Poster Presentation Melbourne Immunotherapy Network Winter Symposium 2021

Targeting immunosuppressive myeloid cells to improve immunotherapy for breast cancer (#117)

Kellie Mouchemore 1 2 , Lap Hing Chi 1 2 , Yang Liao 1 2 , Caroline Bell 1 , Suraya Roslan 1 , Wei Shi 1 2 , Nicole Haynes 3 , John Hamilton 4 , Robin Anderson 1 2
  1. Olivia Newton-John Cancer Research Institute, Heidelberg, VIC, Australia
  2. La Trobe University School of Cancer Medicine, Bundoora, VIC, Australia
  3. Peter MacCallum Cancer Centre, Parkville, VIC, Australia
  4. Department of Medicine, Royal Melbourne Hospital, The University of Melbourne, Parkville, VIC, Australia

The first immunotherapy for metastatic triple negative breast cancer (TNBC) has been approved. However, compared to melanoma and lung cancer, response rates for TNBC remain low. The presence in tumours and metastases of granulocytic myeloid-derived suppressor cells (G-MDSCs), defined by their ability to suppress T cell activity, is a likely mechanism of immunotherapy resistance. Targeting of G-MDSCs has thus far proven problematic, due to lack of specific depletion agents and the inability to easily differentiate G-MDSCs from neutrophils (both are CD11b+Ly6G+). Preclinical studies have implicated granulocyte colony-stimulating factor (G-CSF) in driving the accumulation of G-MDSCs and the development of metastatic disease. However, the receptor for G-CSF (G-CSFR) is also expressed by neutrophils, creating concern that targeting G-CSF will cause neutropenia.

Using the 4T1 mammary tumour model, we have demonstrated that therapeutic blockade of the G-CSFR can partially reduce metastasis without causing neutropenia in mice. Similarly, shRNA-mediated knockdown of G-CSF in 4T1 tumour cells also reduced metastasis. In both scenarios, G-CSF inhibition did not alter primary tumour growth nor primary tumour infiltration with CD11b+Ly6G+ cells. High expression of PD-L1 on tumour and myeloid cells was maintained, indicating that addition of anti-PD-L1 therapy was required for further benefit. However, the combination of G-CSF inhibition with anti-PD-L1 did not sensitize 4T1 primary tumours to this immunotherapy.

Single cell RNA-sequencing analysis on the myeloid infiltrate of 4T1 tumours demonstrated that although G-CSFR blockade reduced prototypical G-MDSC populations, a subpopulation of highly immunosuppressive G-MDSC remained, marked by expression of the receptor for vitamin B3. We hypothesize that a triple combination therapy targeting these remaining myeloid cells alongside G-CSFR blockade and anti-PD-L1 will increase control of both primary and metastatic disease, thus representing a clinically translatable strategy for improving immunotherapy in TNBC patients.