CD22 Chimeric Antigen Receptor Research Articles

Abstract 36: The discovery of a novel CD19xCD22 dual-targeting CAR for the development of an IPSC-derived cell therapy

Abstract Dual-targeting cell therapies to address CD19 antigen loss in heme malignancies are being explored by Century Therapeutics and others for unmet clinical need in hematological malignancies. CD22 chimeric antigen receptor (CAR) T-cell therapies have shown promising efficacy in relapsed or refractory B-lineage acute lymphoblastic leukemia (B-ALL) alone and in combination with CD19 CAR-T cell therapies (1,2). CD22 is a multi-domain glycoprotein expressed on B-cells. Natural splice variants of CD22 can exclude domains near the N-terminus, allowing for therapeutic resistance when these domains are targeted. We identified novel single domain antibodies (VHH) to CD22 that bind to multiple epitopes on different domains of CD22, including membrane proximal binders. The VHHs were affinity matured to enhance cell-based binding and formatted as CARs in monovalent and tandem, bivalent formats. Our data demonstrates the best efficacy against CD22+ tumor cells with a biparatopic, tandem VHH CAR format. Through additional engineering, bispecific CARs that combined three CD22-binding VHH and the anti-CD19 scFv, FMC63, in a loop CAR configuration were designed and tested. This CD19xCD22 bispecific, CD22 biparatopic CAR was engineered into primary T-cells and demonstrated cytotoxicity activity against CD19 and CD22 positive tumor cells as well as CD19 knockout and CD22 knockout cell lines. CD19xCD22 bispecific primary CAR-T cells were efficacious in controlling tumors in mouse xenograft models. This novel CAR is being tested in IPSC-derived iT and iNK cells for off-the-shelf allogeneic cell therapy to expand patient access beyond CD19 CAR-T cell therapies. Ref: 1. Nat Med. 2018;24(1):20-28. doi: 10.1038/nm.4441. 2. Blood. 2018;132(suppl 1). Abstract 27 Citation Format: Jill M. Carton, John Wheeler, Chris Dower, Liam Campion, Hillary J. Millar, Marilda Beqiri, Barry Morse, Buddha Gurung, Rebecca Genovese, Steven DeLuca, Charles Dominick, Michael Naso, Hy Levitsky. The discovery of a novel CD19xCD22 dual-targeting CAR for the development of an IPSC-derived cell therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 36.

Efficacy and Safety of Ino Combined with Reduce Dose R-GDP Regimen in Patients with Relapsed/Refractory B-Cell NHL in China

Background: The combination of immunochemotherapy resulted in satisfactory efficacy and survival for B-cell Non-Hodgkin Lymphoma (NHL). But many patients face no response to combination therapy or the disease progression again. This study aimed to investigate the efficacy and safety of Inotuzumab Ozogamicin (InO) combined with reduce dose R-GDP as a salvage therapy in patients with relapsed/refractory (R/R) B-NLH. Patients and methods: Seven patients with R/R B-NLH or even failure of anti-CD19 and CD22 chimeric antigen receptor (CAR) T cell therapy received InO combined with reduce dose R-GDP combination therapy. All patients received Rituximab (R) 375mg/m 2 intravenously (IV) on day -1, and InO 0.8 mg/m 2 (IV) on day 0. Then Gemcitabine (G) 500-1000mg/m 2(IV), Dexamethasone (D) 40 mg/day (IV) on day 1-4 and Cisplatin (P) 40-75mg/m 2 (IV) on day 1 of each 21-day cycle (±2 days). Treatment was continued for up to two cycles unless progressive disease (PD) or intolerable toxicity occurred. The primary endpoint of this study was the Complete remission (CR) and Overall response rate (ORR). The secondary endpoint of this study was Progression-free survival (PFS), Overall survival (OS) and Adverse events (AE), particularly hematological toxicity. In our study, the response evaluation was detected using computed tomography (CT) or PET-CT. Results: Seven patients with R/R B-NLH received 1-2 cycles of InO combined with reduce dose R-GDP therapy. Three patients did not respond to anti-CD19 and CD22 CAR-T cell therapy or their disease progressed again. The other four patients did not respond to more than 3 lines of combined immunochemotherapy and never achieved Partial response (PR) or CR. Four patients with R/R B-NLH who received 1-2cycles of InO combined with reduce dose R-GDP therapy achieved CR and survived to this day. Three of the four patients had previously received anti-CD19 and CD22 CAR-T cell therapy. Two patients with R/R B-NLH achieved PR after one cycle of InO combined with reduce dose R-GDP therapy, the other one patient achieved Stable disease (SD) only. One R/R B-NLH patient who achieved CR after one cycle of this salvage therapy had a PFS of one year to date without any other treatments. The other three R/R B-NLH patients who achieved CR after 1-2 cycles of this salvage therapy had a PFS of 3-6 months without any other treatments. The other two patients who achieved PR after 1 cycle of this salvage therapy survived 7 months and 2 months, while the only one patient who achieved SD survived one month without any other treatment. The main AE of this study was hematological toxicity. Three patients were defined as 3-4 grade of neutropenia, four patients were defined as 3-4 grade of anemia and three patients were defined as 3-4 grade of thrombocytopenia in this study. Except for one patient who was defined as grade 3 of thrombocytopenia recovered nine months after this salvage therapy, all other patients recovered two to three months after this study. There have been no deaths from severe infections due to hematological toxicity. Conclusions Ino combined with reduce dose R-GDP therapy in patients with R/R B-NLH could be an effective and safety therapy. However, the GDP regimen needs to be reduced appropriately to avoid severe hematological toxicity.

CD22 TCR-Engineered T Cells Exert Anti-Leukemia Cytotoxicity without Causing Inflammatory Responses

