T-ALL relapse usually occurs early but can occur much later, which has been suggested to represent a de novo leukemia. However, we conclusively demonstrate late relapse can evolve from a pre-leukemic subclone harbouring a non-coding mutation that evades initial chemotherapy.

Introduction Minimal residual disease (MRD) monitoring is fundamental for risk stratification and treatment in acute lymphoblastic leukemia (ALL). Sensitive molecular assay requires presentation samples to be screened for immunoglobulin and T-cell receptor (IG/TCR) gene rearrangements. Previously samples were screened by PCR, heteroduplex analysis and Sanger sequencing to identify clone defining IG/TCR gene rearrangements. Subsequently these sequences were used to develop qPCR assays for MRD follow-up monitoring. More recently, targeted IG/TCR high-throughput sequencing (HTS) has been used with similar panels to identify gene rearrangements. Both these methods have limitations, in that only rearrangements covered by the panels are detected, and the techniques are labour intensive and expensive. Whole genome sequencing (WGS) has potential to discover all IG/TCR rearrangements in a single assay giving greater potential to detect markers for follow up MRD monitoring. In England all patients with ALL are eligible for WGS at diagnosis via the National Health Service, Genomic Medicine Service (NHS GMS). Methods Sixty-two patients, presenting with ALL at either Great Ormond Street Hospital or Bristol Children's Hospital, had short-read WGS via the NHS GMS performed concurrently with standard of care (SOC) targeted HTS to identify IG/TCR rearrangements. A bespoke bioinformatic pipeline was developed to identify IG/TCR rearrangements in the WGS samples. SOC targeted HTS analysis was performed using the open source Vidjil bioinformatic software. A cut-off of 5% (clone proportion of total clonotypes) was used to define an acceptable disease-specific clone for monitoring. Results All 62 samples had a greater number of rearrangements detected by WGS than targeted HTS. The mean number of clones >5% detected was 2.7 by targeted HTS and 3.1 by WGS. For rearrangements identified by targeted HTS at a 5% cut-off, 97% (185 out of 190) were identified by WGS. All targets were identified by WGS using a 10% cut-off (Table 1).Despite the lower depth of coverage of the WGS assay, it retrieved 89% of sequences seen in SOC down to a level of 1%, as well as additional re-arrangements not seen by the SOC assay (Figure 1). As well as additional rearrangements being discovered by WGS, we observed novel rearrangements (dd3/ja rearrangements) not previously identified using a targeted approach. To establish if the additional sequences identified by WGS were clinically applicable as MRD targets, we designed allele specific oligonucleotides for a representative 32 rearrangements for testing by qPCR. Over half (n=16) were able to be used to track MRD in follow-up samples. New primer/probe sets were designed for the novel dd3/ja rearrangements identified by WGS, these were also tested by qPCR and performed well enough to be clinically applicable for MRD monitoring. Conclusions WGS approach outperformed targeted HTS, identifying more IG/TCR rearrangements suitable for molecular MRD monitoring in all patients. This is of particularly benefit in cases where targeted HTS only identifies 1 or 2 potential MRD targets. WGS also has the advantage of identifying other key gene rearrangements (BCR-ABL, KMT2A, SIL-TAL, IKZF1) that make excellent targets for MRD monitoring. Beyond this, WGS delivers all diagnostic information, germline and additional somatic findings that aid diagnosis. For ALL, in one test WGS can now replace all molecular diagnostics. In England all patients with leukemia are eligible for centralised WGS. With continued improvements in turnaround times, and reducing costs, we propose that WGS should be the new diagnostic standard for pediatric ALL

Background Minimal residual disease (MRD) measured on end-of-induction bone marrow (BM) is the most important biomarker for guiding therapy in pediatric acute lymphoblastic leukemia (ALL). Due to limited sensitivity of current approaches, peripheral blood (PB) is not a reliable source for identifying patients needing treatment changes. We sought to determine if high-throughput sequencing (HTS) (next-generation sequencing) of rearranged immunoglobulin and T-cell receptor genes can overcome this and be used to measure MRD in PB. Procedure We employed a quantitative HTS approach to accurately measure MRD from one million cell equivalents of DNA from 17 PB samples collected at day 29 after induction therapy in patients with precursor B-cell ALL. We compared these results to the gold-standard real-time PCR result obtained from their paired BM samples, median follow-up 49 months. Results With the increased sensitivity, detecting up to one abnormal cell in a million normal cells, we were able to detect MRD in the PB by HTS in all those patients requiring treatment intensification (MRD >= 0.005% in BM). Conclusion This is proof of principle that using the increased sensitivity of HTS, PB can be used to measure MRD and stratify children with ALL. The method is cost effective, rapid, accurate, and reproducible, with inherent advantages in children. Importantly, increasing the frequency testing by PB as opposed to intermittent BM sampling may allow extension of the dynamic range of MRD, giving a more complete picture of the kinetics of disease remission while improving relapse prediction and speed of detection.

