Immunologically relevant effects of radiation therapy on the tumor microenvironment

Abstract

Focal radiation therapy (RT) has been successfully employed to clinically manage multiple types of cancer for more than a century. Besides being preferentially cytotoxic for malignant cells over their non-transformed counterparts, RT elicits numerous microenvironmental alterations that appear to factor into its therapeutic efficacy. Here, we briefly discuss immunostimulatory and immunosuppressive microenvironmental changes elicited by RT and their impact on tumor recognition by the host immune system.

Introduction

Focal radiation therapy (RT) has been a pillar in modern clinical cancer management for more than a century, both as a curative intervention (e.g., in patients with squamous cell carcinoma or anal carcinoma) and as a palliative measure (e.g., for the control of painful metastatic lesions to the bones from a variety of cancers) (1, 2). While recent technological advances including (but not limited to) stereotactic body RT (SBRT) and intensity-modulated RT (IMRT) have enabled an ever more precise delivery of increasing RT doses to specific anatomical location with improved sparing of adjacent healthy tissues (3, 4), the therapeutic effects of RT mainly build on the differential ability of transformed versus normal cells to repair RT-elicited macromolecular damage (5, 6). Indeed, at odds with their healthy counterparts, malignant cells generally exist in an altered biological state that is often associated with the constitutive activation of stress-response pathways, which ultimately limits their capacity to adapt to additional alterations of homeostasis (7, 8). More specifically, RT mediates cytostatic/cytotoxic effects by eliciting oxidative damage to macromolecules, especially (but not solely) DNA, either directly or – more so – upon the ionization of water molecules that in turn behave as oxidants (5, 6, 9–12). This is generally associated with the activation of the so-called DNA damage response (DDR) (13), a complex series of molecular events that can have diametrically opposed consequences on cellular homeostasis: (1) if DNA damage is sufficiently contained, the DDR promotes repair in the context of a temporary cell cycle arrest, followed by cell cycle re-entry and resumed proliferation; (2) if DNA is damaged beyond repair, the DDR either elicits a permanent proliferative arrest commonly known as cellular senescence (14) or triggers regulated variants of cell death (15), measures that are aimed at protecting global organismal homeostasis in the context of compromised cellular fitness in irradiated tissues (16).

Notably, most solid tumors exhibit a considerable degree of intratumoral heterogeneity (17), implying that not all malignant cells are equally sensitive to standard clinical RT schedules. Moreover, many non-transformed cells are also exposed to RT during irradiation as part of the integral dose exposure. These cells encompass all components of the tumor microenvironment (TME) and adjacent tissues, as well as peripheral blood mononuclear cells (PBMCs) and granulocytes circulating within the target volume while the beam is on (18). Some of these cells, especially circulating lymphocytes, are exquisitely radiosensitive and hence can be exposed to lethal RT doses (19), which explains the lymphopenic side effects associated with some RT procedures (e.g., spleen exposure during RT for the management of gastric carcinoma) (20). This is particularly relevant in the clinical setting as post-RT lymphopenia and/or the consequent increase in circulating neutrophil-to-leukocyte ratio (NLR) have been linked with poor prognosis in multiple cohorts of patients with cancer treated with RT (21, 22). Conversely, most terminally differentiated cells (e.g., skeletal muscle cells, neurons) are relatively radioresistant and thus survive RT employed at dose and fractionation schedules commonly employed in cancer patients (23).

Importantly, both failing and successful adaptation to the effects of RT (and exogenous stress in general) (16) in the TME and adjacent tissues is associated with the activation of numerous signaling pathways involved in tissue remodeling, inflammation and immunity, many of which ultimately factor in the therapeutic effects of RT (24–26). Here, we provide an overview of immunostimulatory and immunosuppressive effects of focal RT and their impact on the recognition and elimination of cancer cells by the host immune system.

Immunostimulatory effects of RT

When used according to specific dose and fractionation schedules, focal RT can exert potent immunostimulatory effects, including (but not limited to): (1) the upregulation of genes encoding potentially immunogenic tumor-associated antigens (TAAs), (2) the overexpression of MHC class I molecules on the surface of irradiated cells, (3) the exposure of NK cell-activating ligands (NKALs), (4) the activation of immunogenic cell death (ICD), and (5) the upregulation of death receptors (DRs).

TAAs and MHC Class I molecules.

