis, grade 2-4 aGVHD in HAPLOs was associated with significantly lower NRM, but this effect was limited to donor-recipient pairs [CumInc: 15 vs 30%, HR=0.5, p=0.002] that were not sex-mismatched (female to male). In sex-mismatched pairs [CumInc: 30% vs 26%, HR=1.2, p=0.6], NRM was comparable between HAPLOs and MUDs. Subset analyses in recipients (N=718) ≥60 years also showed HAPLOs (N=92) to have a lower 6-m CumInc of grade 2-4 (29 vs 44%, p<0.001) and lower prevalence (25 vs 41%, p=0.003) of grade 3-4. Median (range) time to aGVHD diagnosis was relatively delayed [38 (11-200) vs 34 (7-374) days, p=0.04] in HAPLOs. Organ involvement did not differ between the 2 groups; however, stage 4 was less frequent in HAPLOs. This difference was most pronounced for LGI (3 vs 13%, p=0.004). In multivariate analysis, grade 2-4 aGVHD in HAPLOs was associated with significantly lower [CumInc: 11 vs 33%, HR=0.2, p=0.001] NRM only in the absence of sex-mismatch. Conclusions Compared with 8/8 HLA-matched unrelated SCT with standard GVHD prophylaxis, aGVHD tends to be less severe and associated with lower NRM after HAPLO SCT with PTCy GVHD prophylaxis. This effect is more pronounced in recipients ≥60 years of age.

Introduction DPP-IV is a homodimeric type II transmembrane receptor identical to the leucocyte surface antigen CD26. The interaction of CD26/DPP-IV on T cells with its ligand caveolin-1 on antigen-presenting cells (APC) enhances T cell activation, proliferation, and cytokine production. It also upregulates CD86 on APC resulting in co-stimulation. In a xenograft mouse model, reducing CD26 expression by a monoclonal antibody prevented aGVHD and preserved graft-versus-tumor effects (Hatano et al. BJH 2013). In a pilot trial testing sitagliptin for enhancement of cord blood engraftment, we observed a lower than expected rate of aGVHD (Farag et al. Stem Cell Dev 2013) and hypothesized that CD26/DPP-IV inhibition with sitagliptin may prevent aGVHD following allogeneic PBSC transplants. Objective Conduct a prospective phase II trial to test if sitagliptin can reduce the rate of grade II-IV aGVHD by day +100 following myeloablative allogeneic PBSC transplants (ClinicalTrials.gov, NCT02683525). The primary endpoint was grade II-IV aGVHD by day +100. Methods Patients received thiotepa (15 mg/kg) and cyclophosphamide (120 mg/kg), followed by G-CSF mobilized PBSC from matched related (6/6 HLA match) or unrelated (10/10 HLA match) donors. Sitagliptin (600 mg q12 hours) was given on days -1 to day +14, together with tacrolimus and sirolimus through day +100, then tapering until day +180. In a Simon two-stage design, <8 of a total of 36 patients developing grade II-IV aGVHD by day +100 would demonstrate a reduction of aGVHD from a historical rate of 30% to <15% with 80% power and type 1 error = 0.1. Results Thirty-six (17 M; 19 F) patients of median age 46 (20-59) years were enrolled. Patients had AML (n=19), ALL (n=9), MDS (n=4), and CML (n=4). Transplants were from matched related (13 patients) or unrelated (23 patients) donors. Acute GVHD occurred in 2 of 36 patients by day +100 (1 grade II; 1 grade IV), for cumulative incidences of 5.6% (95% confidence interval [CI], 0% - 13.2%) and 2.8% (95% CI, 0% - 8.2%) for grade II-IV aGVHD and grade III-IV aGVHD at day +100, respectively (Fig 1). An additional patient developed late-onset grade II aGVHD at day +140 following abrupt stopping of immunosuppression on day +100 for low donor chimerism in absence of relapse. Significant toxicities though day +30 included grades III-IV mucositis (n=15), acute kidney injury (4 with high tacrolimus levels, 1 sepsis), viral reactivation/infection (4 CMV, 2 BK, 1 EBV, 1 HHV6), and passenger lymphocyte syndrome (n=2). Non-relapse mortality was 0% at 12 months. With median follow-up of 699 (118-1312) days, the relapse-free survival was 77.2% (95% CI, 63.2% - 91.1%) and overall survival was 94.3% (95% CI, 86.7% - 100%) at 12 months (Fig 2a-b). Conclusion CD26/DPP-IV inhibition with sitagliptin is well-tolerated and significantly reduced the rate of aGVHD by day +100 in patients undergoing myeloablative allogeneic PBSCT transplantation.

NLRP6 (NOD-like receptor family pyrin domain containing 6) is an important inflammasome component and is highly expressed in intestinal epithelial and immune cells. NLRP6 mediated inflammasome activation plays a critical role in response to intestinal infection and preventing dysbiosis of gut microbiota. However, we recently found that NLRP6 plays a pathogenic role in GVHD. Interestingly, we also found that activated T cells increased NLRP6 expression, but the T cell autonomous role of NLRP6 in regulating T cell responses is unknown. Because NLRP6 is an important regulator of GVH responses, we tested the hypothesis that NLRP6 deficiency in donor T cells would ameliorate GVHD. To test our hypothesis, we first performed a detailed phenotypic analysis of various T cell subsets and activation markers in naïve NLRP6−/− and wild-type (WT) B6 animals and found a similar distribution of naïve, memory, effector and regulatory T cells. In order to examine whether the absence of NLRP6 in donors affects GVHD, WT-BALB/c animals were lethally irradiated and transplanted on day 0 with 5 × 106 bone marrow and 1.0 × 106 splenic CD90+T cells from either syngeneic WT-BALB/c, allogeneic MHC-mismatched WT-B6 or NLRP6−/− animals. Contrary to our hypothesis, the recipients receiving donor T cells from NLRP6−/− animals showed a significantly worse survival compared to allogeneic WT-B6 animals (p<0.05). GVHD mortality and severity were also increased in an MHC mismatched haploidentical B6 into F1model (p<0.05). In contrast, GVHD severity and mortality were similar in an MHC matched multiple minor antigen mismatched B6 into C3H.sw model. We hypothesized that GVHD severity and mortality was similar in the B6 into C3H.sw model because NLRP6 regulates CD4+ and CD8+ T cell responses, differently. In order to test this, we transplanted C3H.sw recipients as above except we infused either 1 × 106 CD4+ or CD8+ T cells from B6-WT or NLRP6−/− animals. GVHD severity and mortality (P<0.05) were enhanced only when NLRP6−/− CD4+ T cells transplanted. To explore how NLRP6 effects T cell responses independent of inflammasome activation, we tested naïve T cell proliferation in vitro after allogeneic or non-specific TCR stimulation by anti-CD3 and CD28 antibody and found that NLRP6−/− CD4+ but not CD8+ T cells proliferated more than WT-B6 CD4+ or CD8+ T cells, respectively, following either stimulus. In addition, activated NLRP6−/− CD4+ T cells showed less apoptosis than WT. Absence of NLRP6 in T cells promotes naïve CD4 T cells into Th1 cell differentiation. These data suggest that NLRP6 regulates CD4+ T cell mediated immune responses and that NLRP6 in donor T cells is critical for controlling CD4+ T cell mediated GVHD.

α‐defensin‐5 (HD5) accounts for 70% of the bactericidal peptide activity of Paneth cells and regulates gut microbial homeostasis. We previously reported that the G allele of rs4415345, a single nucleotide polymorphism (SNP) in the gene for HD5, was associated with approximately 40% lower risk of grade II-IV aGVHD (PMID: 30004111). Considering the role of microbiota disruptions in aGVHD, we hypothesized that rs4415345G may have a specific gut microbiota signature. First, we analyzed linked host-microbiota genetic data from 248 subjects in the Human Microbiome Project to test our hypothesis in healthy individuals. With the assumption that many microbial taxa have no association with a given SNP, we used 10-fold cross-validated Lasso regression as a test of association between rs4415345 and taxon relative abundance. A single taxon (Odoribacter splanchnicus) was retained in 99% of the simulations, always with a positive association with the G allele. Next, we evaluated whether O. splanchnicus is associated with less aGVHD in 226 patients with a 16S rRNA gene-sequenced research stool sample collected between days -30 and -6 of allo-HCT. O. splanchnicus was present in 65 (29%) patients. The univariate hazard ratio (HR) of O. splanchnicus detectability for grade II-IV aGVHD was 0.70, with a 95% confidence interval (95%CI) of 0.44-1.10 (P = 0.12). Multivariable analysis, adjusted for anti-anaerobic antibiotic exposure between days -7 and +14 yielded a similar result (HR: 0.70, 95%CI: 0.44-1.11, P = 0.13). In summary, the G allele of rs4610776 is associated with the higher relative abundance of O. splanchnicus in healthy individuals and less aGVHD in allo-HCT recipients. In an independent dataset, O. splanchnicus presence was associated with a 30% reduction in risk for aGVHD, though this association did not reach statistical significance. O. splanchnicus is a weakly butyrogenic bacterium also capable of producing indole, and depleted in patients with IBD. Both butyrate and indole can limit aGVHD. In addition, O. splanchnicus correlates with less TNFα production in response to inflammatory stimuli. We propose that O. splanchnicus may reduce aGVHD risk in rs4415345G individuals.

