Search

PURPOSE Nonresponse and relapse after CD19-chimeric antigen receptor (CAR) T-cell therapy continue to challenge survival outcomes. Phase II landmark data from the ELIANA trial demonstrated nonresponse and relapse rates of 14.5% and 28%, respectively, whereas use in the real-world setting showed nonresponse and relapse rates of 15% and 37%. Outcome analyses describing fate after post-CAR nonresponse and relapse remain limited. Here, we aim to establish survival outcomes after nonresponse and both CD19+ and CD19– relapses and explore treatment variables associated with inferior survival. METHODS We conducted a retrospective multi-institutional study of 80 children and young adults with B-cell acute lymphoblastic leukemia experiencing nonresponse (n = 23) or relapse (n = 57) after tisagenlecleucel. We analyze associations between baseline characteristics and these outcomes and establish survival rates and salvage approaches. RESULTS The overall survival (OS) at 12 months was 19% across nonresponders (n = 23; 95% CI, 7 to 50). Ninety-five percent of patients with nonresponse had high preinfusion disease burden. Among 156 morphologic responders, the cumulative incidence of relapse was 37% (95% CI, 30 to 47) at 12 months (CD19+; 21% [15 to 29], CD19–; 16% [11 to 24], median follow-up; 380 days). Across 57 patients experiencing relapse, the OS was 52% (95% CI, 38 to 71) at 12 months after time of relapse. Notably, CD19– relapse was associated with significantly decreased OS as compared with patients who relapsed with conserved CD19 expression (CD19– 12-month OS; 30% [14 to 66], CD19+ 12-month OS; 68% [49 to 92], P = .0068). Inotuzumab, CAR reinfusion, and chemotherapy were used as postrelapse salvage therapy with greatest frequency, yet high variability in treatment sequencing and responses limits efficacy analysis across salvage approaches. CONCLUSION We describe poor survival across patients experiencing nonresponse to tisagenlecleucel. In the post-tisagenlecleucel relapse setting, patients can be salvaged; however, CD19– relapse is distinctly associated with decreased survival outcomes.

Chimeric antigen-receptor (CAR)-associated hemophagocytic lymphohistiocytosis (HLH)-like toxicities involving hyperferritinemia, multi-organ dysfunction, coagulopathy, and/or hemophagocytosis are described as occurring in a subset of patients with cytokine release syndrome (CRS). Case series report poor outcomes for those with B-acute lymphoblastic leukemia (B-ALL) who develop HLH-like toxicities, although larger outcomes analyses of children and young adults (CAYA) with B-ALL who develop these toxicities following commercial tisagenlecleucel are not described. Using a multi-institutional database of 185 CAYA with B-ALL, we conducted a retrospective cohort study including groups that developed HLH-like toxicities, high grade CRS without HLH-like toxicities, or no to low grade CRS without HLH-like toxicities. Primary objectives included characterizing the incidence, outcomes, and pre-infusion factors associated with HLH-like toxicities. Among 185 CAYA infused with tisagenlecleucel, 26 (14.1%) met criteria for HLH-like toxicities. One-year overall survival and relapse-free survival were 25.7% and 4.7% in those with HLH-like toxicities, compared with 80.1% and 57.6% in those without. In multivariable analysis for death, meeting criteria for HLH-like toxicities carried a hazard ratio of 4.61 (95% confidence interval: 2.41-8.83), controlling for disease burden, age, and sex. Patients who developed HLH-like toxicities had higher pre-tisagenlecleucel disease burden, ferritin, C-reactive protein levels, and lower platelet and absolute neutrophil counts than patients with high grade or no/low grade CRS without HLH-like toxicities. Overall, CAYA with B-ALL who developed HLH-like toxicities following tisagenlecleucel experienced high rates of relapse and non-relapse mortality, indicating the urgent need for further investigations into prevention and optimal management of patients who develop HLH-like toxicities following tisagenlecleucel.

Remarkable complete response rates have been shown with tisagenlecleucel, a chimeric antigen receptor (CAR) T-cell therapy targeting CD19, in patients up to age 26 years with refractory/relapsed B-cell acute lymphoblastic leukemia; it is US Food and Drug Administration approved forthis indication. Currently, patients receive a single dose of tisagenlecleucel across a wide doserange of 0.2 to 5.0× 106 and 0.1 to 2.5× 108 CAR T cells per kg for patients ≤50 and >50 kg, respectively. The effect of cell dose on survival and remission is not yet well established. Our primary goal was to determine if CAR T-cell dose affects overall survival (OS), event-free survival (EFS), or relapse-free-survival (RFS) in tisagenlecleucel recipients. Retrospective data were collected from Pediatric Real World CAR Consortium member institutions and included 185 patients infused with commercial tisagenlecleucel. The median dose of viable transduced CAR T cells was 1.7× 106 CAR T cells per kg. To assess the impact of cell dose, we divided responders into dose quartiles: 0.134 to 1.300× 106 (n= 48 [27%]), 1.301 to 1.700× 106 (n= 46 [26%]), 1.701 to 2.400×106 (n= 43 [24%]), and 2.401 to 5.100× 106 (n= 43 [24%]). OS, EFS, and RFS were improved in patients who received higher doses of tisagenlecleucel (P= .031, .0079, and .0045, respectively). Higher doses of tisagenlecleucel were not associated with increased toxicity. Because the current tisagenlecleucel package insert dose range remains broad, this work has implications in regard totargeting higher cell doses, within the approved dose range, to optimize patients' potential for long-standing remission.