Three daily RT fractions of 8 Gy each promoted the expression of 4 proteins containing potentially immunogenic mutations in mouse triple negative breast cancer (TNBC) 4T1 cells, and the antigenic peptides of three of them – namely, cullin associated and neddylation disassociated 1 (CAND1), DEXH (Asp-Glu-X-His) box polypeptide 58 (DHX58) and adhesion G protein-coupled receptor F5 (ADGRF5) – were able to elicit CD8⁺ (CAND1 and DHX58) and CD4⁺ (ADGRF5) T cell responses that could be further boosted by RT. In line with this notion, neo-epitopes from all these proteins induced interferon gamma (IFNG, best known as IFNγ) responses in 4T1-draining lymph nodes, especially when combined with RT. Moreover, vaccination of mice with CAND1-derived, DHX58-derived and ADGRF5-derived epitopes improved the control of both irradiated 4T1 primary lesions as well as non-irradiated lung metastases, an effect that pointed to at least some degree of antigen spread (27). Along similar lines, focal RT delivered as a sub-cytotoxic single dose of 10 or 20 Gy has been shown to induce the expression of various TAAs of the carcinoembryonic antigen protein family as well as of mucin 1, cell surface associated (MUC1) in a panel of 17 human cancer cell lines (28).

Mouse glioma GL261 cells treated with a single dose of 4 Gy exhibited accrued levels of MHC Class I molecules on their surface, an effect that could be attributed to increased MHC Class I heavy chains and beta-2-microglobulin (B2M) expression (29). In line with this notion, whole-brain RT delivered to glioblastoma-bearing mice in 2 fractions of 4 Gy each resulted in recovered MHC Class I expression at the invading edge of the tumor and restored disease control by peripheral, vaccine-elicited CD8⁺ T cell responses (29). Along similar lines, human melanoma MelJuSo cells exposed to single RT dose of 1 Gy, 7 Gy and 25 Gy manifested a dose-dependent increase in surface-exposed MHC class I molecules, as did mouse colorectal carcinoma (CRC) MC38 cells exposed to a single RT of 2 Gy, 8 Gy and 10 Gy (30). Interestingly, such an effect (1) was accompanied by alterations in the MHC Class I immunopeptidome, and (2) could also be documented in vivo, in healthy mouse kidneys exposed to a single RT fraction of 25 Gy (30). Similar findings have been obtained in a panel of human cancer cell lines irradiated with a single dose of 8 Gy, an effect that could be enhanced by the pre-administration of the epigenetic modifier decitabine (31), as well as in MC38 cells responding to 10 Gy in a single fraction (32). Recent findings point to the transactivator NLR family CARD domain containing 5 (NLRC5) as to a transcriptional regulator involved in the ability of RT to promote MHC Class I expression (33). Moreover, it has recently been suggested that RT may induce the presentation of antigens generated during the pioneer round of translation from mRNAs containing nonsense mutations (34). However, whether the latter mechanism has a role in RT-driven anticancer immunity remains to be investigated.

NKALs.

Human lung adenocarcinoma NCI-H23 and A549 cells treated with a single RT dose of 8 Gy exhibited accrued levels of multiple NKALs specific for NK cell-activating receptor killer cell lectin like receptor K1 (KLRK1; best known as NKG2D), including MHC class I polypeptide-related sequence A (MICA), MICB, UL16 binding protein 1 (ULBP1), ULBP2 and ULBP3, an effect that could be exacerbated by histone deacetylase (HDAC) inhibitors (35). Moreover, a panel of mouse and human glioblastoma cells exhibited a dose-dependent increased in multiple NKALs upon irradiation with a single RT fraction of 2 Gy, 4Gy or 8 Gy, culminating with increased susceptibility to natural killer (NK)-mediated cytotoxicity (36). Similar results have been obtained by irradiation a panel of human cancer cell lines of multiple histological derivation with 20 Gy in a single fraction (37). These observations suggest that at least in some cancer cell lines, the upregulation of MHC Class I molecules elicited by RT (which potently inhibit NK cell activity) (38) may be effectively counteracted by an increased exposure of NKALs. That said, a fraction of CD8⁺ T cells express NKG2D (39), and the upregulation of NKALs by RT has been shown to stabilize the immune synapse between effector CD8⁺ T cells and cancer cells in mice treated with a CTLA4 blocker (40), suggesting that T cells may also contribute to tumor control by RT downstream of NKAL upregulation.

ICD and cytokines.