Graft-versus-host disease (GVHD) remains a significant cause of morbidity and mortality in patients receiving allogeneic hematopoietic stem cell transplants (aHSCTs). Pre-HSCT chemoradiation results in the death of dividing cells and release of endogenous danger signals. These molecules drive the activation of antigen presenting cells (APCs) and the differentiation of allo-reactive donor T cells. We identified a role for Stimulator of Interferon Genes (STING), an innate immune sensor, in GVHD using pre-clinical MHC-matched unrelated donor (MUD) HSCT models. Here we show that STING rapidly promotes donor CD8+ T cell activation and recipient APC death early after aHSCT. Post MUD HSCT, we previously found a >2x reduction in IFNβ, TNFα and IL-6 mRNA in STING−/− vs WT recipient colonic mRNA expression (48 hrs) as well as decreased weight loss, GVHD scores and skin pathology (6 wks) vs WT. Chimeric studies showed that STING loss in non-hematopoietic cells induced this effect. Conversely, we recently found a single early (D100), dose of the STING agonist DMXAA increased GVHD scores and lethality in WT, but not STING−/−, recipients. Thus, the activation of this pathway can promote GVHD following MUD HSCT. Furthermore, mice homozygous for a murine homologue of a human allele associated with diminished STING activity (STINGHAQ/HAQ) also exhibited reduced GVHD after MUD HSCT, suggesting potential clinical importance. Interestingly, our findings that STING deficiency ameliorates GVHD in MUD HSCT (Bader CS, et al, in revision Sci. Transl. Med.) contrasts reported observations that STING deficiency can exacerbate GVHD after MHC-mismatched (MMUD) HSCT. Since CD4+ and CD8+ T cells are central in MMUD and MUD GVHD, respectively, we transplanted MMUD BALB/c BM + CD8+ T cells into B6-WT and STING−/− mice and notably, found that STING−/- recipients now developed reduced GVHD clinical scores, skin pathology and frequencies of activated T cells 8 wks post-HSCT vs WT. We also found that STING−/− mice had greater numbers of recipient splenic CD11b+CD11c+ APCs and these cells expressed reduced MHC I protein 1 day after MMUD B6 into BALB/c aHSCT vs WT (Fig. A, B). Moreover, STING−/− recipient spleens contained lower numbers of donor CD8+ T cells producing IFNγ and TNFα (Fig. C), supporting the hypothesis that STING contributes to early activation of donor CD8+ T cells and loss of recipient APCs. Next, to identify if reduced host MHC II+ APCs affected donor CD4+ T cell activation, B6-Nur77GFP T cells were used to explicitly monitor T cell receptor signaling. Indeed, STING−/− spleens had greater numbers of donor Nur77GFP CD4+ T cells expressing GFP, CD69 and IFNγ 6 days post-HSCT (Fig. D). Overall, these data provide a mechanism by which STING could promote CD8+ T cell-mediated GVHD yet diminish CD4+-mediated GVHD. Ongoing studies are exploring the use of STING agonists to augment GVL following aHSCT.

Background Anti-thymocyte globulin-free dCBT is associated with high rates of acute GVHD (aGVHD) especially involving the GI tract. Investigation of augmented aGVHD prophylaxis is indicated. Methods We are investigating the addition of a single dose of tocilizumab 8 mg/kg (day -1) to CSA/ MMF (starting day -3) aGVHD prophylaxis in adult patients (pts) with hematologic malignancies undergoing dCBT with intermediate intensity Cy 50/ Flu 150/ Thio 10/ TBI 400 conditioning (NCT03434730). Outcomes of the first 26 patients (with survivor follow-up of at least 100 days) were analyzed & compared to 54 dCBT historic controls transplanted with identical conditioning & only CSA/ MMF. All patients received letermovir if CMV seropositive. Results 26 pts [median age 47 years (range 26-60), median weight 82 kg (range 59-125), 11 AML, 8 ALL, 2 MPAL, 3 MDS/ CML, 2 NHL, median age-adjusted HCT-CI 2.5 (range 0-6)] received dCB grafts with a median CD34+ cell dose of 1.5 × 105/kg/unit (range 0.23-5.94) & median unit-recipient 8-allele HLA-match of 5/8 (range 3-6). One patient had graft failure in the setting of low graft viability/ disseminated adenovirus & a second pt had early transplant-related mortality (TRM) with incomplete count recovery. The remaining 24 patients engrafted at a median of 24 days (range 18-40) for a cumulative incidence of 92% (95%CI:67-98). 88% (95%CI:59-97) of pts engrafted platelets at a median of 43 days (range 32-78). Of evaluable pts, 68% & 83% had 100% single-unit donor chimerism in the blood at days 30 & 100, respectively; single-unit dominance was observed in the remaining cases. Six of 26 pts (23%) developed pre-engraftment syndrome (PES) at a median of 12.5 days, & only 3 of these (12%) required systemic corticosteroids. Additionally, 16 pts (62%) had no neutropenic fever prior to engraftment. The day 100 cumulative incidence of grade II-III & grade III aGVHD was 62% (95%CI:39-78) & 4% (95%CI:<1-17), respectively (Figure). While 16 pts developed upper GI aGVHD, 6 had stage 0+ (stool volume < 500 cc) lower GI aGVHD & the incidence of more severe lower GI involvement was very low [stage 1 (n = 2), stage 3 (n = 1) & no stage 4]. With a median survivor follow-up of 314 days, 1-year overall survival (OS) is 77% (95%:61-98). Five of 26 patients have died to date (1 graft failure, 2 organ failure, 2 infection) with no GVHD-associated TRM. Comparison with historic controls (Table) showed tocilizumab-based dCBT is associated with similar engraftment, reduced PES, reduced severe aGVHD & similar OS to CSA/ MMF alone dCBT. Conclusions Tocilizumab-based dCBT appears safe with reduced PES & grade III-IV aGVHD. While many pts developed upper GI aGVHD, the marked reduction in lower GI aGVHD supports preferential reduction of T-cell homing to the lower gut. Investigation of this approach including aGVHD treatment responses, immune recovery & correlative biomarker/ microbiota studies are ongoing.

Background We previously demonstrated comparable immune reconstitution after HLA-matched sibling (MSD), HLA-matched unrelated (MUD), and HLA-haploidentical (haplo) bone marrow transplantation (BMT) with posttransplant cyclophosphamide (PTCy). Here, we use machine learning algorithms to identify predictive immunologic variables, providing insight for future translational interventions. Methods We assessed 11 pre-transplant factors, 22 immune subsets, and 7 serum markers at day 28 and day 56 post-BMT in 70 patients who engrafted after myeloablative conditioned (MAC) haplo BMT with PTCy (50mg/kg on days +3 and +4), mycophenolate mofetil (days +5-35), and tacrolimus (days +5-180) and 74 patients who engrafted after MAC HLA-matched BMT with PTCy (50mg/kg on days +3 and +4). Classification and regression tree machine based learning with nodal selection was applied to detect risk groups for four outcomes: overall survival (OS), progression-free survival (PFS), acute graft-versus-host disease (aGVHD), and chronic graft-versus-host disease (cGVHD). Results When compared with cytomegalovirus (CMV) seronegative recipients, CMV seropositive recipients had higher CD3+ T cells, effector memory (EM) CD4+ T cells, T effector memory expressing RA (TEMRA) CD4+ T cells, CD8+ T cells, EM CD8+ T cells, and TEMRA CD8+ T cells at early time points after both BMT platforms (not shown). Pre-BMT disease status and natural killer (NK) cell count at day 28 emerged as predictive variables for survival. Two-year OS and PFS for patients in complete remission (CR) with NK cells >54 cells/µL were 98% and 89%, in CR with NK cells <54 cells/µL were 57% and 53%, and for those with active disease were 48% and 40%, respectively (Fig. 1A, Fig. 1B, p-values both < 0.001). CD4+ T cell counts and CXCL-9 levels at day 28 were predictive for aGVHD. The one-year cumulative incidence (CuI) of grade II-IV aGVHD was 12% in patients with CD4+ T cell counts <55 cells/µL. For patients with CD4+ T cells >55 cells/µL, the CuI of aGVHD was 18% in those with CXCL-9 <7.2 in ng/ml and 75% in those with CXCL-9 ≥ 7.2 ng/ml (Fig. 1C, p-value < 0.001). Lastly, reg3α level at day 56 predicted cGVHD; the 1-year CuI of cGVHD was 60% (95% CI: 29-90%) in recipients with levels ≥ 286 ng/ml compared with 14% with reg3α <286 ng/ml (Fig. 1D, p-value < 0.001). Conclusion Machine learning using immune cellular and soluble markers can be successfully applied to identify risk factors for BMT outcomes. High CD4+ counts and CXCL-9 levels at day 28 predicted aGVHD and high reg3α at day 56 predicted cGVHD. Both CXCL-9 and reg3α represent potential therapeutic targets. Finally, disease status and NK cell counts at day 28 predicted OS and PFS. These data support research to determine what influences NK cell recovery as well as to investigate NK cell therapy for recipients with active disease or for those with poor early NK cell reconstitution.

Chronic GvHD affects about 30-60% of patients after allogeneic hematopoietic cell transplant (allo-HCT), and a majority exhibit ocular manifestations with significant impact on quality of life. In severe cases, oGvHD can lead to permanent vision loss. There is no FDA-approved treatment to date as the current options are limited and mostly ineffective. Forehead application of 1% progesterone gel (Pro-ocularTM) increases resistance to painful corneal stimuli in animal studies, as well as improves tear film in human volunteer studies. While the precise mechanism of action is unclear, a cranial nerve V1 mediated neuro-pathway has been implicated. We conducted a prospective, randomized, placebo-controlled, double-masked phase II clinical trial of adult patients with chronic oGvHD who underwent allo-HCT at Dana-Farber/Brigham and Women's Hospital sponsored by Glia LLC. All subjects had moderate to severe eye symptoms, were free of other major ocular comorbidities, and remained on previous ocular treatments and systemic immune suppression through the trial. Subjects were randomized in a 2:1 treatment to placebo fashion. The trial drug or placebo was self-applied to the forehead in the home settings twice a day for a total of 10 weeks. All subjects kept daily diaries to document their medication use and symptoms, and were evaluated in the clinic at baseline, 2 weeks, 6 weeks and 10 weeks. The endpoints of the study were patient reported ocular symptoms and physician recorded ocular signs such as ocular surface staining and tear film. Safety assessments included blood tests of hematology, chemistry and endocrine panels. A total of 33 patients were enrolled. 21 out of 22 patients in the active arm and all 11 patients in the placebo arm completed the trial. Only 1 was lost to follow-up. Treatment was well tolerated and compliance was high. There was no severe adverse event in the active arm. There was a significant reduction of inferior (-1.38 vs -0.19, p=0.0055) and central cornea (-1.18 vs -0.10, p=0.0010) fluorescein staining in the active arm compared to placebo at 10 weeks (Figure 1). No difference was noted in superior cornea staining. The Symptom Assessment Questionnaire iN Dry Eye (SANDE) quantifies ocular symptoms on a visual analog scale filled by the patients. Strikingly, the reduction of global score (-25.87 vs +1.04, p = 0.0006), symptom frequency (-30.72 vs -2.18, p = 0.0002) and severity (-19.82 vs +1.64, p=0.0048) were all significantly superior in the active arm at 10 weeks (Figure 2). Other endpoints are currently under analysis. All subjects in the placebo arm have chosen to receive active treatment in the ongoing open label study. In conclusion, forehead application of 1% progesterone gel significantly improved ocular symptoms and signs in just 10 weeks. It appears to be a safe and effective novel treatment for chronic oGvHD. A larger randomized phase III trial to confirm the results is warranted.