Although 90% of children with acute lymphoblastic leukemia (ALL) are now cured, the prognosis for infant-ALL remains dismal. Infant-ALL is usually caused by a single genetic hit that arises in utero: an MLL/KMT2A gene rearrangement (MLL-r). This is sufficient to induce a uniquely aggressive and treatment-refractory leukemia compared to older children. The reasons for disparate outcomes in patients of different ages with identical driver mutations are unknown. Using the most common MLL-r in infant-ALL, MLL-AF4, as a disease model, we show that fetal-specific gene expression programs are maintained in MLL-AF4 infant-ALL but not in MLL-AF4 childhood-ALL. We use CRISPR-Cas9 gene editing of primary human fetal liver hematopoietic cells to produce a t(4;11)/MLL-AF4 translocation, which replicates the clinical features of infant-ALL and drives infant-ALL-specific and fetal-specific gene expression programs. These data support the hypothesis that fetal-specific gene expression programs cooperate with MLL-AF4 to initiate and maintain the distinct biology of infant-ALL.

ABSTRACT Although 90% of children with acute lymphoblastic leukemia (ALL) are now cured 1 , the prognosis of infant-ALL (diagnosis within the first year of life) remains dismal 2 . Infant-ALL is usually caused by a single genetic hit that arises in utero : rearrangement of the MLL/KMT2A gene ( MLL-r ). This is sufficient to give rise to a uniquely aggressive and treatment-refractory leukemia compared to older children with the same MLL-r 3–5 . The reasons for disparate outcomes in patients of different ages with identical driver mutations are unknown. This paper addresses the hypothesis that fetal-specific gene expression programs co-operate with MLL-AF4 to initiate and maintain infant-ALL. Using direct comparison of fetal and adult HSC and progenitor transcriptomes we identify fetal-specific gene expression programs in primary human cells. We show that MLL-AF4 -driven infant-ALL, but not MLL-AF4 childhood-ALL, displays expression of fetal-specific genes. In a direct test of this observation, we find that CRISPR-Cas9 gene editing of primary human fetal liver cells to produce a t(4;11)/ MLL-AF4 translocation replicates the clinical features of infant-ALL and drives infant-ALL-specific and fetal-specific gene expression programs. These data strongly support the hypothesis that fetal-specific gene expression programs co-operate with MLL-AF4 to initiate and maintain the distinct biology of infant-ALL.

The reprogramming of a patient's immune system through genetic modification of the T cell compartment with chimeric antigen receptors (CARs) has led to durable remissions in chemotherapy-refractory B cell cancers. Targeting of solid cancers by CAR-T cells is dependent on their infiltration and expansion within the tumor microenvironment, and thus far, fewer clinical responses have been reported. Here, we report a phase 1 study (NCT02761915) in which we treated 12 children with relapsed/refractory neuroblastoma with escalating doses of second-generation GD2-directed CAR-T cells and increasing intensity of preparative lymphodepletion. Overall, no patients had objective clinical response at the evaluation point +28 days after CAR-T cell infusion using standard radiological response criteria. However, of the six patients receiving ≥108/meter2 CAR-T cells after fludarabine/cyclophosphamide conditioning, two experienced grade 2 to 3 cytokine release syndrome, and three demonstrated regression of soft tissue and bone marrow disease. This clinical activity was achieved without on-target off-tumor toxicity. Targeting neuroblastoma with GD2 CAR-T cells appears to be a valid and safe strategy but requires further modification to promote CAR-T cell longevity.

The molecular detection of minimal residual disease (MRD) is standard of care in acute lymphoblastic leukemia to personalize the stratification of patients to appropriate intensity chemotherapy regimens. High-throughput sequencing (HTS) techniques are driving changes to MRD methodologies. Our study demonstrates HTS can identify suitable diagnostic markers, even in cases where traditional screening has been unsuccessful. Markers identified by HTS were used to track MRD using standard real-time quantitative PCR. We show, with six patient examples, clinical benefits of utilizing HTS to screen diagnostic samples and its necessity when traditional screening techniques fail. This is practical evidence that current MRD diagnostic marker screening should be replaced by an HTS approach.