Chimeric antigen receptor (CAR) T cells are remarkably effective, especially against B-cell malignancies. However, CAR-T cell therapy is also associated with cytokine-driven inflammatory toxicities such as cytokine release syndrome. Similar cytokine-related toxicity can occur after bispecific T-cell engager (BiTE) administration. It is unknown whether these inflammatory responses are specifically related to T-cell-mediated therapies that are activated through artificial CAR or BiTE signals. Cytokine-driven inflammatory toxicities were not commonly encountered prior to the wide-spread use of CAR-T cells in the clinic–neither from the adoptive transfer of ex vivo-expanded autologous polyclonal antigen-specific T cells, tumor-infiltrating lymphocytes, nor TCR-engineered T cells (TCR-T cells). However, it is challenging to compare these diverse cell products, utilized in different clinical contexts and against different antigens. Preclinical comparisons of CARs and TCRs consistently showed lower sensitivity of CARs for antigen as compared to that of TCRs, at least partly due to inefficient downstream signaling of the former (Salter AI et al. Science Signaling. 2021; Gudipati V et al. Nat Immunol. 2020; Harris DT et al. J Immunol. 2018, Oren R et al. J Immunol. 2014). Nonetheless, it is not known whether these differences in signaling and T cell effector functions affect the preponderance of eliciting inflammatory responses. Here, we discovered a TCR targeting the CD22 antigen expressed by various B-cell malignancies in order to evaluate whether anti-leukemia/lymphoma T-cell cytotoxicity could be achieved without causing inflammatory responses by using TCR-activated T cells rather than CAR-activated T cells for the same antigen, CD22. A TCR specific to HLA-A*02:01-restricted epitope of human CD22 was isolated from an HLA-A2 negative healthy donor T cells through allogeneic in vitro stimulations. TCR alpha/beta-paired single cell sequencing was performed, and the sequence of the most enriched TCR clonotype was reconstructed in a retroviral expression vector. Primary human T cells transduced with this CD22 TCR recognized malignant B-cell lines only when they expressed both HLA-A*02:01 and CD22, confirming the intended specificity of the TCR. The TCR did not cross-react with other human proteins. CD22 TCR-T cells were then compared with CD22 CAR-T cells that are successful in the treatment of patients with CD22+ leukemia/lymphoma but associated with dose-limiting inflammatory toxicities (Shah NN et al. J Clin Oncol. 2020, Frank MJ et al. Blood. 2021). CD22 TCR-T cells mediated efficient in vitro cytotoxicity against CD22-dim lymphoma cells whereas CD22 CAR-T cells exerted suboptimal cytotoxicity. Importantly, TCR-T cells (1e7-3e7 cells/mouse) also cleared leukemia in xenograft models without inducing systemic proinflammatory cytokine elevation (Figure). In contrast, CD22 CAR-T cells, given at the dose necessary to control in vivo leukemia growth (5e6 cells/mouse), led to significantly increased levels of circulating proinflammatory cytokines such as IFNγ (Figure), TNFα, GM-CSF, MIP-1α and MIP-1β. Systemic cytokine elevation after CAR-T cell infusion exacerbated when NSG xenograft mice were pre-infused (24 hours prior to adoptive T cell transfer) with PBMC containing myeloid cell populations that were autologous to T cells. TCR-T cell therapy was not associated with proinflammatory cytokine elevation even in the presence of myeloid cell populations. Consistent with these in vivo observations, in vitro cytokine production by CAR-T cells were higher compared to TCR-T cells in response to CD22+ leukemia cells. Moreover, T cells activated through the CD22 TCR and CD22 CAR by the identical leukemia cell line demonstrated differential transcriptional signatures; CAR-T cells exhibited a disproportionate and significant upregulation of inflammatory pathways compared to TCR-T cells while effector response pathways were either equivalent or enriched in TCR-T cells. Together, these results highlight the opportunity for TCR-T cells to mediate anti-leukemia responses with a potentially safer toxicity profile compared to CAR-T cells. Our work provides the basis for incorporating TCR-based immunotherapy into the treatment paradigm for patients with relapsed/refractory B-cell malignancies, improving the safety of adoptive cell therapy.

Sequential CD19 and CD22 chimeric antigen receptor T-cell therapy for childhood refractory or relapsed B-cell acute lymphocytic leukaemia: a single-arm, phase 2 study

Relapses frequently occur following CD19-directed chimeric antigen receptor (CAR) T-cell treatment for relapsed or refractory B-cell acute lymphocytic leukaemia in children. We aimed to assess the activity and safety of sequential CD19-directed and CD22-directed CAR T-cell treatments. This single-centre, single-arm, phase 2 trial, done at Beijing GoBroad Boren Hospital, Beijing, China, included patients aged 1-18 years who had relapsed or refractory B-cell acute lymphocytic leukaemia with CD19 and CD22 positivity greater than 95% and an Eastern Cooperative Oncology Group performance status of 0-2. Patients were initially infused with CD19-directed CAR T cells intravenously, followed by CD22-directed CAR T-cell infusion after minimal residual disease-negative complete remission (or complete remission with incomplete haematological recovery) was reached and all adverse events (except haematological adverse events) were grade 2 or better. The target dose for each infusion was 0·5 × 106 to 5·0 × 106 cells per kg. The primary endpoint was objective response rate at 3 months after the first infusion. Secondary endpoints were duration of remission, event-free survival, disease-free survival, overall survival, safety, pharmacokinetics, and B-cell quantification. The prespecified activity analysis included patients who received the target dose and the safety analysis included all treated patients. This study is registered with ClinicalTrials.gov, NCT04340154, and enrolment has ended. Between May 28, 2020, and Aug 16, 2022, 81 participants were enrolled, of whom 31 (38%) were female and 50 (62%) were male. Median age was 8 years (IQR 6-10), all patients were Asian. All 81 patients received the first infusion and 79 (98%) patients received sequential infusions, CD19-directed CAR T cells at a median dose of 2·7 × 106 per kg (IQR 1·1 × 106 to 3·7 × 106) and CD22-directed CAR T cells at a median dose of 2·2 × 106 per kg (1·1 × 106 to 3·7 × 106), with a median interval of 39 days (37-41) between the two infusions. 62 (77%) patients received the target dose, including two patients who did not receive CD22 CAR T cells. At 3 months, 60 (97%, 95% CI 89-100) of the 62 patients who received the target dose had an objective response. Median follow-up was 17·7 months (IQR 11·4-20·9). 18-month event-free survival for patients who received the target dose was 79% (95% CI 66-91), duration of remission was 80% (68-92), and disease-free survival was 80% (68-92) with transplantation censoring; overall survival was 96% (91-100). Common adverse events of grade 3 or 4 between CD19-directed CAR T-cell infusion and 30 days after CD22-directed CAR T-cell infusion included cytopenias (64 [79%] of 81 patients), cytokine release syndrome (15 [19%]), neurotoxicity (four [5%]), and infections (five [6%]). Non-haematological adverse events of grade 3 or worse more than 30 days after CD22-directed CAR T-cell infusion occurred in six (8%) of 79 patients. No treatment-related deaths occurred. CAR T-cell expansion was observed in all patients, with a median peak at 9 days (IQR 7-14) after CD19-directed and 12 days (10-15) after CD22-directed CAR T-cell infusion. At data cutoff, 35 (45%) of 77 evaluable patients had CAR transgenes and 59 (77%) had B-cell aplasia. This sequential strategy induced deep and sustained responses with an acceptable toxicity profile, and thus potentially provides long-term benefits for children with this condition. The National Key Research & Development Program of China, the CAMS Innovation Fund for Medical Sciences (CIFMS), and the Non-Profit Central Research Institute Fund of Chinese Academy of Medical Sciences. For the Chinese translation of the abstract see Supplementary Materials section.

