Bispecific Antibody Construct Research Articles

Bispecific and multispecific antibodies in oncology: opportunities and challenges.

Research into bispecific antibodies, which are designed to simultaneously bind two antigens or epitopes, has advanced enormously over the past two decades. Owing to advances in protein engineering technologies and considerable preclinical research efforts, bispecific antibodies are constantly being developed and optimized to improve their efficacy and to mitigate toxicity. To date, >200 of these agents, the majority of which are bispecific immune cell engagers, are in either preclinical or clinical evaluation. In this Review, we discuss the role of bispecific antibodies in patients with cancer, including history and development, as well as innovative targeting strategies, clinical applications, and adverse events. We also discuss novel alternative bispecific antibody constructs, such as those targeting two antigens expressed by tumour cells or cells located in the tumour microenvironment. Finally, we consider future research directions in this rapidly evolving field, including innovative antibody engineering strategies, which might enable more effective delivery, overcome resistance, and thus optimize clinical outcomes.

Abstract 1866: Discovery of mutation-independent EGFR degrading bispecific antibodies that suppress tumor growth in preclinical tumor models

Abstract Extracellular targeted protein degradation (eTPD) has emerged as a promising new drug modality focused on targeted elimination of extracellular and transmembrane proteins. In contrast to intracellular protein degraders, such as proteolysis targeting chimeras (PROTACs) and molecular glues which require ubiquitin-proteosome cellular degradation pathways, extracellular protein degraders can additionally harness endosomal-lysosomal protein degradation. Two recently published examples of extracellular protein degraders include AbTACs (antibody-based PROTAC) which co-engage a protein of interest (POI) and transmembrane E3 ligases, and KineTACs (cytokine receptor-targeting chimeras) which utilize endogenous cytokine receptors to degrade extracellular POIs. These bispecific antibody degrader platforms not only have advantageous pharmacological and drug-like manufacturing properties, but can also be engineered for tissue-specificity and to address multiple complementary targets, with the goal of increased efficacy and decreased toxicity. To that end, we have greatly expanded the extracellular degrader repertoire beyond AbTACs and KineTACs, with a novel bispecific antibody degrader platform called TrainTACs (tissue receptor antigen internalization targeting chimeras). To demonstrate the potential of this novel extracellular degrading platform, we developed degraders for the canonical receptor tyrosine kinase epidermal growth factor receptor (EGFR). EGFR is an oncogenic driver, that has been clinically validated in lung, colorectal and head and neck cancers, but patient benefit has been limited by treatment-related acquired resistance mutations and on-target/off-tumor dose-limiting toxicities. The potential of our TrainTACs, AbTACs and KineTACs to overcome the limitations of current therapeutic modalities was assessed. Gene expression profiles from tumor and normal tissues were evaluated to identify potential degraders co-expressed with EGFR in tumors. More than 70 unique bispecific antibody constructs spanning 20 receptors were generated and screened in tumor cell-based assays to evaluate EGFR antagonism, internalization, and degradation. TrainTACs, AbTACs and KineTACs degraded EGFR in multiple tumor cell lines covering a range of EGFR mutations. Degradation of EGFR led to deep inhibition of EGFR signaling, robust inhibitory effects on tumor spheroids, and in xenograft mouse tumor models. In conclusion, eTPD represents a promising new drug modality, and TrainTACs, AbTACs and KineTACs have expanded the toolbox of extracellular targeted protein degraders that can be utilized in a target-, tissue- and disease-specific manner. Citation Format: Jonathan Sitrin, Lisa Marshall, Hai Tran, Kenneth Ng, Kimberly Hoi, Josef Gramespacher, Zhong Huang, Andy Goodrich, Filomena Housley, May Dayao, Man-Tzu Wang, Katarina Pance, Aleysha Chen, Kevin Carlin, Lichao Zhang, James Lee, Rami Hannoush, Ken Flanagan, Maia Vinogradova, Isaac Rondon, Shyra Gardai. Discovery of mutation-independent EGFR degrading bispecific antibodies that suppress tumor growth in preclinical tumor models [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 1866.

Brain‐Penetrable Antibody Probes for in vivo Tau Imaging

AbstractBackgroundAntibody‐based PET ligands are desirable due to their high specificity; however, their brain uptakes are limited by the blood‐brain barrier (BBB). We previously demonstrated that transport of antibody across BBB can be facilitated through interaction with the transferrin receptor (TfR)1. We report here the first in vivo PET imaging studies using BBB permeable F‐18 radiolabeled bispecific antibodies specific to tau protein in mice.MethodWe performed multiple PET scans using each of the radio‐probes synthesized via conjugation of [18F]SFB with one of the four bispecific antibody constructs: 1) 6B2G12‐ScFv8D3 (full‐size Tau antibody conjugated with TfR fragment (ScFv8D3); abbreviated as TAUb), 2) ScFv235‐ScFv8D3 (small Tau‐TfR; TAUs), 3) 3D6‐ScFv8D3 (full‐size Aβ‐TfR; Aβb) or 4) ScFv3D6‐ScFv3D6 (small Aβ‐TfR; Aβs). We used a matched control design to compare a WT with one of the two types of transgenic (tg) tauopathy mice [PS19 (P301S) or JNPL3 (P301)] in each scan. Subjects were scanned on a MOLECUBES PET/CT scanner. An initial low‐dose (244 ± 54 μCi) dynamic scan (30 min) was performed, followed by two scans at 8hr and 12hr (20 min each static scans) after injection of a high dose (1832 ± 447 μCi) of the same probe. PMOD and FireVoxel were used for imaging processing. After co‐registration of PET/CT/atlas, data were quantified as SUV and SUVR [CB as a reference], and clearance was estimated as %change/hr.ResultTAUs probe displayed highest brain uptake with more specific binding retained in the brain, as compared to TAUb and Aβs. TAUs probe showed faster brain penetration in tau‐tg mice than in WT, with differences in amygdala, basal forebrain, and hypothalamus of 66%, 62% and 85%, respectively, at 8hr post‐injection (Fig. 1). Further, higher brain uptakes (TAUs > Aβs) in the same tau‐tg mouse (Fig. 2) suggest the specific binding of TAUs probe, while Aβs probe can serve as a control due to its lack of specific binding in tau‐tg mice.ConclusionWe have successfully synthesized and evaluated four novel F‐18 bispecific antibodies via in vivo PET. TAUs probe showed most promising BBB permeability and binding specificity to tau antigen with a reasonable clearance.

Rapid Access to Potent Bispecific T Cell Engagers Using Biogenic Tyrosine Click Chemistry.

