SMAC Mimetic Debio 1143 and Ablative Radiation Therapy Synergize to Enhance Antitumor Immunity against Lung Cancer

In patients with early-stage non–small cell lung cancer (NSCLC), delivery of highly conformal, high-dose ablative radiotherapy (ART) has demonstrated local control and survival rates that rival those of surgery (1). Aside from establishing local control of distant metastases, ART may, at times, contribute to systemic disease control through generating antitumor immunity by stimulating tumor antigen and cytokine release (2–4). As such, ART is also being increasingly utilized in patients with oligometastatic disease (5). However, 3 cell populations, myeloid-derived suppressor cells (MDSCs), tumor-associated macrophages (TAMs), and regulatory T cells (Tregs) can neutralize these favorable host adaptive immune responses following radiation (6–9). Furthermore, the immune response following ART is not entirely proinflammatory in nature, and, at times, ART may recruit and activate the immunosuppressive cell compartment.

Secondary mitochondrial-derived activators of caspase (SMAC) mimetics belong to an investigational class of drugs that antagonize inhibitor of apoptosis proteins (IAP), which inhibit caspase function. By either direct inhibition or by triggering autoubiquitination and IAP degradation, SMAC mimetics restore apoptotic function and sensitize cancer cells to chemotherapy and radiation (10–13). Furthermore, IAP depletion by SMAC mimetics alters NFκB signaling such that autocrine or paracrine TNFα-mediated NFκB induction elicits cancer cell death rather than survival (14). IAP antagonism also modulates innate and adaptive immunity (15), increasing the activation of antigen-presenting cells (16), cell death in monocytes, maturation of monocyte-derived dendritic cells (17), and T-cell activation (18). One study has demonstrated that the combination of oncolytic viruses and SMAC mimetic compounds robustly activates the innate immune system against cancer cells (19). SMAC mimetics also have proved to synergistically promote antitumor immunity alongside programmed death-1 (PD-1) immune checkpoint inhibitors in multiple mouse tumor models (20). These properties have led to early clinical trials investigating the clinical activity of SMAC mimetics alongside chemoradiation in head and neck cancer (21) and the anti-PD-L1 antibody avelumab in patients with advanced solid malignancies (22).

We hypothesized that the combination of ART and a SMAC mimetic may synergize to promote an antitumor adaptive immune response. Here, we report a novel finding that treatment of NSCLC with the combination of ART and the SMAC mimetic Debio 1143 displayed strong antitumor efficacy and generated a proinflammatory cascade, culminating in the CD8⁺ T-cell and TNFα-mediated elimination of immunosuppressive cells from the TME. Combination therapy with ART and Debio 1143 promoted a Tc1 effector, tumor-specific immune response, which was sustained beyond the therapeutic time period.

The murine Lewis lung carcinoma (LLC) cell line was obtained from the ATCC. The LLC-OVA cell line expressing a single polypeptide encoding H-2K^bβ2-M, and the OVA SIINFEKL peptide was kindly provided by Amer A. Beg (Moffitt Cancer Center, Tampa, FL) and was generated as described previously (23). Cells were cultured in DMEM supplemented with 10% FBS, 1 mmol/L sodium pyruvate, 2 mmol/L l-glutamine, 10 μg/mL penicillin–streptomycin, and 0.1 mmol/L nonessential amino acids (Life Technologies). All mouse experiments were approved by the Jefferson Institutional Animal Care & Use Committee. Six- to 8-week-old female C57BL/6 mice were purchased from Jackson Laboratory. All mice were maintained under specific pathogen-free conditions. SMAC mimetic Debio 1143 was provided by Debiopharm International SA, and was prepared in an aqueous solution and stored at 4°C within 1 week of experiments.