Background Patients with rel/ref AML have very few options, despite novel targeted therapies (e.g. venetoclax, IDH inhibitors), other than allogeneic hematopoietic cell transplantation (HCT) to achieve a durable CR. In addition, patients ≥55 years of age cannot tolerate multiple cycles of intensive therapy trying to achieve a CR prior to transplant. Older patients are also unable to tolerate more intense myeloablative conditioning. The SIERRA trial is a prospective, randomized, phase 3 trial addressing the unmet need to improve access of older patients with active, rel/ref AML to HCT, the only potentially curative therapy available to them. Methods Patients are randomized (1:1) to receive Iomab-B (day-12) followed by FLU/TBI and HCT, or to conventional care (CC). CC patients may receive investigator's choice of salvage therapy, including newly approved targeted agents and may proceed to standard HCT if they achieve CR. If patients do not achieve CR, the study allows cross-over (CO) to Iomab-B/HCT. Results Preliminary data are available from 75 (50%) patients (Table 1). Patients (median age 64) were heavily pretreated with 85% failing ≥2 induction therapies and 33% failing targeted therapies. All patients transplanted after Iomab-B engrafted. After randomization, 82% (31/38) patients in the CC arm failed salvage therapy, despite 32% (12/38) receiving targeted therapy. 73% (8/11) of the patients receiving venetoclax with HMA or LDAC did not achieve remission. Of the 31 patients, 9 were not CO transplant candidates. Of the 22 patients that crossed over for transplant, majority (91%, N=20) were able to receive Iomab-B/HCT, despite a median of 35% BM blasts. The most common reason preventing CO to transplant was disease progression. CC patients showed increased incidence of Grade 3 febrile neutropenia, compared to Iomab-B patients prior to CO or HCT (34% vs 8%). Iomab-B administration was generally well-tolerated with 1 Grade 3 and no Grade 4, infusion reaction. There were no Iomab-B-related deaths and 100-day non-relapse TRM was 3% on the Iomab-B arm (Table 1). Conclusion Despite advanced age, active disease and heavily pre-treated patients, 100% (31/31) of all patients in the Iomab-B arm who received treatment with Iomab-B were able to undergo HCT and successfully engrafted. In comparison, only 18% (7/38) of the patients on the CC arm achieved remission and underwent conventional HCT despite a high percentage receiving recently approved agents and targeted therapies. Overall, including patients crossing over to receive Iomab-B/HCT after failure of salvage therapy in the CC arm, 68% (51/75) of all patients enrolled in the trial underwent HCT after Iomab-B. These preliminary results represent significant improvements in the current rates of transplantation in this patient population. This SIERRA trial is currently enrolling (www.sierratrial.com or clinicaltrials.gov NCT02665065).

Allogeneic Hematopoietic Stem Cell Transplantation (allo-HSCT) significantly reduces the rate of relapse in Acute Myeloid Leukemia (AML) but comes at the cost of significant Treatment Related Mortality (TRM). Despite the reduction in relapse overall, it remains common, especially in high risk groups. There is increasing evidence that many cases of relapse are driven by residual AML that can be detected with minimal residual disease techniques. A large proportion of the morbidity that prevents most patients accessing allo-HSCT is due to nonspecific toxic conditioning agents which are required to remove recipient Hematopoietic Stem and Progenitor Cells (HSPC) allowing for successful engraftment. We propose that a targeted conditioning agent against AML and HSPC can facilitate a reduction in relapse and TRM in allo-HSCT. CD300f is expressed evenly across HSPC subtypes. CD300f has equivalent transcription and protein expression as CD33 on AML. We have developed an anti-CD300f antibody which efficiently internalises into target cells. Harnessing this ability to internalise, we have generated a highly potent anti-CD300f Antibody Drug Conjugate (ADC) with a pyrrolobenzodiazepine (PBD) warhead which selectively depletes AML cell lines in vitro. Colony Forming Units were inhibited when cultured in the presence of the ADC. The anti-CD300f ADC has a rapid onset of action with >99% cytotoxicity of the AML cell line HL-60 occurring within 24 hours. The ADC synergises with fludarabine, making it a natural combination to use in a minimal toxicity conditioning regimen. The growth of subcutaneous tumours in mice engrafted with the AML cell line U937 was significantly reduced by a single dose of our ADC (at 150 μg/kg or 300 μg/kg). In a bone marrow engraftment model with HL-60, our ADC prolongs the survival at a single dose of 300 μg/kg (Fig 1A). In a humanised mouse model, 85% of all CD34+ cells and 90% of the primitive CD34+ CD38- CD90+ population were depleted with a single injection of our ADC at 300 μg/kg (Fig1 B). There was significant reduction in primary AML cells with a single injection of our ADC at 300 μg/kg (Fig 1C). This proof of principle work establishes an anti-CD300f ADC as an attractive potential therapeutic to develop which will reduce the toxicity of allo-HSCT and the relapse rate afterwards. We have demonstrated that CD300f is expressed across HSPC and AML. Our antibody to CD300f is efficiently internalised and capable of delivering a PBD warhead. Our ADC synergised with fludarabine and has a rapid onset of action. We have demonstrated that our ADC prolongs survival in mouse cell line models as well as depletes both HSPC and primary AML in vivo. A targeted conditioning agent has the potential to reduce relapse risk and toxicity of allo-HSCT in AML.

Introduction Rabbit anti-thymocyte globulin (ATG) is used in allogeneic hematopoietic cell transplantation (HCT) to prevent graft-versus-host-disease (GvHD) and graft failure (GF). Its main and unpredictable toxicity includes poor immune reconstitution associated with increased relapse, viral reactivations and subsequently higher mortality. Early (<100 days) CD4+ T-cell reconstitution has been associated with improved overall survival1. Based on pharmacokinetic and pharmacodynamic (PKPD) studies we developed an individualized ATG dosing regimen aiming for optimal CD4+ T-cell recovery while maintaining a strong anti-GvHD effect. Methods We performed an open label, phase II historically controlled non-randomized prospective clinical trial investigating individualized versus fixed dosing of ATG (Thymoglobulin). Individualized dosing was based on the results from a previous validated population PKPD model1,2 with cumulative doses varying between 2 to 10 mg/kg starting based on weight, recipient lymphocyte counts before first dose of ATG and stem cell source, starting day -9. Primary endpoint was successful CD4+ T-cell reconstitution (CD4+ >50/mm3 at 2 consecutive timepoints within 100 days after HCT1) in patients alive without relapse or graft failure <100 days. Secondary endpoints were overall survival, event free survival, GvHD and graft failure. Results were compared to a previously described historical cohort receiving a fixed cumulative ATG dose of 10mg/kg starting day -51. Minimal follow-up was 1 year. Power was based on 20% effect size with α=0.05 and β=0.10. Multivariate Cox proportional hazard and Fine-Gray competing risks were used. The study was registered in the Dutch Trial Registry (NTR4960). Results We included 58 patients between 2015-2018, and compared to 112 historical controls (table 1). CD4+ reconstitution was significantly better in the individualized dosing group: 83% versus 54% in the fixed dosing group; HR 2.4 (95% CI 1.6-3.6), p<0.0001 (Figure 1a). The individualized group showed a trend towards improved survival (81% vs. 66%; HR 0.54 [95% CI 0.27-1.07], p = 0.078; Figure 1b) and event free survival (79% vs 61%: HR 0.52 [95% CI 0.27-1.01], p = 0.052). No differences were seen in grade 3-4 acute GvHD (7% vs.6%; HR 1.44 [95% CI 0.47-4.44], p = 0.53), and graft failure (5% vs 5%; HR 1.61 [95% CI 0.38-6.83], p = 0.52). Conclusions Individualized ATG dosing significantly increases the chance of rapid immune reconstitution without affecting the incidence of aGvHD and graft failure. Together, this is an important step towards predictable CD4+ T cell reconstitution and improved survival changes.

