TIGIT blockade in the context of BCMA-CART cell therapy does not augment efficacy in a multiple myeloma mouse model

Abstract

BCMA-directed CAR-T therapies have shown promising results in multiple myeloma (MM). However, patients continue to relapse. T cell exhaustion with increased TIGIT expression is a resistance mechanism which was confirmed in CAR-T cells from ARI0002h trial, an academic CAR-T developed in our institution. We aimed to analyze the impact of blocking TIGIT on the efficacy of ARI0002h. We used three different strategies to block TIGIT: (1) Addition of an external blocking anti-TIGIT-antibody (Ab), (2) Modify ARI0002h into a 4^th generation CAR-T, named ARITIGIT, capable of secreting a soluble TIGIT-blocking scFv and (3) TIGIT knock-out in ARI0002h using CRISPR/Cas9. Each strategy was evaluated in vitro and in vivo. Adding a TIGIT-blocking Ab to ARI0002h improved in vitro cytotoxicity, but failed to enhance mice survival. The new 4^th generation CAR-T, ARITIGIT, was also unable to achieve better survival outcomes despite favoring the in vivo model by using a myeloma cell line with high expression of the TIGIT ligand PVR. Interestingly, when mice were challenged with a second infusion of tumor cells, mimicking a relapse model, a trend for improved survival with ARITIGIT was observed (p = 0.11). Finally, TIGIT-knock-out on ARI0002h (KO-ARI0002h) using CRISPR/Cas9 showed similar in vitro activity to ARI0002h. In an in vivo stress model, TIGIT KO-ARI0002h prolonged survival (p = 0.02). However, this improvement was not significant compared to ARI0002h (p = 0.07). This study failed to demonstrate a significant benefit of TIGIT-blockade on ARI0002h cells despite using three different approaches, suggesting that targeting a single immune checkpoint may be insufficient.

Keywords: 4th generation CAR-T; BCMA-CAR-T; CAR-T cell therapy; CRISPR/Cas9; TIGIT blockade; multiple myeloma.

Figure 1.

(a) TIGIT-blockade approaches on the BCMA-directed CAR-T ARI0002h tested in this study. (b) ARITIGIT 4^th generation CART sequence scheme.

Figure 2.

(a) Cytotoxicity results after co-culture of the MM cell lines ARP-1 and U266 with either ARI0002h or untransduced (UT) T cells obtained from healthy donors, adding or not an anti-TIGIT-Ab (the Ab carrier was added to the control groups). Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) ARP1 + UT vs ARP1 + UT + TIGITAb: T cells from n = 8 healthy donors; (2) ARP1 + ARI0002h vs ARP1 + ARI0002h + TIGITAb: T cells from n = 6 healthy donors; (3) U266 + UT vs U266 + UT + TIGITAb: T cells from n = 7 healthy donors; (4) U266 + ARI0002h vs U266 + ARI0002h + TIGITAb T cells from n = 6 healthy donors. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence. (b) Pro-inflammatory cytokine production of IFN-γ, IL-2 and TNF-α after 24 hours of co-culture. Each dot represents a T-cell donor; 3 repeats were performed for each condition in each donor. The number of T cell donors was: (1) ARP1 + UT vs ARP1 + UT + TIGITAb: T cells from n = 7 healthy donors; (2) ARP1 + ARI0002h vs ARP1 + ARI0002h + TIGITAb: T cells from n = 6 healthy donors; (3) U266 + UT vs U266 + UT + TIGITAb: T cells from n = 6 healthy donors; (4) U266 + ARI0002h vs U266 + ARI0002h + TIGITAb T cells from n = 5 healthy donors. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence. (c) In vivo results of this strategy: mice were administered 1.5x10⁶ MM ARP-1 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with 0.5x10⁶ of a healthy donor-derived ARI0002h CAR+ cells or UT T cells. After that, the anti-TIGIT-Ab was administered intraperitoneally twice a week for 4 weeks in the respective groups. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. The number of mice in each group was: (1) UT only n = 4; (2) UT + TIGITAb n = 5; (3) ARI0002h n = 5; (4) ARI0002h + TIGITAb n = 5. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method. (d) In vivo results of the favored model: mice were administered 1.5x10⁶ MM U266 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with 1.5x10⁶ of patient-derived ARI0002h CAR+ cells or UT T cells. After that, the anti-TIGIT-Ab was administered intraperitoneally three times a week, indefinitely, in the respective groups. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. The number of mice in each group was: (1) UT only n = 2; (2) ARI0002h n = 6; (3) ARI0002h + TIGITAb n = 7. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method.

Figure 3.