Mouse CRC CT26 cells, as well as mouse hormone receptor (HR)⁺ breast carcinoma TS/A cells exposed to γ irradiation in vitro can be effectively used as vaccines upon inoculation in immunocompetent syngeneic animals in the absence of any immunological adjuvant (41, 42), demonstrating that RT can elicit bona fide ICD (43). This reflects RT cytotoxicity as well as its ability to promote the emission of various adjuvant-like signals commonly known as damage-associated molecular patterns (DAMPs). Such signals encompass (1) the exposure of endoplasmic reticulum (ER) chaperones such as calreticulin on the plasma membrane, where it serves as an “eat-me” signal for dendritic cell (DC) precursors upon binding to LDL receptor-related protein 1 (LRP1, best known as CD91) (44–46); (2) the autophagy-dependent secretion of ATP into the extracellular microenvironment, where it operates both as a chemoattractant, upon binding to purinergic receptor P2Y2 (P2RY2), and as an immunostimulatory signal, upon binding to purinergic receptor P2X 7 (P2RX7) (44, 47–49); (3) the release of the non-histone chromatin-binding protein high mobility group box 1 (HMGB1), which mediates immunostimulatory effects upon binding to advanced glycosylation end-product-specific receptor (AGER, best known as RAGE) and Toll-like receptor 4 (TLR4) (44, 50, 51). Interestingly, abundant DAMP emission in response to irradiation has also been documented with carbon ions (52). Whether charged particles are superior to photon-based RT at eliciting ICD, however, remains to be formally demonstrated.

RT is also a potent inducer of type I interferon (IFN) secretion, which is not only a critical component of ICD signaling, but also a key factor for RT to elicit potent tumor-targeting immune responses with systemic outreach (i.e., so-called abscopal responses) (42, 53–55). Interestingly, type I IFN secretion by irradiated malignant cells (and the consequent elicitation of abscopal responses) appears to be elicited by the accumulation of mitochondrial DNA (mtDNA) in the cytosol of irradiated cells (42, 56), culminating with the activation of cyclic GMP–AMP synthase (CGAS) signaling (57). This process is therefore particularly sensitive to mechanisms that limit either mtDNA availability, such as the dose-dependent upregulation of the mtDNA-degrading enzyme three prime repair exonuclease 1 (TREX1) (54) and the activation of autophagy (which degrades permeabilized, mtDNA-releasing mitochondria (42), or CGAS signaling, such as the activation of apoptotic caspases (which cleave CGAS) (58, 59). Interestingly, irradiated cancer cells can also elicit CGAS activation in dendritic cells (57), at least in part upon the delivery of extracellular vesicles containing CGAS-activating double-stranded DNA molecules (60). Importantly, RT has been shown to effectively restore proficient type I IFN signaling in preclinical lung cancer models, thus restoring sensitivity to ICIs (61). Supporting a role for type I IFN signaling in the clinical efficacy of RT, elevations in circulating interferon beta 1 (IFNB1) about two weeks after completion of palliative RT plus an immune checkpoint inhibitor (ICI) specific for cytotoxic T lymphocyte-associated protein 4 (CTLA4) were as strong predictor of response in a cohort of patients with non-small cell lung carcinoma (62).

Mouse 4T1, 67NR and 4T07A mammary carcinoma cell lines with different metastatic potential as well as human breast cancer HTB20 cells (but not non-transformed breast epithelial MCF10A cells) secreted chemokine (C-X-C motif) ligand 16 (CXCL16) in a dose-dependent fashion upon exposure to a single RT dose of 2 Gy, 6 Gy, and 12 Gy (63), exemplifying the notion that RT can promote the release of other cytokines beyond type I IFN (see below). Pointing to a role for CXCL16 signaling in the therapeutic effects of RT, 4T1-bearing mice lacking the CXCL16 receptor chemokine (C-X-C motif) receptor 6 (CXCR6) exhibited not only limited tumor infiltration by CXCR6⁺ T cells upon focal RT with a single dose of 12 Gy, but also reduced tumor control (63).In summary, RT stands out as a potent inducer of ICD, a particularly immunogenic form of cell death coupled to DAMP emission and cytokine secretion with profound clinical relevance.

Death receptors.

