Translational strategies exploiting TNF-alpha that sensitize tumors to radiation therapy

Abstract

TNFerade is a radioinducible adenoviral vector expressing tumor necrosis factor-alpha (TNF-alpha) (Ad.Egr-TNF) currently in a phase III trial for inoperable pancreatic cancer. We studied B16-F1 melanoma tumors in TNF receptor wild-type (C57BL/6) and deficient (TNFR1,2-/- and TNFR1-/-) mice. Ad.Egr-TNF+IR inhibited tumor growth compared with IR in C57BL/6 but not in receptor-deficient mice. Tumors resistant to TNF-alpha were also sensitive to Ad.Egr-TNF+IR in C57BL/6 mice. Ad.Egr-TNF+IR produced an increase in tumor-associated endothelial cell apoptosis not observed in receptor-deficient animals. Also, B16-F1 tumors in mice with germline deletions of TNFR1,2, TNFR1 or TNF-alpha, or in mice receiving anti-TNF-alpha exhibited radiosensitivity. These results show that tumor-associated endothelium is the principal target for Ad.Egr-TNF radiosensitization and implicate TNF-alpha signaling in tumor radiosensitivity.

Figure 1

Growth of B16-F1 tumors in C57BL/6 mice. (a) B16F-1 tumors were treated with Ad.Egr-TNF±20 Gy. A reduction in mean tumor volume was observed in tumors treated with Ad.Egr-TNF+20 Gy compared with 20 Gy (P=0.015, day 11). ■ Control (viral buffer); ▲ Ad.Egr-TNF; ▼ 20 Gy; and ◆ Ad.Egr-TNF+20 Gy. (b) B16-F1 tumors were treated with Ad.Empty±20 Gy. No reduction in mean tumor volume was observed in tumors treated with a null vector (Ad.Empty) +20 Gy compared with 20 Gy (P=0.99, day 11). ■ Control (viral buffer); ▲ Ad.Empty; ▼ 20 Gy; and ◆ Ad.Empty +20 Gy. Data are graphed as mean tumor volume (mm³) ±s.e.m.

Figure 2

Intratumoral concentrations of human tumor necrosis factor-α (TNF-α). B16-F1 tumors were excised, homogenates were prepared and ELISA were performed. A significant increase in human TNF-α was detected 24 h after treatment with Ad.Egr-TNF+20Gy compared with Ad.Egr-TNF alone (P<0.0001). No human TNF-α was detected in the control tumors or tumors treated with 20Gy.

Figure 3

Growth of B16F-1 tumors in mice with a germline deletion in tumor necrosis factor-α (TNF-α) (TNF−/−). Treatment with Ad.Egr-TNF+20 Gy significantly decreased mean tumor volume compared with 20 Gy (P=0.05, day 7). ■ Control (viral buffer); ▲ Ad.Egr-TNF; ▼ 20 Gy and ◆ Ad.Egr-TNF+20 Gy. Data are graphed as mean tumor volume (mm³) ±s.e.m.

Figure 4

Growth of B16F-1 tumors in tumor necrosis factor (TNF) receptor-deficient mice. (a) B16F-1 tumors in mice with a germline deletion in TNF receptors 1 and 2 (TNFR1,2−/−). No significant difference in tumor regrowth was observed in tumors treated with Ad.Egr-TNF+20 Gy compared with tumors treated with 20 Gy (P=0.5, day 12). (b) B16F-1 tumors growing in mice with a germline deletion in TNF receptor 1 (TNFR1−/−). No significant difference in tumor regrowth was observed in tumors treated with Ad.Egr- TNF+20 Gy compared with tumors treated with 20 Gy (P=0.49, day 12). ■ Control (viral buffer); ▲ Ad.Egr-TNF; ▼ 20 Gy and ◆ Ad.Egr-TNF+20 Gy. Data are graphed as mean tumor volume (mm³) ±s.e.m.

Figure 5

Growth of BFS-2C murine fibrosarcoma tumors. (a) BFS-2C (cell line derived from TNFR1−/− mouse) tumors growing in C57BL/6 mice. Tumors treated with Ad.Egr-TNF+20 Gy produced significant reduction in mean tumor volume compared with tumors treated with 20 Gy (*P=0.008, day 4; **P=0.01, day 7). (b) BFS-2C tumors growing in TNFR1,2−/− mice. No significant difference in mean tumor volume was observed in tumors treated with Ad.Egr- TNF+20 Gy compared with tumors treated with 20 Gy. ■ Control (viral buffer); ▲ Ad.Egr-TNF; ▼ 20 Gy and ◆ Ad.Egr-TNF+20 Gy. Data are graphed as mean fractional tumor volume (V/Vo) ±s.e.m.

Figure 6

Effect of treatment with Ad.Egr-TNF+20 Gy on tumor-associated endothelial cell apoptosis. Percent endothelial apoptosis was significantly increased at 4 and 24 h after treatment with Ad.Egr-TNF+20 Gy in tumors grown in C57BL/6 mice, compared to 20 Gy or Ad.Egr- TNF treatment. Tumors grown in TNFR1,2−/− mice did not show a similar increase in endothelial apoptosis. (a) Untreated control tumor in C57BL/6 mice; (b) 4 h post-Ad. Egr-TNF+20 Gy treatment in C57BL/6 mice; (c) untreated tumor in TNFR1,2−/− mice; (d) 4 h post-Ad.Egr- TNF+20 Gy treatment in TNFR1,2−/−. Scale bar, 20 μm. Arrows show double staining of VEGFR2 (blue) and cleaved caspase 3 (red); (e) percent apoptotic endothelial cells per high power field (× 400) ±s.e.m. Twenty random fields of tissue sections from 3–5 tumors were counted.