Introduction: H3K27M-mutated diffuse intrinsic pontine glioma (DIPG) and spinal cord diffuse midline glioma (DMG) are universally lethal central nervous system (CNS) tumors in children and young adults. We previously demonstrated safety and activity of GD2.41BB.z chimeric antigen receptor T cells (CAR-Ts) at dose level 1, 1x106 GD2 CAR-T/kg (Majzner/Ramakrishna et al. Nature 2022) and reported results of dose level 2, 3x106 GD2 CAR-T/kg (Majzner et al. AACR 2022). Here, we present in depth high-dimensional analyses to define the immune states that contribute to CAR-T activity in patients. Methods: Thirteen patients (10 DIPG/3 spinal DMG; 4-30 years old; 7F/6M) were enrolled in this GD2 CAR-T phase 1 clinical trial (NCT04196413). GD2 CAR-Ts were administered to 12/13 enrolled patients. In the first cohort, CAR-Ts were administered initially intravenously (IV), followed by serial intracerebroventricular infusions (ICV; range 0-11 infusions/patient). Patient GD2 CAR-T product, peripheral blood, and cerebrospinal fluid (CSF) samples were evaluated for CAR-T expansion (qPCR; flow cytometry), cytokine signatures (Multiplex Luminex), and immune cell profiles (single cell RNA-sequencing). Data were analyzed in the context of clinical trajectory and patient response. Results: 10/12 infused subjects demonstrated clinical and/or radiographic benefit, with less systemic toxicity following ICV compared to IV infusion. CAR-T expansion was noted in the periphery and CSF of all treated patients and following serial ICV infusions. In peripheral blood, cytokine concentrations, including IFN-gamma, IL6, and CXCL9, were higher after IV compared to ICV CAR-T infusions, correlating with increased systemic inflammation. Conversely in CSF, cytokine concentrations, such as CCL2 and CXCL9, were higher following ICV compared to IV CAR-T infusions. Transcriptomic analysis was conducted on 576,199 single cells from 91 samples, including GD2 CAR-T products and patient CSF. This is the largest CAR-T dataset in CNS tumors. Patient CSF samples were dominated by T cell and myeloid populations. After IV CAR-T infusion, patient CSF exhibited an increased fraction of regulatory T cells and suppressive myeloid populations from baseline. These immune suppressive cells reduced after ICV infusion. Ongoing analyses are underway to explore the relation of these immune populations to patient response. Conclusions: H3K27M-mutated DIPG/DMG patients demonstrate continued clinical response with serial ICV GD2 CAR-T infusions, with heterogeneity in the durability of response across patients. In-depth correlative analyses profile distinct immune populations and demonstrate population shifts depending on route of administration and over the course of treatment. Key findings from these data will allow for iterative improvement in CAR-T therapies for H3K27M+ DIPG/DMG patients, providing hope to shift the paradigm of this fatal disease. Citation Format: Sneha Ramakrishna, Zinaida Good, Moksha Desai, Daniel Zamler, Rebecca Mancusi, Jasia Mahdi, Robbie Majzner, Liora Schultz, Rebecca Richards, Jennifer Kamens, Valentin Barsan, Cynthia Campen, Sonia Partap, Zachary Ehlinger, Warren Reynolds, Yiyun Chen, Mark P. Hamilton, Jennifer Moon, Christina Baggott, Michael Kunicki, Michelle Fujimoto, Amy Li, Sneha Jariwala, Sharon Mavroukakis, Emily Egeler, Ashley Jacobs, Courtney Erickson, Karen Yamabe-Kwong, Snehit Prabhu, Kara Davis, Steve Feldman, Bita Sahaf, Crystal L. Mackall, Michelle Monje. Immune signatures of GD2 CAR T cell activity in H3K27M+ diffuse midline glioma patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 959.

Autologous T cells genetically engineered to express a chimeric antigen receptor (CAR) targeting CD19 and/or CD22 have achieved high complete response rates in patients with hematologic malignancies, but >50% of patients progress following therapy. Here, we sought to understand key T cell intrinsic factors impacting efficacy: CAR T cell expansion, persistence, and homing to the tumor. Using an endogenous T cell receptor (TCR) sequence as a ‘barcode’, we followed individual T cell clonotypes at the single-cell level from pre-manufacture apheresis and infusion products to tumor-involved lymph node and blood at peak and late expansion in 22 adult patients with relapsed or refractory large B cell lymphoma (LBCL) or acute lymphoblastic leukemia (ALL) treated with axicabtagene ciloleucel, an FDA-approved CD19-CAR T cell immunotherapy, or bispecific CD19/CD22 CAR T cells on an investigator-initiated trial (NCT03233854). The resulting CAR T cell atlas comprises matched transcriptome (scRNA-seq) and surface protein expression (CITE-seq) for 846,344 cells from 97 samples, with 215,045 unique TCR clonotypes identified, including 8,747 clonotypes that could be traced across 2+ timepoints in CAR mRNA+ cells. This atlas enabled us to ask: “What were the phenotypes of ‘successful’ CAR T cell clonotypes with optimal homing, expansion, and persistence properties at the time of infusion or pre-manufacture apheresis?” We found that successful T cell clonotypes at apheresis had juvenile features, including IL7R expression. Conversely, successful clonotypes in the infusion product had elevated interferon pathway activity and effector signatures, including GZMB expression. Further, we built a cell-cell interactome using all live cells from on-treatment biopsies and identified a set of 149 specific ligand-receptor pairs significantly enriched in patients who progressed. Finally, we defined dynamics of TCR clonotypes with predicted specificities for viral and self-antigens. These analyses pinpoint the identities of source T cells and infusion CAR T cells with properties impacting efficacy, and also identify ligand-receptor pairs that could be modulated to enhance CAR T cell response in the tumor at the genetic or pharmacological level. This work was supported in part by the Parker Institute for Cancer Immunotherapy, California Institute for Regenerative Medicine, Kite Pharma, and Stanford Cancer Institute. Citation Format: Zinaida Good, Mark P. Hamilton, Jay Y. Spiegel, Sreevidya Kurra, Moksha H. Desai, Snehit Prabhu, Shin-Heng Chiou, Christine Y. Yeh, Yiyun Chen, Eric Yang, Michael G. Ozawa, Fang Wu, Matthew J. Frank, Lori Muffly, Gursharan K. Claire, Juliana Craig, Maria I. Iglesias, Sushma Bharadwaj, Katherine A. Kong, Dhananjay Wagh, John Coller, Mark M. Davis, Sylvia K. Plevritis, Bita Sahaf, David B. Miklos, Crystal L. Mackall. Lineage tracing of CAR T cells in patients with B cell malignancies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1128.