Bispecific antibodies as T cell engagers designed to display binding capabilities to both tumor-associated antigens and antigens on T cells are considered promising agents in the fight against cancer. Even though chemical strategies to develop such constructs have emerged, a method that readily converts a therapeutically applied antibody into a bispecific construct by a fully non-genetic process is not yet available. Herein, we report the application of a biogenic, tyrosine-based click reaction utilizing chemoenzymatic modifications of native IgG1 antibodies to generate a synthetic bispecific antibody construct that exhibits tumor-killing capability at picomolar concentrations. Control experiments revealed that a covalent linkage of the different components is required for the observed biological activities. In view of the highly potent nature of the constructs and the modular approach that relies on convenient synthetic methods utilizing therapeutically approved biomolecules, our method expedites the production of potent bispecific antibody constructs with tunable cell killing efficacy with significant impact on therapeutic properties.

Innate Cell Engager (ICE®) AFM13 Combined with Preactivated and Expanded (P+E) Cord Blood (CB)-Derived Natural Killer (NK) Cells for Patients with Refractory CD30-Positive Lymphomas: Final Results

CONCLUSIONS Pts with refractory Hodgkin (HL) and other CD30+ lymphomas have few effective therapies available and targeted NK cell immunotherapy is an active area of research. Transfer of non-targeted NK cells has shown limited clinical benefit as target recognition of cancer cells by NK cells constitutes a barrier. The innate cell engager AFM13 is a first-in-class CD30/CD16A bispecific antibody construct that induces selective killing of CD30+ tumor cells by engaging and activating NK cells. As a single agent, AFM13 has limited clinical activity against HL, likely due to the reduced effector function of autologous NK cells in these pts, which provides the rationale for combining AFM13 with allogeneic NK cells. Our prior preclinical work identified a promising combination of CB-derived NK cells that were first IL-12/IL-15/IL-18-preactivated followed by ex vivo expansion (P+E) with engineered K562 feeder cells expressing membrane-bound IL-21, 4-1BBL and CD48 and exogenous IL-2, and subsequently complexed with AFM13 just before infusion; these cytokine-induced memory-like NK cells presented chimeric antigen receptor (CAR)-NK-like features and increased in vitro and in vivo antitumor activity compared to either non-AFM13-precomplexed P+E NK cells or to AFM13 alone (Kerbauy et al, Clin Cancer Res 2021). Building on this work we wished to study this CB NK-cell therapy in pts with refractory CD30+ lymphomas. This investigator-initiated single-center phase I-II trial (NCT04074746) evaluated the safety and activity of P+E AFM13-precomplexed CB NK cells followed by systemic AFM13. Pts ages 15-75 with CD30+ lymphomas refractory to brentuximab vedotin were enrolled. The goals were to identify the recommended phase 2 dose (RP2D) of NK cells, estimate the overall (ORR) and complete response (CR), event-free (EFS) and overall survival (OS) rates, and study NK cell activation and persistence. Each treatment cycle consisted of fludarabine/cyclophosphamide (days −5 to −3) followed by infusion (day 0) of the AFM13-precomplexed CB NK cells, cultured for 14 days as described above, and three weekly IV infusions of AFM13 (200 mg, days 7, 14 and 21). Two cycles were planned for pts 1-19 and up to 4 cycles from patient 20 on. Response was evaluated on day 28 of each cycle. Enrollment proceeded at 3 dose levels (DL) of 10 6, 10 7 and 10 8 NK/Kg, respectively. Each patient received the same NK dose in all cycles. Between 09/20-12/22 we enrolled 42 pts (37 HL, 5 NHL), pretreated with a median 7 lines, at DL1 (N=3), DL2 (N=3) and DL3 (N=36), with 1 (N=2), 2 (N=21), 3 (N=3) or 4 cycles (N=16). All had active progressive disease at enrollment and no bridging therapy was given. The cords used for each of the 117 cycles were selected without consideration for HLA match, which was 0/6 (N=53), 1/6 (N=46), 2/6 (N=14), 3/6 (N=3) and 4/6 (N=1). The product infused consisted of >94% NK cells (expanded >2,700 fold), with only 0.02% T cells. The NK cells were 96% viable, 91% bound to AFM13, and 97.6% CD16+. The products were infused fresh. There was 1 grade 2 infusion-related reaction (IRR) in 115 infusions of AFM13-NK (0.8%) and 27 IRR infusion related reactions in 350 infusions of AFM13 (7.7%) (1 G3, 26 G2). We saw no cases of CRS, ICANS or GVHD of any grade. DL3 (10 8 NK/Kg) was established as the RP2D. The ORR and CR were 92.8% and 66.7%, respectively (94.4% and 72.2%, respectively, in 36 pts treated at the RP2D). Nine pts had a response consolidated with SCT (5 allo, 4 auto). At median follow-up of 14 (6-34) months, the EFS/OS rates are 31%/76%; median EFS/OS are 8 months/not reached. No relapses were associated with antigen loss. In the subset of pts planned for 4 cycles at RP2D, 15 of 23 pts remained event free at 6 mo (4 pts with and 11 without SCT consolidation), for 6-month EFS/OS rates of 65%/83%. Overall, 8 of the 9 pts (89%) receiving a consolidation SCT are event free at 10 to 34 months. AFM-bound CB NK cells were detected in blood from day 1 post infusion for up to 3 weeks. Donor NK levels followed similar patterns after each cycle, which argues against a sensitization after the first NK infusion. CyTOF studies showed increased expression of AFM13 and activation markers in donor NK cells. We confirmed trafficking of AFM13-bound donor NK cells to tumor sites shortly after infusion. CB-derived cytokine-induced memory-like NK cells precomplexed with AFM13 have excellent tolerability and activity for pts with heavily pretreated and refractory CD30+ lymphoma.

Targeting GD2 after allogeneic SCT: effector cell composition defines the optimal use of ch14.18 and the bispecific antibody construct NG-CU (GD2-CD3).

We investigated whether T cell-recruiting bispecific anti-CD3/GD2 antibody NG-CU might be an alternative to therapeutic anti-GD2 monoclonal antibody (mAb) ch14.18, mediating complement-dependent cytotoxicity (CDC) and antibody-dependent cell-mediated cytotoxicity (ADCC) through natural killer (NK) cells for immunotherapy in high-risk/relapsed neuroblastoma after autologous/allogeneic stem cell transplantation (auto/alloSCT). Different antibody concentrations and effector-to-target ratios (E:T) were evaluated using xCELLigence RTCA system, peripheral blood mononuclear cells (PBMCs) (healthy donors and patients after alloSCT), and neuroblastoma cell lines (LS/LAN-1). Mean specific lysis of LS cells utilizing PBMCs from healthy donors and ch14.18 (1µg/ml) was 40/66/75% after 12/24/48 h compared to 66/93/100% in the presence of NG-CU (100 ng/ml). NG-CU showed enhanced cytotoxicity compared to ch14.18, even at lower concentrations and E:T ratios, and completely eradicated LS cells after 72 h. To decipher the influence of effector cell subsets on lysis, different ratios of T and NK cells were tested. At a ratio of 1:1, ch14.18 was more effective than NG-CU. Using patient PBMCs taken at different time points posttransplant, significant lysis with both constructs was detectable depending on percentages and total numbers of T and NK cells; in the early posttransplant phase, NK cells were predominant and ch14.18 was superior, whereas later on, T cells represented the majority of immune cells and NG-CU was more effective. Our study highlights the importance of analyzing effector cell subsets in patients before initiating antibody-based therapy. Consequently, we propose an adjusted administration of both antibody constructs, considering the state of posttransplant immune recovery, to optimize anti-tumor activity.

