Suppression of Type I IFN Signaling in Tumors Mediates Resistance to Anti-PD-1 Treatment That Can Be Overcome by Radiotherapy

Abstract

Immune checkpoint therapies exhibit impressive efficacy in some patients with melanoma or lung cancer, but the lack of response in most cases presses the question of how general efficacy can be improved. In addressing this question, we generated a preclinical tumor model to study anti-PD-1 resistance by in vivo passaging of Kras-mutated, p53-deficient murine lung cancer cells (p53^R172HΔg/+K-ras^LA1/+ ) in a syngeneic host exposed to repetitive dosing with anti-mouse PD-1 antibodies. PD-L1 (CD274) expression did not differ between the resistant and parental tumor cells. However, the expression of important molecules in the antigen presentation pathway, including MHC class I and II, as well as β2-microglobulin, were significantly downregulated in the anti-PD-1-resistant tumors compared with parental tumors. Resistant tumors also contained fewer CD8⁺ (CD8α) and CD4⁺ tumor-infiltrating lymphocytes and reduced production of IFNγ. Localized radiotherapy induced IFNβ production, thereby elevating MHC class I expression on both parental and resistant tumor cells and restoring the responsiveness of resistant tumors to anti-PD-1 therapy. Conversely, blockade of type I IFN signaling abolished the effect of radiosensitization in this setting. Collectively, these results identify a mechanism of PD-1 resistance and demonstrate that adjuvant radiotherapy can overcome resistance. These findings have immediate clinical implications for extending the efficacy of anti-PD-1 immune checkpoint therapy in patients. Cancer Res; 77(4); 839-50. ©2016 AACR.

Figure 1. Generation of an anti-PD-1–resistant lung tumor mouse model

(A) Schematic illustration. 344SQ mouse lung cancer cells were subcutaneously injected into the flank the syngeneic 129Sv/ev mice on day 0. Anti-mouse PD-1 or isotype control IgG antibodies (10 mg/kg) were administered intraperitoneally on days 3, 7, 10 and 14. Tumor growth was monitored for up to 4 weeks. A nonresponsive tumor was isolated and digested into a single-cell suspension. Tumor cells were cultured in vitro for 2–3 weeks, and then reinoculated into 129Sv/ev mice, followed by anti-PD-1 treatment. This procedure was repeated for 4 cycles. (B) Representative tumor growth curve of parental 344SQ cells and the anti-PD-1–resistant 344SQ cells upon control IgG or anti-PD-1 treatment. Data represented as mean ± SD from an n of 5. ***P<0.001, multiple t test; experiments were also repeated at least three times. (C) Representative picture of hematoxylin and eosin staining of parental and anti-PD-1–resistant tumors (×200 magnification). Red enclosed area indicates tumor necrosis. Black arrows indicate mitotic tumor cells.

Figure 2. Downregulation of MHC molecule expression on anti-PD-1–resistant tumors

(A) Flow cytometry studies of MHC class I (H-2D^b and H-2K^b) and MHC class II (I-A/I-E) expression on 344SQ_P and 344SQ_R tumor cells (gated on CD45^-) after cocultured with syngeneic splenocytes (gated on CD45⁺). (B) H-2K^b and I-A/I-E expression on the parental and anti-PD-1 resistant tumors isolated from tumor-bearing mice with anti-PD-1 or control IgG treatment. (C) Western blotting of β2 macroglobulin (β2M) in parental and resistant tumor tissues. Vinculin was used as a loading control. *P<0.05, ***P<0.001 in Student’s t tests; Data represent means ± SD for an n of 5, with experiments repeated at least three times.

Figure 3. Reduced infiltration and function of tumor-infiltrating lymphocytes in anti-PD-1–resistant tumors

(A) Representative flow cytometry staining of CD4⁺ and CD8⁺ T cells in immune cells isolated from parental and anti-PD-1–resistant tumors. (B) Percentages of CD45⁺CD4⁺ and CD45⁺CD8⁺ T cells in the gated lymphocytes isolated from tumors. (C) Total numbers of tumor-infiltrating immune cells (CD45⁺) and (D) CD45⁺CD4⁺ and CD45⁺CD8⁺ T cells isolated from parental and anti-PD-1–resistant tumors. (E) Total numbers of IFNγ-producing CD4⁺ and CD8⁺ tumor-infiltrating lymphocytes in tumors. Data represent means ± SD for an n of 5, with experiments repeated at least three times. *P<0.05, **P<0.01 in Student’s t tests; N.S. indicates not statistically significant.

