Local radiation enhances systemic CAR T-cell efficacy by augmenting antigen crosspresentation and T-cell infiltration

Abstract

Chimeric antigen receptor (CAR) T-cell therapy targeting CD19 (CART-19) represents a significant advance in the treatment of patients with relapsed or refractory CD19+ B-cell lymphomas. However, a significant portion of patients either relapse or fail to respond. Moreover, many patients have symptomatic disease, requiring bridging radiation therapy (RT) during the period of CAR T-cell manufacturing. To investigate the impact of 1 to 2 fractions of low-dose RT on CART-19 treatment response, we developed a mouse model using A20 lymphoma cells for CART-19 therapy. We found that low-dose fractionated RT had a positive effect on generating abscopal systemic antitumor responses beyond the irradiated site. The combination of RT with CART-19 therapy resulted in additive effects on tumor growth in irradiated masses. Notably, a significant additional increase in antitumor effect was observed in nonirradiated tumors. Mechanistically, our results validate activation of the cyclic guanosine adenosine synthetase/stimulator of interferon genes pathway, tumor-associated antigen crosspriming, and elicitation of epitope spreading. Collectively, our findings suggest that RT may serve as an optimal priming and bridging modality for CAR T-cell therapy, overcoming treatment resistance and improving clinical outcomes in patients with CD19+ hematologic malignancies.

Graphical abstract

Figure 1.

Enhanced abscopal antitumor effects in A20 lymphoma model with fractionated RT. (A) Working model: time line and schematic representation of in vivo A20 tumor–bearing mice treated with 2 × 4-Gy or 1 × 8-Gy RT in 1 of the 2 tumors. (B-C) A20 tumor growth from irradiated and nonirradiated abscopal tumor as followed by caliper measurements (n = 5-6 mice). (D) Gene ontology terms that are significantly enriched with an adjusted P value <.05 in the differentially expressed gene sets. Data represent irradiated tumors from the fractionated- or single-dose–treated groups. (E-F) Gene expression of granzyme B (gzmb) and perforin 1 (prf1) in irradiated and nonirradiated tumor 5 days after RT. (G-H) CD45⁺/CD3⁺ and CD45⁺/CD3⁺/CD8⁺ T-cell infiltration in irradiated and abscopal tumors 34 days after tumor implantation. Graphs show the mean ± standard error of the mean (SEM). ∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Data are representative of 3 independent experiments. CTR, control; ns, not significant; SARRP, Small Animal Radiation Research Platform.

Figure 2.

Enhancing CART-19 cell infiltration and antitumor response through preceding RT in vivo. (A) Working model: time line and schematic representation of in vivo A20 tumor–bearing mice treated with RT followed by CART-19 cell injection. All mice of all groups were treated with the same cyclophosphamide (CP) lymphodepletion regimen. (B-C) A20 tumor growth from irradiated and abscopal tumor (n = 17-21). (D) Survival curve after treatment administration. (E) CD45.2⁺/CD3⁺ T-cell infiltration in irradiated and abscopal tumors. (F) CD45.2^–/CD3⁺/CD45.1⁺ T-cell infiltration in irradiated and abscopal tumor representing CART-19 cells. (G) CD45.2^–/CD3⁺/CD45.1⁺ T-cell infiltration in the spleen representing CART-19 cells. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Data are representative of 3 independent experiments. CTR, control.

Figure 3.

RT/CART-19 combination treatment enhances crosspresentation of TAAs and T-cell responses to TAAs. (A) CD45.2⁺/CD11c⁺ DC infiltration in irradiated tumor on day 23 and day 34 after tumor challenge. (B-C) AH1-specific T-cell infiltration in irradiated and abscopal tumors on day 23 and day 34 after tumor challenge. (D) IFN-γ spots of an enzyme-linked immunospot (ELISPOT) assay after stimulation of tumor cell suspensions with AH1 peptide. (E) CD45.2⁺/CD3⁺ T cells isolated from treated mice stimulated with AH1 peptide and incubated with A20 tumor cells. (F) CD45.2⁺/CD3⁺ T cells isolated from treated mice stimulated with AH1 peptide and incubated with CT26 tumor cells. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CTR, control.

Figure 4.

Enhanced tumor protection in a colorectal model through adoptive transfer of CD3⁺ T cells from mice treated with RT plus CART-19. (A) Working model: time line and schematic representation of in vivo A20 tumor–bearing mice treated with 2 × 4-Gy RT followed by CAR T cells and adoptive T-cell transfer to CT26 tumor–bearing mice. All mice of all groups were treated with the same CP lymphodepletion regimen. (B) AH1-specific T cells from donor mice after ex vivo expansion, before ACT. (C) CT26 tumor growth of recipient mice (n > 4). (D) CD45⁺/CD3⁺/CD8⁺ T-cell tumor infiltration. (E) CD45⁺/CD3⁺/CD8⁺/CD107⁺ T-cell tumor infiltration. (F) CD45⁺/CD3⁺/CD8⁺/AH1-specific T-cell tumor infiltration. (G) IFN-γ spots of an ELISPOT assay after AH1 peptide stimulation of splenocytes from mice that received ACT. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Data are representative of 2 independent experiments. ACT, Adoptive Cell Therapy.

Figure 5.

RT-induced activation of the cGAS/STING pathway promotes IFN type I gene and chemokine expression upregulation in A20 tumor cells. (A-D) Expression of ifna1, ifnb1, ccl5, and cxcl9 after 24, 48, and 72 hours of RT in vitro. (E-F) Expression of ifna1, inb1, cxcl9, and cxcl11 48 hours post in vivo RT. (G) Activation of STING pathway in A20 tumors treated with CART-19 alone or combination of CART-19 and RT. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. b-TUB, b-tubulin; RT, Radiation Therapy; p-IRF3, phosphorylated-Interferon Regulatory Factor 3; p-STING, phosphorylated-Stimulator of interferon genes; p-TBK1, phosphorylated-TANK-binding kinase 1.

Figure 6.

RT enhances CAR T-cell therapy through STING activation. (A) Working model: time line and schematic representation of in vivo A20 tumor–bearing mice treated with RT followed by CART-19 cell injection with or without the STING antagonist H-151. All mice of all groups were treated with the same CP lymphodepletion regimen. (B-C) Tumor growth from irradiated and abscopal tumors (n = 8). (D) CD45.2^–/CD3⁺/CD45.1⁺ T-cell infiltration in irradiated and abscopal tumors, representing the CART-19 cells. (E) CD45.2⁺/CD11c⁺ DC infiltration in irradiated and abscopal tumors. (F-G) IFN-γ spots after overnight stimulation of tumor or spleen cell suspensions with AH1 peptide with or without anti–MHC I antibody. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Data are representative of 1 experiment.

Figure 7.

Enhancing sensitivity of target cells to CTL function through RT: implications for CAR T-cell efficacy and antigen escape. (A) Percentage of A20 tumor cells alive after RT, CAR T-cell, and RT and CAR T-cell treatments. (B) Flow cytometry for B220 and CD19 markers after in vitro treatment of A20 tumor cells with RT, CAR T-cell, and RT and CAR T-cell treatments. (C) B220⁺ and CD19⁺ alive cells after treatments. (D) B220⁺ alive cells after treatment with RT, CAR T cells, and T cells isolated from CT26 tumor–bearing mouse adoptively transferred with T cells from RT plus CAR T-cell donor. Graphs show the mean ± SEM. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CTR, XXX.