Because of the shortcomings of radiotherapy alone as a double-edged sword—having both immunogenic and immunosuppressive effects in the TME—addition of immunotherapy is a good strategy to overcome the inadequacy of radiation to mount a robust antitumor immune response. Preclinical and clinical evidence have demonstrated improved outcomes of radiotherapy in the presence of various types of immunotherapy that modulates different facets of tumor immunity.
5.1. Toll-Like Receptor Agonists
Toll-like receptor (TLR) signaling is crucial for activating dendritic cells to cross-prime effector T-cells. TLR agonists function to improve the ability of dendritic cells to present tumor antigens released from radiation cell killing. TLR9, the most extensively studied member of the TLR family, binds to unmethylated cytosine-phosphate-guanosine (CpG) oligodeoxynucleotide from bacterial DNA to induce cellular and humoral immunity. In murine model, TLR9 agonist has been shown to enhance therapeutic effects of radiation by increasing tumor-infiltration of natural killer dendritic cells, which led to fewer metastases and longer survival [61
]. Similarly, targeting TLR9 with CpG oligodeoxynucleotide improved tumor response to radiation in preclinical models [62
]. In vivo studies also demonstrated therapeutic synergy through combining TLR7/8 activation and irradiation, with combination treatment resulting in inhibition of local tumor growth and metastatic progression [66
Clinically, combination of TLR9 agonist and radiation has shown some success in treatment of lymphomas. In a phase I/II study, 15 patients with stage III-IV relapsing low-grade B-cell lymphoma were treated with intratumoral injection with CpG DNA PF-3512676 and concurrent 4 Gy radiation to a single lesion, resulting in clinical response in 4 patients and stable disease regression in two [71
]. In a subsequent phase I/II trial, injection with PF-3512676 was used in mycosis fungoides with 33% response rate, with clinical responders showing greater reduction in Tregs [72
]. There are two trials evaluating the combination of TLR7 agonist imiquimod and radiation, one for breast cancer with skin metastases (NCT01421017) and another for diffuse intrinsic pontine glioma (NCT01400672), both with pending results.
Cytokine signaling is the main mode of communication between immune cells to activate or suppress effector immune functions. Using pro-inflammatory cytokines to bolster effector cytotoxic T-cell functions can potentially overcome radiation-induced suppressive Treg accumulation. However, outcomes so far with combining cytokine therapy with radiation are modest at best.
Interleukin-2 (IL-2) is a cytokine that regulates differentiation and proliferation of T-cells into effector and memory cells when stimulated by antigens. Preclinical studies on combining IL-2 and radiation are lacking. However, building upon the evidence that radiation can augment pro-inflammatory and immunogenic changes, a phase I study of stereotactic body radiation therapy (SBRT) in conjunction with high-dose IL-2 was performed in metastatic melanoma or renal cell carcinoma. Of the 12 patients treated, 8 patients had clinically significant response, and immune monitoring revealed greater proliferation of CD4+ T-cells with effector memory phenotype [73
IL-12 is another pro-inflammatory cytokine that activates NK cells and cytotoxic CD8+ T-cells, as well as signaling differentiation of naïve CD4+ cells to T-helper 1 cells that can mediate antitumor immune response. Preclinical evaluation of combining radiation with IL-12 therapy is limited. However, one study showed that intratumoral expression of IL-12 led to increased IFN-γ levels and radiosensitizing effects [74
]. There are no existing clinical trials testing the efficacy of this combination therapy.
