Immune cold tumors arise due to inadequate induction of an innate immune response during early tumor progression (Bonaventura et al., 2019). This is usually due to reduced expression of interferons (IFNs) and interferon stimulated genes (ISGs) (Bonaventura et al., 2019; Galon and Bruni, 2019; Liu and Sun, 2021). Reduced expression of interferons leads to loss of major histocompatibility complex (MHC) class 1 expression which impedes recognition by T-cells (Dhatchinamoorthy et al., 2021). Loss of MHC class 1 is observed in about 60% of TNBCs which may account for the lack of response to immunotherapies in patients (Dusenbery et al., 2021). One of the key regulators of the interferon response is stimulator of interferon genes (STING) (Zhu et al., 2019; Decout et al., 2021). STING is a cytosolic sensor of the cyclic dinucleotide 2′3′-cyclic guanosine monophosphate–adenosine monophosphate (2′3′-cGAMP), which is produced by cyclic GMP-AMP synthase (cGAS) due to cytosolic DNA as a result of cell death or viral infection (Decout et al., 2021). STING activates tank-binding kinase 1 (TBK1), leading to activation of interferon response factor (IRF) three which acts as a transcription factor for ISGs (Decout et al., 2021). Induction of interferons act as a “red flag” for the immune system and results recruitment of anti-tumor immune cells. Blocking this cascade at any stage could result in an immune cold, interferon depleted tumor (Zhu et al., 2019). It is important to note, that 2′3′-cGAMP functions as an immune transmitter between different cancer cells as well as cancer cells and immune cells (such as dendritic cells) whereby one cell can activate STING in a neighbouring cell (Carozza et al., 2020) leading to a positive feedback cascade and increased inflammation. However, excessive inflammation, a hallmark of cancer (Hanahan and Weinberg, 2011), leads to disruption of tissue homeostasis and recruitment of inhibitory immune cells such as regulatory T-cells (T_regs), myeloid derived suppressor cells (MDSCs), macrophages and upregulation of immune checkpoint molecules which sustains an immunosuppressed immune hot TIME.

As a mechanism of stimulating STING in immune cold tumors, either 2′3′-cGAMP or agonists (small molecules that activate STING) have been implemented to stimulate interferon expression and promote an anti-tumor immune response (Deng et al., 2014). Using the 4T1 TNBC model, the STING agonist, cyclic diguanylate, lead to induction of IFNβ expression which was coupled to an increase in activation of STING and its downstream effectors (Yin et al., 2022). Cyclic diguanylate synergized with atezolizumab (anti-PDL1) to reduce tumor growth through activation of cytotoxic T-cells and suppression of T_regs (Yin et al., 2022). Other work on a HER2+ model using the STING activator, ADU-S100, demonstrated that STING activation can induce tumor regression as well as protect against tumor rechallenge in non-tolerized animals (Foote et al., 2017). The effect of ADU-S100 was dependent on both CD8⁺ and CD4⁺ T-cells (Foote et al., 2017). Moreover, blockade of immune checkpoints, OX40 receptor and anti-PDL1 lead to a significant increase in ADU-S100 efficacy and complete regression of tumors (Foote et al., 2017).

Several STING agonists are currently in early phases of clinical trials as an adjuvant or in combination with other immunotherapies (NCT05070247, NCT04144140, NCT03956680, NCT03843359, NCT03010176, NCT03249792, NCT04096638). Additionally, already approved poly ADP ribose polymerase (PARP) inhibitor, Olaparib, in a TNBC mouse model demonstrated its capacity to induce CD8⁺ T-cell infiltration via activation of STING (Pantelidou et al., 2019) as well as promote upregulation of PDL1 in several human derived cell lines (Jiao et al., 2017) suggesting that inducing DNA damage may promote a strong anti-tumor immune response through STING activation. Noteworthy, that increasing the tumor mutational load is associated with favourable immune infiltration and response to immunotherapies (Strickler et al., 2021), potentially due to elevated basal activation of STING. However, only a very small fraction of breast cancers have a high mutational load compared to immunogenic tumors such as lung and melanoma (Thomas et al., 2018; Barroso-Sousa et al., 2020). Additionally. HR + tumors harbour a smaller proportion of tumors with a high mutational burden (Thomas et al., 2018). Mechanisms that increase the mutational load might be clinically beneficial, especially for HR + breast cancer. Indeed, inactivation of the DNA mismatch repair (MMR) by targeting MutL homologue 1 (MLH1) leads to an increased mutational load due to reduced genomic stability, triggering an immune response that sensitizes tumors to anti-PD1 (Germano et al., 2017).