Chimeric antigen receptor (CAR) T cells provide a therapeutic option in hematologic malignancies. However, treatment failure after initial response approaches 50%. In allogeneic hematopoietic cell transplantation, optimal fludarabine exposure improves immune reconstitution, resulting in lower nonrelapse mortality and increased survival. We hypothesized that optimal fludarabine exposure in lymphodepleting chemotherapy before CAR T-cell therapy would improve outcomes. In a retrospective analysis of patients with relapsed/refractory B-cell acute lymphoblastic leukemia undergoing CAR T-cell (tisagenlecleucel) infusion after cyclophosphamide/fludarabine lymphodepleting chemotherapy, we estimated fludarabine exposure as area under the curve (AUC; mg × h/L) using a validated population pharmacokinetic (PK) model. Fludarabine exposure was related to overall survival (OS), cumulative incidence of relapse (CIR), and a composite end point (loss of B-cell aplasia [BCA] or relapse). Eligible patients (n = 152) had a median age of 12.5 years (range

Key Points CD19 CAR therapy for R/R EM disease, particularly CNS, offers a beneficial option with similar toxicity and survival to BM disease.There was no increased cytokine release syndrome or neurotoxicity in patients with R/R EM disease, including active CNS disease at infusion., Visual Abstract, Chimeric antigen receptor (CAR) T cells have transformed the therapeutic options for relapsed/refractory (R/R) B-cell acute lymphoblastic leukemia. Data for CAR therapy in extramedullary (EM) involvement are limited. Retrospective data were abstracted from the Pediatric Real World CAR Consortium (PRWCC) of 184 infused patients from 15 US institutions. Response (complete response) rate, overall survival (OS), relapse-free survival (RFS), and duration of B-cell aplasia (BCA) in patients referred for tisagenlecleucel with EM disease (both central nervous system (CNS)3 and non-CNS EM) were compared with bone marrow (BM) only. Patients with CNS disease were further stratified for comparison. Outcomes are reported on 55 patients with EM disease before CAR therapy (CNS3, n = 40; non-CNS EM, n = 15). The median age at infusion in the CNS cohort was 10 years (range

BACKGROUND: H3K27M-mutated DMGs express high levels of the disialoganglioside GD2 and GD2-CAR T-cells (GD2-CART) regress DMG in preclinical models. METHODS: NCT04196413 is a 3 + 3 Phase I dose escalation trial testing GD2-CART in patients with biopsy-proved H3K27M DMG, with dose-limiting toxicities (DLT) considered independently for DIPG and spinal DMG (sDMG). Arm A tested escalating doses of IV GD2-CART (DL1=1e6 GD2-CART/kg; DL2=3e6 GD2-CART/kg) following lymphodepletion (LD). After the DLT period, patients with clinical and/or radiographic benefit were eligible for subsequent ICV GD2-CART infusions (10-30e6 GD2-CART) administered via Ommaya without LD. RESULTS: Twelve subjects were treated after standard radiotherapy, 7 of whom began treatment at the time of progression [n=4 DL1 (3 DIPG/1 sDMG); n=8 DL2 (6 DIPG/2 sDMG)]. No DLTs were observed on DL1. Three subjects experienced DLT on DL2 (2 DIPG/1 sDMG) due to grade-4 cytokine release syndrome (CRS). On both dose levels, all subjects exhibited transient symptoms related to on-tumor inflammation, termed Tumor Inflammation-Associated Neurotoxicity (TIAN); no DLT due to TIAN has occurred. Ten subjects experienced radiographic and/or clinical benefit after IV infusion and received subsequent ICV infusions (median=4 ICV infusions/pt, range=1-7). ICV infusions were not associated with high-grade CRS. Four patients continue to receive ICV infusions on study and have experienced continued clinical and radiographic benefit, currently 7-11 months following enrollment. Two patients (one sDMG, one DIPG) have experienced near-complete (>95%) tumor volume reduction. CONCLUSIONS: IV treatment of DIPG and sDMG with GD2-CART is safe at a dose of 1e6/kg, but associated with frequent high-grade CRS at 3e6/kg. ICV GD2-CART has been well tolerated and has mediated impressive sustained clinical benefit in some patients with DIPG/sDMG. Given these findings, we are launching a new arm to assess safety and activity and to define the recommended phase 2 dose for ICV delivery of GD2-CART without LD.

Infants with B-cell acute lymphoblastic leukemia (B-ALL) have poor outcomes because of chemotherapy resistance leading to high relapse rates. Tisagenlecleucel, a CD19-directed chimeric antigen receptor T-cell (CART) therapy, is US Food and Drug Administration approved for relapsed or refractory B-ALL in patients ≤25 years; however, the safety and efficacy of this therapy in young patients is largely unknown because children M1 marrow) were refractory to this therapy (n = 5). Overall, tisagenlecleucel was tolerable in this population, with only 3 patients experiencing ≥grade 3 cytokine release syndrome. No neurotoxicity was reported. This is the largest report of tisagenlecleucel use in infant B-ALL and shows that this therapy is safe and can be effective in this population. Incorporating this novel immunotherapy into the treatment of infant B-ALL offers a promising therapy for a highly aggressive leukemia.