(a) Cytotoxicity results after co-culture of the MM cell lines ARP-1 and U266 with either ARI0002h or ARITIGIT manufactured from healthy donor or patient T cells. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) ARP1 + ARI0002h vs ARP1 + ARITIGIT + TIGITAb T cells from n = 7 healthy donors; (2) U266 + ARI0002h vs U266 + ARITIGIT + TIGITAb T cells from n = 9 healthy donors; (3) ARP1 + ARI0002h vs ARP1 + ARITIGIT + TIGITAb T cells from n = 4 patients; (4) U266 + ARI0002h vs U266 + ARITIGIT + TIGITAb T cells from n = 4 patients. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence. (b) Pro-inflammatory cytokine production of IFN-γ, IL-2 and TNF-α after 24 hours of co-culture. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) ARP1 + ARI0002h vs ARP1 + ARITIGIT + TIGITAb T cells from n = 7 healthy donors; (2) U266 + ARI0002h vs U266 + ARITIGIT + TIGITAb T cells from n = 8 healthy donors. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence. (c) In vivo results of the unchallenged group: mice were administered 1.5x106 MM ARP-1 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with 1.5x106 of patient-derived untransduced (UT) (n = 3) vs. ARI0002h (n = 7) vs. ARITIGIT (n = 6) T cells. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method. (d) In vivo results of the re-challenged group: mice were administered 1.5x106 MM ARP-1 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with 3.5x106 of patient-derived untransduced (UT) (n = 3) vs. ARI0002h (n = 6) vs. ARITIGIT (n = 6) T cells. When clearance of the disease was observed in the CART groups, mice were readministered 1.5x106 MM ARP-1 cells. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method. (e) In vivo results of the favored model: mice were administered 1.5x106 MM U266 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with either 0.75 or 1.5x106 of patient-derived untransduced (UT) vs. ARI0002h vs. ARITIGIT T cells. The number of mice in the 0.75x106 dose experiment was: UT only n = 2, ARI0002h n = 8 and ARITIGIT n = 8. The number of mice in the 1.5x106 dose experiment was: UT only n = 2, ARI0002h n = 7 and ARITIGIT n = 7. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method.

Figure 4.

(a) Flow cytometry staining of PVR and nectin-2 (N2) in U266 and ARP-1 cells. Gray = Isotype; orange = PVR-expressing cells; red = nectin-2-expressing cells. (b) The ARP-1 cell line was lentivirally transduced with a viral vector containing the N2 gene. Positive N2 cells were sorted by flow cytometry to obtain a 100% N2-positive ARP-1 population gray = unstained cells; red = nectin-2-expressing cells. (c) Cytotoxicity results after co-culture of the MM cell lines ARP-1 and ARP-1-N2 with ARI0002h vs. ARI0002h + TIGIT-AB vs. ARITIGIT. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) ARP1 + ARI0002h vs ARP1-N2 + ARI0002h: T cells from n = 10 healthy donors; (2) ARP1 + ARI0002h + TIGITAb vs ARP1-N2 + ARI0002h + TIGITAb n = 6 healthy donors; (3) ARP1 + ARITIGIT vs ARP1-N2 + ARITIGIT n = 6 healthy donors. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence. (d) Pro-inflammatory cytokine production of IFN-γ, IL-2 and TNF-α after 24 hours of co-culture. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) ARP1 + ARI0002h vs ARP1-N2 + ARI0002h: T cells from n = 10 healthy donors; (2) ARP1 + ARI0002h + TIGITAb vs ARP1-N2 + ARI0002h + TIGITAb n = 6 healthy donors; (3) ARP1 + ARITIGIT vs ARP1-N2 + ARITIGIT n = 6 healthy donors. Paired students’ t-test and the Wilcoxon ranked sign test were applied according to normality adherence.

Figure 5.

(a) Cytotoxicity results after co-culture of the MM cell lines ARP-1 and U266 with either T52 gRNA-ARI0002h or T45 gRNA-ARI0002h. Results obtained were normalized to the WT-ARI0002h group. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) tumor cells + T52 gRNA-ARI0002h: n = 5 healthy donors; (2) tumor cells + T45 gRNA-ARI0002h: n = 4 healthy donors. (b) Pro-inflammatory cytokine production of IFN-γ, IL-2 and TNF-α after 24 hours of co-culture. Each dot represents a T cell donor; 3 repeats were performed for each condition in each donor. The number of T-cell donors was: (1) tumor cells + WT-ARI0002h: n = 5 healthy donors (2) tumor cells + T52 gRNA-ARI0002h: n = 5 healthy donors; (2) tumor cells + T45 gRNA-ARI0002h: n = 4 healthy donors. (c) In vivo results: mice were administered 1.5x10⁶ MM ARP-1 cells on day 0, 24 h after radiation with 2 Gy. When disease luminescence was detectable, mice were injected with 1.5x10⁶ of untransduced (UT) vs. WT-ARI0002h vs. TIGIT KO-ARI0002h. The number of mice in the experiment was: UT only n = 1, WT n = 9 and TIGIT KO n = 18. Mice were followed weekly using bioluminescence and sacrifice was performed when prespecified criteria were met. To analyze survival the log-rank test was applied and survival curves were obtained using the Kaplan-Meier method.