In a cDNA library screen, HER2, which is highly expressed in ∼25% of breast cancers (Harbeck et al., 2019), was found to impede STING activation (Wu et al., 2019). Detailed analysis of this phenomenon demonstrated that the intracellular domain (ICD) of HER2 physically sequesters STING, preventing interaction with TBK1/IRF3 (Wu et al., 2019). Additionally, AKT1 activation downstream HER2, phosphorylates TBK1 at serine 510 which inhibits its downstream activity and association with STING (Wu et al., 2019). This data strongly argues that oncogene mediated signaling can impact the innate immune response and result in an immune cold TIME, which is amenable to receptor tyrosine kinase inhibitors (RTKi) such as lapatinib (Wu et al., 2019). Interestingly, activation of DNA sensing increased the HER2-AKT1 cascade (Wu et al., 2019). We envision that this may behave as a mechanism of resistance in tumors treated with STING activators.

Novel approaches at regulating STING have led to development of small molecules targeting ecto-nucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1) which is an endoplasmic reticulum and cell surface enzyme that degrades natural STING ligand 2′3′-cGAMP as well as ATP/GTP to AMP and GMP (Li et al., 2014; Carozza et al., 2022; Gangar et al., 2022; Goswami et al., 2022). Inhibition of ENPP1 using a small molecule inhibitor was sufficient to induce an active immune response which synergized with radiation therapy (Carozza et al., 2020) potentially due to reduced degradation of 2′3′-cGAMP produced by irradiated cells. Moreover, other studies demonstrated that chromosomally unstable breast tumors suppress an immune response against them via selective upregulation of ENPP1 which degrades 2′3′-cGAMP that may be produced in response to dsDNA released into the cytosol following nuclear envelope ruptures (Li et al., 2021b). Indeed, knockout of ENPP1 in TNBC syngeneic models reduced metastasis to the lung and led to a massive increase in immune cell infiltration at the primary tumor (Li et al., 2021b). Moreover, knockout of ENPP1 synergized with PD1 blockade (Li et al., 2021b). Both patient tumors and mouse models demonstrated that ENPP1 is elevated in relapsed tumors of the TNBC and HER2+ subtypes (Ruiz-Fernandez de Cordoba et al., 2022). ENPP1 promoted neutrophil extracellular trap (NET) formation via tumor cell expression of haptoglobin, supporting relapse and protecting the tumor from radiotherapy (Ruiz-Fernandez de Cordoba et al., 2022). Additionally, both expression of interferons and STING activation are associated with differentiation of the tumor, reduced stem cell capacity and reduced expression of epithelial-mesenchymal transition (EMT) gene signature (Cheng et al., 2020; Cheon et al., 2022; Goswami et al., 2022). Targeting ENPP1 reduced expression of EMT markers and signatures (Goswami et al., 2022). Only one documented case of first-in-human ENPP1 inhibitor in addition to anti-PD1 was performed (Csiki et al., 2022), but no large scale trials have been preformed thus far. We posit that inhibition of ENPP1 may provide strong clinical benefit especially to immune cold TNBCs and HER2+ breast cancers. Combination of 2′3′-cGAMP (Ohkuri et al., 2017) or radiation (Carozza et al., 2020) coupled to ENPP1 inhibition may demonstrate a synergistic effect due to the prolonged accumulation of 2′3′-cGAMP in the extracellular milieu.