Introduction: The CARPALL study (NCT02443831) employed a novel CD19CAR (CAT-41BBz CAR) with a faster off rate than the Kymriah FMC63-41BBz CAR (CAT 3.1x10-3s-1, FMC 6.8 x 10-5s-1), with equivalent on-rate (CAT 2.2 x 105, FMC 2.1 x 105). We herein report updated outcomes and CAR T cell persistence with an additional 6 months follow up from a submitted manuscript (Ghorashian et al., Nat Med, submitted) Methods: Patients aged <25 years with high risk, relapsed CD19+ B-ALL were eligible on this multi-centre, open label, non-randomised phase I study of autologous CAT-41BBz CAR T cells. Patients were followed to a data cut-off of 07/18/2019. CAT-41BBz CAR T cells were generated by magnetic bead activation of leucapheresed PBMCs, lentiviral transduction, followed by bioreactor expansion and magnetic bead removal prior to cryopreservation. All patients received lymphodepletion (fludarabine + cyclophosphamide) followed by 1x106/kg CAR T cells. Presence of CAR T cells in the blood and bone marrow (BM) was assessed (flow cytometry and qPCR) monthly for 6 months, then 6 weekly to 1 year and then 3 monthly. BM MRD was assessed (IgH qPCR, flow cytometry) at the same time-points up to 2 years to establish durability of responses as a stand-alone therapy. Primary end-points were incidence of grade 3-5 toxicity and the proportion of patients achieving molecular remission. Results: Of 17 patients recruited, 14 were treated due to manufacturing failure in 3 patients.The median age was 9 years (range 1-19 years). All patients had advanced ALL with a median of 4 prior therapy lines. 10 of 14 patients (71%) had relapsed post allogeneic SCT. Prior to lymphodepletion, 4 patients had >5% BM disease, 6 had disease between 5x10-2and 1x10-5, 4 were BM MRD negative having had recurrent isolated CNS disease. Median transduction efficiency was 31% (range 16.5 to 96.4%). 12/14 treated patients received the anticipated dose of 1x106CAR T cells/kg (2 received 0.9x106/kg). Considering all evaluable patients, (n=14 for CAR T cell persistence by qPCR, n=13 by flow) the geometric mean of Cmax was 128 912/µg DNA and of the area under the curve between D0 and D28 was 1,721,355 copies/ µg DNA (Table 1). At the point of maximal expansion, a median of 35% of circulating T cells were CAR+. Median half-life was 34 days (range 3-102). CAR T cells continued to be detectable by qPCR in 11 of 14 (79%) patients at last assessment and by flow cytometry up to 30 months post infusion in 8 of 13(61%). Median duration of CAR T persistence by flow was 261 days (range 7-917). 3 patients failed to have persistence of CAR T cells beyond 1 month. T cell mediated anti-CAR specific cytotoxic activity was detected in 2/2 evaluable patients. Updated persistence data will be presented at the meeting Cytokine release syndrome (CRS) occurred in 13 (93%, grade 1 n=9, grade 2 n=4). None developed ≥grade 3 CRS, had CRS-related ICU admission, or received Tocilizumab. CRS was associated with modest elevations of IL-6, IFN-γand IL-10. Grade 2 neurotoxicity was observed in 3 patients and resolved spontaneously. One patient had grade 4 leucoencephalopathy presumed due to chemotherapy as well as grade 5 sepsis. Ten patients (71%) had grade 3-4 cytopenia persisting beyond day 28 or recurring afterthis. 12/14 (86%) patients achieved molecular complete or continuing complete remission at a median of 30 days post infusion (range 30-90 days, Table 2). At a median follow-up of 20.3 months, 4/14 (29%) evaluable patients remain MRD negative. 5 relapsed with CD19-disease, 1 with CD19+ disease. The median duration of EFS (based on death or morphological relapse) has not been reached, 12 month EFS = 52%, OS = 70% (Figures 1, 2 and Table 3). Conclusion: We noted excellent CAR T cell expansion and persistence in a ALL cohort treated with the fast off-rate CAT-41BBz CAR despite their lower BM disease at treatment compared to other studies. The kinetics documented for all evaluable patients showed a 5-fold greater CAR T cell expansion and 2-fold longer half-life than responders in published series utilising tisagenlecleucel in a similar ALL cohort (Mueller et al., Blood 2017). Patients had a favourable toxicity profile with no severe (grade 3-4) CRS and equivalent disease outcomes to the ELIANA study despite having similarly advanced disease (Maude et al., NEJM 2018292). These data suggest long lived CAR T cell persistence supports stand-alone therapy for ALL with durable responses. Disclosures Ghorashian: Celgene: Honoraria; novartis: Honoraria; UCLB: Patents & Royalties: UCLB. Kramer:UCLB: Patents & Royalties. Ciocarlie:Servier: Other: Financial Support. Farzaneh:Autolus Ltd: Equity Ownership, Research Funding. Pule:Autolus: Employment, Equity Ownership, Patents & Royalties. Amrolia:UCLB: Patents & Royalties.