CD19/CD22 targeting with cotransduced CAR T cells to prevent antigen-negative relapse after CAR T-cell therapy for B-cell ALL

AbstractCD19-negative relapse is a leading cause of treatment failure after chimeric antigen receptor (CAR) T-cell therapy for acute lymphoblastic leukemia. We investigated a CAR T-cell product targeting CD19 and CD22 generated by lentiviral cotransduction with vectors encoding our previously described fast-off rate CD19 CAR (AUTO1) combined with a novel CD22 CAR capable of effective signaling at low antigen density. Twelve patients with advanced B-cell acute lymphoblastic leukemia were treated (CARPALL [Immunotherapy with CD19/22 CAR Redirected T Cells for High Risk/Relapsed Paediatric CD19+ and/or CD22+ Acute Lymphoblastic Leukaemia] study, NCT02443831), a third of whom had failed prior licensed CAR therapy. Toxicity was similar to that of AUTO1 alone, with no cases of severe cytokine release syndrome. Of 12 patients, 10 (83%) achieved a measurable residual disease (MRD)–negative complete remission at 2 months after infusion. Of 10 responding patients, 5 had emergence of MRD (n = 2) or relapse (n = 3) with CD19- and CD22-expressing disease associated with loss of CAR T-cell persistence. With a median follow-up of 8.7 months, there were no cases of relapse due to antigen-negative escape. Overall survival was 75% (95% confidence interval [CI], 41%-91%) at 6 and 12 months. The 6- and 12-month event-free survival rates were 75% (95% CI, 41%-91%) and 60% (95% CI, 23%-84%), respectively. These data suggest dual targeting with cotransduction may prevent antigen-negative relapse after CAR T-cell therapy.

A Predictive Model of Severe Cytokine Release Syndrome After Coadministration of CD19- and CD22-Chimeric Antigen Receptor T-Cell Therapy in Children With B-Cell Hematological Malignancies Based on Patient-Reported Outcomes.

Chimeric antigen receptor T-cell therapy-related severe cytokine release syndrome (sCRS) has seriously affected the life safety of patients. To explore the influencing factors of sCRS in children with B-cell hematological malignancies and build a risk prediction model. The study recruited 115 children with B-cell hematological malignancies who received CD19- and CD22-targeted chimeric antigen receptor T-cell therapy. A nomogram model was established based on symptomatic adverse events and highly accessible clinical variables. The model discrimination was evaluated by the area under the receiver operating characteristic curve. The calibration of our model was evaluated by the calibration curve and Hosmer-Lemeshow test. The bootstrap self-sampling method was used to internally validate. Thirty-seven percent of the children experienced sCRS. Indicators included in the nomogram were tumor burden before treatment, thrombocytopenia before pretreatment, and the mean value of generalized muscle weakness and headache scores. The results showed that the area under the receiver operating characteristic curve was 0.841, and the calibration curve showed that the probability of sCRS predicted by the nomogram was in good agreement with the actual probability of sCRS. The Hosmer-Lemeshow test indicated that the model fit the data well (χ2 = 5.759, P = .674). The concordance index (C-index) obtained by internal validation was 0.841 (0.770, 0.912). The nomogram model constructed has a good degree of discrimination and calibration, which provides a more convenient and visual evaluation tool for identifying the sCRS. Incorporation of patient-reported outcomes into risk prediction models enables early identification of sCRS.

Impact of fludarabine exposure on the depth and duration of lymphopenia and CAR T-cell outcomes.

10033 Background: Chimeric antigen receptor (CAR) T-cells are highly effective for patients with relapsed or refractory B-ALL. Lymphodepletion has been paramount to CAR T-cell expansion and efficacy. Recent data identified an association of fludarabine exposure with reduced relapse rates, longer leukemia free survival, and more prolonged B-cell aplasia following tisagenlecleucel. However, the relationship of fludarabine exposure to lymphopenia and the mechanism by which dosing impacts outcomes following CAR T-cells remains unknown. Methods: We conducted a retrospective review of children and young adults with B-ALL who received fludarabine for lymphodepletion (LD) prior to infusion of anti-CD19, anti-CD22, or anti-CD19/22 bispecific CAR T-cells. The primary objective of this study was to investigate the impact of fludarabine on the depth and duration of lymphopenia. Secondary objectives included examining the impact of fludarabine exposure on: 1) overall clinical response, and 2) CAR T-cell expansion. All patients in the analysis cohort received fludarabine 25 mg/m2/day x 3 days ( = 75 mg/m2 cumulative dosing). Fludarabine exposure (AUC per dose) for each patient was predicted using a validated population PK model and analyzed with various responses/endpoints using the 3-day cumulative exposure/patient. Results: Of 137 children and young adults, the median cumulative AUC was 8.3 mg*hr/mL (range, 3.00-13.83 mg*hr /mL). The overall complete response (CR) rate was: 67.9% (93/137). The median duration of lymphopenia (defined as an absolute lymphocyte count (ALC) £ 150 K/uL) was 5 days (range, 0-15); and the median ALC on Day 0 was 80 K/uL (range, 0-1250 K/uL). When stratified by the median fludarabine exposure, there was no difference in depth or duration of lymphopenia or CAR T-cell expansion collectively or per trial. The median fludarabine exposure also did not differ between responders and non-responders. However, depth and duration of lymphopenia did differ in the CD22 CAR T-cell trial. Specifically, in those achieving CR, the duration of lymphopenia was longer than in non-responders (7 days versus 4 days, p = 0.030). Similarly, CAR T-cell expansion was higher if lymphopenia duration was greater than 5 days (p = 0.019). These findings were not seen across our less persistent CD19 or CD19/22 CAR T-cell trials. Conclusions: Although lymphodepletion impacts CAR T-cell response, we did not find an association between CR rate, depth of lymphopenia, or duration of lymphopenia with varying fludarabine exposures. This raises additional questions regarding the impact of LD therapy on CAR T-cell outcomes—especially since a more prolonged lymphopenia was associated with high CR rates and greater peak expansion in at least one of our studies (CD22 CAR T-cells). Future analysis will evaluate the impact of fludarabine on Day 0 cytokines and other immune parameters. Clinical trial information: NCT03448393 .

A systematic review and meta-analysis of CD22 CAR T-cells alone or in combination with CD19 CAR T-cells.

https://www.crd.york.ac.uk/prospero, identifier CRD42020193027.

CLASH of the Titans: How CAR-T Cells Can Triumph Over Tumors.

CLASH of the Titans: How CAR-T Cells Can Triumph Over Tumors.