Background The combination of Clofarabine + Fludarabine + Busulfan (CloFluBu) was found to have synergistic anti-leukemic activity against ALL and AML blasts in vitro (Andersson et al: BBMT 2011). As TBI induces significant late effects in childhood ALL, and AML patients have high relapse rates, we hypothesized that CloFluBu may be a potential alternative to TBI in ALL, and could add anti-leukemic activity in AML. Within the “Dutch COG HCT Working Group” we prospectively studied the outcomes of a CloFluBu-conditioning regimen for lymphoblastic and myeloid malignancies. Methods Patients from the 2 pediatric HCT programs (LUMC and UMC Utrecht/Princess Máxima Center for Pediatric Oncology) in the Netherlands with a lymphoblastic or myeloid malignancy receiving their first HCT, between August 2011 and April 2019, were included. Over 4 days, Clofarabine 30mg/m2 was given in 1 hour, followed by Fludarabine 10mg/m2 in 1 hour, followed by a 3-hour infusion of once-daily targeted Busulfan (weight-based dosing + therapeutic drug monitoring to a total Bu-exposure of 90mg*h/L). Thymoglobulin was added in unrelated donors (except in AML receiving cord blood). GvHD-prophylaxis was according to standard protocols. Minimal Residual Disease (MRD) negative was defined as < 10e-4. Primary endpoints were Overall Survival (OS) and Leukemia Free Survival (LFS). Secondary endpoints were Non Relapse Mortality (NRM), Relapse, acute and chronic Graft-versus-Host Disease (GvHD), and VOD/SOS. Cox Proportional Hazard and fine and gray competing risk models were used for data analysis. Results 155 children were included; 66 ALL (38 in CR1, 28 ³ CR2), 69 AML (28 in CR1, 40 in CR2, 1 in active disease) and 20 other malignancies (mostly MDS-EB). Median age was 9.7 (0.5-18.6) years. Most donors were unrelated (119 vs 36 related); 79 Bone Marrow (BM), 66 Cord Blood (CB) and 10 Peripheral Blood Stem Cells. Median follow up was 964 (19-2994) days. Overall the 3-yr estimated OS and LFS was 72 ± 4.5% and 65 ± 5% respectively. Estimated 3-yr LFS for MRD-neg ALL, MRD-pos ALL, AML CR1 and AML CR2 was 74 ± 7%, 40 ± 12%, 64 ± 10%, 65 ± 9% respectively; Fig 1). NRM in whole cohort was 10.3 ± 3.0% (with 25 ± 4.3% for AML CR1 (n=28), and 5.8 ± 2.3% in the rest; Fig 2). Other endpoints: only 2 graft-failures were noted, incidence of aGvHD III-IV at 6 months was 11 ± 3%, extensive chronic GvHD at 3-yr was 5.2 ± 2.2%. Relapse at 3-yr was 25 ± 4.3% (MRD-neg ALL 16.2 ± 3.7%, versus 60 ± 4.9% in MRD-pos ALL, in AML 22.5 ± 4.2%) and no VOD/SOS was noted. Conclusion CloFluBu in myeloid- and lymphoblastic malignancies, showed very limited toxicity and encouraging LFS in all groups, in particular for MRD negative ALL. More studies, preferably in randomized controlled clinical trials, are needed to draw firm conclusion with regards to the anti-leukemic effect and late effects.

Introduction Allogeneic stem cell transplantation (SCT) is potentially curative therapy for patients with MDS. Prior studies have shown no overall-survival (OS) advantage of any conditioning regimen over the others. Reduced-intensity conditioning (RIC) is usually associated with lower non-relapse mortality (NRM) compared to myeloablative conditioning (MAC), however, relapse rates are higher, resulting in similar OS. Fludarabine and treosulfan (FT) is a reduced-toxicity regimen with intense anti leukemia activity in patients with myeloid malignancies. We explored outcomes following FT conditioning in comparison with other regimens. Methods We retrospectively analyzed SCT outcomes of MDS patients reported to the chronic malignancies working party of EBMT (n=1722). The conditioning regimens were defined according to standard EBMT criteria and included FT (n=367), RIC (n=687) or MAC (n=668). Results FT and RIC recipients were older than MAC recipients, median age 59, 59 and 51 years, respectively (P < 0.001). More FT recipients had untreated MDS at SCT, 45%, 26% and 36%, respectively (P < 0.001). Disease status at diagnosis, cytogenetics, secondary origin and prior transformation to AML, were not different between the regimens. With a median follow-up of 20 months (1-171), 807 patients are alive, 328 died of relapse and 587 had NRM. The 5-year rates of relapse, NRM and OS for the entire group were 21% (95%CI, 19-23), 35% (95%CI, 33-38) and 44% (95%CI, 42-437), respectively. Relapse rates were similar after FT and MAC, 16% (13-21) and 19% (16-22), respectively, significantly lower than after RIC, 25% (21-28) (P = 0.003). Multivariate analysis (MVA) identified age >55 years (HR 1.42), advanced histology (RAEB 1/2, HR 2.01/ 1.73), prior transformation to AML (HR 1.87), refractory disease to prior therapies (HR 1.75) secondary origin (HR 1.33) and poor cytogenetics (HR 2.83) as associated with higher relapse risk. FT conditioning was associated with a lower risk (HR 0.63, P = 0.006). NRM was lower after FT and RIC than after MAC, 34% (29-39), 33% (30-37) and 38% (35-40), respectively (P = 0.05). MVA identified age >55 years (HR 1.39) and MAC (HR 1.34) as associated with higher NRM, while SCT from sibling donor (HR 0.81) and in CR (HR 0.80) were associated with a lower risk. In all, 5-year OS was 50% (44-55), 43% (38-47), and 43% (39-47), after FT, RIC and MAC, respectively (Figure 1, P = 0.03). MVA identified age >55 years (HR 1.39), advanced histology (RAEB 1/2, HR 1.35/ 1.10), prior transformation to AML (HR 1.37), and poor cytogenetics (HR 1.58) as associated with lower OS. FT conditioning was associated with better OS (HR 0.79, P = 0.01). Conclusions FT is associated with similar low relapse rates as MAC and similar low NRM as RIC, resulting in improved outcome over both RIC and MAC. FT might be the preferred regimen for SCT in MDS. These observations merit further study in randomized prospective trials.

Current approaches to conditioning prior to transplant and gene therapy employ non-specific genotoxic agents that are associated with severe acute and long-term toxicities. Consequently, many transplant eligible patients either refuse transplant or receive reduced intensity regimens, which still have adverse toxicities with increased risk of relapse and graft failure. Importantly, patients with non-malignant disease may choose to forgo a potentially curative transplant due to risks of conditioning-associated toxicities, potentially limiting access of these patients to transplant and emerging HSC-based gene therapy. Thus, safe and effective targeted conditioning agents would have broad application to transplant in both malignant and non-malignant settings as well as gene therapy. We developed a CD117 antibody conjugated to amanitin (anti-CD117-AM) to specifically deplete hematopoietic stem and progenitor cells (HSPCs) as a novel approach to patient preparation for allogeneic and autologous transplant. ADCs have significant single agent anti-tumor activity in the clinic, however, achieving suitable tolerability at efficacious doses has presented a clinical challenge. We engineered the properties of the anti-CD117-AM to provide a broad therapeutic window across pre-clinical models. Through modification of the linker moiety, the maximum tolerated dose (MTD) of the ADC in C57Bl6 mice was improved 17-fold compared to previous versions. The engineered anti-CD117-AM maintained potent cytotoxicity on primary human CD34+ cells and a CD117+ AML cell line (<10 pM IC50). In humanized NSG mice, a single dose (0.1 mg/kg) of the ADC selectively depleted >95% of human HSPCs (Fig. 1A). Single doses of the anti-CD117-AM demonstrated robust tumor control in xenograft models of AML, despite its purposefully shortened half-life as a targeted agent for transplant indications (Fig. 1B). The tolerability of the engineered anti-CD117-AM was assessed in cynomolgus NHPs. The ADC was well-tolerated with an MTD at ≥ 2 mg/kg (Fig. 1C). A single dose of the ADC was able to drive marked depletion of HSPCs and downstream progeny. The pharmacokinetics of the ADC exhibited linear clearance supporting tolerability well above target saturation. In summary, optimization of the chemical linker between antibody and payload yielded a well-tolerated ADC with wide safety margins that could eliminate HSPCs and confer survival benefit in AML xenograft models. ADC-mediated conditioning through selective depletion of CD117+ cells may provide an improved approach to patient preparation for transplant in malignant and non-malignant disease including gene therapy, broadening patient access to potentially curative treatments.

Background In recipients of allogeneic hematopoietic stem cell transplantation (HSCT), organ toxicity is a barrier to administering high-intensity conditioning. We hypothesized that determinants of acute organ toxicity are specific to individual regimens. We sought to characterize these toxicities, evaluate their prognostic implication, and derive predictors of severe toxicity at the regimen level. Methods This retrospective study included adults undergoing first allogeneic HSCT at a single center between 2001 and 2014. Patients received grafts from matched sibling or unrelated donors and received any of the following regimens: Cy/TBI, Bu/Cy, Flu/Bu 12.8 mg, Flu/Bu 6.4 mg, Flu/Treosulfan 36-42 gr/m2, and Flu/Mel 100-140 mg/m2. Studied toxicities included acute kidney injury (AKI) per KDIGO definition and increases in total bilirubin, AST, ALT, and alkaline phosphatase (Alk Phos) per CTCAE grading. The incidence of toxicities from conditioning through 30-days post-HSCT was tabulated. Risk factors for severe organ toxicity were assessed within each regimen using multivariable logistic regressions. Results In a cohort of 707 patients, the main indications were acute leukemia (57%), myelodysplastic syndrome (13%), and aggressive lymphoma (9%). GvHD prophylaxis included methotrexate in 80% of patients, and 56% received ATG. The incidence of AKI and increased serum bilirubin in each regimen is shown in Figure 1A/B, respectively. Sinusoidal-obstructive syndrome (6% overall) accounted for only 17% of gr. ≥ 3 bilirubinemia in the entire cohort. Elevations in AST, ALT, and Alk Phos of gr ≥ 3 were not common (<8%). AKI gr ≥ 2, increased bilirubin gr ≥ 3, AST gr ≥ 3, and Alk Phos gr ≥ 2 were associated with increased 100-day mortality (p < 0.05). Acute severe organ toxicity (ASOT) was defined as the occurrence of any of these toxicities. ASOT had an odds ratio (OR) of 3.4 (95% CI: 2.2-5.3) for 100-day mortality. Within each regimen, we studied the relationship between ASOT and transplant/patient characteristics (Figure 1C). Elevations in baseline bilirubin were predictive of ASOT in Cy/TBI (OR: 1.7 [1.2-2.4]), while increasing creatinine was predictive in patients conditioned with Flu/Mel (OR: 1.4 [1.1-1.9]). In patients treated with Bu/Cy, administration of ATG increased the risk of ASOT (1.3 [1.1-1.6]). Conclusion Allogeneic transplantation recipients are at high risk for acute organ damage. We describe patterns of renal and liver toxicity across several regimens. Determinants of acute severe organ toxicity, defined as those associated with short-term mortality, are regimen dependent. Our findings suggest these factors should be considered in selecting the preparative regimen. While requiring validation, the newly-defined composite endpoint of acute severe organ toxicity (ASOT) may be valuable in studying transplantation strategies.