Introduction Chimeric antigen receptor (CAR) T cell therapy has been extremely efficacious in pediatric patients with multiply relapsed and/or refractory B-cell acute lymphoblastic leukemia (B-ALL) with overall remission rates of 81% by three months post-infusion (Maude et al., N Engl J Med, 2018), and achieved FDA approval for this indication. In order for the product to meet the standards of this approval for commercial release, both the leukapheresis and manufactured products must meet a variety of specific requirements, some of which are more stringent than those in these pivotal clinical trials. The Managed Access Program (MAP) provides access to tisagenlecleucel for patients with B-ALL or diffuse large B-cell lymphoma that is out of specification (OOS) for whom repeat leukapheresis is not feasible. Patients may also receive OOS tisagenlecleucel by applying for a single-patient Investigational New Drug (IND). Previous reports have shown no difference in efficacy or toxicity between patients receiving tisagenlecleucel that meets commercial release specifications and those receiving OOS tisagenlecleucel (Grupp, et al., Blood Abstr 614, 2019; Jaglowski, et al., Blood Abstr 627, 2019). This study seeks to evaluate outcomes in pediatric and young adult patients who received tisagenlecleucel via the MAP or a single-patient IND in our Pediatric Real World CAR Consortium (PRWCC). Methods Retrospective data were abstracted from the PRWCC database of patients with relapsed/refractory B-ALL from fifteen different US institutions who received tisagenlecleucel as an FDA-approved therapy outside the context of a clinical trial. Patients whose cellular product was obtained through the MAP (NCT03601442) or with single patient IND approval due having either a cryopreserved leukapheresis product and/or manufactured tisagenlecleucel that did not meet specifications for commercial release were categorized as MAP/IND and those whose product met all release criteria were categorized as standard of care (SOC). Results Among 185 total infused patients in our database, 24 (13%) received tisagenlecleucel either via the MAP (n=14) or a single patient IND (n=10). Baseline patient and disease characteristics were not significantly different for MAP/IND patients versus the SOC cohorts (Table 1). Median duration of follow-up post-CAR T cell infusion for these infused patients was 342.5 days (range 107-780) versus 318 days (range 6-863) for the SOC cohort (p=0.43). Reasons for products being OOS included cell viability 9 months prior (n=1), and determination of residual beads >50 beads/3x106cells (n=1). Overall survival at 6- and 12-months was 96% versus 83% and 85% versus 70% for the MAP/IND versus SOC, respectively. Event-free survival at 6- and 12-months was 65% versus 63% and 55% versus 51%, respectively. Probability of continued remission at 6- and 12-months among patients who achieved a complete remission (CR) at day 28 was 79% versus 75% and 66% versus 63% for the MAP/IND versus SOC, respectively (Figure 1). Comparing toxicities between patients in the MAP/IND versus SOC cohorts, cytokine release syndrome (CRS, any grade) occurred in 46% versus 61%, CRS (>grade 3) in 17% versus 19%, immune effector cell-associated neurotoxicity syndrome (ICANS) in 8% versus 22%, and infectious complications in 54% vs. 37%, respectively (p=ns for all). Conclusions In our retrospective cohort evaluating the use of tisagenlecleucel to treat pediatric and young adult patients with relapsed/refractory B-ALL in the real-world setting, neither the efficacy nor safety of tisagenlecleucel seem to be compromised by use of products OOS for commercial release. Larger studies are needed to further delineate specific cut-offs outside of which either efficacy and/or safety may truly be impacted. Disclosures Phillips: Novartis: Membership on an entity's Board of Directors or advisory committees. Stefanski:Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Margossian:Novartis: Membership on an entity's Board of Directors or advisory committees; Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees. Verneris:Novartis: Membership on an entity's Board of Directors or advisory committees; Fate Therapeutics: Consultancy, Current equity holder in publicly-traded company; Bmogen: Consultancy, Current equity holder in publicly-traded company; Uptodate: Consultancy. Myers:Novartis: Consultancy, Honoraria, Other: ELIANA trial Steering Committee, Speakers Bureau. Brown:Janssen: Membership on an entity's Board of Directors or advisory committees; Servier: Membership on an entity's Board of Directors or advisory committees; Jazz: Membership on an entity's Board of Directors or advisory committees; Novartis: Membership on an entity's Board of Directors or advisory committees. Qayed:Mesoblast: Consultancy; Novartis: Consultancy. Hermiston:Novartis: Membership on an entity's Board of Directors or advisory committees; Sobi: Membership on an entity's Board of Directors or advisory committees. Satwani:Mesoblast: Consultancy; Takeda: Consultancy. Curran:Novartis: Consultancy, Research Funding; Celgene: Research Funding; Mesoblast: Consultancy. Mackall:Apricity Health: Consultancy, Current equity holder in private company; Lyell Immunopharma: Consultancy, Current equity holder in private company; BMS: Consultancy; Nektar Therapeutics: Consultancy; Allogene: Current equity holder in publicly-traded company; NeoImmune Tech: Consultancy. Laetsch:Novartis: Consultancy, Research Funding; Bayer: Research Funding; Pfizer: Research Funding; Cellectis: Consultancy.