AMP produced by ENPP1 is subjected to further hydrolysis by CD73 (Allard et al., 2020). The resultant adenosine may then bind adenosine receptors found on stromal or cancer cells to reinforce an immunosuppressive microenvironment (Allard et al., 2020). CD73 expression was found to be highest in TNBC compared to other breast cancer subtypes and was significantly associated with worse prognosis (Loi et al., 2013). Activity of CD73 in TNBC induced immunosuppression and promoted resistance to anthracyclines (Loi et al., 2013). Targeting CD73 using a monoclonal antibody increased the number of tumor specific CD8⁺ T-cells and induced INFγ expression (Loi et al., 2013). Combination of anthracycline (doxorubicin) with anti-CD73 delayed death due to metastasis of animals post surgical resection of their primary tumor (Loi et al., 2013). Moreover, anti-CD73 synergized with both anti-PD1 and anti-CTLA4 (Loi et al., 2013). Expression of CD73 has also been correlated with resistance to anti-HER2 targeted therapy, Trastuzumab (Turcotte et al., 2017). Neutralization of CD73 synergized with anti-HER2 monoclonal therapy, eliminated HER2+ tumor growth in syngeneic models as well as reduced pulmonary metastasis (Turcotte et al., 2017). Anti-CD73 targeting antibodies as well as CD73 inhibitors are currently in early phase clinical trials in combination with immune checkpoint blockades and Trastuzumab (NCT03616886, NCT03875573, NCT03454451, NCT05431270, NCT05143970, NCT05001347, NCT04148937).

Chemotherapy in combination with anti-PD1 is approved for PDL1+ TNBC (Cortes et al., 2022). While the idea of eventually phasing out chemotherapy and radiation is highly tempting due to their systemic adverse effects; they still remain viable options to prime an immune response and potentially induce upregulation PDL1 and other co-inhibitory molecules (Zitvogel et al., 2008).

DNA damage following radiation induces STING activation which promotes recruitment of dendritic cells which act as antigen presenting cells priming T-cell mediated immunity (Dhatchinamoorthy et al., 2021) (impact is under clinical assessment NCT03978663). Both chemotherapy and radiation can induce cell and DNA damage, generating neoantigens and inducing antigenicity which may sensitize tumors to immunotherapies (Zitvogel et al., 2008; McGranahan et al., 2016; Galon and Bruni, 2019). Additionally, chemotherapy has been demonstrated to selectively deplete MDSCs, T_regs and restores T- and natural (NK) cell functions (Ghiringhelli et al., 2007; Vincent et al., 2010). DNA damage following radiation induces STING activation which promotes recruitment of dendritic cells that act as antigen presenting cells to prime T-cell mediated immunity (Deng et al., 2014). Clinical studies (NCT03978663) are currently underway to correlate radiation treatment with immune cell recruitment and patient outcome.

Resistance to radiation and chemotherapy posses an additional challenge, as tumors need to respond to radiation or chemotherapy to effectively prime an immune response. Resistance to radiation has been linked to hypoxia (briefly discussed in the next section) where lack of molecular oxygen prevents generation of free radicals to react with DNA (Wang et al., 2021). Strategies aimed at promoting tumor oxygenation by vasculature normalization have demonstrated the capacity to sensitize tumors to radiation and improve chemotherapy delivery in multiple tumor models (Telarovic et al., 2021). Tyrosine kinase inhibitor Sunitinib is under clinical investigation in combination with chemotherapy to induce vasculature normalization to enhance chemotherapy delivery (NCT01803503). Blood vessel normalization and perfusion is driven by IFNγ expressing cytotoxic T-cells and is a predictor of response to ICB, anti-PD1 and anti-CTLA4, and the anti-HER2 monoclonal, Trastuzumab (Baker et al., 2018; Zheng et al., 2018). Therefore, strategies at improving vasculature perfusion may permit increased cytotoxic T-cell infiltration, which establishes a positive feedback loop, increasing vasculature perfusion further allowing better delivery of chemotherapy and anti-tumor immune cells. Thus, improved vasculature perfusion may also allow better tumor oxygenation and improved immune responses.