SOHO State of the Art Updates and Next Questions | Optimal Timing of Blinatumomab for the Treatment of B-Acute Lymphoblastic Leukemia.

Blinatumomab is a CD19 targeting bi-specific T-cell engager antibody construct developed for the treatment of CD19 expressing B-cell malignancies. Numerous adult and pediatric B-ALL clinical trials have demonstrated blinatumomab's efficacy in the relapse setting as well as in patients with residual disease after upfront chemotherapy. The safety profile of blinatumomab is also favorable, making it a feasible option for most patients. Several key questions remain, including the role of blinatumomab as a replacement for toxic elements of standard chemotherapy regimens in the upfront setting, its role as a bridge to hematopoietic stem cell transplantation, or whether previous blinatumomab impacts the efficacy of subsequent CAR-T cell therapy.

Safety of AFM11 in the treatment of patients with B-cell malignancies: findings from two phase 1 studies

BackgroundThe prognosis for patients with relapsed and/or refractory (R/R) non-Hodgkin’s lymphoma (NHL) or acute lymphoblastic leukaemia (ALL) remains poor, with existing treatments having significant side effects. Developed for the treatment of these cancers, AFM11 is a tetravalent, bispecific humanised recombinant antibody construct (TandAb®) designed to bind to human CD19 and CD3 and lead to the activation of T cells inducing apoptosis and killing of malignant B cells.MethodsTwo open-label, multicentre, dose-escalation phase 1 studies evaluated the safety, pharmacokinetics and activity of AFM11 in patients with R/R CD19-positive B cell NHL (AFM11-101) and in patients with CD19 + B-precursor Philadelphia-chromosome negative ALL (AFM11-102). Adverse events (AEs) were assessed and recorded; imaging (NHL) or bone marrow assessment (ALL) were used to evaluate response. Additional pharmacodynamic assays undertaken included cytokine release analysis and B-cell and T-cell depletion.ResultsIn AFM11-101, 16 patients with R/R NHL received AFM11 in five different dose cohorts. Of which, 14 experienced drug-related treatment-emergent AEs (TEAEs) [including five serious AEs (SAEs)], five patients experienced dose-limiting toxicity (DLT) and ten patients discontinued the study. The high number of neurological events led to a decrease in infusion frequency during the study. No objective response to treatment was observed. In AFM11-102, 17 patients with R/R ALL received AFM11 in six different dose cohorts. Thirteen patients experienced drug-related TEAEs (including four SAEs), DLTs occurred in two patients and five patients discontinued the study. An objective response was recorded in three patients. The maximum tolerated dose could not be determined in either study due to early termination.ConclusionsAFM11 treatment was associated with frequent neurological adverse reactions that were severe in some patients. In ALL, some signs of activity, albeit short-lived, were observed whereas no activity was observed in patients with NHL; therefore, further clinical development was terminated.Trial registrationNCT02106091. Safety Study to Assess AFM11 in Patients With Relapsed and/or Refractory CD19 Positive B-cell NHL. Registered April 2014. NCT02848911. Safety Study to Assess AFM11 in Patients With Relapsed or Refractory Adult B-precursor ALL. Registered July 2016.

Emerging Antibodies in Cancer Therapy

Since the first monoclonal antibody, Orthoclone OKT3, was approved in 1986 for the treatment of acute allograft rejection in renal transplant recipients, the number of antibody‐based therapies has increased remarkably, with more than 130 monoclonal antibodies being currently available. In particular, due to their robust binding affinity and high specificity, various monoclonal antibodies have achieved success as complementary antitumor drugs. Moreover, therapeutic potency can be improved by engineering Y‐shaped binding sites of the antibodies to simultaneously recognize two distinct antigens; such molecules are called bispecific antibodies and function as a bridge, with one site binding to a tumor‐specific antigen and the other to immune cells to activate host immune responses. In addition, the development of trispecific antibodies has led to further enhancement of therapeutic specificity and effects. Indeed, a variety of structurally engineered antibodies have proven effective in tumor therapy. This article provides insights into the clinical translation of emerging antibody constructs, firstly describing conventional T cell‐redirecting bispecific antibodies and subsequently discussing the application of bispecific and trispecific antibody constructs based on innate immune cell recruitment. Finally, bispecific antibodies against mutant p53 and KRAS that have been actively studied in cancer treatment are introduced.

Innate Cell Engager AFM13 Combined with Preactivated and Expanded Cord Blood-Derived NK Cells for Patients with Double Refractory CD30+ Lymphoma

