Complementarity-determining region clustering may cause CAR-T cell dysfunction.

Chimeric antigen receptor (CAR)-T cell therapy is rapidly advancing as cancer treatment, however, designing an optimal CAR remains challenging. A single-chain variable fragment (scFv) is generally used as CAR targeting moiety, wherein the complementarity-determining regions (CDRs) define its specificity. We report here that the CDR loops can cause CAR clustering, leading to antigen-independent tonic signalling and subsequent CAR-T cell dysfunction. We show via CARs incorporating scFvs with identical framework and varying CDR sequences that CARs may cluster on the T cell surface, which leads to antigen-independent CAR-T cell activation, characterized by increased cell size and interferon (IFN)-γ secretion. This results in CAR-T cell exhaustion, activation-induced cell death and reduced responsiveness to target-antigen-expressing tumour cells. CDR mutagenesis confirms that the CAR-clustering is mediated by CDR-loops. In summary, antigen-independent tonic signalling can be induced by CDR-mediated CAR clustering, which could not be predicted from the scFv sequences, but could be tested for by evaluating the activity of unstimulated CAR-T cells.

Single-Cell RNA Analysis Reveals Cell-Intrinsic Functions of CAR T Cells Correlating with Response in a Phase II Study of Lymphoma Patients.

PURPOSE: Although CD19 chimeric antigen receptor T cells (CAR-T) therapy has shown remarkable success in B-cell malignancies, a substantial fraction of patients do not obtain a long-term clinical response. This could be influenced by the quality of the individual CAR-T infusion product. To shed some light on this, clinical outcome was correlated to characteristics of CAR-T infusion products. PATIENTS AND METHODS: In this phase II study, patients with B-cell lymphoma (n = 23) or leukemia (n = 1) received one or two infusions of third-generation CD19-directed CAR-Ts (2 × 108/m2). The clinical trial was registered at clinicaltrials.gov: NCT03068416. We investigated the transcriptional profile of individual CD19 CAR-T infusion products using targeted single-cell RNA sequencing and multicolor flow cytometry. RESULTS: Two CAR-T infusions were not better than one in the settings used in this study. As for the CAR-T infusion products, we found that effector-like CD8+CAR-Ts with a high polyfunctionality, high cytotoxic and cytokine production profile, and low dysfunctional signature were associated with clinical response. An extended ex vivo expansion time during CAR-T manufacturing negatively influenced the proportion of effector CD8+CAR-Ts in the infusion product. CONCLUSIONS: We identified cell-intrinsic characteristics of effector CD8+CAR-Ts correlating with response that could be used as an indicator for clinical outcome. The results in the study also serve as a guide to CAR-T manufacturing practices.

Whole body FDG PET/MR for progression free and overall survival prediction in patients with relapsed/refractory large B-cell lymphomas undergoing CAR T-cell therapy.