Background: H3K27M-mutated DMGs are universally lethal central nervous system tumors that express high levels of the disialoganglioside GD2. IV administered GD2-CAR T cells (GD2-CART) regress DMG in preclinical models, and locoregionally delivered CARs demonstrate enhanced activity in xenograft models of brain tumors. Methods: NCT04196413 is a 3+3 Phase I dose escalation trial testing GD2-CART in patients with H3K27M DMG, with dose-limiting toxicities (DLT) considered independently for DIPG and spinal DMG (sDMG). Arm A tested escalating doses of IV GD2-CART (DL1: 1e6 GD2-CART/kg; DL2=3e6 GD2-CART/kg) following lymphodepletion (LD). After the DLT period, patients with clinical and/or radiographic benefit were eligible for subsequent ICV GD2-CART (10-30e6 GD2-CART) administered via Ommaya catheter without LD every 4-8 weeks for a maximum of 12 doses. We previously reported early results from 4 patients treated on DL1, which demonstrated clinical activity and manageable toxicity. Here we provide updated results for DL1 and DL2. Results: Thirteen subjects were enrolled and 11 treated [n=4 DL1 (3 DIPG/1 sDMG); n=9 DL2 (7 DIPG/2 sDMG)]. Two subjects were removed prior to treatment due to rapid progression. No DLTs were observed on DL1. Three subjects experienced DLT on DL2 (2 DIPG/1 sDMG) due to grade 4 cytokine release syndrome (CRS), successfully managed with tocilizumab, anakinra, and corticosteroids. CRS occurred earlier on DL2 vs. DL1 (Day 3 vs 7). On both dose levels, all subjects exhibited transient symptoms related to on-tumor inflammation, termed Tumor Inflammation-Associated Neurotoxicity (TIAN), which was successfully managed with anakinra and, in some cases, CSF drainage and dexamethasone. No DLT due to TIAN has occurred. Ten patients have had adequate follow-up to assess benefit. Nine experienced radiographic and/or clinical benefit after IV infusion, and they received subsequent ICV GD2-CART infusions (median= 4 ICV infusions/pt, range 1-6). ICV infusions were not associated with high-grade CRS, although some subjects developed transient fever, headache, meningismus, nausea, and/or vomiting, and several subjects developed TIAN. Four patients continue to receive ICV infusions on study and have experienced continued clinical and radiographic benefit at 11+, 9.5+, 8+ and 7+ months following enrollment. A 31-year-old with sDMG has experienced a near-complete (>95%) reduction in tumor volume and a 17-year-old with DIPG experienced a near-complete (>98%) reduction in volume of a pontine tumor. Conclusions: IV treatment of DIPG and sDMG with GD2-CART is safe at a dose of 1e6/kg, but associated with unacceptable rates of high-grade CRS at 3e6/kg. ICV GD2-CART without LD, administered following a previous course of IV GD2-CART with LD, has been well tolerated and has mediated impressive sustained clinical benefit in some patients with DIPG/sDMG. Given these findings, we are launching a new arm to assess safety and activity and to define the recommended phase 2 dose for ICV delivery of GD2-CART without LD. Patients are eligible for up to 12 ICV infusions of GD2-CART administered every 4-6 weeks. Clinical benefit will be formally assessed using patient-reported outcomes. GD2-CART has the potential to transform therapy for patients with H3K27M+ DIPG/sDMG. Citation Format: Robbie G. Majzner, Jasia Mahdi, Sneha Ramakrishna, Shabnum Patel, Harshini Chinnasamy, Kristen Yeom, Liora Schultz, Valentin Barsan, Rebecca Richards, Cynthia Campen, Agnes Reschke, Angus Martin Shaw Toland, Christina Baggott, Sharon Mavroukakis, Emily Egeler, Jennifer Moon, Ashley Jacobs, Karen Yamabe-Kwong, Lindsey Rasmussen, Esther Nie, Sean Green, Michael Kunicki, Michelle Fujimoto, Zach Ehlinger, Warren Reynolds, Snehit Prabhu, Katherine E. Warren, Tim Cornell, Sonia Partap, Paul Fisher, Gerald Grant, Hannes Vogel, Bita Sahaf, Kara Davis, Steven Feldman, Michelle Monje, Crystal L. Mackall. Major tumor regressions in H3K27M-mutated diffuse midline glioma (DMG) following sequential intravenous (IV) and intracerebroventricular (ICV) delivery of GD2-CAR T cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT001.

Autologous T cells genetically engineered to express a chimeric antigen receptor targeting CD19 (CD19-CAR) have achieved high complete response rates in patients with hematologic malignancies, but >50% of patients progress following therapy. Here, we sought to understand key T-cell intrinsic factors impacting efficacy, namely CAR T-cell expansion, persistence, and homing to the tumor. Using an approach called reverse fate mapping, we followed individual T-cell clones at the single-cell level from pre-manufacture apheresis to the infusion product, tumor-involved lymph node, and blood at peak and late expansion in 12 adult patients with relapsed or refractory large B-cell lymphoma treated with axicabtagene ciloleucel, an FDA-approved CD19-CAR T-cell immunotherapy. The resulting CAR T-cell atlas comprises matched transcriptome (scRNA-seq) and surface protein expression (CITE-seq) for 322,028 cells from 44 samples, with 119,397 unique T-cell receptor (TCR) clonotypes identified. This atlas enabled us to ask questions like: “What were the phenotypes of the most successful CAR T-cell clones at the time of infusion or pre-manufacture apheresis?” We found that T-cell clonotypes with juvenile features at apheresis, including IL7R expression, were the most successful at expansion to higher frequencies in the infusion product, while clones with effector gene expression programs, such as those encoding perforin and granzymes, contracted between apheresis and product. Conversely, it was GZMK-expressing T cells in pre-manufacture apheresis that were dominant in the tumor early following CAR T-cell infusion. Further, T-cell clonotypes with active effector programs at infusion dominated at peak expansion. Finally, we defined active expression modules and pathways in the infusion product for CAR T-cell clones that homed to the tumor or became dominant at late expansion. These analyses pinpoint the molecular mechanisms that could be modulated to rationally steer CAR T-cell differentiation trajectories at the genetic or pharmacological level. This work was supported in part by the Parker Institute for Cancer Immunotherapy, California Institute for Regenerative Medicine, Kite Pharma, and Stanford Cancer Institute. Citation Format: Zinaida Good, Mark P. Hamilton, Jay Y. Spiegel, Sreevidya Kurra, Moksha Desai, Snehit Prabhu, Eric Yang, Michael G. Ozawa, Paul J. Hanson, Fang Wu, Matthew J. Frank, John H. Baird, Lori Muffly, Gursharan K. Claire, Juliana Craig, Katherine A. Kong, Dhananjay Wagh, John Coller, Sylvia K. Plevritis, Bita Sahaf, David B. Miklos, Crystal L. Mackall. Reverse fate mapping of CD19-targeted CAR T cells in patients with large B-cell lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3603.