Hypoxia is regarded as a critical phenomenon in breast cancer (Zhang et al., 2021). Indeed, an abundance of work has demonstrated that hypoxia and upregulation of hypoxia inducible factor 1α (HIF1α) in the epithelium regulates breast cancer progression and metastasis via promoting angiogenesis, induction of cancer stem phenotypes and EMT (Muz et al., 2015). Additionally, hypoxia promotes immune suppression either by inhibiting T-cell activity or through exclusion of immune cells by a barrier comprised of low pH, low glucose, and low oxygen conditions (Wang et al., 2021; Zhang et al., 2021). Interestingly, CD73 is under the control of hypoxia and potentially contributes to hypoxia mediated immunosuppression (Allard et al., 2020). Spatial analysis of 4T1 tumors demonstrated that hypoxic areas (HIF1α+) are poorly infiltrated with T-cells as expected (Ma et al., 2022). This work identified that through HIF1α, hypoxia induced silencing of INFγ and tumor necrosis factor (TNF) through histone deacetylase (HDAC) and enhancer of zeste homolog 2 (EZH2) mediated epigenetic silencing in T-cells (Ma et al., 2022). Inhibition or genetic ablation of HIF1α, HDAC or EZH2 sensitized anti-PD1 resistant 4T1 tumors to anti-PD1 (Ma et al., 2022). Knockout of HIF1α in the macrophages of polyoma middle T (PyMT) tumors increased proliferation and cytotoxic activity of T-cells, emphasizing that hypoxia may promote immunosuppression through macrophage function as well (Doedens et al., 2010).

While low oxygen stabilizes HIFs and regulates a hypoxia induced response, receptor tyrosine kinases (RTKs) such as the epidermal growth factor (EGFR) family, which are commonly highly expressed in breast cancers can stabilize HIFs (Lauzier et al., 2007; Lee et al., 2009; Nilsson et al., 2010; Turpin et al., 2016). Interestingly, it was demonstrated that through HIF2α, hypoxia can promote translation of EGFR mRNA (Franovic et al., 2007; Uniacke et al., 2012) implying the presence of a positive feedback loop between HIFs and EGFR. Knockout of AXL RTK in a HER2+ driven mouse model reduced expression of hypoxic markers, reactivated the immune response and improved vasculature perfusion in the primary tumors leading to strong reduction in lung metastasis (Goyette et al., 2021). Inhibition of AXL using R428 synergized with PD1 blockade at blocking metastasis (Goyette et al., 2021). Clinical trials assessing the AXL inhibitor, Bemcentinib, with Pembrolizumab (anti-PD1) on TNBC was attempted on a small cohort of patients (NCT03184558) but none achieved complete response leading to termination of the trial. Newer trials are currently underway to assess impact of chimeric antigen receptor (CAR) T-cells against AXL as well as other RTKs including HER2 and EGFR on advanced tumors (NCT04842812).

EZH2 acts as the catalytic subunit of polycomb repressive complex 2 (PRC2) and places a repressive methyl mark on lysine 27 of histone 3, H3K27me3 (van Mierlo et al., 2019). Progression of PyMT-driven tumors is correlated with an increase in H3K27me3 repressive marks (Cai et al., 2018). EZH2 knockout in PyMT and HER2-driven models demonstrated profound impact on tumor initiation, progression, and metastasis, supporting that EZH2 behaves as on oncogene by suppressing expression of tumor suppressors (Hirukawa et al., 2018; Smith et al., 2019). We have previously demonstrated that inhibition of EZH2 in a HER2+ mouse model induces tumor cell expression of endogenous retroviral elements that promote expression of IFNs and sensitize tumors to anti-HER2 targeted therapy (Hirukawa et al., 2019). Work using ovarian cancer models demonstrated that inhibition of EZH2 induced expression of inflammatory cytokines C-X-C ligand (CXCL) 9 and CXCL10 that promoted recruitment of cytotoxic T-cells into the tumor (Peng et al., 2015). Studies from other tumor models have demonstrated that EZH2 inhibits expression of PDL1 (Xiao et al., 2019) and the MHC class 1 antigen processing pathway (Burr et al., 2019). These studies support that epigenetic inhibition may impact induction of a robust immune response both through the tumor cell proper as well as through direct action on immune cells. The EZH2 inhibitor, SHR2554 is currently in clinical trials in combination with PARP, androgen receptor and cyclin dependent kinases four and 6 (CDK4/6) inhibitors (NCT04355858 and NCT03805399).