Introduction To assess the therapeutic potential of non-TBI conditioning and T-cell depleted (TCD) allo HCT for high risk malignancies, we conducted a prospective trial comparing: A. our standard 1375 cGy HFTBI + Thiotepa, 5mg/kg/dx2 + cyclophosphamide, 60mg/kg/dx2 vs B. Busulfex, 0.8mg/kg/6hx12(PK adjusted) + Melphalan 70mg/kg/dx2 + Fludarabine 25mg/m2/dx5 vs C. Clofarabine 20mg/m2/dx5 + Melphalan 70mg/m2/dx2 + Thiotepa 5mg/kg/dx2 as preparation for T-cell depleted CD34+ PBSC allografts isolated by the CliniMACS system. All pts received Thymoglobulin pre-transplant to prevent infection. No GVHD prophylaxis was given post transplant. Pts were stratified to arms A, B or C based on disease stage and clinical factors enhancing TBI risks (e.g. age >60m prior therapy or comorbidities), with Arm B the non-TBI arm used for myeloid malignancies and Arm C for lymphoid malignancies. Results From 5/13/10 to 12/31/16, 283 consenting patients were transplanted (106 in Arm A, 144 in B, 33 in C). All pts engrafted by day 10-11; 3 pts in Arm B and 1 in C had late graft failure with sustained T-cell chimerism. The 3 in Arm B reconstituted after TCD PBSC boost alone (N=2) or after CTX/ATG (N=1) conditioning. Cumulative Incidence (C.I.) of Acute Grade 3-4 GVHD was higher in Arm A (12%) than in B (4.2%) or C (3.0%) (p = 0.03), as was C. GVHD (Arm A 8.5%; B (1.5% p = 0.02); C (3%)). CI of relapse at 3 yrs was similar in all arms (A=28%; B=21%; C15%) (p = 0.22). Although pts in Arm B were older (med. 58.5 yrs) than in A (32 yrs) or C(32 yrs), NRM at 3 yrs for Arm B (16%) and A (14.8) were similar. However, NRM in Arm C (32.7%) was higher (p = 0.09). Because the median age of patients in Arm B (58.5 yrs) differed significantly from Arms A (31.9 yrs) and C (32.3 yrs); we evaluated DFS for each arm in 20 year cohorts of increasing age (Fig 1 A,B,C). Because of low cohort size, pts in Arm C aged <40 were compared to those ≥ 40. In Arm A, 5 year DFS rates were 54.2% (Age ≤20), 62.9% (Age>20-40) and 54% (Age 40-60). In Arm B, 5 year DFS for pts ≤ 20 was 69.3% and 85.7% for those >20-40, 48.6% for those >40-60 and 46.8% for those 60+ respectively (p = 0.02). In Arm C, 5 year DFS for pts ≤40 was 65% and for those >40, 30.7% (p = 0.01). Conclusion These results demonstrate that these chemotherapy-based regimens secure engraftment and hematopoietic reconstitution and similarly low rates of GVHD and relapse. They also suggest an advantage for patients ≤40 treated on Arm B. In >40-60yo patients, DFS for Arm B is equivalent to Arm A and also yields favorable results in older pts (>60-73 yrs). However, while patients ≤40 yrs of age fared well on Arm C, older patients were less likely to achieve extended OS and DFS than those on Arms A or B, likely as a consequence of increased TRM.

Cerebral adrenoleukodystrophy (CALD) is a rare metabolic disorder in which rapid and progressive inflammatory demyelination leads to irreversible loss of neurologic function and death. There is no approved treatment for CALD, although Lenti-D Drug Product (DP) is in development as an autologous hematopoietic stem cell (HSC) gene therapy. We have previously shown promising results from the ALD-102 clinical study of the safety and efficacy of Lenti-D DP in boys ≤17 years old with CALD; based on data from 29 patients, Lenti-D DP was generally safe and led to stabilization of disease progression. The primary efficacy endpoint in ALD-102 is the proportion of patients who are alive and free of major functional disabilities (MFD) at Month 24. The primary safety endpoint is the proportion of patients who experience acute (≥Grade 2) or chronic graft-versus-host disease (GVHD) by Month 24. ALD-102 has completed enrollment, with 32 boys having received Lenti-D DP as of April 2019. The median DP cell dose was 11.4 (min - max, 5.0 - 20.1) × 106 CD34 cells/kg; median DP vector copy number was 1.2 (min - max, 0.5 - 2.7) copies/diploid genome. Median day of neutrophil and platelet engraftment was 13.0 days (min-max, 11.0 - 41.0) and 32.0 days (min - max, 16.0 - 60.0), respectively. The median follow-up time for the cohort of 32 patients is 21.2 months (min - max, 0.0 – 60.2). Fifteen patients have completed 24 months of follow-up in ALD-102 and continue to be free of major functional disabilities (MFD) through their last follow-up in a long-term extension study. Fourteen patients remain in ALD-102; the longest follow-up among these patients is 20.4 months. Two patients were withdrawn and referred for allo-HSCT before their Month 24 visit; another experienced early and rapid disease progression while on-study resulting in MFDs and death. All other Lenti-D DP-treated patients showed stabilization of neurologic function score at their last follow-up (stable NFS was defined as NFS ≤4 without a change of >3 from baseline). There have been no reports of graft failure, GVHD, or transplant-related mortality; recorded adverse events are consistent with myeloablative conditioning. There is no evidence of replication competent lentivirus or insertional oncogenesis. Lenti-D DP appears to stabilize neurologic disease progression and continues to show favorable safety with the longest follow-up at 60.2 months. Additional follow-up is ongoing to assess durability of effects and long-term safety.

Introduction Allogeneic hematopoietic stem cell (HSC) transplantation (allo-HSCT) is a treatment option for several monogenic diseases; however, its use is limited by the need for a matched donor and risk of HSCT-related complications. Autologous HSC gene addition does not have some of these limitations and may have similar efficacy with an improved safety profile. Ex vivo gene addition therapy using lentiviral vectors (LVV) is being evaluated in patients with transfusion-dependent β-thalassemia (TDT) using betibeglogene autotemcel (beti-cel, LentiGlobin for TDT) in the HGB-204, -205, -207, and -212 studies, sickle cell disease (SCD) using LentiGlobin for SCD in HGB-205 and -206, and cerebral adrenoleukodystrophy (CALD) using Lenti-D in ALD-102. The safety outcomes following autologous gene modified HSCT in these ongoing studies are summarized. Methods HSCs are collected using disease-appropriate procedures, then CD34+ cells are transduced with LVV encoding disease-specific therapeutic transgenes. After myeloablation with busulfan (SCD, TDT) or busulfan/cyclophosphamide (CALD), patients are infused with LVV-transduced CD34+ HSCs. Patients are followed for two years and offered participation in long term follow-up studies (LTF-303 [NCT02633943] or LTF-304 [NCT02698579]). Results Across all 6 studies, 110 patients have been treated as of most recent analyses (Table 1). Patients have been followed for <1 to 60.6 months and 47 patients have >2 years follow-up. No patient experienced primary or secondary graft failure. One patient with CALD experienced disease progression and died 22 months after drug product (DP) infusion of disease complications. Two additional patients with CALD withdrew from the study after DP infusion and were referred for allo-HSCT. All other patients with CALD, TDT, and SCD remain alive. Most (107/110) patients had ≥ one grade 3 or 4 adverse event (AE) attributed to conditioning; most common AEs were cytopenia, febrile neutropenia, and stomatitis. Myelodysplastic syndrome was reported in one patient with SCD. After investigation for LVV insertion in malignant cells, the AE was assessed as not related to LentiGlobin insertion or transgene expression. AEs reported as related to the DP are shown in Table 1. There was no GVHD and no clinically relevant clonal dominance or LVV-mediated replication competent lentivirus. Summary Data from 110 patients followed for up to 5 years supports that the safety profile of gene-modified autologous HSCT does not carry the risks of GVHD, graft rejection, and long-term immunosuppression attendant to allo-HSCT. While the safety profile beyond 5-years is still being established, these data suggest that HSC gene therapy may be an acceptable therapy for patients with TDT, SCD, and CALD, particularly in patients at increased risk of complications from allo-HSCT, such as those who lack a suitable donor or are more advanced in age.