Chimeric antigen receptor (CAR)-modified T cells targeting CD19 demonstrate unparalleled responses in relapsed/refractory acute lymphoblastic leukemia (ALL)(1-5), but toxicity, including cytokine-release syndrome (CRS) and neurotoxicity, limits broader application. Moreover, 40-60% of patients relapse owing to poor CAR T cell persistence or emergence of CD19(-) clones. Some factors, including the choice of single-chain spacer(6) and extracellular(7) and costimulatory domains(8), have a profound effect on CAR T cell function and persistence. However, little is known about the impact of CAR binding affinity. There is evidence of a ceiling above which increased immunoreceptor affinity may adversely affect T cell responses(9-11). We generated a novel CD19 CAR (CAT) with a lower affinity than FMC63, the high-affinity binder used in many clinical studies(1-4). CAT CAR T cells showed increased proliferation and cytotoxicity in vitro and had enhanced proliferative and in vivo antitumor activity compared with FMC63 CAR T cells. In a clinical study (CARPALL, NCT02443831), 12/14 patients with relapsed/refractory pediatric B cell acute lymphoblastic leukemia treated with CAT CAR T cells achieved molecular remission. Persistence was demonstrated in 11 of 14 patients at last follow-up, with enhanced CAR T cell expansion compared with published data. Toxicity was low, with no severe CRS. One-year overall and event-free survival were 63% and 46%, respectively.

Human lymphopoiesis is a dynamic lifelong process that starts in utero 6 weeks post-conception. Although fetal B-lymphopoiesis remains poorly defined, it is key to understanding leukemia initiation in early life. Here, we provide a comprehensive analysis of the human fetal B-cell developmental hierarchy. We report the presence in fetal tissues of 2 distinct CD191 B-progenitors, an adult-type CD10+ve ProB-progenitor and a new CD10-ve PreProB-progenitor, and describe their molecular and functional characteristics. PreProB-progenitors and ProB-progenitors appear early in the first trimester in embryonic liver, followed by a sustained second wave of B-progenitor development in fetal bone marrow (BM), where together they form > 40% of the total hematopoietic stem cell/progenitor pool. Almost one-third of fetal B-progenitors are CD10-ve PreProB-progenitors, whereas, by contrast, PreProB-progenitors are almost undetectable (0.53% +/- 0.24%) in adult BM. Single-cell transcriptomics and functional assays place fetal PreProB-progenitors upstream of ProB-progenitors, identifying them as the first B-lymphoid-restricted progenitor in human fetal life. Although fetalBMPreProB-progenitors and ProB-progenitors both give rise solely to B-lineage cells, they are transcriptionally distinct. As with their fetal counterparts, adult BM PreProB-progenitors give rise only to B-lineage cells in vitro and express the expected B-lineage gene expression program. However, fetal PreProB-progenitors display a distinct, ontogeny-related gene expression pattern that is not seen in adult PreProB-progenitors, and they share transcriptomic signatures with CD10-ve B-progenitor infant acute lymphoblastic leukemia blast cells. These data identify PreProB-progenitors as the earliest B-lymphoid-restricted progenitor in human fetal life and suggest that this fetal-restricted committed B-progenitor might provide a permissive cellular context for prenatal B-progenitor leukemia initiation.

Amplicon-based next-generation sequencing (NGS) of immunoglobulin (IG) and T-cell receptor (TR) gene rearrangements for clonality assessment, marker identification and quantification of minimal residual disease (MRD) in lymphoid neoplasms has been the focus of intense research, development and application. However, standardization and validation in a scientifically controlled multicentre setting is still lacking. Therefore, IG/TR assay development and design, including bioinformatics, was performed within the EuroClonality-NGS working group and validated for MRD marker identification in acute lymphoblastic leukaemia (ALL). Five EuroMRD ALL reference laboratories performed IG/TR NGS in 50 diagnostic ALL samples, and compared results with those generated through routine IG/TR Sanger sequencing. A central polytarget quality control (cPT-QC) was used to monitor primer performance, and a central in-tube quality control (cIT-QC) was spiked into each sample as a library-specific quality control and calibrator. NGS identified 259 (average 5.2/sample, range 0–14) clonal sequences vs. Sanger-sequencing 248 (average 5.0/sample, range 0–14). NGS primers covered possible IG/TR rearrangement types more completely compared with local multiplex PCR sets and enabled sequencing of bi-allelic rearrangements and weak PCR products. The cPT-QC showed high reproducibility across all laboratories. These validated and reproducible quality-controlled EuroClonality-NGS assays can be used for standardized NGS-based identification of IG/TR markers in lymphoid malignancies.