Oncolytic viruses (OVs) are naturally occurring or genetically modified viruses that are capable of selectively infecting and destroying tumor cells (Jin et al., 2021). OVs induce tumor cell death, resulting in release of pathogen-associated molecular patterns (PAMPs) and danger-associated molecular patterns (DAMPs) that can stimulate the anti-tumor immune response (Jin et al., 2021). While OVs are not currently approved, several are currently under pre-clinical and clinical evaluation (Carter et al., 2021; Jin et al., 2021). A herpes simplex virus (HSV)-based OV designated G47Δ, reduced pulmonary metastatic growth (Wang et al., 2012). Addition of IL-12 coding sequence to G47Δ (designated G47Δ-mIL12) killed both murine and human breast cancer cells in vitro and in syngeneic models. It also led to increased infiltration of cytotoxic T-cells and decreased MDSCs infiltration into the tumors (Ghouse et al., 2020). Several OVs are currently in clinical assessment in combination with other immunotherapies such as anti-PDL1 and CAR-T cells (NCT05600582, NCT05081492, NCT03004183, NCT04445844, NCT02779855, NCT04185311, NCT05076760, NCT04215146, NCT03740256, NCT05180851, NCT04521764).

Given that T-cells are known for their ability to recognize and kill cancer cells, efforts at expanding them ex vivo and reintroducing them back into the patient in a process known as adoptive cell transfer (ACT), has been successful in highly immunogenic tumors (Luo et al., 2022). Recently, ACT of T-cells targeting mutant proteins (SLC3A2, KIAA0368, CADPS2 and CTSB) detected in a patient tumor in combination with interleukin two have been demonstrated to induce tumor regression (Zacharakis et al., 2018). Clinical investigations aimed at expanding anti-HER2 specific T-cells for reintroduction into patients in combination with Trastuzumab (NCT05378464) are currently actively recruiting. Moreover, genetically engineered CAR-T cells have been developed targeting commonly expressed breast cancer antigens such as HER2 (Szoor et al., 2020), EGFR (Liu et al., 2019), EpCAM (Osta et al., 2004), AXL (Zhao et al., 2020a) and c-MET (Tchou et al., 2017). Unfortunately, CAR-T cells are met with the same challenges as regular T-cells where the tumor microenvironment may not be permissive for their survival or activity, presence of inhibitory immune cells as well as lack of homing cues (Li et al., 2020). Studies have addressed this by stimulating STING using 2′3′-cGAMP or DMXAA which improved trafficking and CAR-T cells targeting HER2 in the TIME of breast cancer (Xu et al., 2021). Tumor heterogeneity, whereby some tumor cells may lack the antigen targeted by the CAR or loss of MHC class 1 which is required for recognition of the tumor cell may hinder the impact of CAR-T cells (Li et al., 2020; Luo et al., 2022).

Development of CAR-natural killer (CAR-NK) cells targeting breast cancer tumor antigens (Liu et al., 2020; Portillo et al., 2021) harbours several advantages over CAR-T cells, importantly not requiring MHC matching (Khawar and Sun, 2021). Therefore, CAR-NK need not be from the same patient, have a low risk of graft vs. host disease and do not require MHC to be expressed on the tumor cell (Khawar and Sun, 2021). Additionally, CAR-NKs demonstrate enhanced tumor killing potential compared to CAR-T cells coupled with reduced risk of cytokine release syndrome, which is acute systemic inflammation that could be life threatening (Shimabukuro-Vornhagen et al., 2018; Khawar and Sun, 2021; Portillo et al., 2021).