Background Patients (pts) receiving CTL019 (tisagenlecleucel) for B-ALL with high tumor burden (HTB) are at high risk for developing severe CRS. Tocilizumab, an IL-6 receptor antibody, is a vital component of severe CRS management; however, its role in preventing severe CRS is not known. We sought to determine the effectiveness of preemptive tocilizumab (PT) administration in decreasing the rate of grade (gr) 4 CRS in HTB pts. Methods We conducted a pilot trial of risk-adapted PT after CTL019 (NCT02906371). HTB pts, defined as ≥40% bone marrow (BM) blasts immediately before infusion, received a single dose of PT for high persistent fever (2 temperatures ≥38.5C in 24hr). The primary endpoint was frequency of gr4 CRS (Penn scale), with an observed rate of ≤5/15 predefined as clinically meaningful. Secondary endpoints included complete remission (CR) rate and ICU length of stay (LOS). A comparator cohort with HTB who received standard CRS management (stdCRS) was identified from the initial CTL019 trial (NCT01626495). Results Characteristics of the PT (n=15) and stdCRS (n=26) HTB cohorts are shown in Table 1. All pts developed gr≥2 CRS; median time to fever was longer in the PT cohort [PT, 3d (IQR 2-9); stdCRS, 2d (IQR 1-7), p=0.03]. Gr4 CRS was observed in 4/15 (27%) pts in the PT cohort vs. 13/26 (50%) in the prior stdCRS cohort [RR 0.53 (95% CI, 0.21-1.34), p=0.18]. In the stdCRS cohort, gr4 CRS was associated with earlier onset of fever (p=0.04). In patients with earlier CRS onset (fever by day +4), gr4 CRS was observed in 4/9 (44%) vs. 13/21 (62%) in the PT and stdCRS cohorts [RR 0.72 (95% CI, 0.32-1.60), p=0.42]. Except for a trend toward fewer vasoactive days in the PT cohort, ICU LOS and resource utilization were not significantly different (Table 2). The CR rate at day 28 was similar in the PT and stdCRS cohorts (87% vs. 85%, p=1.00). Conclusion Risk-adapted PT administration reduced gr4 CRS, meeting the predefined study endpoint, without impacting the CR rate. A secondary comparison to a prior trial showed a clinically meaningful decrease in the rate of gr4 CRS from 50% to 27%; however, the analysis was not powered to detect a statistically significant difference. Ongoing analyses will evaluate CAR T cell expansion, duration of remission, and additional safety endpoints, including rates of neurotoxicity.

Background Fanconi anemia (FA) is a rare genetic disorder characterized by defective cellular DNA repair, associated developmental abnormalities, progressive bone marrow failure (BMF), and a predisposition to hematologic malignancies and solid tumors. 80% of FA patients develop BMF due to progressive depletion of their BM stem cells. Although allogeneic HSCT is a curative treatment for BMF, its utilization and efficacy is limited by availability of donors, risk of GVHD and transplant-related toxicities. Pre-clinical studies showed that ex-vivo insertion of a functional FANCA gene into autologous FA-A CD34+ HSPCs provides a survival advantage to the gene-modified stem cells, leading to correction of BMF. Feasibility of this approach was established in the FANCOLEN-1 clinical trial (Spain). Modifications to the collection and manufacturing processes were made in the Phase I study in the U.S. Design and Methods RP-L102-0418 (clinicaltrials.gov # NCT03814408) is a Phase I clinical trial evaluating the feasibility and safety of autologous CD34+ cells transduced with a lentiviral vector (LV) carrying the FANCA gene (PGK-FANCA-WPRE) in children with FA-A. Patients with early evidence of cytopenias, and with BM CD34+ count >30/µL were eligible for treatment. PBMCs were collected via leucocytapheresis after mobilization with G-CSF and Plerixafor. CD34+ HSPCs were enriched, placed in culture with cytokines, and transduced with PGK-FANCA-WPRE LV. The investigational drug product (DP) (RP-L102) was infused fresh in patients within 4 hours of release, without any conditioning regimen. Patients are being followed for safety assessments (replication competent lentivirus, insertion site analysis) and efficacy (increasing peripheral blood vector copy number and BM mitomycin-C resistance), along with monitoring of cytopenias. Results Two FA-A patients (aged 5 and 6 years) were consented and enrolled on the study at Stanford University. Mobilization and apheresis procedures were performed successfully without any serious adverse events. Both the patients needed RBC priming of the apheresis circuit because of their weight <20 kgs and developed thrombocytopenia post-apheresis. DP was successfully manufactured and infused in both the patients without any adverse events. 6 months post-infusion, the progressive cytopenias have stabilized. Detailed safety and efficacy data will be presented. Conclusions DP was successfully manufactured and infused in the Phase I study and both FA-A patients are showing early signs of stabilization of cytopenias. Plans for opening the Phase II study are in progress. Preventing BMF by gene therapy without any exposure to chemo-radiotherapy has the potential to change the natural history of FA.

Introduction Cytokine release syndrome (CRS) and immune effector cells associated neurotoxicity syndrome (ICANS) are common CAR T cells toxicities. Various grading systems that attribute different grades of severity to symptoms are currently used, preventing comparisons of different products and possibly leading to management implications. Methods We included 53 patients (pts) with B acute lymphoblastic leukemia (B-ALL) treated with 1928z CAR T cells (NCT01044069) and 49 pts with diffuse large B cell lymphoma (DLBCL) who received axicabtagene ciloleucel (axi-cel, n=36) or tisagenlecleucel (tisa-cel, n=13) after FDA approval. CRS and ICANS were retrospectively graded according to the new ASTCT consensus grading system, for comparison with other available systems: Lee, Penn, Memorial Sloan Kettering Cancer Center, CARTOX and CTCAEv5.0 for CRS; CTCAEv4.03 and CARTOX for ICANS. We then used ASTCT grades to predict management according to current guidelines for DLBCL pts (axi-cel and tisa-cel insert packages, CARTOX and NCCN guidelines) and compare it to the actual treatment received at our center. Results According to the ASTCT grading, B-ALL pts had 87% CRS (28% grade ≥3), while DLBCL pts had 86% CRS (14% grade ≥3) with axi-cel, and 54% (no grade ≥3) with tisa-cel. B-ALL pts had 55% ICANS (45% grade ≥3), while DLBCL pts had 55% ICANS (33% grade ≥3) with axi-cel, and 15% with tisa-cel (no grade ≥3) (fig. 1). When comparing grading systems, agreement on CRS and ICANS diagnosis was found in 99% and 91% cases, respectively. However, when evaluating toxicities grade by grade, only 27% pts had the same grade in each system for CRS, and 55% for ICANS (fig. 2). When predicting management for DLBCL pts, we found some relevant differences across current guidelines (fig. 3). At our center 58% and 13% of patients with CRS received tocilizumab (toci) and steroids, respectively, similarly to what predicted according to axi-cel's, CARTOX and NCCN guidelines, but differently from tisa-cel's label (10% and 5%). For ICANS, while tisa-cel's label does not provide any recommendations, other guidelines-based predictions were mostly overlapping. Indications for treatment vary across current guidelines, which were developed on single products and different grading systems, thus, should not be universally applied. This discrepancy becomes particularly relevant for cases with discordant CRS/ICANS grades. As such, we gave toci to 66% pts upgraded to grade 3 by Penn (grade 2 by ASTCT), while they would receive both toci and steroids according to axi-cel/CARTOX/NCCN guidelines, but would not be treated according to tisa-cel's. Conclusions Different grading systems provide inconsistent CRS/ICANS scores. To avoid discrepancies in assessing and managing toxicities of different products, a unified grading should be used and paired management guidelines with product-specific indications should be developed.

Introduction CD19 directed CAR T cells have potent activity in relapsed/refractory (R/R) aggressive B-cell lymphomas (B-NHL) leading to the FDA approval of axicabtagene ciloleucel (axi-cel) and tisagenlecleucel (tisa-cel). Initial reports on commercial use of axi-cel suggest many patients (pts) would not have met eligibility criteria for ZUMA-1, yet outcomes and toxicities appeared similar. Limited data on the “real world” use of tisa-cel is available. We performed a multicenter retrospective study to evaluate efficacy, safety, and patterns of use in 8 US centers that had the option of prescribing either commercial product for R/R B-NHL. Results As of 7/31/2019, 242 pts underwent apheresis for axi-cel or tisa-cel. Of these, 163 (67%) underwent apheresis for axi-cel, and 79 (33%) for tisa-cel. 14 (9%) axi-cel and 3 (4%) tisa-cel pts died prior to CAR T-cell infusion from lymphoma progression, and 1 (1%) tisa-cel pt was not infused for other reasons. Detailed baseline pt characteristics were available for 180/242 pts (Figure 1). Median age at apheresis was 58 years (range: 18-85) for axi-cel and 67 years (range: 36-88) for tisa-cel. ECOG PS was 0-1 in 86% of axi-cel and 94% of tisa-cel pts. The median number of prior therapies was 3 (range: 2-11) for axi-cel and 4 (range: 2-9) for tisa-cel pts. Bridging therapy was given in 61% of axi-cel and 72% of tisa-cel pts. Median time from apheresis to CAR T-cell infusion was 28 days for axi-cel and 44 days for tisa-cel. CAR T-cell infusion was inpatient in 100% of axi-cel and 39% of tisa-cel pts. Safety was evaluable in 213 pts. CRS was graded according to institutional practices (CARTOX (38%), Penn scale (31%), ASTCT (19%), and Lee scale (11%)). NEs were graded per CARTOX (80%), ASTCT (19%), or CTCAE V4.03 (1%). Grade ≥3 CRS and NEs occurred in 13% and 41% of axi-cel pts and 1% and 3% of tisa-cel pts. The median onset of CRS and NEs was 2 and 6 days in axi-cel, and 3 and 5 days in tisa-cel treated pts, respectively. Tocilizumab was administered in 62% of axi-cel pts with 57% receiving steroids. In tisa-cel pts, tocilizumab was administered in 13% of cases, with 7% receiving steroids. Response assessment was performed for infused pts at day 30 and/or day 90, or in those with clinical progression. Of 120 axi-cel pts evaluable at day 30, the ORR was 72% with 43% achieving a CR. Of the 32 tisa-cel pts evaluable at day 30, the ORR was 59% with 44% achieving a CR. At day 90, the ORR for axi-cel was 52% with 39% achieving a CR, while for tisa-cel the ORR was 48% with 39% achieving a CR. Conclusions Efficacy outcomes in the commercial setting appear similar to responses seen in the pivotal clinical trials. Though different toxicity grading scales were employed, tisa-cel appears to be associated with less CRS and NEs. Updated analyses of usage patterns and outcomes with uniform ASTCT toxicity grading will be presented at the meeting.