Optimal management of central nervous system (CNS) infiltration is a key remaining challenge in delivering precision therapy for childhood acute lymphoblastic leukemia (ALL).[1][1] Most CNS relapses occur in children without high-risk features, and minimal residual disease (MRD) does not reliably

Abstract Background: Treatment of high risk neuroblastoma remains challenging; current multimodal treatment regimens achieve long term survival in <50% of patients and are associated with significant morbidity. Ganglioside GD2 is abundantly expressed on almost all neuroblastomas whilst expression on normal tissue is highly limited, providing a suitable CAR target. Here, we report the preliminary results of a Phase I clinical study of GD2-CART for refractory/relapsed neuroblastoma (NCT02761915). Trial design: The therapeutic (1RG-CART) is autologous T-cells transduced with a gamma-retroviral vector encoding both an anti-GD2 CAR and the RQR8 suicide gene. The CAR comprises a humanized anti-GD2 single chain variable fragment derived from the K666 antibody and CD28/CD3ζ signalling domains. Both lymphodepletion and CART dose were escalated as follows: dose level (DL) 1 without lymphodepletion, DL2 with 1.2 g/m2 cyclophosphamide and DL3 and beyond 1.2 g/m2 cyclophosphamide and 125 mg/m2 fludarabine followed by administration of a single intravenous dose of 1x107/m2 (DL1-3) or 1x108/m2 (DL4) 1RG-CART. Primary objectives are to assess safety and tolerability. Results: To date, 12 patients have been enrolled. All had relapsed/refractory neuroblastoma with measurable disease in bone (n=11), bone marrow (n=7) and/or soft tissue sites (n=9). Cell products were successfully manufactured for all patients. Median transduction efficiency was 34.5% (range 16-54%). Nine patients have been treated on DL1 (n=4), DL2 (n=1), DL3 (n=1) and DL4 (n=3) respectively. No dose limiting toxicity (DLT) was seen. For patients treated on DL1-3 (1x107/m2), 1RG-CART could not be detected in peripheral blood, and no clinical responses were seen. In contrast, expansion of 1RG-CART cells as detected by flow cytometry and qPCR was seen in the 3 patients treated on DL4 (1x108/m2). In two DL4 patients, 1RG-CART expansion was still limited and transient (marking levels <10,000 copies/μg DNA). These patients had disease progression as measured at Day +28. In one DL4 patient however, 1RG-CART marking levels of >40,000 copies/μg DNA were achieved. This patient developed Grade 2 cytokine release syndrome (Day +5) and biochemical evidence of tumour lysis (Day +21). Disease reassessment on Day +28 showed response in many sites of bone/marrow disease as measured by mIBG scintigraphy, and near complete tumour clearance in bone marrow which at baseline was heavily infiltrated with neuroblastoma. Disease progression occurred on Day +45 at which time 1RG-CART were no longer detectable. In the absence of DLT this prompted us to continue with DL5 (1x109/m2). Conclusions: These preliminary results are the first to demonstrate on-target activity in bone and bone marrow of GD2-CART in this childhood solid tumour. Further 1RG-CART dose escalation is warranted, and under way. Citation Format: Karin Straathof, Barry Flutter, Rebecca Wallace, Simon Thomas, Gordon Cheung, Angela Collura, Talia Gileadi, Jack Barton, Gary Wright, Sarah Inglott, David Edwards, Claire Barton, Karen Dyer, Nigel Westwood, Thalia Loka, Sarita Depani, Karen Howe, Giuseppe Barone, Martin Pule, John Anderson. A Cancer Research UK phase I trial of anti-GD2 chimeric antigen receptor (CAR) transduced T-cells (1RG-CART) in patients with relapsed or refractory neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr CT145.

Universal gene-edited CAR19 T cells eliminate infant leukemia.