BACKGROUND: To find semi-quantitative and quantitative Positron Emission Tomography/Magnetic Resonance (PET/MR) imaging metrics of both tumor and non-malignant lymphoid tissue (bone marrow and spleen) for Progression Free Survival (PFS) and Overall Survival (OS) prediction in patients with relapsed/refractory (r/r) large B-cell lymphoma (LBCL) undergoing Chimeric Antigen Receptor (CAR) T-cell therapy. METHODS: A single-center prospective study of 16 r/r LBCL patients undergoing CD19-targeted CAR T-cell therapy. Whole body 18F-fluorodeoxyglucose (FDG) PET/MR imaging pre-therapy and 3 weeks post-therapy were followed by manual segmentation of tumors and lymphoid tissues. Semi-quantitative and quantitative metrics were extracted, and the metric-wise rate of change (Δ) between post-therapy and pre-therapy calculated. Tumor metrics included maximum Standardized Uptake Value (SUVmax), mean SUV (SUVmean), Metabolic Tumor Volume (MTV), Tumor Lesion Glycolysis (TLG), structural volume (V), total structural tumor burden (Vtotal) and mean Apparent Diffusion Coefficient (ADCmean). For lymphoid tissues, metrics extracted were SUVmean, mean Fat Fraction (FFmean) and ADCmean for bone marrow, and SUVmean, V and ADCmean for spleen. Univariate Cox regression analysis tested the relationship between extracted metrics and PFS and OS. Survival curves were produced using Kaplan-Meier analysis and compared using the log-rank test, with the median used for dichotomization. Uncorrected p-values < 0.05 were considered statistically significant. Correction for multiple comparisons was performed, with a False Discovery Rate (FDR) < 0.05 considered statistically significant. RESULTS: Pre-therapy (p < 0.05, FDR < 0.05) and Δ (p < 0.05, FDR > 0.05) total tumor burden structural and metabolic metrics were associated with PFS and/or OS. According to Kaplan-Meier analysis, a longer PFS was reached for patients with pre-therapy MTV ≤ 39.5 ml, ΔMTV≤1.35 and ΔTLG≤1.35. ΔSUVmax was associated with PFS (p < 0.05, FDR > 0.05), while ΔADCmean was associated with both PFS and OS (p < 0.05, FDR > 0.05). ΔADCmean > 0.92 gave longer PFS and OS in the Kaplan-Meier analysis. Pre-therapy bone marrow SUVmean was associated with PFS (p < 0.05, FDR < 0.05) and OS (p < 0.05, FDR > 0.05). For bone marrow FDG uptake, patient stratification was possible pre-therapy (SUVmean ≤ 1.8). CONCLUSIONS: MTV, tumor ADCmean and FDG uptake in bone marrow unaffected by tumor infiltration are possible PET/MR parameters for prediction of PFS and OS in r/r LBCL treated with CAR T-cells. TRIAL REGISTRATION: EudraCT 2016-004043-36.

Tumor-Targeted Delivery of the p53-Activating Peptide VIP116 with PEG-Stabilized Lipodisks.

Stapled peptides targeting the interaction between p53 and its negative regulators MDM2 and MDM4 have exhibited great potential as anti-cancer drugs, albeit with room for improvement in formulation and tumor specificity. Lipid bilayer disks (lipodisks) have emerged as promising drug nanocarriers and can by attachment of targeting moieties be directed selectively towards tumor cells. Tumor-targeted delivery of stapled peptides by use of lipodisks may therefore increase the uptake in the tumors and limit toxicity in healthy tissue. Here, we utilized epidermal growth factor receptor (EGFR)-targeted lipodisks to deliver p53-activating stapled peptide VIP116 to EGFR-expressing tumor cells. We demonstrate that VIP116 can be stably formulated in lipodisks (maximum peptide/lipid molar ratio 0.11). In vitro cell studies verify specific binding of EGF-decorated lipodisks to tumor cells and confirm that targeted delivery of VIP116 significantly decreases tumor cell viability.

Insertion of the Type-I IFN Decoy Receptor B18R in a miRNA-Tagged Semliki Forest Virus Improves Oncolytic Capacity but Results in Neurotoxicity.

Oncolytic Semliki Forest virus (SFV) has been suggested as a potential candidate for the treatment of glioblastoma and neuroblastoma. However, the oncolytic capacity of SFV is restricted by the anti-viral type-I interferon (IFN) response. The aim of this study was to increase the oncolytic capacity of a microRNA target tagged SFV against glioblastoma by arming it with the Vaccinia-virus-encoded type-I IFN decoy receptor B18R (SFV4B18RmiRT) to neutralize type-I IFN response. Expression of B18R by SFV4B18RmiRT aided neutralization of IFN-ß, which was shown by reduced STAT-1 phosphorylation and improved virus spread in plaque assays. B18R expression by SFV4 increased its oncolytic capacity in vitro against murine glioblastoma (CT-2A), regardless of the presence of exogenous IFN-ß. Both SFV4B18RmiRT and SFV4miRT treatments controlled tumor growth in mice with syngeneic orthotopic gliomablastoma (CT-2A). However, treatment with SFV4B18RmiRT induced severe neurological symptoms in some mice because of virus replication in the healthy brain. Neither neurotoxicity nor virus replication in the brain was observed when SFV4miRT was administered. In summary, our results indicate that the oncolytic capacity of SFV4 was improved in vitro and in vivo by incorporation of B18R, but neurotoxicity of the virus was increased, possibly due to loss of microRNA targets.