No abstract

The majority of patients with TNBC fail to derive benefit from PD-L1 blockade alone. Recent phase III trial results show that PD-L1 blockade increases survival when combined with chemotherapy, warranting exploration of more effective treatment combinations. One potential new agent is the V-domain Ig suppressor of T cell activation (VISTA), a negative checkpoint ligand that inhibits T cell activation via a mechanism that is non-redundant with PD-1. In this study, the highly metastatic 4T1 mouse model of TNBC is used to test the combination of focal tumor radiation (RT) with low dose cyclophosphamide (CYP), VISTA and anti-PD-1 therapy, and explore its impact on the microbiome (MB). Materials/Methods: Mice were inoculated with 4T1 cells (day 0) and randomly assigned to one of several treatment groups (6-10 per group). RT was 12 Gy x 2 on d13 and 14. Anti-VISTA was given 3x/wk starting d14; anti-PD-1 on d14, 17 and 20; and CYP on day 9. Mice were followed for tumor growth and survival or euthanized at day 21 or 32 to assess tumorinfiltrating lymphocytes/ T-cell responses or mets. MB samples were obtained at d8, 12, 13 and 15. Results: RT significantly delayed tumor growth (p<0.05) but did not impact mets. Anti-VISTA or anti-PD-1, alone or combined, did not reduce primary tumor growth or mets, but VISTA blockade was as effective as anti-PD-1 when combined with RT. No further improvement in met control was seen with RT+dual PD-1/VISTA blockade. Surprisingly, addition of CYP 4 days before RT led to further significant improvement in met control in RT+ dual PD-1/VISTA treated mice (30% met-free on d32 (p<0.001)), increased priming of tumor-specific CD8+ T cells, and increased tumor infiltration of CD8+ T cells (53.5% vs 14.0% in control vs 16.5% in RT+anti-VISTA vs 21.1% in RT+anti-PD1, p<0.05). Strikingly, the quadruple treatment group displayed the highest median survival (54.5d vs 34d in control vs 45.5d in RT+anti-PD1, p<0.01). The therapeutic effects of anti-VISTA were in part due to reduced intra-tumoral granulocytic MDSC. The impact of CYP and RT on the composition, diversity and abundance of the MB were analyzed for distinctive changes. Conclusion: PD-1 blockade has shown remarkable benefits in a small subset of patients, prompting investigations to identify promising combinations such as the use of RT + anti-PD-1. Here we show that RT can also synergize with anti-VISTA in inducing systemic anti-tumor responses, providing the first evidence that RT sensitizes refractory tumors to VISTA blockade. Remarkably, treatment with anti-VISTA markedly improved the therapeutic effect of RT+PD-1 blockade, but only in the presence of lowdose CYP prior to RT. We are currently investigating the mechanisms of this very promising combination, particularly on the changes in gut MB. Overall, these results suggest that a multipronged approach may be necessary to enhance response to IT in TNBC.