4T1 cells exposed to 3 RT fractions of 8 Gy expressed increased levels of the DRs Fas cell surface death receptor (FAS) and tumor necrosis factor receptor superfamily, member 10b (TNFRSF10B, best known as DR5 or TRAIL-R2) on their membrane, which along with an increase in surface-exposed MHC Class II molecules rendered them more susceptible to lysis by cytotoxic CD4⁺ T cells (27). Similar results have been obtained with human pancreatic adenocarcinoma Capan 2 cells engineered to express a model antigen (sLeA) upon exposure to a single dose of 2 Gy in vitro (64), as well as with human CD19⁺ acute lymphoblastic leukemia NALM cells irradiated with 1 Gy in vitro or in vivo (upon inoculation in immunodeficient hosts) (65). Importantly, in both these settings, low-dose RT delivered either focally (64) or as total body irradiation (TBI) (65) extended the therapeutic potential of TAA-targeting CAR T cells. However, while in the former setting such an increased response derived from TAA-independent, DR-mediated CAR T cells cytotoxicity (64), in the latter it reflected an improved expansion and persistence of CAR T cells (65). These observations delineate DR upregulation on malignant cells as yet another therapeutically relevant immunostimulatory effect of RT.

Immunosuppressive effects of RT

Depending on a number of factors, including (but not limited to) specifical experimental or clinical setting as well as dose and fractionation regimen, focal RT can also elicit immunosuppressive effects such as: (1) the secretion and activation of transforming growth factor beta 1 (TGFB1), inhibin subunit beta A (INHBA) and other cytokines/chemokines; (2) the accrued exposure on the plasma membrane of immunosuppressive factors such as CD274 (best known as PD-L1), CD47 and 5’-nucleotidase ecto (NT5E, best known as CD73); (3) vascular disruption coupled to decreased oxygen tension; and (4) chronic, indolent type I IFN signaling.

TGFB1, INHBA and other cytokines.

TGFB1 and INHBA are cytokines from the TGFβ family with potent immunosuppressive activity, at least in part reflecting their capacity to recruit CD4⁺CD25⁺FOXP3⁺ regulatory T (T_REG) cells to the irradiated TME and to generate a potent fibrotic reaction (24, 66). In immunocompetent mice bearing bilateral MC38 tumors, TGFβ neutralization with a monoclonal antibody synergized with focal RT delivered to a single tumor in 3 fractions of 8 Gy coupled with systemic TNF receptor superfamily member 9 (TNFRSF9, best known as CD137 or 4–1BB agonism and programmed cell death 1 (PDCD1, best known as PD-1) blockage, ultimately enabling systemic disease eradication in a majority of mice (67). Importantly, this was paralleled by increased recruitment of cytotoxic CD8⁺ T cells to non-irradiated tumors and elevations in circulating IFNγ in the absence of remarkable toxicity for the hosts (67). Similar results have been obtained in immunocompetent mice bearing bilateral TS/A lesions or unilateral 4T1 lesions (which metastasize rapidly to the lungs) that were treated with 6 RT fraction of 5 Gy each at one disease site, plus a TGFβ-blocking antibody (68), as well as in immunocompetent mice bearing syngeneic SCC VII head and neck cancers, TC-1 lung carcinomas or LL2 lung carcinomas receiving a single RT dose of 8 Gy in combination with transforming growth factor beta receptor 1 (TGFBR1) or TGFBR2 inhibitors (69). Moreover, administration of the TGFβ blocker fresolimumab along with hypo-fractionated RT was well tolerated in a cohort of patients with metastatic breast carcinoma, and subjects receiving fresolimumab at the 10 mg/Kg dose had a favorable systemic immune response and experienced longer median overall survival than patients treated with fresolimumab 1 mg/Kg (70). Interestingly, TGFβ neutralization has been shown to exacerbate the ability of RT to elicit the secretion of INHBA by a panel of mouse and human breast cancer cell lines (71, 72). In line with this notion, Inhba deletion from 4T1 cells improved the ability of RT combined with TGFβ neutralization to control 4T1 in vivo, correlating with a strong reduction in tumor-infiltrating T_REG cells (71, 72). Of note, T_REG cells as well as immunosuppressive myeloid cells can also be recruited to the irradiated TME upon the secretion of C-C motif chemokine ligand 2 (CCL2) (73). Accordingly, the therapeutic effects of a single RT fraction of 7.5 Gy against mouse TC-1 lung carcinomas growing in immunocompetent hosts were improved in Ccr2^−/− mice (which lack the main CCL2 receptor) as compared to wild-type mice (73). In summary, multiple cytokines secreted by irradiated cancer cells support the establishment of an immunosuppressive TME that limit therapeutic responses to RT.

PD-L1, CD47 and CD73.