Introduction Chimeric antigen receptor (CAR) T cells have shown impressive results in B-cell malignancies but, unfortunately, their use is still limited by the logistical and financial burdens related to the need of generating autologous cell products. The development of universal allogeneic CAR T cells to be used off-the-shelf has faced major limitations, namely the risk of Graft-versus-Host-Disease (GvHD) induction and the rejection of the cells by the host immune system. Invariant Natural Killer-T (iNKT) cells are innate lymphocytes deprived of GvHD induction potential and displaying antitumor effects, both directly and indirectly, through enhancement of CD8 T cell responses. Preclinical studies in xenogeneic mouse models demonstrated the feasibility of using iNKT cells as a platform for CAR-based therapies, and two clinical trials are currently ongoing. Objective To assess the immunoadjuvant effect of allogeneic CD19-specific iNKT CAR on the host immune system. Methods We transduced murine iNKT cells from FVB/N (H-2Kq) mice with a CD19-specific CD28/CD3ζ CAR and assessed their antitumor effect in vitro and in vivo. Results CD19-iNKT CAR had a direct cytotoxic effect in vitro against the CD19+ A20 lymphoma cells and significantly improved survival of mice after administration to major histocompatibility complex (MHC)–mismatched, immunodeficient BALB/c (H-2Kd) Rag2−/− gamma-chain−/− mice receiving A20 cells (2 × 10e4; Figure 1A) without inducing any signs of GvHD. To test the efficacy of iNKT CAR cells in the presence of host immune cells, we employed BALB/c mice receiving sublethal irradiation (4.4 Gy), resulting in only a partial and transient lymphopenia. The antitumor effect of allogeneic iNKT CAR cells was greatly enhanced (Figure 1B) suggesting the participation of host cells in the antitumor effect. Interestingly, the iNKT CAR effect was partially abrogated when we employed as recipients BALB/c BATF3−/− mice, in which CD8 T cell cross-priming is impaired as a result of the absence of BATF3-dependent CD103+ CD8a+ dendritic cells known to play a role in iNKT cell interactions with other immune effector cells (Figure 1C). Co-administration of allogeneic FVB/N iNKT CAR with autologous BALB/c CD8 T cells at the time of transfer of T-cell-depleted autologous bone marrow cells and A20 lymphoma cells into lethally irradiated BALB/c recipients resulted in a synergistic effect (Figure 1D). Due to their immuno-adjuvant effect on host CD8 T cells, low numbers of allogeneic iNKT CAR outperformed allogeneic conventional CAR T (Tcon CAR) when administered to sublethally irradiated BALB/c mice receiving A20 cells (Figure 1E-F). Conclusion Collectively, these results demonstrate the potent immunoadjuvant effect exerted by allogeneic iNKT CAR cells toward the host immune system suggesting that iNKT CAR cells are an attractive off the shelf product for adoptive immunotherapy.

Introduction CAR T cells have demonstrated unique potency for tumor cytoreduction and durable response in hematological malignancies. However, disease relapse remains a concern due to the emergence of antigen negative variants, tolerization of CAR T cell populations and lack of T cell persistence. We postulated that vaccination with a personalized cancer vaccine, that we developed, in which patient derived tumor cells are fused with autologous dendritic cells would enhance CAR T cell efficacy through the expansion of T cell clonal populations and the vaccine mediated enhancement of T cell activation and persistence. In the present study, we examined the potential synergy between CAR T cells targeting CD19 and syngeneic DC/tumor fusions. Methods/Results CD19 CAR T cells and DC/tumor fusions were studied in the context of a murine A20 lymphoma model, and their combination effectively lysed A20 cells in a CTL assay in a higher percentage than CAR T cells alone (20% vs 34%). Moreover, we examined the interaction of vaccine and CAR T cells ex vivo using the IncuCyte S3 Live-Cell Analysis System which allows for live cell visualization of lysis of A20 cells over time. We studied the impact of combining vaccine educated and CAR T cells as well as an individual T cell population that underwent sequential vaccine mediated stimulation followed by transduction with the CD19 CAR. Even in this assay, coculture with either combined vaccine educated and CAR T cells or sequentially vaccine educated, and transduced T cells demonstrated the highest levels of cytotoxicity that was maintained over time. Enhanced lysis by combined vaccine stimulation and CAR T cells was similarly demonstrated in another tumor cell line, 5TGM1, a multiple myeloma cell line transduced to express CD19. On the contrary wild type 5TGM1 cells were recognized by the DC/tumor fusion stimulated cells in contrast to CAR T cells alone (40% vs. 8%). Finally, we examined the capacity of vaccine educated T cells in conjunction with CAR T cells to target A20 cells in an immunocompetent murine model. Mice were challenged with 1 × 106 A20 Mcherry-Luc and after lymphoma engraftment animals were treated with 3 × 106 T cells consisting of CAR T cells, vaccine educated T cells or the combination. Serial bioluminescence imaging demonstrated greatest reduction in tumor burden using combined CAR T and vaccine educated T cells with 4/5 animals still disease free at day 13 after tumor challenge. Conclusions Combined therapy with T cells stimulated by DC/tumor fusions and CAR T cells exhibited potent lysis of murine lymphoma and myeloma cells compared to CAR T cells or vaccine educated T cells alone in in vitro and immunocompetent murine models. These findings suggest potent synergy between these modalities that may overcome recognized pathways of resistance including the broadening of the tumor specific response and vaccine mediated activation of CAR T cell populations.

Introduction Sickle cell disease (SCD) is the most common inherited hemoglobin disorder affecting approximately 100,000 people in the United States (U.S.). Allogeneic hematopoietic cell transplantation (alloHCT) is currently the only curative option for SCD. However, alloHCT is an optional benefit under Medicaid, the largest health program in the US. This means that individual state Medicaid programs can chose to cover alloHCT, define the indications, clinical trial coverage, and/or determine the scope of key health benefits, including donor search, cell procurement, medication, travel and lodging. This study of coverage for alloHCT for patients with SCD aims to understand 1) scope of state Medicaid coverage benefits 2) transplant center (TC) financial counseling practices and experience working with state Medicaid programs. Methods Financial coordinators at active Blood and Marrow Transplant Clinical Trial Network (BMT CTN) 1503 clinical trial TCs in 12 states (CA, FL, GA, IL, LA, MI, NC, NY, OH, PA, TX and VA) estimated to have more than 50 newborns diagnosed with SCD in 2016 were contacted. Qualitative, semi-structured interviews 30-60 minutes in duration were conducted via telephone using WebEx teleconference by trained interviewers from Be The Match – NMDP, between May and October 2019. Interviews were recorded, transcribed and analyzed using content analysis, using computer assisted qualitative data analysis software (CAQDAS), NVivo 10. Results Of the 26 TCs contacted, a total of 9 TCs representing 7 states participated in semi-structured interviews (Table 1). Preliminary data analysis of interviews revealed that travel and lodging was the most common benefit with limited or absent coverage for alloHCT patients with SCD but showed wide variation by state. Clinical trials were inconsistently covered with some states following Medicaid guidelines for covering alloHCT as part of a clinical trial but not as standard of care while other states covered alloHCT as a part of standard care but not as part of a clinical trial. Conclusions There is wide variation in state Medicaid coverage benefits particularly for clinical trial coverage and travel and lodging. These data provide insight into potential areas to influence changes in policy in order to enhance access to life altering curative therapy for SCD.

Introduction Historically, there have been limited curative treatment options for patients (pts) who relapse or have refractory Large B Cell Lymphoma (LBCL). Recently, autologous anti-CD19 chimeric antigen receptor T-Cell (CAR T) therapies were approved for the treatment of pts with relapsed or refractory LBCL after ≥ 2 prior systemic therapies. Objectives To describe the demographic and clinical characteristics of Medicare pts receiving CAR T therapy (axicabtagene ciloleucel or tisagenlecleucel), and compare healthcare utilization, costs, and outcomes pre- and post-CAR T therapy. Methods The study utilized a single-group pre-test/post-test design and Center for Medicare and Medicaid Services 100% Medicare Fee-for-Service (FFS) Part A & B claims data. (Part D has not been released.) Pts with LBCL, received CAR T therapy between 10/1/2017 and 9/30/2018, and were continuously enrolled in Medicare FFS for 6 months prior to and 100 days after CAR T infusion. Index episode of care was CAR T infusion and associated inpatient stay, if any. Baseline characteristics included age, gender, race, specific LBCL diagnosis, and comorbidities. Measures of utilization and cost pre- and post-CAR T were standardized as per patient per month (PPPM) to account for different follow-up durations, and included hospitalizations, intensive care unit (ICU) transfers, and emergency department (ED) visits. Pre- and post- CAR T statistical analyses excluded the index episode. Results 177 pts are included with an average age of 70 years; male (58.8%); white (87.6%); a primary diagnosis of diffuse LBCL (91.5%) and infrequent autologous stem cell transplant (< 5%). Charlson Comorbidity Index score ≥ 3 were common (74.6%) and included 1 or more comorbidities that would have disqualified them from CAR T clinical trials (43%) (e.g. renal failure, heart failure, recent history of DVT/PE). Pts spent a median of 16 days in hospital during their index episode of care and nearly half (45.5%) were transferred to ICU during their stay. During the 6-month pre-index period, over half the pts had ≥ 1 hospitalization, and nearly 20% had ≥3. Of these, 27.1% were re-admitted during the post-index period. For those hospitalized, the median length of stay (LOS) pre- and post-index was 7 and 5 days, respectively. The number of pts with an ED visit was reduced by half during post- vs. pre-index (15.8% vs. 29.9%). There was no evidence of subsequent intravenous outpatient during the 100-day post-index period although claims may lag for some pts. Exclusive of index episode costs, total healthcare costs during the pre-index period were $9,749 PPPM pre- vs. $7,121 post-index, a 27% decrease. Conclusions The results of this real-world study indicate that older pts with multiple comorbidities can be treated successfully with CAR T therapy, and that post-index care was associated with lower hospitalization rates, bed days, ED visits, and total costs.