Patients with multiple relapsed or refractory (R/R) Hodgkin lymphoma (HL) and other CD30+ lymphoma have few effective therapies available and targeted adoptive natural killer (NK) cell immunotherapy is an active area of investigation. Transfer of non-targeted NK cells has shown limited clinical benefit as target recognition of cancers by NK cells remains a substantial barrier. The innate cell engager (ICE®) construct, AFM13, is a first-in-class CD30/CD16A bispecific antibody construct that induces specific and selective killing of CD30+ tumor cells by engaging and activating NK cells. As a single agent, AFM13 has limited clinical activity against HL, likely due to the reduced effector function of autologous NK cells in these patients, which provides the rationale for the combination of AFM13 with allogeneic NK cells. Our prior preclinical work identified a promising combination of cord blood (CB)-derived NK cells that were first IL-12/IL-15/IL-18-preactivated followed by expansion (P+E) with engineered K562 feeder cells expressing membrane-bound IL-21, 4-1BB ligand and CD48 ex vivo, and subsequently complexed with AFM13 just prior to infusion (Kerbauy et al, Clin Cancer Res 2021; 27:3744-56). These cytokine-induced memory-like CB-derived NK cells precomplexed with AFM13 (AFM13-NK) presented chimeric antigen receptor (CAR)-NK cell-like features and showed an increased in vitro and in vivo antitumor activity compared to either P+E NK or AFM13 alone. Building on this work we wished to clinically study adoptive AFM13-NK cell therapy in pts with R/R CD30+ lymphomas. METHODS: This single-center phase I-II trial (NCT04074746) evaluates the safety and activity of AFM13-NK, followed by AFM13 monotherapy. Patients aged 15-75 years with R/R CD30+ lymphomas refractory to brentuximab vedotin are enrolled. The trial goals are to identify the recommended phase 2 dose (RP2D) of AFM13-NK cells, estimate the overall (ORR) and complete response (CR) rates, event-free (EFS) and overall survival (OS) rates, and studies of NK cell activation and persistence. Each treatment cycle consisted of fludarabine/cyclophosphamide (days −5 to −3), followed by infusion of the AFM13-NK cells, cultured for 14 days as described above (day 0), and three weekly IV infusions of AFM13 (200 mg, days 7, 14 and 21). Two cycles were administered to patients 1-19 and 4 cycles were given starting with patient 20. Response was evaluated on day 28 of each cycle. Patients were enrolled at 3 dose levels: DL1 (106 NK/Kg), DL2 (107NK/Kg) and DL3 (108 NK/Kg). Each patient received the same NK dose in all cycles. RESULTS: As of 7/31/22, 30 pts have been enrolled and treated (Table) at DL1 (N=3), DL2 (N=3) and DL3 (N=24), with 1 (N=1), 2 (N=24), 3 (N=3) or 4 cycles (N=3). All patients had active progressive disease at enrollment and no bridging therapy was given. The cords used for each of the 69 cycles were selected with no consideration for HLA match, which was 0/6 (N=31), 1/6 (N=31), 2/6 (N=5), 3/6 (N=1) and 4/6 (N=1). The product infused consisted of >99% NK cells (expanded >2,700 fold), with only 0.02% T cells. The NK cells were >98% viable and 92% of them were bound to AFM13. The products were infused fresh. There were no infusion-related reactions (IRR) after infusing AFM13-NK and 11 IRR in 182 infusions of AFM13 alone (6%) (1 grade 3, 10 grade 2). We saw no cases of CRS, ICANS or GVHD of any grade. DL3 (108 NK/Kg) was established as the RP2D. The ORR in the entire study is 97% with 63% CRs. All 24 patients treated at the RP2D responded (100% ORR) with 70.8% (17/24) CRs. Five patients had a response consolidated with a SCT. At median follow-up of 8 (range 1-23) months, the EFS and OS rates of all 30 patients are 57% and 83%, respectively. AFM-NK cells were detected in serum from day 1 post infusion and persisted for 3-4 weeks. Donor NK levels followed similar patterns after each cycle, which argues against a sensitization effect in the patient after the first NK infusion. Donor NK cells showed by CyTOF increased binding to AFM13 and expression of activation markers. CONCLUSIONS: This is the first clinical trial to date using an ICE® construct precomplexed with cytokine-induced memory-like CB-NK cells to treat patients with CD30+ R/R HL and NHL. Our preliminary results indicate that this ICE® precomplexed CB-NK cell therapy has an excellent tolerability profile and is highly active in patients with heavily pretreated R/R CD30+ lymphomas and warrants further investigation. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Incidence of Central Nervous System Toxicity and Cytokine Release Syndrome in Leukemia Patients Treated with Blinatumomab

INTRODUCTION Blinatumomab (BLINA) is a bispecific antibody construct that directs cytolytic T-cells to CD19 expressing B-cells resulting in proliferation of T-cells. It has potent activity in B-cell acute lymphoblastic leukemia (B-ALL) and chronic myeloid leukemia in lymphoid blast phase (CML-BP) and is associated with a risk of central nervous system toxicity (CNS tox) and cytokine release syndrome (CRS). Here we report the incidence, severity, and management of CNS tox and CRS in patients with B-ALL and CML-BP who received BLINA in a real-world setting. METHODS We retrospectively evaluated pts ≥18 years old with B-ALL or CML-BP treated with BLINA from 2016 to 2020 at MD Anderson Cancer Center. CNS tox and CRS were graded per CTACEv5 (Common Terminology Criteria for Adverse Events). CNS tox excluded headache. The primary objective was to evaluate the incidence of CNS tox and CRS. Secondary objectives included evaluation of toxicity severity, time to toxicity, duration of toxicity, and management of toxicity. Risk factors for toxicity were assessed in a multivariate analysis. RESULTS A total of 265 patients (pt) met eligibility criteria. Their median age was 45 years (range, 18-84 years) and 155 (58%) were male (Figures 1A). The cumulative number of BLINA cycles was 754 with a median of 2 cycles per pt (range, 1-11). Sixty-five pts (25%) received BLINA as part of frontline therapy with an overall median number of 1 prior therapy. Prior to starting BLINA, 43 pts (24%) had ≥20% bone marrow blasts and 124 pts (47%) were minimal residual disease (MRD) negative. Ten pts (4%) had active CNS disease before BLINA initiation, 27 pts (10%) had a history of prior CNS disease. A dose ramp-up was performed in 279 BLINA cycles (37%). During these cycles, dose escalation occurred on a median of day 8 (range, 2-17). Eighty-eight pts (33%) experienced CNS tox while receiving BLINA. CNS tox occurred during 15% (114) of all BLINA cycles and were classified as grade 1, 2, 3, and 4 or greater during 46 (40%), 37 (32%), 28 (25%), and 3 cycles (3%), respectively. Seventy CNS tox cases (61%) occurred during cycle 1, 29 (25%) during cycle 2, and 15 (13%) during cycle 3 or beyond (Figure 1B). Of the CNS tox cases occurring after cycle 1, 23 (52%) had CNS tox during a previous cycle. Median time to onset of CNS tox was 4 days (range, 1-25), median duration of CNS tox was 3 days (range, 1-28). Common symptoms of CNS tox included tremor (n=57), confusion (n=47), weakness (n=11), aphasia (n=9), and gait disturbance (n=8). Of the 114 cycles in which CNS tox occurred, 97 (85%) were managed with steroids. BLINA was held during 71 cycles due to CNS tox (62%) and dose-reduced during 44 cycles (39%). Thirteen cycles with CNS tox (11%) required no intervention. Of the 114 cycles with CNS tox, 66 (58%) were followed by additional cycles of BLINA. Pts who experienced CNS tox received a similar total number of BLINA cycles (median, 2) as pts who did not (median, 2). Pts who experienced CNS tox had a similar median overall survival (1.5 years) to those who did not (1.6 years). Notably, 51% of pts with active CNS disease or a prior history of CNS disease experienced CNS tox compared with 30% of pts without CNS disease. Preliminary multivariate logistic regression identified age, gender, and MRD positivity as risk factors for CNS tox. Ninety-four pts (35%) experienced CRS. CRS occurred during 115 cycles (15%), of which 36 (31%) were in the setting of a confirmed infection. CRS was classified as grade 1, 2, or 3 during 85 (74%), 23 (20%), and 7 cycles (6%), respectively. Eighty-two CRS cases (71%) occurred during cycle 1, 24 (21%) during cycle 2, and 9 (8%) during cycle 3 or beyond (Figure 1B). Of the CRS cases occurring after cycle 1, 19 (58%) had CRS during a previous cycle. Median onset of CRS was 3 days (range, 1-22), median duration of CRS was 2 days (range, 1-14). CRS was managed with steroids during 83 cycles (72%). BLINA was held during 56 cycles due to CRS (49%) and dose-reduced during 21 cycles (18%). Twenty-six cycles with CRS (23%) required no intervention. Of the 115 cycles with CRS, 79 (69%) were followed by additional cycles of BLINA. CONCLUSION CNS tox and CRS occurred in 33% and 35% of pts treated with BLINA and in 15% and 15% of BLINA cycles, respectively. With steroids and therapy modification, most pts who experienced toxicity were able to continue BLINA treatment while exhibiting similar outcomes to those without toxicity. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Determining Binding Affinity (KD) of Radiolabeled Antibodies to Immobilized Antigens.