Digital polymerase chain reaction strategies for accurate and precise detection of vector copy number in chimeric antigen receptor T-cell products

Background aimsVector copy number (VCN), an average quantification of transgene copies unique to a chimeric antigen receptor (CAR) T-cell product, is a characteristic that must be reported prior to patient administration, as high VCN increases the risk of insertional mutagenesis. Historically, VCN assessment in CAR T-cell products has been performed via quantitative polymerase chain reaction (qPCR). qPCR is reliable along a broad range of concentrations, but quantification requires use of a standard curve and precision is limited. Digital PCR (dPCR) methods were developed for absolute quantification of target sequences by counting nucleic acid molecules encapsulated in discrete, volumetrically defined partitions. Advantages of dPCR compared with qPCR include simplicity, reproducibility, sensitivity and lack of dependency on a standard curve for definitive quantification. In the present study, the authors describe a dPCR assay developed for analysis of the novel bicistronic CD19 × CD22 CAR T-cell construct. MethodsThe authors compared the performance of the dPCR assay with qPCR on both the QX200 droplet dPCR (ddPCR) system (Bio-Rad Laboratories, Inc, Hercules, CA, USA) and the QIAcuity nanoplate-based dPCR (ndPCR) system (QIAGEN Sciences, Inc, Germantown, MD, USA). The primer–probe assay was validated with qPCR, ndPCR and ddPCR using patient samples from pre-clinical CAR T-cell manufacturing production runs as well as Jurkat cell subclones, which stably express this bicistronic CAR construct. ResultsddPCR confirmed the specificity of this assay to detect only the bicistronic CAR product. Additionally, the authors’ assay gave accurate, precise and reproducible CAR T-cell VCN measurements across qPCR, ndPCR and ddPCR modalities. ConclusionsThe authors demonstrate that dPCR strategies can be utilized for absolute quantification of CAR transgenes and VCN measurements, with improved test–retest reliability, and that specific assays can be developed for detection of unique constructs.

Programmed death-ligand 1 expression on CD22-specific chimeric antigen receptor-modified T cells weakens antitumor potential.

The molecules of programmed cell death protein‐1 (PD‐1) and ligand‐1 (PD‐L1) become new therapeutic targets for cancer therapy. Although tumor‐expressed PD‐L1 molecule is frequently dispensable for checkpoint blockade in some cancer patients, recent studies suggest that T cell‐expressed PD‐L1 molecule might play a crucial role in antitumor immunity. Here, to investigate CD22 chimeric antigen receptor (CAR)‐T cell therapy, we have generated the different CD22 CAR‐T constructs. We noticed that tumor cells induced PD‐L1 expression on the surface of CD22 CAR‐T cells. The induced PD‐L1 might limit immunogenic responses of CAR‐T cells. T cell‐expressed PD‐L1 leads to a suppressive signal by PD‐1/PD‐L1 engagement of CD22 CAR‐T cells. Meanwhile, PD‐L1 suppresses CD22 CAR‐T cell differentiation into memory T cells and negatively affected secretions of several essential cytokines, such as interleukin‐2 (IL‐2) and tumor necrosis factor (TNF)‐α. We further observed that anti‐PD‐L1 monoclonal antibodies rescued cytokine secretion of CD22 CAR‐T cells rather than anti‐PD‐1 monoclonal antibodies. Our current studies provide a potential mechanism to understand the functions and roles of T cell‐expressed PD‐L1 in tumor microenvironment. These results will encourage the physicians to re‐recognize the important roles of PD‐L1 in cancer immunotherapy studies and provide the helpful guidance for clinical operation of PD‐L1 inhibition drugs.

A novel and efficient CD22 CAR-T therapy induced a robust antitumor effect in relapsed/refractory leukemia patients when combined with CD19 CAR-T treatment as a sequential therapy

BackgroundCD19 chimeric antigen receptor (CAR) therapy has achieved impressive success in relapsed or refractory (R/R) B-cell malignancies, but relapse due to antigen escape is increasingly appearing reported. As the expression profile of CD22 is similar to that of CD19, CD22 has become a candidate target when CD19 CAR-T therapy fails.MethodsA novel CD22 CAR incorporating scFv derived from an HIB22 hybridoma which bound the first and second Ig-like extracellular domains of CD22 antigen was constructed. Preclinical investigation of the CD22 CAR-T therapy against B-cell malignancies was evaluated by coculturing CD22 CAR-T cells with tumor cell lines or primary blasts from patients in vitro and using a xenograft mouse model in vivo. Further clinical study of CD22/CD19 CAR-T sequential therapy was conducted in 4 R/R adult B-cell acute lymphoblastic leukemia (B-ALL) patients.ResultsThe novel CD22 CAR-T treatment had specific cytotoxicity to CD22 + target cells, and the survival time of mice in the CD22 CAR-T treatment group was significantly prolonged. Furthermore, it’s validated that sequential CD22/CD19 CAR-T therapy was significantly superior than single CD19 or CD22 CAR-T treatment in a relapse xenograft model. All 4 patients achieved complete remission (CR) with negative minimal residual disease (MRD), including 3 patients who had received prior CD19-related immunotherapy. The proliferation of CD19 and CD22 CAR-T cells was observed respectively in vivo, and 3 of the 4 patients experienced cytokine release syndrome (CRS); 2 of these patients had grade 1 CRS and 1 had grade 3 CRS. Long term follow-up showed that 3 of the 4 (75%) patients had sustained CR for up to 1 year. Analysis of antigen expression in the relapsed patients demonstrated that loss or diminution of CD19 and CD22 expression might cause antigen escape from CAR-T surveillance.ConclusionsIn summary, the novel CD22 CAR-T therapy was validated with antitumor effects both in vitro and in vivo. Furthermore, our study demonstrated the safety and robust efficacy of sequential CD22/CD19 CAR-T therapy in xenograft models and clinical trials, especially as the salvage treatment for R/R B-ALL patients with antigen loss or in whom anti-CD19 related immunotherapy failure failed.Trial registration: Chinese Clinical Trial Registry (ChiCTR): ChiCTR1900025419, Supplementarily registered 26 August, 2019.