Interferon-α (IFN-α) has broad immunological activities that modulate tumor immunity, including activation of dendritic cells and promotion of survival and expansion of natural killer (NK) cells and cytotoxic T-cells. IFN-α increases radiosensitivity of tumor cells in early in vitro studies [75
]. A phase II trial showed improved survival in patients with resected pancreatic adenocarcinoma receiving adjuvant combination of chemoradiation and IFN-α compared to those receiving chemoradiation alone [77
]. However, IFN-α therapy is highly toxic, leading to premature closure of phase II ACOSOG Trial Z05031 due to grade ≥3 toxicity of 95% [78
]. Most recent phase III trial utilizing IFN-α in adjuvant chemoradiation for pancreatic cancer resulted in significant treatment toxicity without improvement in survival [79
Tumor necrosis factor-α (TNF-α) is a potent inflammatory cytokine that has tumoricidal properties. However, earlier use of systemic TNF-α with concomitant radiotherapy has caused significant immune-related adverse effects and low patient tolerability [80
]. TNFerade, a form of gene therapy in which human TNF-α gene controlled by a radiation-inducible promoter is delivered into cancer cells via replication-deficient adenoviral vector, has since been tested with radiotherapy in phase I/II trials with improved toxicity profile [81
]. These promising results led to a phase III multicenter randomized trial for locally advanced pancreatic cancer patients treated with concurrent fluorouracil and radiation with or without intratumoral TNFerade. Despite being safe and well-tolerated, the addition of TNFerade did not improve overall or progression-free survival in this patient cohort [86
A novel approach to cytokine therapy is to conjugate cytokines to antibodies or antibody fragments that specifically target tumor-associated antigens. The resulting class of fusion proteins, also known as immunocytokines, is capable of delivering cytokines directly to the tumor sites and avoiding systemic adverse effects that often limit the use of cytokine therapies [87
]. Several recent studies have investigated the use of IL-2 immunocytokines with radiotherapy. L19-IL2, an immunocytokine with L19 antibody targeting the EDB-domain of fibronectin that is frequently overexpressed in solid tumors, has been shown to synergize with radiation against C51 murine colon carcinoma in a CD8-dependent manner [88
]. Subsequent study by the same group demonstrated abscopal response in the non-irradiated lesions after L19-IL2 and radiation treatment to the index lesions, as well as long-lasting antitumor immunity in cured mice [89
]. These encouraging results led to a phase I clinical study of combining L19-IL2 and SBRT in oligometastatic solid tumors (NCT02086721) with pending results. There are also recent reports of success with combining NHS-IL12 immunocytokine that targets necrotic cells with radiotherapy in preclinical models. Eckert et al. first showed that radiation-induced tumor necrosis can enhance intratumoral accumulation of necrosis-targeting NHS-IL12 immunocytokine, followed by a functional study demonstrating abscopal effect and improved survival in humanized mouse model bearing rhabdomyosarcoma xenografts [90
]. While immunocytokines with TNF-α conjugates are also been tested in preclinical and clinical studies, combination of those agents with radiotherapy has yet to be studied.
5.3. Co-Stimulatory Molecules
After successful priming of tumor antigen-specific T-cells from antigen-presenting dendritic cells, co-stimulatory signaling are required to activate these T-cells to eradicate cancer cells harboring the corresponding antigens. There are two families of ligand/receptor proteins involved in T-cell co-stimulation: (1) B7/CD28 family that includes CD80/CD86 (B7-1/B7-2) ligand binding to CD28 receptor and CD275 (B7-H2) binding to CD278 (ICOS) receptor; and (2) TNF/TNF receptor family that includes ligands (CD40L, OX40L, CD70, and 4-1BBL) and its respective receptors (CD40, OX40, CD27, and 4-1BB). Intriguingly, cancer cells can evolve to hinder these essential stimuli via inhibitory signaling of T-cells (to be further discussed below under Section 5.4
Recent in vivo studies have shown promising results of combining CD40 stimulation and radiation. Using a pancreatic ductal adenocarcinoma mouse model, Rech et al. revealed synergy between radiation and an agonist αCD40 antibody through distinct mechanisms. Ablative dose of radiation triggers early inflammatory stimulus through upregulation of MHC class I and CD86, while αCD40 causes a late response of altering intratumoral and systemic immunosuppressive myeloid cells, collectively yielding abscopal effect and long-term tumor immunity [92
]. In another study using pancreatic cancer models, single fraction of SBRT with agonist αCD40 led to regression of non-irradiated tumor and durable immune memory [93
OX40 is a potent co-stimulatory molecule on activated T-cells, and OX40 signaling can promote effector T-cell survival and inhibit Treg function, which can be achieved via OX40 ligand binding or stimulation via antibody agonists. Combination of single dose of 20 Gy and intratumoral delivery of activating OX40 antibody in murine lung cancer model resulted in CD8 T-cell dependent tumor killing and tumor immunity [94
]. Gough et al. corroborated this finding in murine 3LL-tumor model using high doses of radiation and αOX40 antibody to achieve extended survival and decreased tumor recurrence compared to single treatments alone [95
]. Recently, a combination of radiation and OX40 agonist has demonstrated efficacy in anti-PD-1-resistent murine lung tumors to inhibit local and systemic tumor growth [96
4-1BB, also known as CD137, is the first member of the TNFR family identified as a potential target for cancer immunotherapy. Ligation of 4-1BB receptor on activated T-cells with 4-1BBL or antibody agonist prompts anti-apoptotic signaling to prevent activation-induced cell death and reverse T-cell tolerance [97
]. When combined with 4-1BB agonist, antitumor effect of radiation has been shown to be enhanced. Treatment with 4-1BB antibody agonist with radiation in murine lung (M109) and breast (EMT6) carcinoma models significantly delayed tumor progression [98
]. With the addition of cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) blockade, the combination of 4-1BB activation and radiation further improved survival in GL261 murine glioma model, which is associated with greater infiltration of CD4+ and CD8+ lymphocytes [99
]. Similarly, concomitant inhibition of PD-1 with radiation and 4-1BB agonist enhanced antitumor effect against human BRAF-mutant melanoma [100
]. Interestingly, one study showed decreased off-target immune cell activation with 4-1BB aptamer compared to 4-1BB antibody when combined with radiation, while having similar therapeutic effect with both agents, suggesting potential differences in treatment toxicity of targeting the same receptor with varying forms of agonists [101
To date, clinical experience with combining radiation with co-stimulatory molecules is limited. However, there are several ongoing clinical trials (discussed below, Section 7
) that will shed light on the clinical utility of this therapeutic strategy.
