Pharmacokinetics, Biodistribution and Anti-CD19 CAR Expression Kinetics in Vitro and after In Vivo Delivery of a CD8-Targeted Fusosome in NSG Mice

Despite the recent success of exvivo manufactured chimeric antigen receptor (CAR) T cell therapies for treating B cell malignancies, the time-consuming manufacturing process and cost limit access. To address these limitations, a novel in vivo gene therapy platform has been developed to deliver CAR transgenes directly to T cells through systemic administration of a fusosome- a novel, CD8α-targeted, integrating vector (hereafter SG295). Here, we describe the in vitro binding and transduction kinetics, the in vivo blood clearance, tissue biodistribution, and kinetics of SG295, (i.e., a fusosome encoding an anti-CD19 CAR) in naïve and human peripheral blood mononuclear cell (PBMC)-engrafted NOD SCID gamma (NSG) mice.

The kinetics of CD19modCAR, a precursor to SG295, binding to CD8+ T cells was evaluated in vitro by co-incubating the fusosome (10, 3.3, 1.1 integrating units (IU)/cell) and human pan T cells over a time course of 24 hours. SG295 binding to CD8+ T cells was measured by flow cytometry using anti-CD8 antibodies. Binding was observed at time zero (T₀) following the addition of SG295, as evidenced by a ~15% decrease on average in binding to CD8. This was consistent with masking of the antigen due to fusosome blockade. We observed a dose dependent CD8 masking at indicated time points with maximum effect observed after 1 hour of incubation. Complete recovery of the CD8 masking was observed at 2 hours post SG295 addition suggesting receptor disengagement, SG295 internalization into the cell, or CD8 recycling to cell membranes.

In vivo kinetic studies were performed in human PBMC-engrafted-NSG mice dosed with 4x10⁸ IU/Kg SG295 intravenously (IV). Plasma and tissue samples were taken at several time points from 10 min to 1 week post SG295 injection. SG295 viral particles were measured by RT-ddPCR using a dENV amplicon (genome quantification assay, GQA). Viral integration in the host genome was measured by ddPCR using two amplicons: delU3 for viral vector and CD19 CAR for the transgene (vector copy number, VCN). The resulting generation of CD8+ CAR+ T cells in blood and tissues was measured by flow cytometry.

SG295 fusosomes were cleared rapidly from the plasma by 1 hour post IV dose (t½~5 min) in the absence of target CD8+ T cells and demonstrated linear clearance kinetics. In the presence of target CD8+ T cells, the concentration of unbound SG295 was reduced and SG295 retention in the plasma was prolonged. Similarly, the presence of target CD8+ T cells prolonged retention of SG295 in tissues (lung, liver and spleen).

Target-mediated transduction of SG295 in CD8+ T cells was assessed in human PBMC-engrafted NSG mice using VCN. Samples were assessed from 10 min to 1 week. Both delU3 and CAR VCN were quantifiable in spleen, bone marrow and lung after one week. Corresponding CAR expression on CD8+ T cells was also detected by flow cytometry by 1 week in the blood and bone marrow. However, VCN was not detected in the blood or any other tissues at these timepoints.

Detectable CAR VCN levels were observed in multiple tissues as soon as 10 min after SG295 dosing and for the following 24-72 hours. Similarly, CAR expression assessed by flow was also detected at timepoints up to 6 hours post SG295 dose in blood, liver, and lungs. Taken together, these data suggest that CAR VCN and protein detected at early time points (up to 72 hours) are likely due to the presence of material carried over from production which may be detected for up to 72 hours due to binding to the CD8+ cells in the human PBMC pool. Detectable integration of vector in cell genome and associated CAR expression on CD8+ cells in blood, spleen and bone marrow were observed after 1 week post SG295. This demonstrates productive transduction of CD8+ cells at 1 week or longer timepoints.

In vivo administration of SG295 demonstrates rapid blood clearance (t_1/2 ~5 minutes) and the intended CD8-target engagement and transgene integration resulting in expression of the CD19 CAR after 1 week in target organs of spleen and bone marrow. In addition, there is evidence of material carried over from production at early time points which should be considered when determining early pharmacokinetic parameters. Studies are underway to examine these SG295 pharmacokinetics in non-human primates.

These studies represent the first data assessing the pharmacokinetics and biodistribution of a CD8α-targeted fusosome in vivo and have important implications for dosing and treating patients in the clinic.

No relevant conflicts of interest to declare.