High efficiency closed-system gene transfer using automated spinoculation

BackgroundGene transfer is an important tool for cellular therapies. Lentiviral vectors are most effectively transferred into lymphocytes or hematopoietic progenitor cells using spinoculation. To enable cGMP (current Good Manufacturing Practice)-compliant cell therapy production, we developed and compared a closed-system spinoculation method that uses cell culture bags, and an automated closed system spinoculation method to decrease technician hands on time and reduce the likelihood for microbial contamination.MethodsSepax spinoculation, bag spinoculation, and static bag transduction without spinoculation were compared for lentiviral gene transfer in lymphocytes collected by apheresis. The lymphocytes were transduced once and cultured for 9 days. The lentiviral vectors tested encoded a CD19/CD22 Bispecific Chimeric Antigen Receptor (CAR), a FGFR4-CAR, or a CD22-CAR. Sepax spinoculation times were evaluated by testing against bag spinoculation and static transduction to optimize the Sepax spin time. The Sepax spinoculation was then used to test the transduction of different CAR vectors. The performance of the process using healthy donor and a patient sample was evaluated. Functional assessment was performed of the CD19/22 and CD22 CAR T-cells using killing assays against the NALM6 tumor cell line and cytokine secretion analysis. Finally, gene expression of the transduced T-cells was examined to determine if there were any major changes that may have occurred as a result of the spinoculation process.ResultsThe process of spinoculation lead to significant enhancement in gene transfer. Sepax spinoculation using a 1-h spin time showed comparable transduction efficiency to the bag spinoculation, and much greater than the static bag transduction method (83.4%, 72.8%, 35.7% n = 3). The performance of three different methods were consistent for all lentiviral vectors tested and no significant difference was observed when using starting cells from healthy donor versus a patient sample. Sepax spinoculation does not affect the function of the CAR T-cells against tumor cells, as these cells appeared to kill target cells equally well. Spinoculation also does not appear to affect gene expression patterns that are necessary for imparting function on the cell.ConclusionsClosed system-bag spinoculation resulted in more efficient lymphocyte gene transfer than standard bag transductions without spinoculation. This method is effective for both retroviral and lentiviral vector gene transfer in lymphocytes and may be a feasible approach for gene transfer into other cell types including hematopoietic and myeloid progenitors. Sepax spinoculation further improved upon the process by offering an automated, closed system approach that significantly decreased hands-on time while also decreasing the risk of culture bag tears and microbial contamination.

CD22 CAR Optimization for Improved in-Human Activity Following Inadequate CD22 CAR Activity in Phase 1 Clinical Trial PLAT-04

Background: CD19 targeting chimeric antigen receptor (CAR) T cells have induced unprecedented remission rates in high-risk precursor B Acute Lymphoblastic Leukemia (ALL); however recurrent disease with CD19 antigen escape variants is not uncommon. Therefore, we developed a novel CD22 targeting CAR, and following preclinical validation, tested it in a first-in-human pediatric and young adult phase 1 clinical trial, PLAT-04 (NCT03244306). Four subjects were treated at 2 dose levels (DL) (1x10 6/kg (DL1) and 3x10 6/kg (DL2)). The CD22 CAR T cell product (SCRI-CAR22v1) was successfully manufactured (n=4) and no dose limiting toxicity (DLTs), cytokine release syndrome (CRS) or neurotoxicity was observed. However, all subjects had minimal CAR T cell expansion, with 3 of 4 subjects demonstrating persistent or progressive disease at day 21 evaluation despite continued CD22 expression on leukemic blasts. Based on the poor in vivo expansion and lack of activity, enrollment was voluntarily halted to interrogate and optimize the CAR construct for enhanced performance.Methods: Human T cells were transduced to express one of two CD22 CAR constructs. We designed SCRI-CAR22v2, a CD22 CAR that utilizes the same scFv as SCRI-CAR22v1 but with a shorter linker between M971 VH and VL and a shorter hinge with differing transmembrane region, and both using CD8 alpha (Figure A). This construct maintained the truncated EGFR extracellular tag (EGFRt) for tracking and potential in vivo suicide mechanism. The two transduced CAR T cell products were compared preclinically by flow cytometry, chromium release assay and in an in vivo murine model to understand differing T cell activity between the CAR constructs. Additionally, SCRI-CAR22v2 is currently under investigation in a dose finding phase 1 clinical trial, PLAT-07 (NCT04571138).Results: Following use of cetuximab-APC and biotinylated anti-human Fab antibody for surface EGFRt and CAR detection, the SCRI-CAR22v1 expresses lower levels of EGFRt but similar CAR levels on the cell surface demonstrated by MFI (Figure B). Biotinylated, soluble CD22 antigen was also used to evaluate CD22 CAR receptor activity and, as measured by MFI, a higher affinity is suggested via SCRI-CAR22v2 as compared to SCRI-CAR22v1 (Figure B). K562 cells expressing low, medium or high CD22 were used to evaluate the impact of surface antigen expression on the CAR activity level. SCRI-CAR22v2 demonstrates improved targeted cell lysis at all 3 antigen quantity levels by chromium release assay (Figure C). In NSG mice inoculated with Raji tumor cells expressing ffluc, SCRI-CAR22v2 demonstrated improved survival compared to SCRI-CAR22v1 (Figure D) and clearance of Raji tumor cells (Figure E). Based on this promising preclinical data, we initiated enrollment onto PLAT-07, a phase 1 dose finding trial (2x10 5cells/kg (DL1), 5x10 5cells/kg (DL2) and 1x10 6cells/kg (DL3)) of SCRI-CAR22v2. To date, 3 subjects have been enrolled and successfully infused at DL1. All had prior CD19-CAR therapy and 2 lacked CD19 leukemic expression at the time of SCRI-CAR22v2 infusion. At the time of cell infusion, one subject had only extramedullary disease, one had MRD of <1% and one subject had a larger disease burden of 30% ALL. None experienced a DLT and all were MRD negative in the bone marrow at day 28 and the subject with EMD demonstrated a complete metabolic response by PET scan. Figure F exhibits the improved expansion and engraftment of the SCRI-CAR22v2 cells as compared to SCRI-CAR22v1 DL1 (n=3) and DL2 (n=1), and higher peak levels of CD22 CAR T cells as compared to SCRI-CAR22v1 DL1 and DL2 (Figure G).Conclusions: Despite encouraging preclinical data, SCRI-CAR22v1 demonstrated poor expansion and engraftment in a Phase 1 trial. Notably, minor CAR alterations lead to encouraging in-human activity in early clinical findings. Our experience suggests a shorter linker and hinge as well as incorporation of an CD8 alpha transmembrane region improves the clinical activity of CD22 targeted CAR T cells in subjects with recurrent disease following CD19 CAR T cells. Further evaluation is needed to elucidate the critical CAR components and/or assays at the preclinical level that can best predict which CAR should be brought to the clinic for further evaluation. [Display omitted] DisclosuresOrentas: Lentigen: Patents & Royalties. Jensen: BMS: Patents & Royalties; Umoja Biopharma: Current holder of stock options in a privately-held company, Membership on an entity's Board of Directors or advisory committees, Research Funding; Bluebird Bio: Research Funding. Gardner: Novartis: Consultancy; BMS: Patents & Royalties.