Determining binding affinity (KD) is an important aspect of the characterization of radiolabeled antibodies (rAb). Typically, binding affinity is represented by the equilibrium dissociation constant, KD, and can be calculated as the concentration of antibody at which half the antibody binding sites are occupied at equilibrium. This method can be generalized to any radiolabeled antibody or other protein and peptide scaffolds. In contrast to cell-based methods, the choice of immobilized antigens is particularly useful for validating binding affinities after long-term storage of antibodies, distinguishing binding affinities of fragment antigen-binding region (Fab) arms in bispecific antibody constructs, and determining if there is variability in antigen expression between different cell lines. This method involves immobilizing a fixed amount of antigen to specified wells on a breakable 96-well plate. Then, nonspecific binding was blocked in all wells with bovine serum albumin (BSA). Subsequently, the rAb was added in a concentration gradient to all wells. A range of concentrations was chosen to allow the rAb to reach saturation, i.e., a concentration of antibody at which all antigens are continuously bound by the rAb. In designated wells without immobilized antigen, nonspecific binding of the rAb can be determined. By subtracting nonspecific binding from total binding in the wells with immobilized antigen, specific binding of the rAb to the antigen can be determined. The KD of the rAb was calculated from the resulting saturation binding curve. As an example, binding affinity was determined using radiolabeled amivantamab, a bispecific antibody for epidermal growth factor receptor (EGFR) and cytoplasmic mesenchymal-epithelial transition (cMET) proteins.

Abstract CT003: Innate cell engager (ICE®) AFM13 combined with preactivated and expanded cord blood (CB)-derived NK cells for patients with refractory/relapsed CD30+ lymphoma

Abstract Pts with R/R CD30+ lymphoma have few effective therapies available. We previously developed a promising therapy using CB-derived NK that were first IL-12/IL-15/IL-18-preactivated followed by expansion (P+E) with K562 feeder cells ex vivo and subsequently complexed with ICE® AFM13 (a first-in-class CD30/CD16A bispecific antibody construct) prior to infusion. These cytokine-induced memory-like NK showed greater in vivo antitumor activity than either P+E NK or AFM13 alone. This single-center phase I-II trial (NCT04074746) evaluates AFM13-precomplexed CB-NK cells in pts ages 15-75 with R/R CD30+ lymphoma. Pts receive 2 cycles of fludarabine/cyclophosphamide (days −5 to −3) followed by AFM13-CB NK cells (day 0) and 3 weekly intravenous infusions of AFM13 (200 mg, days 7, 14 and 21). Pts were enrolled at 3 dose levels: DL1 (106 NK/Kg), DL2 (107 NK/Kg) and DL3 (108 NK/Kg). 19 pts have been treated to date at DL1 (N=3), DL2 (N=3) and DL3 (N=13); 18 have completed both planned cycles. All pts had active progressive disease at enrollment and no bridging therapy was given. Cords were selected for the 37 cycles without consideration for HLA match, which was 0/6 (N=16), 1/6 (N=17), 2/6 (N=2), 3/6 (N=1) and 4/6 (N=1). There were no cases of CRS, ICANS or GVHD; 6 infusion-related reactions (1 G3, 5 G2) in 110 infusions of AFM13 alone and no reactions to the AFM13-loaded NK cells. DL3 was established as the RP2D. There were 17/19 responses (8 CR, 9 PR, 2 PD). All 13 pts treated at the RP2D responded (6 CR, 7 PR). 2 pts subsequently received an autologous SCT. At median follow-up of 6 (2-16) months PFS/OS across all dose levels are 58%/79%. Expansion of CB NK cells occurred as early as 3 days post infusion and persisted for up to a month. In conclusion, the preliminary results of this first clinical trial of ICE®-precomplexed NK cells for R/R CD30+ lymphoma indicate excellent tolerability and high activity and warrant further investigation of this approach. Baseline patient characteristics (N=19) Age, median (range) 37 (20-68) Gender (male/female) 13/6 Diagnosis (Hodgkin/T-NHL) 17/2 No. prior lines therapy, median (range) 6 (1-14) Prior brentuximab vedotin 19 Prior anti-PD-1 18 Prior SCT (autologous/allogeneic) 13 (8/5) Prior cellular therapy (CAR-T) 2 No. prior relapses/progressive disease, median (range 5 (1-14) Citation Format: Yago Nieto, Pinaki Banerjee, Indreshpal Kaur, Roland Bassett, Lucila Kerbauy, Rafet Basar, Mecit Kaplan, Lori Griffin, Daniel Esqueda, Christina Ganesh, Melissa Barnett, Amin Alousi, Chitra Hosing, Jeremy Ramdial, Neeraj Saini, Samer Srour, Karenza Alexis, Andreas Harstrick, Elizabeth J. Shpall, Katayoun Rezvani. Innate cell engager (ICE®) AFM13 combined with preactivated and expanded cord blood (CB)-derived NK cells for patients with refractory/relapsed CD30+ 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 CT003.

Abstract IAP0301: Bispecific T cell engagers (TCEs) for treatment of solid tumors: Challenges and opportunities