CAR-macrophages (CAR-Ms) have similar advantages to CAR-NKs with the addition of macrophages having better capacity at infiltrating the tumor (Klichinsky et al., 2020). CAR-M targeting HER2, decreased tumor burden and increased overall survival of tumor bearing mice (Klichinsky et al., 2020). Moreover, they induced expression of inflammatory cytokines capable of converting tumor resident M2 macrophages to the M1 phenotype and boosted anti-tumor T-cell activity (Klichinsky et al., 2020). To our knowledge, no clinical trials are currently underway assessing CAR-NKs or CAR-Ms in breast cancer.

Overall, while immune cold tumors represent a difficult to treat tumor and are associated with poor patient outcome, there is encouraging data suggesting that reactivation of the tumor immunity cycle via STING agonists or targeting ENPP1 or CD73 is sufficient to increase immune cell infiltration into the tumor especially in combination with either chemotherapy or radiotherapy. Targeting hypoxia both directly (through HIF inhibition) and indirectly (through RTKs) may relieve T-cell exclusion therefore, sensitizing patients to traditional therapies and immunotherapies. Epigenetic rewiring of the tumor as well as the immune cells themselves may increase their activation and infiltration into the tumor. Oncolytic virus-based therapies may induce hotness in the tumor due to DAMP and PAMP generation. Finally, to increase the number of anti-tumor immune cells at the tumor generation of genetically engineered T-cells, NKs and macrophages may help reshape the TIME (Figure 3).

Activation of the nuclear factor kappaB (NFΚB) inflammatory pathway is observed predominantly in ER negative tumors through activation of RTKs and/or cellular stress (Biswas et al., 2004; Frasor et al., 2015; Wang et al., 2015). Moreover, active NFKB is correlated with larger, more advanced tumors and correlates with lung and brain metastasis (Wang et al., 2015). NFKB activity is associated with expression of inflammatory cytokines that promote Th2 polarization and maintaining a cancer stem cell population that facilitates oncogenesis (Liu et al., 2010; Wang et al., 2015; Zhang et al., 2022). Inhibition of NFKB delayed tumor onset and abrogated tumor progression in a PyMT driven model of breast cancer (Connelly et al., 2011). Recent work demonstrated that the reticuloendotheliosis viral oncogene homolog A (RELA) subunit of NFKB recruits cat eye syndrome chromosome region candidate 2 (CECR2), an epigenetic factor that induces gene expression of several inflammatory cytokines including colony stimulating factor 1 (CSF1) (Zhang et al., 2022). This interaction was important in recruitment of macrophages as well as their polarization to the M2 state in the lung metastatic niche (Zhang et al., 2022). Targeting CECR2 in mice harbouring orthotopically transplanted 4T1 cells reduced metastasis through reduced M2 macrophage levels and concomitant increase in cytotoxic T-cell activity (Zhang et al., 2022).

NFKB components have been shown to interact with STAT3 and coregulate inflammatory oncogenic pathways in breast cancer (Fan et al., 2013). In addition to STAT3’s ability to regulate the immune microenvironment, STAT3 transcriptional targets regulate anti-apoptotic pathways, EMT and expression of metalloproteinases (MPPs) that act in concert to regulate growth and metastasis (Ma et al., 2020). Interestingly, mammary specific knockout of STAT3 in the PyMT model did not impact nascent tumors, but they were eventually eliminated by an active innate and adaptive immune response (Jones et al., 2016). This finding contends that targeting STAT3 or one of its essential transcriptional targets may be sufficient to induce a robust anti-tumor immune response that leads to complete clearance of tumors. It is noteworthy that our STAT3 knockout was restricted to the mammary epithelium (Jones et al., 2016), future studies using STAT3 inhibitors (Zou et al., 2020) to address the role of STAT3 in other cell types of the TIME are essential for understanding the full role of STAT3 in immunosuppression. To this end, STAT3 activation (assessed using expression of phosphorylated STAT3) in astrocytes of the brain microenvironment supported invasive growth of breast-to-brain metastasis in part through reduced activation of T-cell activity and upregulation of CHI3L1 expression (Priego et al., 2018; Dankner et al., 2022). TTI-101, a STAT3 inhibitor, is currently in early phase clinical trials (NCT03195699 and NCT05384119). Of note, active STAT3 downstream interleukin (IL) six was demonstrated to bind estrogen response elements and promote gene expression independent of the ER or its coactivator forkhead box A1 (FOXA1) promoting metastasis and resistance to anti-estrogen-based therapy (Siersbaek et al., 2020).