Cells were harvested after being cultured for less than 2 weeks and counted. Six- to 8-week-old female C57BL/6 mice were injected subcutaneously at the right flank with 5 × 10⁵ LLC-OVA cells in 200-μL serum-free DMEM. Tumors were allowed to develop for 9 days, until tumor diameter was approximately 0.1–0.2 cm³, after which the mice were divided into 4 treatment groups (n = 12 per group). On day 9, the C57BL/6 mice treated with either Debio 1143 (100 mg/kg for consecutive 10 days) by oral gavage, ART of 30 Gy by using a PanTak 310keV X-ray machine at 0.25-mm Cu plus 1-mm Al added filtration at 125 cGy/minute, or the combination. Tumor dimensions were measured every 3 days, and volumes were calculated using the formula: V = L × W² × 0.52, where L and W are the long and short diameters of the tumor, respectively. Tumor growth delay experiments were repeated 3 times to verify their accuracy. For the survival assay, mice were sacrificed when they showed signs of morbidity or when subcutaneous tumors reached 2,000 mm³ in size.

To obtain single-cell suspensions, 12 days after ART, tumor tissues were minced into small fragments, and digested using 1 mg/mL collagenase IV (Sigma-Aldrich) and 0.2 mg/mL DNaseI (Life Technologies) at 37°C for 45 minutes. Dissociated cells were passed through a 70-μm cell strainer twice and washed 3 times in DMEM. Spleens from tumor-bearing mice were homogenized and pass through a 70-μm cell strainer in ice-cold PBS to achieve single-cell suspension. Red blood cells were lysed using ACK Lysis Buffer (Life Technologies). All samples were pretreated with CD16/CD32 FcR blocker (BD Biosciences) before staining. Cells were stained with the following antibodies obtained from eBioscience: CD11b-PE (clone M1/70, catalog No. 12-0112-83), Gr-1 APC (clone RB6-8C5, catalog No. 17-5931), F4/80 (clone BM8, catalog No. 17-4801), CD8a-PE (clone 53-6.7, catalog No. 12-0081), Mouse regulatory T-cell staining kit (clone FJK-16s, catalog No. 88-8118), OVA peptide (SIINFEKL)-loaded H-2Kb tetramer linked to APC (clone 25-D1.16, catalog No. 17-5743). The following antibodies were obtained from BioLegend: CD45-PerCP/Cy5.5 (clone 30-F11, catalog No. 103132), CD8a-FITC (clone 53-6.7, catalog No. 100706), CD8a-Alexa Fluor 594 (clone 53-6.7, catalog No. 100758), CD4-FITC (clone GK1.5, catalog No. 100406), CD11b-PE (clone M1/70, catalog No. 101208). For intracellular cytokine staining, cells were stimulated with 50 ng/mL PMA, 500 ng/mL ionomycin, and 10 μg/mL GolgiPlug (BD Biosciences) at 37°C for 4 hours. Cells were then harvested, fixed, and permeabilized using Cytofix/Cytoperm Fixation/Permeabilization Kit (BD Biosciences) and stained with IFNγ-FITC (eBioscience; clone XMG1.2, catalog No. 11-7311), and TNFα-Alexa Fluor 647 (BioLegend; clone MP6-XT22, catalog No. 506314) to identify cytokine-secreting CD8⁺ T cells. Data were collected on a BD LSRII Flow Cytometer, and analyzed using FlowJo software (Tree Star Inc.).

Total RNA was isolated using TRIzol reagent (Life Technologies) according to the manufacturer's instructions. cDNAs were synthesized with TaqMan Reverse Transcription Reagents (Applied Biosystems), following the manufacturer's instructions. The primer pairs used for iNOS, Arginase I, TNFα, IL10, and GAPDH are obtained from Taqman RT reagents (Applied Biosystems). The fold stimulation was determined using the comparative cycle threshold method (2^−ΔΔC_t). All experiments were performed in triplicate.

Quantification of serum cytokines (TNFα, IFNγ, IL10, IL1β) was performed using R&D Quantikine ELISA Kit according to the manufacturer's instructions. The plates were read at 450 nm within 30 minutes. All standards, controls, and samples were run in triplicate.