Understanding the CD8+ CAR T Cell Response to Antigen-Low Leukemias

Chimeric antigen receptor (CAR) T cell therapy has emerged as a highly efficacious treatment for B-lineage acute lymphoblastic leukemias (B-ALL). However, downregulation of the CAR-targeted antigen on leukemia cells, predicted to reduce cellular avidity, is associated with post-CAR T cell leukemic relapse following CD22 CAR treatment (Fry et al., Nat. Med., 2017). We have observed reduced function of human CAR T cells against low target antigen site density (Ag Lo) human leukemia in immunodeficient mouse models, relative to CARs responding to high-antigen expressing leukemia. Thus, a better understanding of CAR responses to Ag Lo leukemia could help to increase the durability of remissions. We set out to develop a model system in which we could further interrogate the consequences of low-avidity interactions on CAR immunobiology, generating variants of a murine B-ALL driven by the E2A-PBX fusion protein (E2A) with different levels of target antigen to use in an immunocompetent syngeneic mouse model.We observed impaired expansion (p<0.0001) and tumor clearance (p<0.001) of CAR T cells responding to low-antigen variants of E2A (E2A-Ag Lo) as compared to wildtype E2A expressing high levels of antigen (E2A-WT). While CD8+ CAR T cell (CAR8) transcription factor (TF) expression in response to E2A-Ag Lo versus E2A-WT was largely similar early after CAR infusion, by day 9 post-CAR, CAR8s responding to E2A-Ag Lo exhibited decreased expression of multiple TFs, with Eomes (p<0.01), Irf4 (p<0.001) and Blimp1 (p<0.01) showing the largest magnitude change relative to CAR8s responding to E2A-WT. Additionally, CAR8s from mice bearing E2A-Ag Lo became enriched for cells of a “terminally exhausted” phenotype (Eomes+/PD1 Hi/TOX Hi) by day 11 post-CAR, and negatively-enriched for the “progenitor exhausted” (Tcf1+/PD1 Int) phenotype which can be functionally rescued by anti-PD1 therapy (p<0.0001, p<0.01). These data suggest that continual stimulation by low density antigen leads to a gradual reduction in the ability of CAR8s to mount an effector response, and eventually to T cell states with sub-optimal anti-tumor efficacy.Following in vitro stimulation of human CD22 CARs across a range of leukemic antigen densities, we saw that the percentage of CAR+ cells capable of producing IFNγ and IL2 corresponded to target antigen density (p<0.01, p<0.001). As human CARs are commonly manufactured from heterogenous bulk donor T cells, we hypothesized that antigen sensitivity is impacted by the prior antigen-experience of a given T cell. We predicted that T cells which had encountered cognate antigen through their TCR prior to CAR manufacturing (CAR8 AgEx) would have enhanced capacity to respond to low-avidity stimulation compared to CARs manufactured from naïve CD8+ T cells (CAR8 Naïve). We used a well-characterized ovalbumin vaccination model with OT-I TCR-transgenic T cells, allowing defined control of T cell antigen experience, to generate CAR8 AgEx. We found that CAR8 AgEx were highly antigen-sensitive relative to CAR8 Naïve, showing almost no reduction in numbers of cells capable of producing IFNγ and TNFα in vitro against E2A-Ag Lo as compared to E2A-WT. In vivo, CAR8 AgEx showed near complete depletion of E2A-Ag Lo in bone marrow by day 11 post-CAR, while mice treated with CAR8 Naïve maintained a substantial tumor burden (p<0.01). To test our hypothesis in human cells, we manufactured CD22 CAR T cells from naïve (CD45RO-) versus non-naïve (CD45RO+) starting T cell populations, and again found that CAR AgEx outperformed CAR Naïve against Ag Lo leukemia in production of IFNγ and IL2 in vitro (p<0.001, p<0.01) and in early leukemic clearance in vivo (p<0.0001, day 13).In conclusion, we have established a model to study the immunobiology of the CAR T cell response to Ag Lo B-ALL in an intact host. Preliminary findings indicate impaired expansion and tumor clearance of Ag Lo leukemia, associated with altered CAR T cell transcriptional profiles and features of T cell exhaustion. Furthermore, T cell history prior to CAR manufacturing has a drastic impact on the capacity to respond to Ag Lo leukemia. Future studies with this model will expand our characterization of CAR T cells responding to Ag Lo leukemia, with the goal of optimizing antigen sensitivity.We expect that advancing our understanding on the interplay of antigen density and CAR differentiation status will prove useful in developing more effective iterations of this therapy. DisclosuresFry: Sana Biotechnology: Current Employment, Current equity holder in publicly-traded company.

Infectious complications of CAR T-cell therapy across novel antigen targets in the first 30 days

AbstractInfections are a known complication of chimeric antigen receptor (CAR) T-cell therapy with data largely emerging from CD19 CAR T-cell targeting. As CAR T-cell therapy continues to evolve, infection risks and management thereof will become increasingly important to optimize outcomes across the spectrum of antigens and disease targeted. We retrospectively characterized infectious complications occurring in 162 children and adults treated among 5 phase 1 CAR T-cell clinical trials. Trials included targeting of CD19, CD22, disialoganglioside (GD2) or B-cell maturation antigen (BCMA). Fifty-three patients (32.7%) had 76 infections between lymphocyte depleting (LD) chemotherapy and day 30 (D30); with the majority of infections (61, 80.3%) occurring between day 0 (D0) and D30. By trial, the highest proportion of infections was seen with CD22 CAR T cells (n = 23/53; 43.4%), followed by BCMA CAR T cells (n = 9/24; 37.5%). By disease, patients with multiple myeloma had the highest proportion of infections (9/24; 37.5%) followed by acute lymphoblastic leukemia (36/102; 35.3%). Grade 4 infections were rare (n = 4; 2.5%). Between D0 and D30, bacteremia and bacterial site infections were the most common infection type. In univariate analysis, increasing prior lines of therapy, recent infection within 100 days of LD chemotherapy, corticosteroid or tocilizumab use, and fever and neutropenia were associated with a higher risk of infection. In a multivariable analysis, only prior lines of therapy and recent infection were associated with higher risk of infection. In conclusion, we provide a broad overview of infection risk within the first 30 days post infusion across a host of multiple targets and diseases, elucidating both unique characteristics and commonalities highlighting aspects important to improving patient outcomes.