Macrophages are the most abundant immune cell type in the TIME, thus representing an attractive therapeutic target. Work from the PyMT mouse model demonstrated that as the tumor progresses, there is a decrease in mammary resident macrophages and an increase in tumor associated macrophages (Attalla et al., 2021). Macrophages are recruited to the tumor site due to expression of inflammatory cytokines, especially CSF1 and chemokine ligand (CCL) 2 (Lin et al., 2001; Franklin et al., 2014; Linde et al., 2018; Neubert et al., 2018). The importance of CSF1 was demonstrated using CSF1 knockout PyMT mice which exhibited delayed disease progression as well as prevented pulmonary metastasis through loss of vascular endothelial growth factor (VEGF) and epidermal growth factor (EGF) expression (Lin et al., 2001). Moreover, macrophages promote tumor cell migration and metastatic dissemination through secretion of EGF and Wnt1 (Wyckoff et al., 2007; Linde et al., 2018). Macrophages, especially those M2 polarized, establish the immunosuppressive microenvironment in immune hot subtypes through expression of cytokines that promote Th2 T-helper cell polarization and inhibition of cytotoxic T-cell function (Liu et al., 2021) (Figure 4A). Moreover, macrophages have been demonstrated to expresses PDL1 to promote exhaustion of T-cells (Ahmed et al., 2020). Inhibition of CSF1 receptor (CSFR) by small molecule (BLZ945) has demonstrated anti-tumor effects in glioma and breast cancer models through loss of M2 polarized macrophages and enhancement of T-cell recruitment to the tumor (Pyonteck et al., 2013; Strachan et al., 2013). Phase II clinical trial for TNBC combining Pembrolizumab (anti-PD1), radiation and Axatilimab (anti-CSFR) is currently ongoing (NCT05491226).

Another macrophage marker that has demonstrated pre-clinical interest is the “do not eat me” signaling axis consisting of signal regulatory protein-alpha (SIRPα) expressed on the macrophage (and other phagocytes) and CD47 expressed on cancer cells (Chen et al., 2022). CD47 is transcriptionally regulated by inflammation, though NFKB and TNF (Betancur et al., 2017) as well as hypoxia, through HIF1α (Zhang et al., 2015). CD47 has also been demonstrated to be an essential breast cancer stem cell marker (Zhang et al., 2015; Kaur et al., 2016; Chen et al., 2022). Engagement of CD47 with SIRPα inhibits the phagocytic capacity of macrophages (Chen et al., 2022). Antibody neutralization of CD47, induced a shift towards M1 macrophages and enhanced antigen presentation (Latour et al., 2001; Zhang et al., 2016; Chen et al., 2022). CD47 blockade also synergized with trastuzumab through enhancing antibody dependent cellular phagocytosis (Tsao et al., 2019). Multiple anti-CD47 antibodies have been developed and are under clinical assessment (Chen et al., 2022).

Neutrophils are the most common cell type in human blood and have recently gained significant research interest. The neutrophil to lymphocyte ratio in many solid tumors is correlated with patient survival with a higher ratio being associated with worse patient survival (Bartlett et al., 2020; Gago-Dominguez et al., 2020). Neutrophils are recruited to the tumor (site of injury) through secretion of inflammatory cytokines such as tumor growth factor β (TGFβ), CXCL8 (also known as IL-8), CXCL1, CXCL2 as well as IFNs (SenGupta et al., 2021). Furthermore, neutrophils secrete MMP9 (SenGupta et al., 2021) a prominent enzyme that degrades the extracellular matrix, promoting cancer progression and tumor cell dissemination (Owyong et al., 2019). Moreover, expression of chemokine ligand CCL20 by the tumor cell induced expression of PDL1 on neutrophils which reinforces an immunosuppressive microenvironment via exhaustion of T-cells (Kwantwi et al., 2021).

