Combination of Cisplatin and Irradiation Induces Immunogenic Cell Death and Potentiates Postirradiation Anti-PD-1 Treatment Efficacy in Urothelial Carcinoma

Abstract

The therapeutic benefit of immune checkpoint inhibitor monotherapy is limited to a subset of patients in urothelial carcinoma (UC). Previous studies showed the immunogenicity of cisplatin and irradiation. Here, we investigated whether chemoradiotherapy (CRT), a combination of cisplatin and irradiation, could improve the efficacy of postirradiation anti-programmed cell death 1 (PD-1) treatment in UC. In our advanced UC patient cohort, patients with CRT showed a significantly better objective response rate (75%/22%) and overall survival (88%/30% at 12 months) following later pembrolizumab therapy compared to those without. Then, we created syngeneic UC mouse models by inoculating MB49 cells s.c. in C57BL/6J mice to examine the potential of CRT to enhance antitumor immunity in conjunction with postirradiation anti-PD-1 treatment. Nonirradiated tumors of the mice treated with CRT/postirradiation anti-PD-1 treatment had a significantly slower growth rate and a significantly higher expression of cytotoxic T cells compared to those of the mice treated with anti-PD-1 treatment alone. The mice treated with CRT/postirradiation anti-PD-1 treatment showed the best survival. Mechanistically, CRT provoked strong direct cytotoxicity and increased expressions of immunogenic cell death markers in MB49 cells. Therefore, the combination of cisplatin and irradiation induces immunogenic cell death and potentiates postirradiation anti-PD-1 treatment efficacy in UC.

Keywords: carcinoma; chemoradiotherapy; cisplatin; immunogenic cell death; immunotherapy; transitional cell.

Figure 1

Associations of cisplatin-based CRT with therapeutic response and survival after pembrolizumab therapy in patients with advanced urothelial carcinoma (UC). (A) Objective response rate (ORR), (B) progression-free survival (PFS), and (C) overall survival (OS) are compared between CRT (n = 8) and Non-CRT (n = 18) groups. (D) Comparison of OS curves between CRT (n = 8) and Non-CRT (n = 10) groups in 18 bladder cancer patients. The p-value is calculated by the chi-square test for ORR and the log-rank test for PFS and OS. Asterisks indicate p-values comparing two groups, as indicated in the figure. NS, not significant; * p < 0.05.

Figure 2

Concurrent anti–PD-1 treatment facilitated abscopal effects induced by a single dose of irradiation in MB49 tumor-bearing mice. (A) Scheme for tumor inoculation and treatments. Six- to seven-week-old mice are inoculated subcutaneously with MB49 cells in the left hindlimb and in the right flank, and seven days later, receive irradiation with a single fraction of 10 Gy to the left hindlimb and anti–PD-1 treatment (or its isotype controls). (B) MB49 tumor growth curves of irradiated tumors in the left hindlimb (left) and nonirradiated tumors in the right flank (right). (C) Survival curves of mice. Log-rank test is used to compare survival curves. All data are shown as the mean ± standard error of the mean (SEM). Asterisks indicate p-values comparing two groups, as indicated in the figure. * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 3

The combination of cisplatin and irradiation potentiated the efficacy of postirradiation anti–PD-1 treatment in MB49 tumor-bearing mice. (A) Scheme for tumor inoculation, treatments, and T cell analysis. For the CRT model (top), six- to seven-week-old mice are injected with MB49 cells in the left hindlimb, and seven days later, receive cisplatin at 3 mg/kg and irradiation with a single fraction of 10 Gy to the left hindlimb. The mice are injected with MB49 cells in the right flank 14 days after irradiation, and seven days later, receive anti–PD-1 treatment (or its isotype controls). For the Non-CRT model (bottom), only a right flank tumor is established in the mice, and cisplatin and irradiation are not given. (B) MB49 tumor growth curves of irradiated tumors in the left hindlimb (left) and nonirradiated tumors in the right flank (right). (C) Survival curves of mice. The log-rank test is used to compare survival curves. (D–G) Proportions of CD45⁺ cells in the live cells (D), CD3⁺ cells in the CD45⁺ subpopulation (E), CD8⁺CD4⁻ cells in the CD3⁺ subpopulation (F), and IFNγ⁺ cells in the CD8⁺CD4⁻ subpopulation (G) in single-cell suspensions of the irradiated and nonirradiated tumors on day seven in mice treated with CRT/postirradiation anti–PD-1 treatment compared to those of the nonirradiated tumors on day seven in mice treated with anti–PD-1 treatment alone (n = 5/group). All data are shown as the mean ± standard error of the mean (SEM). Asterisks indicate p-values comparing two groups, as indicated in the figure. NS, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 4

The combination of cisplatin and irradiation increases HMGB1 protein secretion and cell surface expression of calreticulin protein in MB49 cells. (A) MB49 cells are treated with cisplatin at 0.6 mg/L and/or irradiated with a single fraction of 10 Gy. Medium is collected two and five days after treatment, and HMGB1 protein expression levels in each medium are examined in duplicate by ELISA (n = 3/group). (B) MB49 cells are treated with cisplatin at 0.6 mg/L and/or irradiated with a single fraction of 10 Gy. Cells are collected, and cell surface calreticulin protein expression levels are examined by flow cytometry (n = 3/group). (C,D) Seven days after inoculation MB49 cells in the left hindlimb, mice are treated with cisplatin at 3 mg/kg and/or irradiation with a single fraction of 10 Gy. Tumor sections before and after CRT (n = 3/group) are stained for HMGB1 (C) and calreticulin (D). Left, representative images; right, comparison of the mean (± standard error of the mean [SEM]) number of HMGB1- or calreticulin-positive cells counted in five fields of view (FOV). All data are shown as the mean ± SEM. Asterisks indicate p values comparing two groups, as indicated in the figure. NS, not significant; * p < 0.05; ** p < 0.01.