Introduction: Chimeric Antigen Receptor (CAR) T cell therapy targeting CD19 has shifted our treatment approach for relapsed and refractory (r/r) pediatric B cell acute lymphoblastic leukemia (ALL). The landmark ELIANA pediatric trial studying tisagenlecleucel, CD19-specific CAR T cells, demonstrated a complete response (CR) rate of 81% in 75 infused patients and 12 month overall survival (OS) and event-free survival (EFS) rates of 76% and 50% respectively. Cytokine release syndrome (CRS) and neurotoxicity rates of 77% and 40% were respectively reported. In August 2017, the FDA approved tisagenlecleucel for B-cell ALL that is refractory or in second or greater relapse in patients up to age 25. With CAR commercialization, institutions deliver tisagenlecleucel without the regulation of a clinical study and practices relating to CAR delivery and reporting remain heterogeneous. Here, we report real world clinical outcomes using commercially available tisagenlecleucel for pediatric r/r B-ALL. Methods and Results: Retrospective data were collected from PRWCC member institutions (n=15) and included 200 patients. This includes 15 (7.5%) patients not infused due to manufacturing failure (n=6), death from disease progression and/or toxicity (n=7), or physician discretion following disease remission from prior therapy(n=2). The remaining 185 patients (92.5%) were infused with tisagenlecleucel, including 87% (161) receiving standard-of-care CAR T cell products meeting manufacturing release criteria and 13% (24) receiving CD19-CAR T cells manufactured by Novartis and provided on the managed access program (NCT03601442; n=14) or with single-patient IND approval (n=10). At time of CAR T cell infusion, median age was 12 years (range 0-26) with 40% females and 60% males. Median duration of follow-up at time of analysis was 11.2 months (range 0.2-28.8). The CR rate at 1 month follow up was 79% (156/198) on an intent-to-treat basis and 85% (156/184) among evaluable infused patients. Of infused patients achieving morphologic CR with available testing, 97% (148/153) were negative for MRD by flow cytometry. Duration of remission at 6 and 12 months among patients who achieved CR was 75% and 63% respectively, with 35% (55/156) of responders experiencing relapse. At time of relapse, 41% (21/51) of evaluable patients had relapse with CD19- disease and 59% (30/51) had continued CD19 expression. OS and EFS rates were 85% and 64% at 6 months and 72% and 51% at 12 months, respectively. CRS and neurotoxicity of any grade were seen in 60% (111/184) and 22% (39/181) of evaluable patients with ≥ grade 3 CRS and neurotoxicity rates of 19% (35/184) and 7% (12/181) respectively. One grade 5 CRS and 1 grade 5 neurotoxicity (intracranial hemorrhage) were reported. Post infusion toxicity management included tocilizumab in 26% (47/184) and systemic steroids in 14% (25/184) of patients. Among 181 infused patients with documented disease burden, 51% (95) had high burden (HB) disease , as defined by >5% bone marrow lymphoblasts, peripheral blood lymphoblasts, CNS3 status or non-CNS extramedullary (EM) site of disease; 22% (40) had low burden (LB) disease, defined by detectable disease not meeting the HB criteria; and 25% (46) had no detectable disease (NDD) at time of last evaluation prior to CAR infusion. The morphologic CR rate was lower at day 28 in HB vs. LB and NDD (74% vs. 98% and 96%) and the OS and EFS were lower among patients with HB at 6 mo [OS; 75% (HB), 94%(LB), 98% (NDD), EFS; 50% (HB), 86% (LB), 75%(NDD), p Conclusions: This retrospective, multi-institutional analysis describes real world outcomes using tisagenlecleucel to treat pediatric r/r B-ALL. Early responses at 1 month and OS and EFS at 6 and 12 months are comparable to reported ELIANA trial outcomes. Safety is demonstrated in this cohort with lower rates or CRS and neurotoxicity, likely related to a lower disease burden cohort. Continued relapse and decrease in OS without evident plateau is seen following 6 months post-infusion warranting expanded follow up. Comparative analysis of outcomes in patient cohorts with varying disease burden demonstrate decreased CR, EFS and OS in patients with high disease burden as compared to patients with lower disease burden or no detectable disease at last evaluation prior to CAR infusion. Disclosures Phillips: Novartis: Membership on an entity's Board of Directors or advisory committees. Stefanski:Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Margossian:Novartis: Membership on an entity's Board of Directors or advisory committees; Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees. Verneris:Fate Therapeutics: Consultancy, Current equity holder in publicly-traded company; Novartis: Membership on an entity's Board of Directors or advisory committees; Bmogen: Consultancy, Current equity holder in publicly-traded company; Uptodate: Consultancy. Myers:Novartis: Consultancy, Honoraria, Other: ELIANA trial Steering Committee, Speakers Bureau. Brown:Jazz: Honoraria; Servier: Honoraria; Janssen: Consultancy; Novartis: Consultancy. Qayed:Novartis: Consultancy; Mesoblast: Consultancy. Hermiston:Novartis: Membership on an entity's Board of Directors or advisory committees; Sobi: Membership on an entity's Board of Directors or advisory committees. Satwani:Takeda: Consultancy; Mesoblast: Consultancy. Curran:Novartis: Consultancy, Research Funding; Mesoblast: Consultancy; Celgene: Research Funding. Mackall:Lyell Immunopharma: Consultancy, Current equity holder in private company; Nektar Therapeutics: Consultancy; NeoImmune Tech: Consultancy; Apricity Health: Consultancy, Current equity holder in private company; BMS: Consultancy; Allogene: Current equity holder in publicly-traded company. Laetsch:Cellectis: Consultancy; Novartis: Consultancy, Research Funding; Pfizer: Research Funding; Bayer: Consultancy, Research Funding.