Abstract T cell-engaging bispecific antibody constructs (TCEs) have the potential to engage all cytotoxic T cells for redirected cancer cell lysis, including CD8, CD4, gamma/delta, NKT and regulatory T cells. TCEs transiently connect T cells with target cells leading to formation of a cytolytic synapse. Once activated by TCEs, T cells can adopt a serial lysis mode and increase troop size by their local proliferation. In numerous animal models, TCEs can eradicate large tumors as single-agent therapy. TCEs show a remarkable single-agent activity in hematological diseases, including ALL, NHL (DLBCL and FL) and AML. A CD19/CD3-bispecific TCE called blinatumomab (Blincyto) has been approved in 2014, and six other TCEs targeting CD20, BCMA and CD123 are in late-stage development. With complete response rates >50% and durability of responses, TCEs are in par with autologous CD19 CAR-T cells in NHL but are easier to manufacture and allow for a faster treatment start. Treatment of solid tumor indications with TCEs is not as successful as treatment of blood-borne cancers. This may be due to issues with tumor penetration, the hostile tumor microenvironment and the scarcity of T cells in cold tumors. A variety of TCEs explore their single-agent activity against solid tumors by targeting surface antigen like PSMA, GPC3, GD2, or HER-2. Enhancements in the form of combination therapy with checkpoint inhibitors, or 4-1BB and CD28 agonists are being explored as ways of augmenting solid tumor activity of TCEs. The first TCE showing robust anti-tumor activity in a solid tumor indication is targeting a gp100 peptide/MHC (developed by Immunocore). It is made from a soluble TCR fragment of very high affinity for the gp100 peptide HLA*0201 complex fused to an anti-CD3 scFv fragment. Although only few partial responses were observed in uveal melanoma patients, the single-agent treatment showed a very robust improvement in overall survival with a hazard ratio of 0.51 compared to investigators’ choice. This data will inspire the development of more TCEs that are specific for peptide/MHC complexes. This class of targets can also be addresses by TCR-mimetic antibodies. TCEs using such antibodies showed high anti-tumor activity in mouse models against MAGE-A4, p53 mutant or tyrosinase peptide/MHC complexes, all targets that are otherwise not amenable to classical TCEs. Another emerging opportunity for treating cancers are conditional TCEs. These are infused as inactive precursors. Inside tumors, the locally produced proteases convert the precursors into highly active TCEs. One example is the so called ProTriTACs, where a half-life extending albumin-binding domain masks the CD3-binding domain. The mask confers a 500-fold reduction in the cytolytic activity of the TCE. Inside tumors, the mask is cleaved off and a TCE is activated that has a very short serum half-life once it is released from the tumor. Conditional TCEs promise not only a wider therapeutic index but can also leverage as targets tumor-associated antigens that are widely expressed on tumors but also on healthy normal tissues, such as EGFR or EpCAM/TROP1. On-target toxicity of TCEs relates to the quality of target antigens. Finding more tumor-specific surface antigens, appropriate pMHC targets (e.g., mutant oncogene epitopes) and conditional T cell engagers are ways to address this challenge. Off-target toxicity of TCEs relates to T cell activation because of redirected target cell lysis. In the worst case, a cytokine release syndrome is induced. In the best case, a low-level cytokine release is contributing to the efficacy of TCEs. CRS can be managed by steroids, step dosing, anti-IL6 antibody and/or dasatinib. Another promising means is subcutaneous administration of TCEs. A dozen TCE programs currently explore this route, which allows reduction of Cmax and of the initial cytokine release. TCEs will soon become a major weapon in the arsenal of cancer therapy. They will not only work in blood-borne cancers but will soon conquer select solid tumor indications. Peptide/MHC-targeting and conditional TCEs are very promising approaches to achieve solid tumor activity of TCEs. Citation Format: Patrick A. Baeuerle. Bispecific T cell engagers (TCEs) for treatment of solid tumors: Challenges and opportunities [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr IAP0301.

Transcriptomic Features of Immune Exhaustion and Senescence Predict Outcomes and Define Checkpoint Blockade-Unresponsive Microenvironments in Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is a molecularly and clinically heterogeneous disease. Reinstating immunological control of AML is highly desirable to eradicate chemotherapy-resistant clones and provide long-term disease control. We recently identified bone marrow (BM) microenvironmental transcriptomic profiles that stratify patients with newly diagnosed AML into an immune-infiltrated and an immune-depleted subtype and that refine the accuracy of survival prediction beyond that afforded by current prognosticators (Vadakekolathu J et al., 2020). We have also shown that CD8 + T cells from patients with AML exhibit features of immune exhaustion and senescence (IES), including heightened expression of killer cell lectin-like receptor subfamily G member 1 (KLRG1) and B3GAT1 (encoding CD57) (Knaus H et al., 2018). Whether deranged T-cell functions affect the likelihood of responding to antitumor therapy, including immune checkpoint blockade (ICB), is an outstanding question in AML.In the current study, we analyzed 183 BM samples collected longitudinally at time of AML onset, response assessment and disease relapse from multiple cohorts of patients with AML treated with standard-of-care induction chemotherapy, and from 33 elderly AML patients with newly diagnosed or chemotherapy-refractory/relapsed AML treated with azacitidine, and the PD-1 checkpoint inhibitor pembrolizumab (NCT02845297). Primary patient specimens and associated clinical data were obtained via informed consent in accordance with the Declaration of Helsinki on research protocols approved by the Institutional Review Boards of the participating Institutions. RNA (150-200 ng) was extracted from BM aspirates and was processed on the nCounter FLEX analysis system (NanoString Technologies, Seattle, WA) using the PanCancer Immune profiling panel, as previously published (Vadakekolathu J et al., 2020). The correlation between transcriptomic features of IES, clinical characteristics, therapeutic response and patient outcome was validated using publicly available RNA-sequencing and NanoString data from 1,698 patients with AML, including samples from the TCGA-AML (n=147 cases), Beat-AML Master Trial (n=264 cases, of which 240 with survival data and 195 with chemotherapy response data) and Children's Oncology Group (COG)-TARGET AML series (n=145 cases).We initially showed that, compared with their non-senescent CD8 +CD57 -KLRG1 - counterpart, senescent CD8 +CD57 +KLRG1 + T cells are functionally impaired in terms of their ability to effect AML-blast killing mediated by an anti-CD33/CD3 bi-specific T-cell engager antibody construct (kindly provided by Amgen, USA; effector/target [E/T] ratio = 1:5). We then used gene set enrichment analysis (GSEA) to derive a transcriptomic signature of IES encompassing natural killer (NK)-cell and stem-like CD8 + T-cell markers, and showed that IES states correlate with lymphoid infiltration, adverse-risk molecular lesions (TP53 and RUNX1 mutations), experimental gene signatures of leukemia stemness (LSC17 score; Ng et al., 2016) and poor outcome in response to standard induction chemotherapy (Fig. 1A). In independent validation cohorts of children and adults with AML, the IES score was higher at baseline in patients with primary induction failure (following a standard 2 cycles of chemotherapy) compared with complete remission, increased in post-chemotherapy BM specimens, and predicted survival with greater accuracy than the ELN cytogenetic risk classifier (Fig. 1B). In the immunotherapy setting, high IES scores at baseline defined a checkpoint blockade-unresponsive AML tumor microenvironment and correlated with significantly shorter overall survival (9.1 versus 15.56 months in patients with high and low IES scores, respectively; HR = 3.32 (95% CI = 1.19-9.25); log-rank P = 0.021; Fig. 1C). Finally, the IES-related gene set also predicted for long-term outcomes and objective responses, based on RECIST criteria, to single-agent nivolumab or pembrolizumab, or combination anti-PD-1 + anti-CTLA-4, in 106 patients with melanoma (PRJEB23709 and GSE93157 series), a tumor type known to derive durable clinical benefit from ICB (Fig. 1D).Our findings encourage the pursuit of immune senescence reversal as a strategy to functionally reinvigorate T cells and could inform the delivery of ICB and other T cell-targeting immunotherapies to patients who are likely to benefit. [Display omitted] DisclosuresRadojcic: Syndax Pharmaceuticals: Research Funding; Regeneron Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees; Allakos: Membership on an entity's Board of Directors or advisory committees. Minden: Astellas: Consultancy. Tasian: Aleta Biotherapeutics: Consultancy; Gilead Sciences: Research Funding; Kura Oncology: Consultancy; Incyte Corporation: Research Funding.