5.4. Immune Checkpoint Inhibition
Immune checkpoints are regulatory mechanisms that serve to prevent over-stimulation of activated T-cells, which can lead to autoimmunity. CTLA-4 and PD-1, also members of the B7/CD28 family, are expressed on activated T-cells to act as an “off” switch when bound by ligands CD80/CD86 and PD-L1/PD-L2, respectively. Cancer cells often over-express PD-L1/PD-L2 to exploit the intrinsic mechanism of T-cell inhibition. Antibody antagonists against CTLA-4 and PD-1/PD-L1 attenuate tumor-induced inhibitory signaling, thereby shifting towards T-cell stimulation and bolstering adaptive tumor immunity. Moreover, as previously mentioned, radiotherapy can upregulate PD-1/PD-L1 on tumor and immune cells in the TME, and as such, combining checkpoint inhibition with radiation may nullify this undesired immunosuppressive sequela.
Demaria et al. first reported synergy of radiation and CTLA-4 blockade in a preclinical setting. Using the poorly immunogenic murine 4T1 mammary carcinoma model, this study showed that only combinatory treatment of radiation and CTLA-4 inhibition, but not either treatment alone, exhibited significant survival advantage over control. Furthermore, systemic antitumor immunity was provoked with combined treatment resulting in decreased lung metastases, which required the presence of CD8+ T-cells [102
]. In a follow-up study by the same group, abscopal effect of combined radiation and CTLA-4 blockade was demonstrated using bilateral tumor models in which the unirradiated tumors displayed significant growth delay after irradiation of the primary tumors. This systemic effect is CD8-dependent and correlates with increased TILs and tumor-specific IFN-γ-producing T-cells in the unirradiated tumors [7
The number of prospective clinical investigations assessing safety and efficacy of combined CTLA-4 inhibition and radiotherapy is on the rise. So far, only a few trials have reported results. In a phase I/II study, patients with metastatic castrate-resistant prostate cancer were treated with escalating doses of ipilimumab with or without radiotherapy. Maximum dose tested of 10 mg/kg ipilimumab with 8 Gy radiation to one to three bony metastases showed acceptable toxicity profile, with one-third of patients having stable disease or better [103
]. Using the same ipilimumab and radiation dose, a phase III multicenter trial randomized 799 patients with metastatic prostate cancer to ipilimumab versus placebo after radiotherapy to osseous metastases. Although median overall survival of patients receiving ipilimumab only trended higher than those receiving placebo (11.2 vs. 10.0 months, p
= 0.053), post-hoc subgroup analysis of patients with good prognostic features demonstrated significant survival benefit with ipilimumab (22.7 vs. 15.8 months, p
= 0.0038) [104
]. There are also several small prospective studies that reported abscopal responses and improved overall survival in metastatic melanoma patients treated with ipilimumab and radiation to brain and/or visceral metastases [105
]. Of interest, a joint clinical and preclinical study in patients with metastatic melanoma implicated T-cell exhaustion from upregulation of tumor PD-L1 expression in the resistance towards treatment with radiation and CTLA-4 antibody. Treatment-induced increase in the PD-1/PD-L1 axis was reproduced in murine melanoma models, and the addition of PD-L1 blockade significantly improved response to radiation and CTLA-4 inhibition [109
]. In addition, a recent clinical study shed light on plausible mechanisms underlying favorable responses to combination treatment with radiation and CTLA-4 blockade. A cohort of 39 patients with metastatic NSCLC were treated with radiotherapy to one metastasis with concurrent ipilimumab, with response rate of 18% and disease control in 31% of patients. Increase in IFN-β and T-cell receptor clonal dynamics predicted response to combination therapy, and further characterization of a single responder revealed expansion of two specific T-cell clones that target an immunogenic mutation on a radiation-induced gene, KPNA2. While this intriguing finding supports the hypothesis that radiotherapy can enhance neoantigen exposure to the host immune system, validation in an expanded cohort is warranted. Furthermore, the single-arm nature of this trial precludes the determination of the degree of contribution from either radiotherapy or ipilimumab in the observed immunological effects [110
