The efficacy versus toxicity profile of combination virotherapy and TLR immunotherapy highlights the danger of administering TLR agonists to oncolytic virus-treated mice

Abstract

Injection of oncolytic vesicular stomatitis virus (VSV) into established B16ova melanomas results in tumor regression, in large part by inducing innate immune reactivity against the viral infection, mediated by MyD88- and type III interferon (IFN)-, but not TLR-4-, signaling. We show here that intratumoral (IT) treatment with lipopolysaccharide (LPS), a TLR-4 agonist, significantly enhanced the local therapy induced by VSV by combining activation of different innate immune pathways. Therapy was further enhanced by co-recruiting a potent antitumor, adaptive T-cell response by using a VSV engineered to express the ovalbumin tumor-associated antigen ova, in combination with LPS. However, the combination of IT LPS with systemically delivered VSV resulted in rapid morbidity and mortality in the majority of mice. Decreasing the intravenous (IV) dose of VSV to levels at which toxicity was ameliorated did not enhance therapy compared with IT LPS alone. Toxicity of the systemic VSV + IT LPS regimen was associated with rapidly elevated levels of serum tumor necrosis factor-α (TNF-α) and interleukin (IL)-6, which neither systemic VSV, nor IT LPS, alone induced. These data show that therapy associated with direct IT injections of oncolytic viruses can be significantly enhanced by combination with agonists of innate immune activation pathways, which are not themselves activated by the virus alone. Importantly, they also highlight possible, unforeseen dangers of combination therapies in which an immunotherapy, even delivered locally at the tumor site, may systemically sensitize the patient to a cytokine shock-like response triggered by IV delivery of oncolytic virus.

Figure 1

Type-I interferon signaling is critical for antitumor effects of intratumoral LPS. Survival (tumor <1.0 cm in any diameter) of (a) C57BL/6 or (b) IFNAR^−/− mice (n = 8/group) bearing subcutaneous (s.c.) tumors following treatment with three intratumoral doses of LPS (200 µg/50 µl) or PBS (50 µl), starting on day 7 (days 7, 9, 11). (c) Tumors from either C57BL/6 (in a) or IFNAR^−/− (in b) were harvested 1 day after the second injection (day 10) and analyzed for CD11b⁺F4/80⁺IAb⁺ macrophages by flow cytometry (n = 3/group). Macrophage percentages were determined relative to the total number of CD45⁺ cells in the tumor and are shown as mean ± SD. LPS, lipopolysaccharide; N.S., not significant; PBS, phosphate-buffered saline.

Figure 2

Enhanced therapy and immune cell infiltration following combination intratumoral VSV-ova + LPS treatment. (a) Survival (tumor <1.0 cm in any diameter) of C57BL/6 mice (n = 8/group) bearing subcutaneous tumors following treatment with three intratumoral doses of VSV-GFP (5 × 10⁸ PFU/50 µl), PBS (50 µl), and/or LPS (200 µg/50 µl) starting on day 7 (days 7, 9, 11). LPS (200 µg/50 µl), when given in combination with VSV-GFP, was given on day 8. (b) Survival (as in a) following treatment with three intratumoral doses of VSV-ova (5 × 10⁸ PFU/50 µl), PBS (50 µl), and/or LPS (200 µg/50 µl) starting on day 7 (days 7, 9, 11). LPS (200 µg/50 µl), when given in combination with VSV-ova, was given on day 8. Tumors were harvested following a single round of treatment (n = 3/group) and analyzed for infiltrating (c) CD11b⁺Gr1⁺F4/80⁻ neutrophils or (d) CD3⁺CD8⁺ T lymphocytes. Percentages were determined relative to the total number of CD45⁺ cells in the tumor and are shown as mean ± SD. GFP, green fluorescent protein; LPS, lipopolysaccharide; PBS, phosphate-buffered saline; ova, ovalbumin; PFU, plaque-forming unit; VSV, vesicular stomatitis virus.