Abstract Introduction: Published studies of CD19 CAR T cells have shown unprecedented response rates in ALL but with a 23-27% incidence of severe Cytokine Release Syndrome (CRS) and 27-50% incidence of severe neurotoxicity which may limit broader application. We developed a novel second generation CD19CAR (CAT-41BBz CAR) with a lower affinity and faster off-rate but equivalent on-rate than the FMC63-41BBz CAR (Kd 116 nM vs 0.9 nM, T1/2 10s vs 1260s) utilised in CTL019 currently under consideration by the FDA. Pre-clinical studies indicated T-cells transduced with CAT-41BBz mediate enhanced tumor clearance and show increased expansion in an NSG-NALM6 stress test model (Kramer et al., submitted). We here report interim results from a multi-centre, Phase I clinical study of autologous CAT-41BBz CAR T cells as therapy for high risk/relapsed paediatric ALL, CARPALL (NCT02443831) demonstrating efficacy with an excellent safety profile. Methods: Autologous T cells were activated with anti-CD3/CD28 beads, transduced with a SIN lentiviral vector encoding CAT-41BBz CAR and expanded for 4 days prior to magnetic bead removal and cryopreservation. Transduction efficiency was assessed using an anti-idiotype antibody. Serum levels of cytokines associated with CRS were measured using cytometric bead array. All patients received lymphodepletion with fludarabine 150 mg/m2 + cyclophosphamide 1.5g/m2 followed by a single infusion of CAR T cells at a dose of 1x106 CAR+ T cells. Patients were monitored for the presence of CAR T cells in the blood by flow cytometry and by qPCR for the 41BBz junctional region, as well as circulating B cell count monthly for 6 months and then 6 weekly to 1 year. Disease status was assessed in the bone marrow morphologically, by IgH qPCR, as well as by flow cytometric assessment of MRD at the same time-points to establish durability of responses as a stand-alone therapy. The primary end-points were incidence of grade 3-5 toxicity related to CAR T cells within 30 days and the proportion of patients achieving molecular remission. Results: We have enrolled 10 patients and treated 8 to date. Six of 8 had relapsed post myeloablative SCT. The median disease burden prior to lymphodepletion was 9% blasts (ranging from molecular CR to 74% blasts, Table 1). It was possible to generate a product meeting release criteria in all but 1 patient (90% feasibility). Median transduction efficiency was 18.1% (range 6.7 to 76.3%). All treated patients received the anticipated dose of 1x106 CAR T cells/kg. Cytokine release syndrome occurred in all patients (grade 1 n=4, grade 2 n=4), but to date none have developed ≥ grade 3 CRS, required ICU admission or therapy with Tocilizumab. CRS was associated with modest elevations of IL-6, IFN-γ and IL-10 and resolved spontaneously in all. Grade 2 neurotoxicity was observed in 3 patients and resolved spontaneously, but no severe (≥grade 3) neurotoxicity was seen. Five patients had prolonged grade 4 neutropenia lasting > 30 days but this resolved in all by 2 months. Only 1 patient experienced significant infective complications in the context of pre-existing poor marrow reserve following allogeneic SCT. 6/7 (86%) evaluable patients achieved molecular remission at a median of 30 days post infusion (range 30-60 days, Table 1). One patient did not respond and died of CD19+ disease progression. At a median follow-up of 5.9 months (range 28-328 days), 4/7 evaluable patients remain in flow MRD negative remission of whom 3 show no evidence of molecular MRD at 1, 7.5 and 9 months. Two patients relapsed with CD19- disease at 3 and 4 months post infusion: 1 of these remains alive with disease at 11 months and the other died of disease progression. Reflecting our pre-clinical data with CAT-41BBz CAR, we have seen excellent CAR T cell expansion (median 65459 copies/µg DNA at 1 month, range 609 to 230112) and persistence at up to 11 months post-infusion (Figure 1). All 7 evaluable patients have ongoing CAR T cell persistence detectable by both flow and qPCR as well as ongoing B cell aplasia at last follow-up. Conclusions: These interim results with a novel low affinity CD19 CAR show similar remission rates to those reported by US studies in paediatric ALL with an improved safety profile. No severe (grade ≥3) CRS or neurotoxicity has occurred to date despite high tumour burden in 4 patients. Excellent CAR T cell expansion has been documented, as well as long duration of CAR T cell persistence and associated B cell aplasia. Disclosures Ghorashian: UCL: Patents & Royalties: UCL Business. Kramer: UCL: Patents & Royalties: UCL Business. Lucchini: Alexion: Membership on an entity's Board of Directors or advisory committees. Pule: Autolus Ltd: Employment, Equity Ownership, Research Funding; UCL: Patents & Royalties: UCL Business.