A Systematic Review and Meta-Analysis of CD22 Car T-Cells Alone or in Combination with CD19 Car T-Cells

Background: Chimeric antigen receptor (CAR) T-cells targeting CD22 may benefit patients with B-cell malignancies, especially those that have failed CD19 CAR T-cells. CAR T-cells co-targeting CD19 and CD22 may prevent relapse due to antigen loss. We conducted a systematic review and meta-analysis evaluating the efficacy and toxicity of CD22-targeting CAR T-cells. Methods: We searched MEDLINE, EMBASE, Web of Science, and the Cochrane Central Register of Controlled Trials from inception to July 8, 2020 for full-length articles and conference abstracts of clinical trials employing CD22-targetting CAR T-cells in acute lymphocytic leukemia (ALL) and non-Hodgkin’s lymphoma (NHL); primary outcome was best complete response (bCR). A DerSimonian and Laird random-effects model with arcsine transformation was used to pool outcome proportions. Findings: Fifty-nine of 601 references were included, representing 14 early phase studies with 339 patients, investigating CD22 or CD19/CD22 CAR T-cells. CD22 CAR T-cells had a bCR of 66% [95% CI, 52-79%] in ALL (n= 100), with 73% of patients having received anti-CD19 CART previously. CD19/CD22 CAR T-cells had a bCR rate of 95% [95% CI, 87-99%] in ALL (n= 150) and 51% [95% CI, 38-63%] in NHL (n= 61). The estimated incidence of total and severe (grade ≥3) CRS were 85% [95% CI, 79-91] and 43% [95% CI, 1%-8%]. ICANS and severe ICANS had an estimated incidence of 14% [95% CI, 8-22%] and 1% [95% CI, 0-3%]. Interpretations: Early phase trials of CD22 and CD19/CD22 CAR T-cells show high remission rates in ALL, and CD19/CD22 CAR T-cells have also shown substantial remissions in NHL. Severe CRS or ICANS were rare and dual-targeting did not increase toxicity. Variability in CAR construct, dose, and patient factors between studies limits comparisons. Long-term outcomes are yet to be reported and will reveal whether responses translate into durable long-term remission. Funding: BioCanRx.Systematic Review Registration: PROSPERO CRD42020193027 Declaration of Interest: KAH serves on advisory board for Kite/Gilead, Cellgene/BMS, Novaratis, and receives payments for Jazz Pharmaceutical educational programs. KAH has also received travel support for meetings from BioCanRx. NJF, KA, NK, and HA report no competing interests.

Modulation of CD22 Antigen Density Improves Efficacy of CD22 Chimeric Antigen Receptor (CAR) T Cells Against CD22lo B-Lineage Leukemia and Lymphoma

While overall survival for acute lymphoblastic leukemia (ALL) in children treated with risk-adapted, multi-agent chemotherapeutic regimens is greater than 85%, relapsed ALL remains a major challenge. Chimeric antigen receptor (CAR) T cell therapy targeting CD19 provides these patients with a remission induction rate of 70-90% across multiple clinical trials, but loss of the targeted epitope results in relapse following CD19 CAR T cells. In the setting of these relapses, our group conducted a CD22 CAR T cell trial, which induces remission in approximately 70% percent of patients with relapsed/refractory CD22-positive pre-B cell ALL, including those relapsing after CD19 CAR therapy (NCT02315612). However, relapse associated with reduced CD22 site density has shortened remission duration in some patients. To test the impact of low CD22 site density on CD22 CAR T cell functionality, we modified Nalm6 ALL cell lines using CRISPR/Cas9 to eliminate the CD22 gene, followed by lentiviral transduction of full-length CD22 and single-cell cloning based on CD22 expression levels. CD22 site density was assessed with the Quantibrite-PE Assay Kit and four cell lines were identified for a spectrum of CD22 expression: CD22 negative (CD22neg); CD22 low (CD22lo) - site density comparable to that seen in ALL patients relapsing following CD22 CAR therapy; CD22 normal (Nalm6) - site density comparable to ALL CD22 expression; and CD22 high (CD22hi) - site density comparable to CD19 levels. As CD22 site density decreases, CD22 CAR T cells produce incrementally less interferon-gamma and do not produce IL-2 against CD22lo ALL. CD22 CAR delayed, but was unable to prevent, progression of CD22lo ALL in murine xenografts. We next evaluated two potential approaches to overcoming low site density antigen escape: improving CAR activity or increasing tumor sensitivity. To improve CAR activity, we generated a high-binding affinity CD22 CAR (CD22 CARH1) using an affinity-enhanced version of the scFv from the CD22 CAR construct in our ongoing trial. Despite higher affinity, the CD22 CARH1 T cell did not improve in vitro killing or in vivo clearance of CD22lo ALL. To increase tumor sensitivity, we sought to increase CD22 site density using the PKC inhibitor Bryostatin1, a drug previously shown to increase CD22 expression in CLL. Bryostatin1 increases CD22 site density on a panel of pre-B cell ALL and diffuse large B cell lymphoma (DLBCL) cell lines. Interestingly, the CD22 increase is durable for at least 1 week after exposure to Bryostatin1. Bryostatin1-mediated CD22 increased site density improves cytokine production by CD22 CAR T cells. To evaluate the mechanism of Bryostatin1-mediated upregulation of CD22 expression, RNAseq was performed on Bryostatin1-exposed ALL. Preliminary analysis demonstrates that Bryostatin1 does not directly affect CD22 mRNA expression levels, but modulates early transcription factors within the B-cell receptor pathway. Based on these results, we are further evaluating pathways affected by Bryostatin1 through protein expression analysis and assessing in vivo function of Bryostatin1 in the setting of a patient-derived CD22 CAR-resistant xenograft model. In conclusion, these results demonstrate a clear threshold antigen site density, below which the CD22 CAR is unable to clear CD22-expressing ALL. We further demonstrate that increased CD22 CAR affinity does not seem to improve CD22 CAR activity against CD22lo ALL. Bryostatin1-mediated CD22 modulation can improve CD22 CAR response against leukemia and lymphoma providing a potential rational combination to rescue CD22 CAR activity and remission durability. DisclosuresNo relevant conflicts of interest to declare.