CAR-T cells rest to get back in the race Chimeric antigen receptor (CAR)–T cells, which are engineered to target specific tumor antigens, are increasingly used as an immunotherapy. CAR-T cells have shown promising results in patients, particularly in hematologic cancers, but their anticancer activity can be limited by the onset of exhaustion and the loss of effectiveness. Weber et al. characterized the phenotypic and epigenomic changes associated with CAR-T cell exhaustion caused by continuous activity and the beneficial effects of transient rest periods (see the Perspective by Mamonkin and Brenner). The authors tested different approaches for providing these rest periods, such as using the drug dasatinib to temporarily suppress T cell activity, which helped to prevent exhaustion and improved antitumor activity in mouse models. Science , this issue p. eaba1786 ; see also p. 34

Genome-wide association studies have been effective at revealing the genetic architecture of simple traits. Extending this approach to more complex phenotypes has necessitated a massive increase in cohort size. To achieve sufficient power, participants are recruited across multiple collaborating institutions, leaving researchers with two choices: either collect all the raw data at a single institution or rely on meta-analyses to test for association. In this work, we present a third alternative. Here, we implement an entire GWAS workflow (quality control, population structure control, and association) in a fully decentralized setting. Our iterative approach (a) does not rely on consolidating the raw data at a single coordination center, and (b) does not hinge upon large sample size assumptions at each silo. As we show, our approach overcomes challenges faced by meta-studies when it comes to associating rare alleles and when case/control proportions are wildly imbalanced at each silo. We demonstrate the feasibility of our method in cohorts ranging in size from 2K (small) to 500K (large), and recruited across 2 to 10 collaborating institutions.

Purpose We evaluated the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) variant pathogenicity guidelines for internal consistency and compatibility with Bayesian statistical reasoning. Methods The ACMG/AMP criteria were translated into a naive Bayesian classifier, assuming four levels of evidence and exponentially scaled odds of pathogenicity. We tested this framework with a range of prior probabilities and odds of pathogenicity. Results We modeled the ACMG/AMP guidelines using biologically plausible assumptions. Most ACMG/AMP combining criteria were compatible. One ACMG/AMP likely pathogenic combination was mathematically equivalent to pathogenic and one ACMG/AMP pathogenic combination was actually likely pathogenic. We modeled combinations that include evidence for and against pathogenicity, showing that our approach scored some combinations as pathogenic or likely pathogenic that ACMG/AMP would designate as variant of uncertain significance (VUS). Conclusion By transforming the ACMG/AMP guidelines into a Bayesian framework, we provide a mathematical foundation for what was a qualitative heuristic. Only 2 of the 18 existing ACMG/AMP evidence combinations were mathematically inconsistent with the overall framework. Mixed combinations of pathogenic and benign evidence could yield a likely pathogenic, likely benign, or VUS result. This quantitative framework validates the approach adopted by the ACMG/AMP, provides opportunities to further refine evidence categories and combining rules, and supports efforts to automate components of variant pathogenicity assessments.

Next-generation sequencing technologies are fueling a wave of new diagnostic tests. Progress on a key set of nine research challenge areas will help generate the knowledge required to advance effectively these diagnostics to the clinic.

Divergent adaptation can be associated with reproductive isolation in the process of speciation [1]. We recently demonstrated the link between divergent adaptation and the onset of reproductive isolation in experimental populations of the yeast Saccharomyces cerevisiae evolved from a single progenitor in either a high-salt or a low-glucose environment [2]. Here, we used whole-genome re-sequencing of representatives of three populations to identify 17 candidate mutations, six of which explained the adaptive increases in mitotic fitness in the two environments. In two populations evolved in high salt, two different mutations occurred in the proton efflux pump gene PMA1 and the global transcriptional repressor gene CYC8; the ENA genes encoding sodium efflux pumps were over-expressed once through expansion of this gene cluster and once due to mutation in the regulator CYC8. In the population from low glucose, one mutation occurred in MDS3, which modulates growth at high pH, and one in MKT1, a global regulator of mRNAs encoding mitochondrial proteins, the latter recapitulating a naturally-occurring variant. A Dobzhansky-Muller (DM) incompatibility between the evolved alleles of PMA1 and MKT1 strongly depressed fitness in the low-glucose environment. This DM interaction is the first reported between experimentally evolved alleles of known genes and shows how reproductive isolation can arise rapidly when divergent selection is strong.