Several preclinical models have also revealed therapeutic synergy of radiation and PD-1/PD-L1 blockade. Treatment of radiation and anti-PD-1 antibody in mouse glioma model improved survival compared to either treatment modality alone. Combined treatment group exhibited increased tumor infiltration of cytotoxic T-cells and decreased Tregs, and glioma tumor cells rechallenged in treated mice failed to grow [111
]. Deng et al. observed increase in PD-L1 expression in the TME after radiation, and the addition of PD-L1 inhibition augmented antitumor effect of radiation. Specifically, efficacy of combined treatment is dependent on CD8+ T-cells and correlated with reduction in the immunosuppressive MDSCs [8
]. Similar findings were reported by Dovedi et al., which also revealed that upregulation of PD-L1 on tumor cells is induced by IFN-γ secretion from CD8+ T-cells [51
]. Subsequent studies further demonstrated that PD-1 blockade enhanced antigen-specific and tumor-specific immunity triggered by radiation [112
]. In a recent mechanistic study, the authors showed that both the preexisting resident T-cells and infiltrating lymphocytes after combination treatment contributed to tumor regression in in-field and out-of-field tumors [114
]. While most studies examined the role of PD-1/PD-L1 axis inhibition in bolstering radiation efficacy, Wang et al. reported that radiotherapy can reverse tumor resistance towards anti-PD-1 therapy through induction of IFN-β and MHC-I expression on tumor cells [115
]. Together, the above evidence underscored the importance of modulating the immune status within the TME to optimize efficacy of cancer therapy in the clinic.
Given that ample evidence lent support to the efficacy of combining PD-1/PD-L1 blockade with radiation, an abundant of clinical trials are now ongoing to investigate the utility of this combination in the clinical setting. Several phase I/II trials have established that concomitant PD-1/PD-L1 inhibition with radiotherapy is generally well-tolerated without dose-limiting toxicities [116
]. Furthermore, with increasing adoption of ablative radiotherapy for treatment of multiple metastatic foci, Luke et al. demonstrated that multisite SBRT to up to four lesions followed by pembrolizumab within 7 days of SBRT completion was well tolerated [120
]. Recently, the phase III PACIFIC trial that randomized locally-advanced unresectable non-small cell lung cancer (NSCLC) patients to either adjuvant durvalumab (anti-PD-L1 antibody) or placebo after chemoradiation demonstrated significant improvement in progression-free survival with durvalumab (median survival 16.8 vs. 5.6 months) [121
]. Most recent update of the trial results also showed improved overall survival with patients receiving durvalumab compared to those receiving placebo (2-year overall survival 66.3% vs. 55.6%) [122
]. It is interesting to note that the secondary analysis of KEYNOTE-001 trial, a study in which locally advanced or metastatic NSCLC patients were treated with anti-PD-1 pembrolizumab, showed improved survival in patients who received prior radiotherapy. Although this analysis was retrospective in nature and hypothesis-generating, it nevertheless shed light on the potential therapeutic synergy of radiotherapy and PD-1/PD-L1 blockade in the clinical setting [123
5.5. Macrophage Polarization
Tumor-associated macrophages (TAMs) play important roles in tumorigenesis and contribute to maintaining an immunosuppressive TME in their default state. TAMs are typically pro-tumorigenic and phenotypically resemble M2 macrophages, and reducing tumor-infiltration of macrophages or modifying the polarity of immunosuppressive TAMs towards pro-inflammatory M1 phenotype have shown to impair tumor growth [124
]. As previously discussed, radiation also promotes M2 polarization within the irradiated tissues. As such, by reversing the immunosuppressive phenotype of TAMs should augment the immunogenic effects of radiation.
CSF1 is a key cytokine responsible for promoting M2 polarization through CSF1 receptor (CSF1R)-mediated signaling and CSF1R blockade in tumor models led to repolarizing of TAMs to the M1 phenotype [125
]. Xu et al. provided evidence that CSF1R blockade improved efficacy of radiotherapy against prostate cancer in a murine model, with associated decrease in intratumoral MDSCs and TAMs populations [50
]. Furthermore, recent study demonstrated that macrophages can express PD-1 and that PD-1 expression correlated with M2 polarization. Blockade of PD-1/PD-L1 axis enhanced phagocytosis of tumor cells by intratumoral PD-1+ macrophages in vivo and decreased tumor burden [127
]. Given that checkpoint inhibition of PD-1/PD-L1 axis is already being widely used, this finding underscores a novel mechanism in which PD-1/PD-L1 blockade can bolster therapeutic effects of radiotherapy.