C57BL/6 mice were inoculated subcutaneously with LLC-OVA cells and treated with radiation and Debio 1143 according to the regimen described above. For CD8⁺ T-cell depletion experiments, 200 μg per mouse of anti-CD8 antibody (clone 2.43; Bio-XCell) or rat IgG2B isotype control were delivered via intraperitoneal injection (i.p.) every 3 days beginning on day 9 for 4 times. For the TNFα blockade experiment, 500-μg anti-TNFα antibody (clone: XT3.11; Bio-XCell) or rat IgG1 isotype control was administered intraperitoneally every 3 days beginning on day 9 for 4 times (4). For the IFNγ-neutralizing experiment, 300-μg anti-IFNγ antibody (clone: R4-6A2; Bio-XCell) or rat IgG1 isotype control was administered intraperitoneally every 3 days beginning on day 9 for 4 times (20). All antibody treatments were started from the day of radiation or 1 day before radiation.

Survival analysis was done using the Kaplan–Meier method with log-rank univariate analysis and Cox regression multivariate analysis, with P < 0.05 considered statistically significant. Data were analyzed using Prism 5.0 software (GraphPad Software). Data are represented as the mean ± SD for all figure panels in which error bars are shown. The P values were assessed using 2-tailed unpaired Student t tests. A P value of less than 0.05 was considered statistically significant.

We hypothesized that the SMAC mimetic compound Debio 1143 could potentiate antitumor adaptive immune responses to radiation. LLC cells stably transfected with OVA-tetramer antigen in subcutaneous syngeneic tumor models were used to quantify tumor-specific immune responses (23). To assess for therapeutic synergy between Debio 1143 and ART, LLC-OVA tumor-bearing C57BL/6 mice were treated with Debio 1143, ART, or both as shown in Fig. 1A. ART was delivered as a 30 Gy dose in a single fraction, a regimen shown to maximize adaptive antitumor immunity in multiple tumor models (24). Moderate decreases in tumor size were observed following ART alone; however, treatment with the combination potently inhibited LLC-OVA tumor growth compared with Debio 1143 (P < 0.001) and ART alone (P < 0.05), and 4 mice had a complete response to combination therapy (Fig. 1B). Furthermore, mouse survival was significantly increased in the ART (P < 0.01) and Debio 1143 + ART groups (P < 0.001) compared with controls (Fig. 1C). SMAC mimetics, in the presence of type I IFN or oncolytic viruses, may stimulate long-term antitumor immunologic memory (20). To assess for immune memory after treatment with ART and Debio 1143, 4 complete responders in the ART + Debio 1143 group were rechallenged with 5 × 10⁶ (i.e., 10 times the number of cells of the initial challenge) LLC-OVA cells in the opposite flank 37 days after the initial treatment with ART with 5 treatment-naïve C57BL/6 mice serving as controls. While tumors grew rapidly in control mice after 12 days, none of the complete responders established tumors 23 days after the rechallenge (Fig. 1D), suggesting an induction of sustained, systemic antitumor immunity with combination therapy.

To assess for tumor antigen–specific immunity and the recruitment of adaptive immune cells after treatment with ART and Debio 1143, LLC-OVA tumors were probed for CD8⁺ T lymphocytes expressing OVA 12 days after ART, as earlier time points did not fully characterize the magnitude of changes in immune cell infiltration (data not shown; ref. 24). Tumor-specific OVA⁺CD8⁺ T cells were increased with ART alone (Fig. 2A), consistent with previous reports (24). In the ART + Debio 1143 group, OVA⁺CD8⁺ T cells were increased 2-fold compared with the ART group (P < 0.05) and 7-fold over the Debio 1143 group (P < 0.001), characterizing a robust tumor antigen–specific adaptive immune response (Fig. 2A). To assess whether combination therapy increased CD8⁺ T-cell activation and differentiation, we analyzed tumors for OVA⁺CD8⁺ cells expressing Tc1 subtype markers IFNγ and TNFα by flow cytometry. OVA⁺CD8⁺ T cells from tumors treated with both Debio 1143 and ART expressed increased levels of IFNγ (P < 0.001 for Debio and ART vs. Debio, P < 0.01 for Debio and ART vs. ART; Fig. 2B) and TNFα (P < 0.001 for Debio and ART vs. Debio, P < 0.05 for Debio and ART vs. ART; Fig. 2C). Given this induction in antigen-specific Tc1 effector responses in mice treated with ART and Debio 1143, we next evaluated whether combination therapy affected antigen-presenting cell infiltration by assessing tumors for dendritic cell markers CD45/CD11b/CD11c/CD86. Not surprisingly, we found significantly more dendritic cells in mice treated with ART and Debio 1143 compared with Debio 1143 (P < 0.001) and ART (P < 0.01) alone (Supplementary Fig. S1). Taken together, these data demonstrate that Debio 1143 potentiates the capability of ART to induce tumor-specific Tc1 effector responses at the level of antigen presentation and T-cell priming (25–28).