Background UCART19 is a genetically modified T-cell product manufactured from non-HLA matched healthy donor cells. Lentiviral-transduced CAR T-cells express (1) an anti-CD19 CAR (anti-CD19 scFv- 41BB- CD3ζ) and (2) an RQR8 switch that is intended to allow targeted elimination of RQR8 + cells by rituximab. UCART19 has been additionally modified to disrupt the T-cell receptor alpha constant (TRAC) and CD52 genes. The preliminary results of this off-the-shelf allogeneic CAR T-cell therapy in a phase I, dose-escalation trial of UCART19 in CD19 + R/R B-ALL adult patients (pts) are described. Methods The primary objective of this study is to determine the maximum tolerated dose of UCART19 by investigating up to four dose levels (DL) in separate sequential cohorts. Adult pts (age ≥16 years) with CD19 + R/R B-ALL who have exhausted available treatment options are eligible. Disease burden must be quantifiable morphologically or with a minimal residual disease (MRD) load ≥1x10 -3 at the end of the last anti-leukemic treatment. The lymphodepletion regimen combines cyclophosphamide and fludarabine, with or without alemtuzumab (FC or FCA). A single dose of UCART19 is administered on Day 0, and pts are closely monitored for safety and anti-leukemic activity until the end of study, 3 months after UCART19 administration. Pts are then rolled-over into a 15-years long-term follow-up study. The dose escalation follows a modified Toxicity Probability Interval (mTPI) design based on the occurrence of dose-limiting toxicity (DLT) assessed at the end of the 28-day evaluation period post UCART19 (D28). Results As of 24 June 2017, the 2 first cohorts (3 pts each) who received the first DL (DL1=6x10 6 total CAR + cells) have been completed. Median age was 22.5 years (range 18-42). Pts received 1 to 5 previous lines of treatment with 5 out of 6 pts having undergone an allogeneic stem cell transplant (allo-SCT). Four of them had relapsed within 4-6 months post-transplant. Prior to UCART19 infusion, 4 pts had low disease burden ( All pts experienced cytokine release syndrome (CRS): 1 G1, 4 G2 and 1 G4. CRS G1 and G2 were manageable by supportive care ± tocilizumab. CRS G4, assessed as a DLT, occurred in the context of neutropenic sepsis, and was considered to be a contributory factor in the patient9s death from multiple organ failure at D15. Time to onset of first CRS symptoms ranged between D5 and D10. CRS correlated with serum cytokine increase (IL-6; IL-10 and INFγ) and UCART19 expansion in the blood. One patient was reported to have probable skin GvHD G1. Only G1 neurotoxic events were observed in 1 patient. Asymptomatic viral reactivations (CMV and/or adenovirus) were seen in 3 pts and resolved with antiviral therapy. Among the 6 pts, 4 achieved a CRi with MRD negativity at D28 (MRD-ve, defined as a tumor burden All 4 pts achieving MRD-ve remission underwent a subsequent allo-SCT, 3 of them within 3 months of UCART19 infusion and 1 following retreatment with FC lymphodepletion and the same dose of UCART19, this patient having relapsed with CD19+ disease 2 months post initial UCART19 infusion. Post allo-SCT, 1 patient relapsed at 100 days with CD19+ disease, 1 died from infection and 2 remain in complete remission. Three pts remain alive at 2.4, 5.3 and 10.2 months respectively post UCART19 treatment. UCART19 (both cells and transgene levels) peaked between D12 and D17 in blood (flow cytometry [figure 1] and qPCR, respectively). UCART19 was detectable in blood from D10 to D28 (up to D42 in 1 patient) and in BM aspirates performed at D14 and D28. In-vivo cell expansion in BM occurred in all but the refractory patient. Conclusion Preliminary results of this first-in-human trial of UCART19 treatment in a high risk R/R B-ALL adult population revealed no unexpected toxicities. Asymptomatic lymphodepletion-related viral reactivations and a probable skin GvHD G1 were encountered. CRi with MRD-ve was achieved in 4 out of 5 pts who reached D28. The 2 first cohorts treated at DL1 have been completed and DL2 will now be investigated on which further results may be presented. The study is active in the UK and will be expanded to other EU countries and the US (NCT 02746952). Disclosures Graham: Servier: Research Funding; Pfizer: Other: Educational meeting attendance; Gilead: Other: Educational meeting attendance; Sanofi: Other: Educational meeting attendance. Yallop: Jazz Pharmaceuticals: Honoraria; Amgen: Honoraria; Pfizer: Other: Advisory board. Jozwik: Servier: Research Funding. Patten: Gilead Inc: Honoraria, Research Funding; Roche: Honoraria; Abbvie: Honoraria. Ellard: Moldmed: Honoraria. Potter: Pfizer: Other: Advisory board; Jazz: Honoraria. Devereux: AbbVie: Consultancy, Honoraria; MSD: Consultancy, Honoraria; Roche: Consultancy, Other: travel expenses; GSK: Consultancy; Gilead: Consultancy, Honoraria, Other: travel expenses, Speakers Bureau; Janssen: Consultancy, Honoraria, Other: travel expenses, Speakers Bureau; Servier: Other: Advisory board. Pagliuca: Jazz: Honoraria; Merck: Honoraria, Research Funding; Bluebird: Honoraria; Pfizer: Honoraria; Basilea: Honoraria; Astellas: Consultancy, Speakers Bureau; Gilead: Honoraria. Zinai: Servier: Employment. Binlich: Servier: Employment. Dupouy: Servier: Employment. Philippe: Servier: Employment. Balandraud: Servier: Employment. Dubois: Servier: Employment. Konto: Bristol-Myers Squibb: Employment, Equity Ownership; Pfizer: Employment, Equity Ownership. Patel: Pfizer: Employment, Equity Ownership. Benjamin: Pfizer: Other: Participated in Adboard meeting, Research Funding; Servier: Research Funding; Celgene: Honoraria.