Strikingly, neutrophils have been demonstrated to undergo a process known as NETosis, during which a neutrophil secretes its DNA into a weblike structure (Vorobjeva and Chernyak, 2020). NET formation is promoted by cancer derived factors such as granulocyte-CSF (G-CSF), CXCL5, CXCL6 (Ronchetti et al., 2021). NETosis has been further demonstrated to contribute to cancer progression and metastasis (Ronchetti et al., 2021). Interestingly, NETs have been demonstrated to sequester circulating tumor cells to ensure their successful metastatic colonization (Cools-Lartigue et al., 2013). Depletion of neutrophils reduced metastatic dissemination via reliving inhibition of cytotoxic T-cell activity (Coffelt et al., 2015). Targeting NETosis using DNase I to dissolve the extracellular DNA or inhibiting protein arginine deiminase 4 (PAD4), an enzyme required for NET formation, using small molecules inhibited cancer growth, metastasis and recurrence in models of breast cancer (Park et al., 2016; Ruiz-Fernandez de Cordoba et al., 2022). Moreover, inhibition of PAD4 using GSK484 sensitizes TNBC xenografts to radiotherapy (Wei et al., 2021) and synergizes with anti-PD1/anti-CTLA dual therapy (Teijeira et al., 2020). Interestingly, NETs were demonstrated to encapsulate tumors cells, preventing physical contact between T-cells and tumor cells (Teijeira et al., 2020; Shinde-Jadhav et al., 2021) and thus may promote T-cell exclusion from the tumor (Figure 4B).

Accumulating evidence suggests that targeting pathway associated with inflammation or cells accumulated at the primary tumor site due to the inflammation can act in synergy with other already approved immunotherapies to lead to collapse of the primary tumor or inhibition of metastasis.

Driven by the idea of combinational therapies, bispecific antibodies (bsAbs) that simultaneously bind multiple tumor specific antigens and/or immune cells leading to an anti-tumor immune response have been generated (Dees et al., 2021). Most bsAbs function as CD3⁺ T-cell engagers, that direct T-cells to the tumor epithelium to elicit their anti-tumor function (Dees et al., 2021). Therefore, bsAbs that target T-cells and commonly expressed breast cancer cell surface antigens such as HER2 (Sen et al., 2001), EGFR (Stamm et al., 2019), epithelial cell adhesion molecule (EpCAM) (Kubo et al., 2018), trophoblast cell-surface antigen 2 (Trop2) and carcinoembryonic antigen-related cell adhesion molecule 5 (CEACAM5) (Chang et al., 2017) have been developed to specifically target tumors expressing these markers. Prominently, CD3/HER2 targeting bsAb (Runimotamab) is under clinical investigation for HER2+ breast cancer (NCT03448042). In addition to T-cell engagers, novel natural killer (NK) cell redirectors have been generated through the use of CD16 (Fc receptor expressed on natural killers) to target the breast cancer antigen, mesothelin known as MesobsFab (Del Bano et al., 2019). MesobsFab promoted recruitment and penetration of NK cells into spheroids as well as demonstrating anti-tumor effects in humanized mice (Del Bano et al., 2019).

It is important to note that the success of these antibodies is highly dependent on T-cells being capable of fully infiltrating the tumor. A cold TIME or restriction of T-cells to the margin or stroma may dampen the effect of these antibodies and render the tumors resistant. Efforts aimed at improving the immunogenicity of tumors and immune cell trafficking into the tumor (discussed in “Approaches to Treat Immune Cold Tumors”) may enhance the effects of these therapies. Additionally, bsAbs have demonstrated to work additively with anti-PD1 treatment (Chang et al., 2017) especially because some bsAbs have been demonstrated to induce expression of IFNγ (Sen et al., 2001) which may increase expression of PDL1, suggesting that bsAbs monotherapy may not be suffice due to T-cell exhaustion.