Genome-wide association studies (GWAS) have successfully identified hundreds of genetic markers associated with a wide range of diseases and quantitative traits (Hindorff et al. 2009; Manolio et al. 2009). Unfortunately, for most common diseases, nearly all associated variants have small effect sizes and taken together explain very little of the genetically heritable variation of the phenotype (Craddock 2007)—a phenomenon often posed as the conundrum of “missing heritability” (Maher 2008). Furthermore, single-locus association methods tend to implicate individual genes in a particular disease or trait, which in turn highlight a single biological entity involved (Saunders et al. 1993; Hugot et al. 2001; Neale et al. 2010). They do not, by definition, seek to implicate links between the functional elements of a system or elucidate pathway connections that may be broken. Investigation of joint gene–gene effects can therefore improve the explanatory ability of genetics twofold. Firstly, interaction—or statistical epistasis, as defined by Fisher (1918)—is hypothesized to explain a part of disease heritability (Marchini et al. 2005; Evans et al. 2006). Secondly, finding significant statistical links (epistatic or otherwise) between genes could provide strong indications of molecular-level interactions that differ between cases and controls. However, an all-pairs (or all-triples) scan of SNPs genome-wide still poses widely discussed computational challenges due to the sheer size of the combinatorial space (Marchini et al. 2005), both for data sets typed on genotyping arrays ( SNPs) and sequencing technologies ( SNVs). Some methods address this problem by restricting the analysis to a small subset of candidate markers—those identified through single-locus analysis or those of biological interest (Emily et al. 2009), or by only checking for interactions between SNPs that are physically close to one another (Slavin et al. 2011). Others like EPIBLASTER (Kam-Thong et al. 2010) and SHIsisEPI (Hu et al. 2010) make use of specialized hardware like multiple Graphical Processing Units (GPUs) to finish computation on genome-wide data sets on the order of days, rather than weeks or months. While it is known that reductionist, candidate SNP-based approaches can miss many real interactions (Culverhouse et al. 2002; Evans et al. 2006) and fail to provide novel biological insights in an unbiased manner, brute-force approaches that rely on hardware for speedup may also scale poorly as data sets increase in size and interaction tests increase in complexity. For genome-wide interaction analysis to become pervasive, there is a pressing need for algorithmic insights that make interaction testing on large data sets a scalable proposition, without placing undue computing or hardware demands on the investigator. The contribution of our work is such a method. Recently, others had exploited the fact that contrasting the linkage disequilibrium (Zhao et al. 2006), Pearson correlation (Kam-Thong et al. 2010) and log-odds ratio (Plink “--fast-epistasis” option) between a pair of SNPs in cases and controls could be computed more efficiently than maximum likelihood estimates in a logistic regression. Usefully, these computationally efficient contrast tests showed high congruence with statistical epistasis under a variety of genetic models. In this study, we do not devise a new statistical test; rather, we use a simplified version of the LD-contrast test for interaction (Zhao et al. 2006) to demonstrate our computational principles. Our version seeks pairs of physically unlinked (often interchromosomal) SNPs that are in strong LD in cases, but in weak LD, no LD, or reverse LD in controls.2 Our computational approach is driven by the intuition that most genome-wide interaction methodologies only report SNP pairs that are statistically significant (as per the test used) after correcting for the number of tests. The question we ask is this: Given a statistical test, is it possible to identify all the significant SNP pairs with high probability (power), without actually applying the test to all possible combinations genome-wide? In other words, can we design a search algorithm that accepts an arbitrary significance cutoff (as input from the user), and then finds all SNP pairs that will pass this cutoff without a brute-force search? We show here that for some contrast tests, this is indeed possible. At this juncture, it is imperative that we point out the two distinct meanings of “power”: Here, unless otherwise specified, we mean the power of an algorithm to identify SNP pairs for which a test statistic is large (i.e., significant), whereas in the broader context of genome-wide interaction mapping literature, power is the ability of a statistical test to detect a real interaction in the data set. Our work focuses on addressing the computational issues that plague an exhaustive search for interaction, leaving issues of statistical power for a separate discussion. The rest of this study is structured as follows. First, we briefly review a simple LD-contrast test that compares LD between binary allelic states (rather than 0/1/2 genotypes) in cases and controls. Next, we present a novel computational framework—probably approximately complete (PAC) testing—that quantifies the power of a search done by an algorithm. PAC is an intuitive concept: For example, a brute-force method that tests all-pairs of SNPs genome-wide is considered fully powered at finding all significant pairs in our framework (i.e., 100% probability of finding all pairs whose test statistic clears the significance cutoff) and have no element of approximation at all (i.e., 100% complete scan of the interaction space in the case-control data set). In this study, we design a two-stage PAC test for common complex diseases that is guaranteed to find all significant pairwise interactions with high power (e.g., probability >95% of finding all pairs with a significant statistic) by looking at almost the entire space of possibilities (e.g., ∼99% complete scan of interaction space). In return for accepting a small loss of certainty and power, we show that algorithms that offer tremendous computational gains can be designed. We evaluate the performance of our implementation of this framework (SIXPAC) on genome-scale data and then present the results of our analysis on bipolar disorder (BD) in the Wellcome Trust Case Control Consortium (WTCCC) data set (Craddock 2007).

Resequencing genomic DNA from pools of individuals is an effective strategy to detect new variants in targeted regions and compare them between cases and controls. There are numerous ways to assign individuals to the pools on which they are to be sequenced. The naive, disjoint pooling scheme (many individuals to one pool) in predominant use today, offers insight into allele frequencies, but does not offer the identity of an allele carrier. We present a framework for overlapping pool design, where each individual sample is resequenced in several pools (many individuals to many pools). Upon discovering a variant, the set of pools where this variant is observed reveals the identity of its carrier. We formalize the mathematical framework for such pool designs, and list the requirements from such designs. Next, we build on the theory of error-correcting codes to design arrangements that overcome pitfalls of pooled sequencing. Specifically, three practical concerns of low coverage sequencing are investigated: (1) False positives due to errors introduced during amplification and sequencing; (2) False negatives due to undersampling particular alleles aggravated by non-uniform coverage; and consequently (3) Ambiguous identification of individual carriers in the presence of errors. We show that in practical parameters of resequencing studies, our designs guarantee high probability of unambiguous singleton carrier identification, while maintaining the features of naive pools in terms of sensitivity, specificity, and the ability to estimate allele frequencies. We demonstrate the ability of our designs by extracting rare variations on pooled short read data of 12 individuals from the 1000 Genome Pilot 3 project.

Identification of the molecular lesion in Caenorhabditis elegans mutants isolated through forward genetic screens usually involves time-consuming genetic mapping. We used Illumina deep sequencing technology to sequence a complete, mutant C. elegans genome and thus pinpointed a single-nucleotide mutation in the genome that affects a neuronal cell fate decision. This constitutes a proof-of-principle for using whole-genome sequencing to analyze C. elegans mutants.