Intensification of Lymphodepletion Optimizes CAR Re-Treatment Efficacy

▪Background: Post chimeric antigen receptor (CAR) therapy relapse has generally occurred via two mechanisms; emergence of antigen-negative leukemia, or via loss of CAR T cells. Anecdotally, CAR re-treatment has had limited success, potentially due to immune-mediated rejection of CAR T cells, or from immunogenicity of epitopes on murine constructs-which may be less applicable to humanized constructs. After several re-infusion attempts on our phase I dose-escalation study of CD22 CAR, in an effort to improve response, we incorporated use of an intensified lymphodepletion regimen prior to CAR re-infusion and report on our findings. Furthermore, preclinical models suggest that modulation of antigen expression, despite persistent positivity, may also impact the CAR response. Therefore, in this re-treatment analysis, we also explore the role of antigen expression on treatment response.Design: Children and young adults with relapsed/refractory CD22+ acute lymphoblastic leukemia (ALL) enrolled on our phase I dose escalation anti-CD22 CAR therapy protocol (NCT02315612) who met eligibility criteria for a second infusion were evaluated. All patients had bone marrow evaluations by morphology and flow cytometry prior to second infusion to demonstrate persistent CD22+ ALL and loss of CD22 CAR T cells, and subsequently had restaging at day 28 (+/- 4 days) post second CAR infusion. Additionally, lymphodepleting chemotherapy, either with the original dosing schema (Fludarabine 25 mg/m2 x 3 days and Cyclophosphamide 900 mg/m2 x 1 day) or at intensified dosing (Fludarabine 30 mg/m2 x 4 days and Cyclophosphamide 600 mg/m2 x 2 days) was utilized at the discretion of the treating team. All patients were re-infused with their original cell dose; 3 x 105 CAR T cells/kg (n=1), 1 X 106 CAR T cells/kg (n=4), or 3 X 106 CAR T cells/kg (n=1). Flow cytometry was utilized to assess CD22 antigen-binding capacity (ABC) of leukemia cells pre and post infusion. Patients were analyzed for response and expansion of CAR T cells.Results: From June 2015 to July 2017, 6 patients with ALL were eligible for a second infusion of CD22 CAR T cells. This included 4 patients who relapsed with CD22+ disease after attaining a complete remission and 2 patients who had stable disease with limited CAR expansion and evidence of clinical benefit following the first infusion. The median time to relapse after a first infusion was 5 months (range: 3-11 months) with a median time to re-infusion of 4 months (range: 2-11 months) from the first infusion. Prior to re-infusion, all patients had <5% circulating CAR T cells, and all had significant disease burden, > M2 marrow (>5% blasts). Our initial re-infusion experience with 4 subjects, which utilized the identical lymphodepletion regimen as with the first infusion, demonstrated CAR expansion in only 2 patients. Following this, the pre-infusion lymphodepleting chemotherapy was intensified, with 2/2 patients having CAR expansion-including one patient who had a minimal expansion with the initial infusion, but robust expansion following the intensified lymphodepletion prior to second infusion. Both patients tolerated the intensified regimen well without any significant chemotherapy related toxicities. Following re-infusion, 4/6 patients had complete clearance of CD22+ leukemia. Notably, in this cohort, 4/6 patients were found to have a drop in their CD22 ABC, between the time from first to second infusion, with a median drop in site density of 3,141 sites/cell (range: 703-6460) with 2 of the re-infusion non-responders having approximately 50% decrease in the ABC from the CAR pre-treatment value.Conclusion: Our preliminary experience with CAR re-infusions demonstrates that in the presence of preserved antigen expression, re-infusions can be effective, and that an intensified lymphodepletion regimen may optimize re-treatment response, both in patients with limited CAR expansion following a first infusion or with post-CAR relapse. Despite antigen positivity, decreased antigen density post-relapse may be one potential mechanism contributing to lack of CAR efficacy with second infusions, and higher level of antigen density may be needed to optimize anti-tumor response. Future directions will include incorporating an intensified lymphodepletion regimen for all second infusions and exploring methods to enhance antigen expression at the time of re-infusion. DisclosuresNo relevant conflicts of interest to declare.

Chimeric Antigen Receptor Induced Cytopenia Differs from Chemotherapy Induced Myelosuppression

▪Background: Cytopenias are a potential complication of chimeric antigen receptor (CAR) therapy. The specific degree, duration, or implication of these cytopenias is not well established, and is further confounded by baseline cytopenias in heavily pre-treated patients with active leukemia. Preclinical models suggest that these cytopenias may be due to a cytokine rich milieu which may contribute to nonspecific suppression of marrow progenitors-independent of the CAR target (Sauter, ASH 2016, abstract 3357); which differs from standard cytotoxic chemotherapy based myelosuppression. We systematically analyzed CAR-related cytopenias on our phase I dose-escalation study of CD22 CAR.Design: Children and young adults with relapsed/refractory CD22+ acute lymphoblastic leukemia (ALL) treated on our phase I dose escalation anti-CD22 CAR therapy protocol (NCT02315612) were evaluated for time to neutrophil and platelet recovery (defined by absolute neutrophil count (ANC) of > 500 k/mcL x 3 days and achievement of transfusion independent platelet count > 20 k/mcL, respectively). All patients had bone marrow evaluations by morphology and flow cytometry at baseline, prior to lymphodepleting chemotherapy (Fludarabine 25 mg/m2 x 3 days and Cyclophosphamide 900 mg/m2 x 1 day) and again at day 28 (+/- 4 days) post-CAR. Infections were closely monitored for. All patients were on anti-fungal prophylaxis and were empirically treated for fever during periods of neutropenia.Results: From December 2014 to July 2017, 30 patients with ALL were treated on study at 3 dose levels. All patients had active bone marrow involvement at baseline with the majority (n=21) having an M2 marrow (>5% blasts) or higher. Twenty-six patients had prior transplants, and 21 patients had prior CD19-directed CAR therapy. Cytokine release syndrome (CRS) was seen in 24/30 (80%), with most patients having grade 1-2 CRS. 18/30 patients (60%) attained a complete remission following CD22 CAR therapy. Baseline hematologic parameters for responders prior to lymphodepletion included grade 2-4 neutropenia and grade 2 and 3 thrombocytopenia, with five patients requiring platelet transfusions (Table 1). Amongst complete responders, the bone marrow at day 28 demonstrated a median cellularity of 28% with trilineage hematopoiesis noted in 16/18 patients. The median CAR T cell % in bone marrow at 1 month was 30%. The median time to ANC nadir was 15 days post CAR infusion (range:10-23 days).17/18 patients had ANC recovery >500 k/mcL, at a median of 22 days post-infusion (range: 15-44 days). One patient had ANC recovery to 260 k/mcL but subsequently developed disease relapse. Eight patients received intermittent GCSF, with clinically relevant responses seen in seven of these patients with a median increase in ANC of 2,180 k/mcL (range: 850-4,100 k/mcL) with limited dosing. The median number of days of GCSF administration to time of ANC recovery was 4, with 3 patients requiring only 1 injection to achieve robust ANC response. Interestingly, only 1/18 patient (6%) had a new onset of infection (clostridium difficile colitis) during the period of CAR associated neutropenia, which was confounded by administration of intravenous antibiotic therapy for fevers that was initiated 24 hours prior to onset of diarrhea. There was no new development of fungal infections or worsening of any pre-existing infections. Platelet recovery to > 20 k/mcL was at a median of 36 days (range: 21-55) post-CAR and was seen in 16 of 18 responders. Two patients who remained platelet transfusion dependent post-CAR included one patient who was later found to have donor-cell monosomy 7 MDS and another patient whose disease relapsed prior to recovery.Conclusion: In contrast to chemotherapy associated myelosuppression, CAR-therapy associated peripheral blood cytopenias existed despite evidence for trilineage hematopoiesis on the bone marrow. Although the period of CAR associated neutropenia is substantial, with prophylactic antifungals and intermittent growth factor support, this heavily pre-treated population was not at an increased risk of developing serious bacterial or fungal infections over baseline. The rapid response to G-CSF is suggestive of an alternative modality of bone marrow inhibition distinct from chemotherapy induced cytotoxicity. Further evaluation in the context of multiple CAR targets and constructs are needed to further explore this complication. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.