Info
Background:
IL13Rα2-directed CAR-T cells are a promising therapy for glioblastoma, but efficacy is limited by an immunosuppressive tumor microenvironment and poor brain delivery following systemic administration. Intracranial delivery bypasses the blood-brain barrier (BBB), but it is invasive and associated with higher risks compared to IV administration. Focused ultrasound (FUS)–mediated BBB opening could increase delivery after IV administration and may enhance immune activation. Moreover, development of such therapies is hindered by a lack of methods to assess their fate as CAR-T cell concentrations in the cerebrospinal fluid or peripheral blood do not reliably correlate with intratumoral concentrations. Reporter gene based CAR-T cell tracking with PET imaging is a valuable tool to assess safety and efficacy of CAR-T cell therapy by enabling their longitudinal tracking, providing a critical link between cell distribution and therapeutic response.
Hypothesis/Objective:
We aim to pair FUS with PET-based CAR-T cell imaging to quantitatively assess CAR-T cell delivery to glioblastoma.
Methods:
Second generation anti-IL13Rα2-scFv CAR-T cells were engineered to co-express the membrane-bound anti-DOTA scFv huC825 (KLG3BBz-huC825). We assessed cellular cytotoxicity using human glioblastoma U87MG GFP-FFluc (IL13Rα2+) cells and assessed reporter mediated [86Y]Y-Aminobenzyl-DOTA ([86Y]Y-ABD) binding in vitro by incubating KLG3BBz-huC825 cells with activity ranging from 1 to 7 MBq. Orthotopic glioblastoma xenografts were established in female NSG mice via stereotactic injection of 2×105 U87MG.GFP-FFluc cells into the right striatum. Tumor engraftment was confirmed on day 10 via bioluminescence imaging (BLI). On day 11, mice received intravenously 2×106 KLG3BBz-huC825 or non-transduced (NT) T cells. On day 7 and 14 post-T cell administration mice were IV injected with [86Y]Y-ABD (6.7 ±1.1 MBq; specific activity: 15.8 MBq/nmol) and imaged at16 h post-tracer injection. To determine optimal sonication parameters, FUS BBB opening was performed on the right cerebral hemisphere of non-tumor-bearing NSG mice using the RK-50 FUS system (FUS Instruments). Following IV administration of Definity® microbubbles, FUS was applied at 1.65 MHz (50 bursts of 10 ms) with peak negative pressures of 0.3 or 0.6 MPa.
Results:
We successfully generated KLG3BBz-huC825 CAR-T cells and confirmed antigen-dependent killing in vitro; 50% lysis was reached at an effector:tumor ratio of 1:7. We verified specific radiotracer binding to huC825-expressing CAR-T cells, which supports our aim of PET-based monitoring of ex vivo labelled KLG3BBz-huC825 CAR-T cell brain entry following FUS-mediated BBB opening. In vitro, radiotracer uptake was highly specific to huC825 expression, reaching 5.0 ± 0.5% of the total dose, compared with 0.05 ± 0.01% in non-transduced T cells. The optimal ex vivo labeling for administration seems to be incubation of 2×106 KLG3BBz-huC825 CAR-T with 3 MBq (Figure 1B). FUS-induced BBB opening was confirmed by Evans Blue extravasation in healthy mice. Histological analyses are ongoing to confirm the absence of FUS-related brain damage. Preliminary PET tracking of the KLG3BBz-huC825 cells in a glioblastoma orthotopic model shows their tumor localization at day 7 post intravenous administration. 14 day post administration, [86Y]Y-ABD uptake was significantly lower compared to day 7 suggestion a decrease in KLG3BBz-huC825 persistence in the tumor microenvironment (Figure 1D).
Conclusion:
Our preliminary results in an orthotopic glioblastoma model support the feasibility of PET-based tracking of CAR-T cells’ localization in GBM. Combination with FUS may enhance the CAR-T cells’ migration across the BBB and tumoral localization after I.V. administration.
Image/Figure:
Click to view full size
Click to view full size
Image/Figure Caption:
Figure 1 (A) Expression of huC825 by KLG3BBz-huC825 T cells measured by flow cytometry. (B) In vitro uptake of [86Y]Y-ABD by 1,000,000, 2,000,000 or 5,000,000 KLG3BBz-huC825 T cells and non-transduced (NT) T cells incubated with 1, 3 or 7 MBq of the radiohapten. (C) Experimental design (D) Left: BLI of U87MG.GFP_FFLuc tumor bearing mice at day 10, 17 and 24 post-implantation. 11 days post implantation mice were injected with either 2×106 KLG3BBs-huC825 or NT cells IV. Right: MIP images at 16 h post injection of [86Y]Y-ABD at 7 and 14 days post I.V. injection of T cells.
Author
Céline Chevaleyre
The University of Texas MD Anderson Cancer Center