Figure 3

Shift in generalized and antigen-specific T-cell activation in mice treated with VSV-ova + LPS. (a,b) C57BL/6 mice bearing 7-day subcutaneous tumors (n = 8/group) were administered intratumorally VSV-ova (5 × 10⁸ PFU/50 µl), PBS (50 µl), and/or LPS (200 µg/50 µl). LPS (200 µg/50 µl), when given in combination with VSV-ova, was given on day 8. (a) Tumor-draining lymph nodes and (b) tumors were harvested on day 14 (n = 3/group). Samples were pulsed with either no peptide (medium), ova or VSV-specific peptides for 48 hours (in a) or 1 hour (in b) with T-cell activation being assessed by IFN-γ ELISA of cell-free supernates (in a) or intracellular IFN-γ staining (in b). Data are shown as mean ± SD where appropriate. ELISA, enzyme-linked immunosorbent assay; IFN, interferon; LPS, lipopolysaccharide; PBS, phosphate-buffered saline; ova, ovalbumin; PFU, plaque-forming unit; VSV, vesicular stomatitis virus.

Figure 4

High-dose, systemic VSV, when combined with local LPS treatment, results in acute toxicity. (a) Illness-free survival of C57BL/6 mice (mice not showing any signs of physical distress) bearing subcutaneous tumors after treatment with intravenous (IV) VSV-Δ51 (5 × 10⁸ PFU/100 µl) or PBS (100 µl) followed by intratumoral (IT) PBS (50 µl) or LPS (200 µg/50 µl) 6 hours later (n = 7/group). Injections started on day 7 and continued every other day for a total of four (days 7, 9, 11, 13). (b,c) Mice (as in a) were treated with IV VSV-Δ51 (5 × 10⁸ PFU/100 µl or 5 × 10⁵ PFU/100 µl) or PBS (100 µl) followed by IT PBS (50 µl) or LPS (200 µg/50 µl) 6 hours later (n = 7/group). Cardiac terminal bleeds were done 1.5 hours after a single combination treatment. Serum (b) TNF-α or (c) IL-6 levels were determined by ELISA and are shown as mean ± SD. (d) Illness-free survival (as in a) of tumor-bearing C57BL/6 or TNFKO mice after a single treatment with IV VSV-Δ51 (5 × 10⁸ PFU/100 µl) or PBS (100 µl) followed by IT PBS (50 µl) or LPS (200 µg/50 µl) 6 hours later (LPS only groups: n = 5/group; VSV + LPS groups: n = 8/group). ELISA, enzyme-linked immunosorbent assay; IL, interleukin; KO, knockout; LPS, lipopolysaccharide; PBS, phosphate-buffered saline; PFU, plaque-forming unit; TNF-α, tumor necrosis factor-α VSV, vesicular stomatitis virus; WT, wild-type.

Figure 5

Systemic VSV combined with LPS does not improve local, LPS-only therapy or increase immune cell infiltration. (a) Survival (tumor <1.0 cm in any diameter) of C57BL/6 mice (n = 7/group) bearing subcutaneous tumors after treatment with intravenous (IV) VSV-Δ51 (5 × 10⁶ PFU/100 µl) or PBS (100 µl) followed by intratumoral (IT) PBS (50 µl) or LPS (200 µg/50 µl) 6 hours later. Injections started on day 7 and continued every other day for a total of four (days 7, 9, 11, 13). (b–i) H&E stained tumors harvested 1 day after the last treatment (day 14). Arrows identify (b–e) representative areas of necrosis (×4 magnification) or (f–i) immune cell infiltration (×40 magnification). Bar equals 10 mm for all images. H&E, hematoxylin and eosin; LPS, lipopolysaccharide; PBS, phosphate-buffered saline; VSV, vesicular stomatitis virus.

Figure 6

Neither efficacy nor toxicity observed following treatment with BCG or Coley's toxin in combination with systemic VSV. (a) Survival (tumor <1.0 cm in any diameter) of C57BL/6 mice (n = 7/group) bearing subcutaneous tumors after treatment with intravenous (IV) VSV-Δ51 (5 × 10⁶ PFU/100 µl) or PBS (100 µl) followed by intratumoral (IT) PBS (50 µl) or BCG (1 mg/50 µl) 6 hours later. Injections started on day 7 and continued every other day for a total of four (days 7, 9, 11, 13). (b) Survival (as in a) after treatment with IV VSV-Δ51 (5 × 10⁶ PFU/100 µl) or PBS (100 µl) followed by IT PBS (50 µl) or Coley's toxin (final dilution of 0.05 Coley's toxin in 50 µl PBS) 6 hours later. Injections started on day 7 and continued every other day for a total of four (days 7, 9, 11, 13). BCG, bacillus Calmette–Guérin; PBS, phosphate-buffered saline; PFU, plaque-forming unit; VSV, vesicular stomatitis virus.