One single RT dose of 12 Gy delivered to BALB/c mice bearing TUBO mammary carcinomas promoted an accrued exposure of PD-L1 on the surface of malignant cells, an immunosuppressive effect that could be effectively offset by the co-administration of ICIs targeting PD-1 (74). Along similar lines, mouse Res177, Res499, and B16F10 melanomas established in immunocompetent, syngeneic hosts and treated with RT in a single fraction of 20 Gy plus a CTLA4-targeting ICI evaded immune control via PD-L1 upregulation, and this could be effectively limited by PD-1 blockage (75). Accordingly, cancer cells from patients with melanoma failing to respond to SBRT plus a CTLA4 blocker in the context of a clinical trial exhibited high levels of PD-L1 and signs of persistence CD8⁺ T cell exhaustion (75). Of note, PD-L1 upregulation on the surface of malignant cells from irradiated tumors have also been shown to reflect CD8+ T cell infiltration and IFNγ secretion (76).

Exposure of human HCT116 and HT29 CRC cells to a single RT dose of 8 Gy elicited not only increased levels of membrane-exposed PD-L1, but elevations of the “don’t eat-me signal” CD47, an effect that was ascribed to ATR serine/threonine kinase (ATR) signaling as elicited by DNA damage (77). In line with these observations, immunocompetent mice bearing bilateral MC38 lesions experienced superior tumor control at both disease site upon irradiation of one lesion with a single RT fraction of 8 Gy combined with systemic PD-1 and CD47 blockage (77). Similar results have been obtained in wild-type mice bearing bilateral small cell lung carcinoma (SCLC) lesions receiving RT in one fraction of 5 Gy or 5 fractions of 4 Gy each to a single lesion combined with systemic CD47 blockage (78, 79). Finally, RT has been shown to promote the dose-dependent upregulation of CD73 – a plasma membrane enzyme that facilitates the conversion of extracellular ATP into the immunosuppressive metabolite adenosine (80) – in a panel of human and mouse breast cancer cell lines (81). Supporting the therapeutic relevance of this pathway, CD73 blockage with a monoclonal antibody improved the capacity of RT delivered as a single dose of 20 Gy along with systemic CTLA4 inhibition to elicit abscopal responses in immunocompetent mice bearing bilateral TS/A lesions (81). These observations delineate PD-L1, CD47 and CD73 as potential targets to limit RT-driven immunosuppression in support of superior therapeutic responses.

Vascular disruption and hypoxia.

Focal RT is relatively cytotoxic to endothelial cells, which – combined with an increase in interstitial pressure due to RT-driven inflammatory reactions – can further limit oxygen availability in (at least some regions of) the TME (82). This is relevant not only because pre-existing as well newly formed hypoxic TME regions are particularly resistant to the cytotoxic effects of RT (83), but also because vascular disruption and hypoxia promote local immunosuppression by a number of mechanisms. These include (but are not limited to): (1) the downregulation of MHC class I molecules on the surface of malignant cells (84); (2) the upregulation of immunosuppressive factors such as PD-L1, vascular endothelial growth factor A (VEGFA) and colony stimulating factor 1 (CSF1) (85–87); (3) the activation of bona fide CD4⁺CD25⁺FOXP3⁺ T_REG cells and immunosuppressive DC populations (88); (4) the inhibition of CD8⁺ T cell effector functions coupled to the conversion of exhausted T (T_EX) cells into T_REG-like cells (89, 90); and (5) the repolarization of TAMs towards an immunosuppressive (M2-like) phenotype (91, 92). In line with these observations, various strategies to limit hypoxia have been shown to increase the efficacy of RT in preclinical tumor models, although the actual contribution of anticancer immunity to this effect remains to be formally elucidated. Despite this unknown, the ability of RT to kill cancer cells in well-perfused tumor regions promotes the selection of hypoxic areas of the TME, where immunosuppression further limits therapeutic responses.

Chronic type I IFN.

While acute, robust and resolving type I IFN responses have been shown to support clinically relevant anticancer immunity as driven by RT (54, 62), intratumoral type I IFN levels as well as genetic signatures of type I IFN signaling have also been associated with treatment resistance in various cohorts of cancer patients (58, 93). The reasons for such an apparent discrepancy remain to be completely explained. Suboptimal and chronic type I IFN signaling has been shown to support the accumulation of cancer stem cell (CSC)-like cells that enable accelerated tumor progression in immunocompetent mouse models of fibrosarcoma exposed to ICD-chemotherapeutics (94), pointing to the intensity and duration of type I signaling as to key determinant of ultimate therapeutic outcome. In line with this possibility, chronic type I IFN signaling has been consistently associated with CD8⁺ T cell exhaustion and immunotherapy resistance (95–97). Taken together, these observations suggest efficient immunostimulation by RT should induce robust and resolving (as opposed to indolent and chronic) type I IFN responses. Whether dose and fractionation can affect the magnitude and kinetic of type IFN signaling as driven by RT remains to be defined.