MDSCs, TAMs, and Tregs are 3 cell populations known to suppress antitumor immunity or even promote tumor progression (29–31). To determine the effects of ART and Debio 1143 on the immunosuppressive cell compartment, LLC-OVA tumors were harvested and assessed for cell-surface markers of MDSC-like cells (CD45⁺CD11b⁺Gr-1⁺), TAMs (CD45⁺CD11b⁺F4/80⁺), and Tregs (CD4⁺CD25⁺FOXP3⁺) by flow cytometry 12 days after ART. ART alone decreased Gr-1⁺CD11b⁺ cell infiltration, but the addition of Debio 1143 dramatically decreased Gr-1⁺CD11b⁺ populations among CD45⁺ cells in the TME (P < 0.05 for Debio and ART vs. ART, P < 0.001 for Debio and ART vs. Debio; Fig. 2D). Treatment with ART alone appeared to increase the TAM phenotype among CD45⁺ cells, which was reversible with the addition of Debio 1143 (P < 0.05; Fig. 2E). As shown in Fig. 2F, the reductions in CD11b⁺Gr-1⁺cells and TAMs with combination therapy coincided with a decline in Treg markers among CD4⁺ cells (P < 0.001 for Debio and ART vs. Debio, P < 0.01 for Debio and ART vs. ART), consistent with the known role of MDSCs in Treg recruitment (32). Taken together, combination therapy with ART and Debio 1143 facilitated the elimination of the immunosuppressive cell compartment within the TME.

We next sought to correlate local cytokine expression with these observations of Tc1 cell expansion and immunosuppressive cell compartment reductions after ART and Debio 1143. Using RT-PCR, we assessed LLC-OVA tumors for expression of the Tc1 cytokines TNFα and IFNγ as well as the immunosuppressive enzyme Arg-1 and cytokine IL10 (8, 33). As shown in Fig. 3A, we detected increases TNFα and IFNγ mRNA transcripts after ART alone that were potentiated by the addition of Debio 1143 (P < 0.01 for Debio and ART vs. Debio, P < 0.05 for Debio and ART vs. ART). ART alone increased IL10 and Arg-1 mRNA, which was reversible with the combination of ART and Debio 1143 (Fig. 3A). To characterize systemic cytokine responses, we measured serum levels of TNFα, IFNγ, IL10, and IL1β, the latter of which is a cytokine crucial to the priming of T cells with tumor antigen by ELISA (34). Again, serum TNFα and IFNγ as well as IL1β were increased in the ART + Debio 1143 group compared with the Debio 1143 (P < 0.01) and ART (P < 0.05) groups (Fig. 3B), which were remarkably comparable in magnitude to the increases in cytokine transcripts. Similar to our observations of LLC-OVA tumors, ART alone increased serum IL10 levels compared with vehicle control, but the addition of Debio 1143 reversed this to levels below both those of Debio 1143 (P < 0.05) and ART alone (P < 0.01). These data are consistent with previous studies describing the mixed proinflammatory and anti-inflammatory cytokine response to radiation (7, 9, 33). The coadministration of Debio 1143 with ART, in contrast, augmented proinflammatory cytokine production and curtailed anti-inflammatory cytokines both locally and systemically.