PCR bias is a potential confounder for PCR-based NGS-MRD quantitation. A method was published using artificial DNA constructs (gBlocks, IDT Technologies, Coralville, IO) to assess PCR primer bias and correct for it(1). We sought to confirm those findings by assessing PCR bias of the Biomed-2 TRG primer set.

Abstract Introduction: Recent clinical trials with T cells engineered to express 2nd generation CD19 chimeric antigen receptors (CARs) unprecedented anti-leukemic responses. We have developed a novel CD19CAR with a new scFv in the 41BBz format (CAT-41BBz CAR) which confers enhanced cytotoxicity and cytokine secretion in response to stimulation with CD19+ targets in vitro as well as equivalent in vivo anti-tumour efficacy to the FMC63 41BBZ CAR in use in clinical studies. We have designed, optimized and validated GMP-grade CAR T cell production using this novel CAR. Based on these data, we have recently initiated a Phase I clinical study (CARPALL) of this novel CAR in pediatric patients with relapsed ALL and other CD19+ hematological malignancies to determine the safety profile and durability of responses to CD19CART therapy. This will be critical in determining whether CD19CAR T cells are best used as a stand-alone therapy or as a bridge to stem cell transplant (SCT). Methods: We initially optimized our GMP production methodology in terms of activation method, cytokine milieu and expansion conditions on healthy donor peripheral blood mononuclear cells (PBMCs) to give optimal transduction efficiency and preserve early memory subsets within the CAR T cell product. We have subsequently validated this methodology using unstimulated leucaphereses from 5 lymphopenic patients with ALL. PBMCs were activated with anti-CD3/CD28 microbeads (Dynabeads CTS) and then lentivirally transduced with the CAT CAR vector. T cells were then expanded in the WAVE bioreactor before bead removal on a magnetic system and cryopreservation. Patients on study receive lymphodepletion with fludarabine and cyclophosphamide followed by a single dose of 106 CAR+ T cells/kg and are then monitored as an in-patient for 14 days post infusion for toxicities such as cytokine release syndrome or neurotoxicity. The primary end-points of the study are toxicity and the proportion of patients achieving molecular CR at 1 month post CD19CAR T cell infusion. Following this, patients undergo intensive monitoring of disease status for a total of 2 years post infusion. To determine the durability of responses, patients achieving a molecular CR will be monitored closely for the re-emergence of molecular level disease without additional consolidative therapy or SCT Results: We were able to generate the target dose of 1x106 CAR+ T cells/kg in 6 of 7 production runs (involving 2 healthy donors and 5 patients) to date, all of which met sterility release criteria. Transduction efficiency was on average 37% (range 7-84%, see table 1). Mean viral copy was 4.2 (range 1.2-5.8). Memory T cells of stem cell-like phenotype (CAR+ CCR7+ CD45RA+ CD95+ CD127+) formed a mean of 9% (range 0-31%), central memory T cells (CAR+ CCR7+ CD45RA-) formed a mean of 43% (range 16-70%) and effector memory T cells formed a mean of 31% (range 0-77%) of the final CAR T cell product. The percentage of CAR T cells expressing dual exhaustion markers (TIM3+ PD-1+) was on average 5% (range 2-8%). So far 2 patients have been treated. Conclusions We have optimized and successfully validated a robust GMP production method for CD19CAR T cells lentivirally transduced with a novel CD19CAR. Preliminary results of therapy with CAT-41BBz CAR T cells in initial patients on the clinical study will be presented. Disclosures Qasim: Autolus: Consultancy, Equity Ownership, Research Funding; Cellectis: Research Funding; Calimmune: Research Funding; Catapult: Research Funding. Pule:Autolus Ltd: Employment, Equity Ownership, Research Funding; UCL Business: Patents & Royalties; Amgen: Honoraria; Roche: Honoraria.