Conclusions

It is now clear that RT imposes immunologically relevant alterations to the TME that (at least in some settings) influence local and/or systemic disease control (Figure 1). In line with this notion, RT has been shown to synergize with numerous immunotherapeutic or immunomodulatory agents in a variety of immunocompetent mouse tumor models. These agents include (but are not limited to): (1) ICIs targeting CTLA4, PD-1, PD-L1, TGFβ or INHBA (54, 68, 71, 74, 98); (2) immunostimulatory monoclonal antibodies specific for TNF receptor superfamily member 9 (TNFRSF9; best known as CD137 or 4–1BB), TNFRSF4 (best known as OX40) or CD40 (67, 99); (3) recombinant cytokines that stimulate lymphoid cells, such as IL15 (100); and (4) other immunomodulatory molecules with diverse mechanisms of action, such as Toll-like receptor 7 (TLR7) agonists, CDK4/6 inhibitors and 5’-nucleotidase ecto (NT5E; best known as CD73) blockers (81, 101, 102). However, combining RT with IT in the clinic has been less straightforward than expected. Indeed, while some prospective randomized clinical trials have demonstrated a survival advantage for patients receiving IT in addition to RT-based standard of care (SOC) over SOC alone (103), others failed to document such a benefit (104–106). While many reasons can be invoked to explain the lack of synergy between RT and IT in some of these studies, it is tempting to speculate that the use of RT according to conventional doses, fractionation, schedules and target volumes may not always be optimal to preserve the immune fitness of the host in support of IT responsiveness (107). Additional work is required to delineate improved RT approaches that optimally cooperate with IT in patients with cancer.

Figure 1. Main immunomodulatory mechanisms elicited by RT in the TME.

Focal radiation therapy (RT) has been shown to impose various immunologically relevant changes to the tumor microenvironment (TME) that may influence local and systemic cancer control. A. On the one hand, focal RT has been shown to (1) render malignant cells more immunogenic upon the derepression of genes encoding tumor-associated antigens (TAAs); (2) cause a potently immunostimulatory form of cell deaths coupled with the acute release of danger signals, cytokines and chemokines; (3) favor the recognition of malignant cells by the immune system via the upregulation of MHC Class I molecules or natural killer (NK) cell-activating ligands (NKALs) on the cell surface; and (4) increase the sensitivity of cancer cells to immune effectors as consequence of death receptor (DR) upregulation. B. On the other hand, focal RT has been associated with the establishment of local immunosuppression as a consequence of: (1) three prime repair exonuclease 1 (TREX1) upregulation and hence degradation of potentially interferogenic double-stranded DNA (dsDNA) species in the cytosol of irradiated cells; (2) transforming growth factor beta 1 (TGFB1, best known as TGFβ) and inhibin subunit beta A (INHBA) signaling, resulting in regulatory T (T_REG) recruitment and fibrosis; (3) CD274 (best known as PD-L1) upregulation on the surface of malignant cells; (4) repolarization of tumor-associated macrophages (TAMs) towards an M2-like, immunosuppressive phenotype; and (5) endothelial cell cytotoxicity coupled with accrued interstitial pressure, culminating with dynamic changes in oxygenation and potential selection of hypoxic tumor areas with accrued immunosuppressive features. A variety of therapeutic strategies impinging on these and other immunologically relevant microenvironmental alterations elicited by RT has been shown to ameliorate the therapeutic profile of RT in immunocompetent tumor models.

Summary points.

• Focal radiation therapy (RT) is an established treatment modality for multiple tumor types

• RT mediates cytostatic/cytotoxic effects preferentially on malignant cells, largely reflecting their reduced capacity to repair RT-induced macromolecular damage compared to normal cells

• RT also causes microenvironmental alterations that may factor into its therapeutic activity

• Immunostimulatory mechanisms engaged by RT, such as the activation of immunogenic cell death (ICD) coupled to the release of type I IFN in the microenvironment of irradiated tumors, may support anticancer immunity

• Conversely, RT-driven immunosuppressive pathways including TGFβ signaling and the recruitment of T_REG cells to the tumor microenvironment (TME) may favor cancer immune escape

• Immunostimulatory and immunosuppressive alterations elicited by radiation therapy in the TME may be harnessed to improve the therapeutic effects of RT