IMMU-05. A PHASE I STUDY IN PROGRESS OF HEGFRVIII-CD3 BI-SCFV (BRITE) IN PATIENTS WITH WHO GRADE IV MALIGNANT GLIOMA

Abstract INTRODUCTION Approximately 30% of glioblastomas harbor the tumor specific EGFRvIII mutation. hEGFRvIII-CD3 bi-scFv (Brain Bi-Specific T Cell Engager - BRiTE) is a novel bispecific antibody construct which can redirect a patient’s entire repertoire of T cells (TCR-agnostic) to specifically lyse EGFRvIII-positive tumor cells. Compared to CAR-T therapy, it offers a highly potent yet off-the-shelf approach for glioblastoma. BRiTE is now entering Phase I clinical trials (NCT04903795) and will be trialed as a monotherapy or with autologous T cell infusion. Migration of T cell binding macromolecules across the blood-brain barrier may be facilitated by activated T cells which adhere to the microvascular endothelium and enter the brain parenchyma. Concurrent administration of activated T cells could therefore enhance trafficking of BRiTE and other antibody-based macromolecules into the intracerebral compartment. HYPOTHESIS We hypothesize that treatment of EGFRvIII-positive WHO grade IV malignant glioma with BRiTE alone or with peripheral T-cell infusion is safe and can induce objective tumor shrinkage at tolerable doses. TRIAL DESIGN/OBJECTIVES A maximum of 18 patients with newly diagnosed or recurrent WHO grade IV malignant EGFRvIII+ glioma will be enrolled after undergoing standard of care treatment. The primary objective will be evaluating the safety of BRiTE alone and with autologous T cell infusion. Patients will receive a single BRiTE infusion, followed 14 days later by an infusion of activated autologous T cells and a second bolus BRiTE infusion. Dose escalation and de-escalation will be managed using a Bayesian optimal interval design. Secondary objectives will be to describe clinical benefit as determined by objective response rate per mRANO criteria, and to evaluate BRiTE pharmacokinetics in serum. CONCLUSION We describe a first-in-human trial of bispecific T cell engager therapy for glioblastoma. We also describe our novel approach for enhancing intracranial penetration of BRiTE using autologous T cell infusion.

Optimal Timing of Blinatumomab for the Treatment of B-Lymphoblastic Leukemia

Introduction Blinatumomab is a CD19-targeting bi-specific T-cell engager antibody construct developed for the treatment of CD19-expressing B-cell malignancies. After early impressive in vitro and in vivo pre-clinical results,1–3 blinatumomab was quickly incorporated into clinical investigations. After a phase I non-Hodgkin lymphoma (NHL) trial established a safe and effective dose, delivered via continuous intravenous (IV) infusion,4 trials exploring its use in B-lymphoblastic leukemia (B-ALL) soon followed. The results of these trials have been largely positive, though the extent of response and benefit appears to vary based on disease status and burden at the time of treatment. Thus, the optimal incorporation of blinatumomab into B-ALL therapy is still being refined.

Targeting BCMA in Multiple Myeloma.

Despite considerable advances in the treatment of multiple myeloma (MM) in the last decade, a significant number of patients still progress on current available therapies. Here, we review treatment modalities used to target BCMA in the treatment of MM, specifically antibody-drug conjugates (ADC), bispecific antibody constructs, and chimeric antibody receptor (CAR) modified T-cell therapies. We will provide an overview of therapies from these classes that have presented or published clinical data, as well as data on mechanisms of resistance to these novel agents. Clinical trials exploring different BCMA-targeting modalities to treat multiple myeloma are underway and demonstrate promising results. In relapsed/refractory multiple myeloma, anti-BCMA ADCs and bispecific antibody constructs are showing impressive efficacy with manageable side effect profiles. In parallel, adoptive cellular therapy has induced dramatic durable responses in multiply relapsed and refractory myeloma patients. Therapeutic approaches targeting BCMA hold significant potential in the management of multiple myeloma and will soon be incorporated in combination with current standard therapies to improve outcomes for patients with multiple myeloma. In addition, novel approaches are being evaluated to overcome resistance mechanisms to anti-BCMA therapies.

Abstract 1737: Understanding the geometry and valency of bispecific antibodies in the optimization of tumor-dependent activation of 4-1BB

Abstract 4-1BB is a TNF family receptor expressed on the surface of tumor-infiltrating T cells. Activation of 4-1BB enhances the activation, metabolism and function of tumor-infiltrating T cells and promotes tumor regression. Several anti-4-1BB antibodies have entered the clinic and have suffered from liver toxicity or lack of activity. To address the clinical liabilities with 4-1BB targeting, we designed bispecific 4-1BB x tumor associated antigen (TAA) antibodies to selectively activate T cells within the tumor microenvironment. To understand the impact of antibody format on potency of T cell activation, we generated a panel of 4-1BBxTAA bispecific antibodies in different formats with the AzymetricTM and EFECTTM platforms. We compared potency and maximal activity of these constructs in co-culture assays with tumor cells and T or Jurkat reporter cells, in comparison to urelumab as a clinical benchmark. The AzymetricTM platform allowed the production and evaluation of unique antibody formats from which an extensive structure activity relationship analysis was performed. We also examined the accessible epitope space by sampling the possible paratope binding conformations of the molecule. From this we determined the geometric limits of antigen engagement. Bispecific antibody constructs utilizing monovalent 4-1BB targeting showed inferior activity than urelumab in co-culture assays with TAAhi tumor cells. Activity greater than urelumab was seen with constructs containing two 4-1BB arms with one or two TAA binding scFv on the C-terminus of the Fc. Bivalent 4-1BB antibodies with an anti-TAA scFv fused to the N-terminus of the 4-1BB Fab showed decreased activity, potentially due to reduced ability to engage 4-1BB in a bivalent manner. Critically, the activity seen was transferable across multiple TAAs. The activity of the 4-1BB bispecific constructs was dependent on the expression level of the TAA. Without the presence of tumor cells, or with constructs where the TAA scFv was replaced by an irrelevant scFv, minimal activity was seen. The modelling of the space accessible to the 4-1BB paratopes showed that the constructs with which we saw the most activity had a distance between 4-1BB and TAA binding paratopes similar to that seen between T and APC membranes as part of an immune synapse, suggesting that this may be a key characteristic for the development of conditional T cell agonist bispecifics. This work represents an investigation of 4-1BBxTAA formats which would be difficult to make without an efficient bispecific technology. We identified a series of 4-1BBxTAA bispecific agonist antibody formats which were transferable between multiple TAAs. These formats also allowed optimization of activity and selectivity to promote maximal therapeutic index and efficacy, key factors which are potentially able to contribute to improved clinical outcomes. Citation Format: Anna von Rossum, Harsh Pratap, Lisa Newhook, Elizabeth C. Halvorsen, Lee Freiburger, Renee Duan, Charles Stevens, Dimitri Tcaciuc, Gesa Volkers, Duncan Browman, Brandon Clavette, David Mills, Thomas Spreter, Daniel T. Patton. Understanding the geometry and valency of bispecific antibodies in the optimization of tumor-dependent activation of 4-1BB [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1737.