Background Allogeneic stem cell transplantation (AlloSCT) is a curative procedure for many hematologic neoplasms. Majority of the AlloSCT require an initial hospitalization and a mortality rate of up to 8% during the SCT hospitalization has been previously reported (Ballen et al, 2014 BBMT). Role of patient-, disease- and transplant-related factors in predicting early mortality during transplant hospitalization has not been previously investigated. We sought to analyze the predictors of inpatient mortality in patients undergoing AlloSCT from a large administrative database, National Inpatient Sample (NIS). Methods International Classification of Diseases, Ninth Revision (ICD-9) procedure codes were used to identify patients undergoing AlloSCT in the 2002-2015 NIS database whereas ICD-9 diagnosis codes were used to assess complications of Allo-SCT. Early transplant-related mortality was defined as death during the hospitalization for SCT. Chi-square was used for categorical data comparisons (p<0.05 was considered significant). Multivariate logistic regression analysis (MVA) was used to identify predictors for early mortality (inpatient mortality). Results During the study period of 2002-2015, we identified 68,296 hospitalizations (median age 50 years and 58% male) undergoing inpatient AlloSCT (baseline characteristics in table 1 & figure 1); death occurred in 5,198 (7.6%). Age <65, cord or bone marrow graft, total body irradiation (TBI), non-white race and higher comorbidity index were associated with higher mortality on the MVA (figure 2). Complications associated with AlloSCT hospitalizations included: invasive fungal infections (IFI) in 2,605 (4%), acute graft vs. host disease (GVHD) in 3,628 (5%), CMV in 3,949 (6%), acute kidney injury (AKI) in 10,191 (15%) and bacterial sepsis in 14,603 (21%). Occurrence of these complications resulted in higher mortality (table 2). TBI, bacterial sepsis, IFI, CMV, AKI, lung (non-infectious) and gastrointestinal (ulcer, bleeding) complications were also predictors of higher mortality on the MVA. Acute GVHD, acute venous thromboembolism (VTE), and cardiac complications (arrhythmia, ischemia) did not predict for higher mortality on the MVA. Conclusion Our analysis identified novel prognostic and predictive factors associated with higher inpatient mortality during SCT hospitalization. Identifying at risk patients can provide an opportunity for early intervention and improving post-transplant outcomes.

Introduction Use of Axi-cel in patients with refractory and relapsed (R/R) large B-cell lymphoma (BCL) is associated with significant inpatient and outpatient resource utilization, which has not been systematically evaluated. We characterized resources used around Axi-cel infusion and identified factors that lead to longer hospitalization during the first 100 days of therapy. Methods We reviewed medical records of consecutive adult patients with R/R large BCL treated with Axi-cel at our center from May 2018 to June 2019. Resource utilization data were collected across 3 time points and sorted into categories that reflect inpatient and outpatient resources (Table 1). We evaluated the number of “days alive and out-of-hospital through Day 100 from cell infusion” (DAOH100), as a surrogate for institutional resource utilization and to identify factors leading to prolonged hospitalization. Results Axi-cel was used in 27 patients with R/R BCL; 18 were male (67%) with median age 63 years (IQR 48-68), median IPI 2 (IQR 2-3), 20 (74%) had ECOG PS < 2 and 24 (89%) had stage ≥ 3 disease. Median number of prior therapies was 3 (IQR 3-5) and HCT-CI score was 2 (IQR 0-3). The median length of stay for the initial hospitalization was 13 days (IQR 9-16) and through Day 100 was 16 days (IQR 12-30). Eight (30%) patients required ICU admission for a median of 2.5 days (IQR 1-9.8), while vasopressors, mechanical ventilation and dialysis were used in 6 (22%), 3 (11%) and 2 (7%) patients respectively. Patients had a median of 5 (IQR 3-7) outpatient clinic visits through Day 100. The median number of radiological studies and cardiac/neurological evaluation through Day 100 were 6 (IQR 4-11) and 1 (IQR 0-6) respectively. Pharmaceutical resources are summarized in Table 1. Cytokine release syndrome (CRS) and CAR-related encephalopathy syndrome (CRES) were seen in 22 (82%) and 20 (74%) patients, with ≥ Grade 2 CRS (Lee, et al) and CRES (CARTOX-10) in 16 (59%) and 13 (48%) patients respectively. By Day 100, 24 (89%) patients were alive and 4 (15%) had disease progression. The median DAOH100 were 84 days (IQR 69-88). DAOH100 were higher in patients with favorable ECOG PS (< 2) compared to unfavorable (≥ 2) (median DAOH100 86 vs 70 days, P=0.04). Patients with no or mild CRES (< grade 2) had higher DAOH100 compared to those with moderate to severe CRES (≥ grade 2) (median DAOH100 86 vs 70 days, P=0.01). DAOH100 did not differ with respect to age, HCT-CI score, IPI, number of prior therapies or CRS grade (Table 2). Conclusion In this single institutional experience of using Axi-cel therapy for R/R BCL, we demonstrated utilization of substantial resources in terms of hospitalization, ICU stay, diagnostic studies and pharmaceutical products. Patients with favorable PS and no or minimal CRES spend a higher number of days at home (alive and out-of-hospital), in first 100 days of Axi-cel therapy.

Background Despite improvements in conditioning regimens and supportive care having expanded the curative potential of HCT, underutilization of HCT in older adults persists (Bhatt VR et al, BMT 2017). Therefore, we conducted a survey of transplant physicians (TP) to determine their perceptions of the impact of older age (≥60 years) on HCT candidacy and utilization of tools to gauge candidacy. Methods We conducted a 23-item, online cross-sectional survey of adult physicians recruited from the Center for International Blood and Marrow Transplant Research between May and July 2019. Results 175/770 (22.7%) TP completed the survey; majority of respondents were 41-60 years old, male, and practicing in a teaching hospital. Over 75% were at centers performing ≥50 HCT per year. When considering regimen intensity, most (96%, n=168) had an upper age limit (UAL) for using a myeloablative regimen (MAC), with only 29 physicians (17%) stating they would consider MAC for patients ≥70 years. In contrast, when considering a reduced intensity/non-myeloablative conditioning (RIC/NMA), 8%, (n=13), 54% (n=93), and 20% (n=35) stated that age 70, 75, and 80 years respectively would be the UAL to use this approach, with 18% (n=31) reporting no UAL. TP agreed that Karnofsky Performance Score (KPS) could exclude older pts for HCT, with 39.1% (n=66), 42.6% (n=72), and 11.4% (n=20) requiring KPS of ≥70, 80, and 90, respectively. The majority (n=92, 52.5%) indicated an HCT-comorbidity index threshold for exclusion, mostly ranging from ≥3 to ≥ 5. Almost all (89.7%) endorsed the need for a better health assessment of pre-HCT vulnerabilities to guide candidacy for pts ≥60 with varied assessments being utilized beyond KPS (Figure 1). However, the majority of centers rarely (33.1%) or never (45.7%) utilize a dedicated geriatrician/geriatric-oncologist to assess alloHCT candidates ≥60 yrs. The largest barriers to performing GA included uncertainty about which tools to use, lack of knowledge and training, and lack of appropriate clinical support staff (Figure 2). Approximately half (n=78, 45%) endorsed GA now routinely influences candidacy. Conclusions The vast majority of TP will consider RIC/NMA alloHCT for patients ≥70 years. However, there is heterogeneity in assessing candidacy. Incorporation of GA into a standardized and easily applied health assessment tool for risk stratification is an unmet need. The recently opened BMT CTN 1704 may aid in addressing this gap.

Advent of novel anti-myeloma agents and broader use of stem cell transplant has led to significant improvement in survival of patients (pts) with Multiple Myeloma (MM). However, there are well-described issues with affordability of novel drugs and rapidly escalating price of these agents (Shih et al. JCO 2017), leading to significant disparity among different sociodemographic groups. Hereby, we interrogated the National Cancer Database (NCDB) (which covers 70% of MM patient diagnosed nationwide) to assess impact of insurance type on survival. We also sought to investigate if autologous transplant may overcome socioeconomic effects on survival, by potentially minimizing the need for chronic use of expensive drugs. Methods: Data from 117,926 MM pts diagnosed with MM (ICD-O 9732) between 2005 and 2014 were analyzed. Results: Median age at diagnosis was 67 (19-90); 55% were males. 57% of pts lived in areas where the median income was < $46k/year (individual income data was not available); Primary insurance was Medicare (52%), private insurance (35%) or Medicaid (5%), and 3% were uninsured. 40% were treated in academic institutions. Median follow up was 30.2 (range, 0-145.2) months. By univariate analysis, better OS was observed in pts with primary MM, lower Charlson Comorbidity Index (CCI), treatment in academic institutions, higher median regional income, or private insurance (p<0.001 for all) (Table-1). 96% of pts were treated in facilities located ≤ 120 miles from area of residence. Amongst pts younger than 65 years, 33% of pts with private insurance received transplant compared to 20% of those on Medicare (p<0.001). For those 65 years and older, 11% of privately insured pts were transplanted compared to only 6 % for those on Medicare (p<0.001). Median age of pts on Medicare and private insurance, was 74 and 57 years old, respectively. When restricting the analysis to pts ≥ 65 years old, pts with private insurance had longer OS compared to Medicare pts (p<0.001). Table-2 shows the results of multivariate analysis. There was a statistically significant difference in survival between patients with private insurance and those with Medicare in favor of the private insurance among pts older than 65 years old (41.9 vs. 30.8 months, p<0.001 (Fig-1)). Similarly insurance type was a significant predictor of survival among pts who received therapies other than transplant among pts younger or older than 65 (Fig-2), however when considering pts who received transplant, there was no difference in survival between privately insured pts and Medicare in both age group (Fig-3). Conclusions: Although insurance type and regional income are associated with MM survival among pts who relied on non-transplant modalities, there was no significant impact of these socioeconomic factors on survival of pts that received an autologous transplant in this large database. This finding merits further investigation.