Figure 4. Radiation increases MHC class I expression and overcomes anti-PD-1 resistance

Mice bearing anti-PD-1–resistant 344SQ tumors were either untreated or irradiated with three 12-Gy fractions. At 5 days after radiation, tumors were isolated, digested into single cells, and stained with live/dead dye, fluorochrome-congugated CD45, and H-2D^b and H-2K^b antibodies. Tumor cells were gated on a CD45^- population. (A) Radiation significantly increased MHC class I (H-2D^b) expression on the anti-PD-1–resistant tumors. (B) Radiation resensitized tumors, which allowed the tumors to respond to anti-PD-1 treatment. 344SQ_R cells were inoculated into syngeneic 129Sv/ev mice. At 8 days after inoculation, tumors had reached about 100 mm³ and were irradiated with three 12-Gy factions over 3 days. The first dose of anti-PD-1 (10 mg/kg) was given on the same day as the first radiation dose, and continued for 3 more doses (twice per week). P<0.001 for XRT vs control groups at Day 19, 21, 23, and 26; *P<0.05 or ***P<0.001 for XRT+anti-PD1 vs XRT groups at Day 21, 23, 26, and 28. Data represent means ± SD for an n of 7–8 per group, with experiments repeated three times.

Figure 5. Radiation plus anti-PD-1 induces tumor regression and increases proportions of CD8+ T cells in both irradiated and non-irradiated tumors

(A) Radiation plus anti-PD-1 synergistically enhanced the antitumor response by controlling both irradiated tumors and other, nonirradiated, tumors and by reducing the number of spontaneous lung metastases in the 344SQ parental tumor model. Primary tumors were established in the right leg by subcutaneous injection of 1×10⁶ 344SQ_P cells into 129Sv/ev mice. Other nonirradiated tumors were established on the left leg by subcutaneous inoculation of the same amount of 344SQ_P cells in the same mouse on day 10 after primary tumor inoculation. On day 14, when the primary tumors had reached about 100 mm³, primary tumors were irradiated with three 12-Gy fractions over 3 days. Anti-PD-1 treatment was given as described for Fig. 4B; n=7–8 per group. In the primary irradiated tumors, P<0.01 for XRT+anti-PD1 vs anti-PD1 groups from day 21 to 42 (the end point), and P<0.05 or <0.01 for XRT+anti-PD1 vs XRT groups from day 33 to 42. In the nonirradiated tumors, P values were <0.05 or <0.01 for control vs anti-PD1 groups from day 13 to 31 (the end point), and P values were <0.01 or <0.001 for XRT vs XRT+anti-PD1 groups from day 13 until the end of the experiment. (B) Radiation plus anti-PD-1 treatment increased the proportion of CD8⁺ T cells in both irradiated and nonirradiated tumors. For these experiments, 344SQ_P cancer cells (1×10⁶) were injected in the right leg and 2×10⁵ cells in the left leg of mice on the same day. Radiation and anti-PD-1 treatments were the same as described for Fig. 5A. Tumors and spleen were collected on day 7 after radiation treatment. n=5 per group. Data represent means ± SD, with experiments repeated three times.

Figure 6. Radiation sensitizes tumors to anti-PD-1 via activating the IFNβ/IFNAR-MHC class I pathway

(A) Radiation induced IFNβ, but not IFNγ, production. Mice bearing 344SQ_R tumors were either untreated or irradiated to a total dose of 36 Gy, given in three daily 12-Gy fractions. Five days later, plasma was collected for cytokine detection (via Multiplex). (B) Both histograms and (C) bar graphs of mean fluorescence intensity from flow cytometry show dose-dependent induction of MHC-I (H-2K^b and H-2D^b) and MHC class II (I-A/I-E) by IFNβ. For these studies, 344SQ_P and 344SQ_R cells were cultured in vitro and treated with IFNβ (100–10,000 U/mL) for 24 h. Cells were then stained for cell-surface expression of H-2K^b, H-2D^b, and I-A/I-E and subjected to flow cytometry. Experiments were repeated twice. (D) Radiation-induced sensitization to anti-PD-1 is blunted by antibody blockade of the IFN α/β receptor I (IFNAR1). Mice bearing anti-PD-1–resistant 344SQ tumors were treated with radiation (36 Gy in three daily 12-Gy fractions) and anti-PD-1 (10 mg/kg, intraperitoneal injection), with or without anti-IFNAR1 (1 mg/kg, intratumoral injection). Tumor growth was monitored. P values were <0.05 or <0.01 for XRT+anti-PD1 vs XRT+anti-PD1+anti-IFNAR1 groups from day 17 to 24 (the end point). Data represent means ± SD for an n of 7, with experiments repeated twice. (E) Model of anti-PD-1 resistance. Tumors that respond to treatment with anti-PD-1 express higher levels of MHC class I, which can stimulate tumor-infiltrating lymphocyte proliferation and activation, promoting strong antitumor immune responses that cause tumor regression. In contrast, anti-PD-1–resistant tumors express relatively lower levels of MHC class I, and thus have less T cell proliferation and activation. As a result, defective antitumor immunity cannot control tumors, and tumors continue progressing. Radiation induces type I IFN, which enhances MHC I expression, resulting in resensitization to anti-PD-1 treatment.