Baseline Biomarkers Influencing Clinical Outcomes in Adults with Relapsed/Refractory B-Precursor Acute Lymphoblastic Leukemia (R/R ALL) Treated with Blinatumomab Versus Standard-of-Care Chemotherapy (SOC) from a Randomized Phase 3 Study

Introduction: Blinatumomab, an anti-CD3/CD19 bispecific T-cell engager antibody construct, showed antileukemic activity and significantly longer overall survival (OS) vs SOC in adults with advanced Philadelphia chromosome-negative R/R ALL (median OS 7.7 vs 4.0 months; P=0.01) (N Engl J Med 2017;376:836−847). Here we aimed to identify baseline biomarkers that were prognostic (disease-related) or predictive (treatment-related) of clinical outcomes.Methods: Adults with R/R ALL (refractory to primary induction therapy or salvage therapy, first relapse <1 year, second or later relapse, or relapse after allogeneic hematopoietic cell transplantation [alloHCT]) were randomized to receive either blinatumomab or SOC. Blinatumomab was given by continuous infusion (4 wks on/2 wks off) for up to 5 cycles (9 μg/d on d1‒7 of cycle 1 and 28 μg/d thereafter). Up to 4 additional maintenance cycles (4 wks on/8 wks off) were allowed for up to 1 year. Evaluated biomarkers included leukocytes, monocytes, lymphocytes, granulocytes, platelets, neutrophils, CD45+ CD3+ CD4+ T cells, CD45+ CD3+ CD8+ T cells, CD45+ CD3- CD19+ B cells, CD3- CD16+ CD56+ cells, bone marrow (BM) blasts, CD3+ T cells, CD19+ cells, and ratios of CD3+/CD19+ and CD3+%/BM blast%. A multivariate model was used to select potential biomarkers associated with treatment outcomes. Effect on outcome was determined if either the main effect or interaction term between the biomarker and treatment had a P value ≤0.3. Prognostic and predictive associations of baseline biomarkers with OS, hematologic response (complete remission [CR] or CR with full, partial, or incomplete hematologic recovery [CR/CRh/CRi]) within the first 12 wks, and minimal residual disease (MRD) response were assessed by Cox regression for OS and multivariate logistic regression for hematologic response and MRD response. Predictive biomarkers were identified using interaction tests with treatment group.Results: A total of 405 patients received blinatumomab (n=271) or SOC (n=134). Baseline characteristics including age, prior alloHCT, salvage status, granulocytes, platelets, and lymphocyte subpopulations (CD45+ CD3+ CD8+, CD45+ CD3- CD19+, and CD3+) were generally balanced between treatment arms. Associations of baseline biomarkers with clinical outcomes are presented in the table. Higher platelet counts were associated with decreased risk of death (HR 0.43; 95% CI 0.29-0.65) and increased CR (odds ratio [OR] 4.12; 95% CI 2.04-8.33) and CR/CRh/CRi (HR 1.64; 95% CI 0.90-2.99). Higher granulocyte counts were associated with decreased risk of death (HR 0.75; 95% CI 0.57-0.99), whereas total leukocyte counts were not (HR 1.04; 95% CI 0.70-1.54). Lower BM blast percentage was associated with increased CR (OR 0.90; 95% CI 0.82-0.97) and CR/CRh/CRi (OR 0.87; 95% CI 0.80-0.95) but not OS (HR 1.04; 95% CI 0.98-1.09). Higher CD3+ T cells were associated with increased MRD response (OR 1.75; 95% CI 0.99-3.08). Higher counts of CD45+ CD3- CD19+ cells increased the risk of death for blinatumomab patients (HR 1.19; 95% CI 1.09-1.29) but also was not associated with a survival advantage for SOC patients (HR 1.03; 95% CI 0.93-1.15). Higher CD45+ CD3+ CD8+ T cells were associated with increased CR (OR 1.39; 95% CI 1.18-1.65) and CR/CRh/CRi (OR 1.44; 95% CI 1.22-1.70) in patients treated with blinatumomab but did not correlate with increased hematologic response in SOC patients. Multiple logistic regression modeling for adverse events (AEs) of interest showed that BM blasts and lymphocyte markers (CD3+ and CD45+ CD3+ CD8+) were not associated with grade ≥3 neurologic events, infections, or cytokine release syndrome.Conclusions: Higher platelet and granulocyte counts, lower BM blasts, and higher numbers of CD3+ T cells were prognostic for improved OS, hematologic response, and MRD response, respectively, in patients with R/R ALL who received blinatumomab vs SOC. A higher frequency of CD45+ CD3+ CD8+ cells was predictive of increased hematologic response to blinatumomab, whereas a higher frequency of CD45+ CD3- CD19+ cells was predictive of decreased OS. BM blast percentage and lymphocyte markers were not associated with AEs of interest. Overall, these data demonstrate that responses to blinatumomab can be seen across a variety of subgroups, including those expressing specific baseline biomarkers. Further study is needed to identify subgroups most likely to benefit from blinatumomab. [Display omitted] DisclosuresWei:AbbVie, Celgene, Servier: Research Funding; AbbVie, Celgene, Novartis, Amgen, Servier: Membership on an entity's Board of Directors or advisory committees; AbbVie, Celgene, Novartis, Amgen, Servier: Honoraria. Ribera:Janssen: Honoraria; Gilead: Honoraria; Celgene: Honoraria; Pfizer: Research Funding, Speakers Bureau; Amgen Inc.: Research Funding, Speakers Bureau; Incyte: Research Funding, Speakers Bureau; ARIAD: Research Funding, Speakers Bureau; Roche: Honoraria. Larson:Daiichi Sankyo: Consultancy, Research Funding; Astellas: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Amgen Inc.: Research Funding. Ritchie:Amgen Inc.: Honoraria. Chen:Amgen Inc.: Employment, Other: Stock holder. Dos Santos:Amgen Inc.: Employment. Zimmerman:Amgen Inc.: Employment, Equity Ownership. Kantarjian:Novartis: Research Funding; Bristol-Meyers Squibb: Research Funding; Amgen: Research Funding; Pfizer: Research Funding; Delta-Fly Pharma: Research Funding; ARIAD: Research Funding.