To evaluate the extent to which CD8⁺ T cells mediate therapeutic synergy between ART and Debio 1143, LLC-OVA tumors were again grown in C57BL/6 mice and treated with ART, Debio 1143, or both with or with anti-CD8 antibody as shown in Fig. 4A. Analysis of tumors treated with anti-CD8 antibody confirmed effective CD8 depletion (Supplementary Fig. S2). Four of 12 mice in the Debio 1143 group had a complete response; however, CD8⁺ T-cell depletion nullified synergy between Deb 1143 and ART. The addition of anti-CD8 antibody resulted in significantly larger tumors than those treated with combination therapy (P < 0.05) and similar in size to those treated with ART alone (Fig. 4B). Deng and colleagues demonstrated that the clearance of CD11b⁺Gr-1⁺ cells from the TME was dependent upon CD8⁺ cells and TNFα (4), and others have shown that the depletion of CD11b⁺Gr-1⁺ cells is adequate to reinstate T-cell immunity (6).To study the effects of CD8 depletion on the immunosuppressive cell infiltrate, we again studied the CD11b⁺Gr-1⁺ cell population, which was the most sensitive to combination therapy, using flow cytometry. While combination therapy with ART and Debio 1143 reduced CD11b⁺Gr-1⁺ cells compared with controls, CD8 depletion partially restored this population (P < 0.05; Fig. 4C). TAM and Treg infiltration, however, was unaffected by CD8 depletion (data not shown). Collectively, these data suggest not only that CD8⁺ T cells mediate the immunologic synergy between ART and Debio 1143, but they also contribute to the elimination of CD11b⁺Gr-1⁺ cells from the TME.

To characterize the role of TNFα in combination therapy, we performed neutralization experiments using anti-TNFα antibody (Fig. 5A). Again, 4 of 12 mice in the ART + Debio 1143 group had a complete response. As expected, systemic blockade of TNFα attenuated the therapeutic synergy between ART and Debio 1143 (P < 0.01; Fig. 5B). TNFα has been shown to promote CD11b⁺Gr-1⁺ cell apoptosis in tumors treated with anti-PD-1 antibodies and ART (4). We hypothesized the surge in expression of TNFα following ART in the presence of Debio 1143 would have similar effects, which we evaluated by assessing tumors treated with and without anti-TNFα antibody for CD11b⁺Gr-1⁺ cells. While TAM and Treg infiltration were again unchanged in the presence of anti-TNFα antibody (data not shown), TNFα depletion partially restored infiltration of tumors by CD11b⁺Gr-1⁺ cells treated with ART and Debio 1143 (P < 0.01; Fig. 5C). Thus, both the therapeutic synergy between ART and Debio 1143 as well as the reductions in CD11b⁺Gr-1⁺ cell infiltration were mediated by TNFα.

Others have shown that IFNγ plays a key role in priming an antitumor adaptive immune response after ART (24). To study the role of IFNγ in the therapeutic synergy between ART and Debio 1143, we depleted IFNγ from treated or vehicle LLC-OVA tumor-bearing mice using blocking antibody as shown in Fig. 6A. Indeed, tumors from mice treated with ART, Debio 1143, and anti-IFNγ were significantly larger (P < 0.05) than tumors from mice treated with just ART and Debio 1143 without anti-IFNγ antibody (Fig. 6B). We next analyzed tumors for infiltration by CD11b⁺Gr-1⁺cells to determine whether elimination of the immunosuppressive cell compartment is dependent upon IFNγ. Again, the addition of anti-IFNγ to Debio 1143 and ART partially restored the CD11b⁺Gr-1⁺ cells in the TME (P < 0.05; Fig. 6C). Collectively, these data indicate that both the therapeutic synergy between ART and Debio 1143 and the associated clearance of CD11b⁺Gr-1⁺ cells from the TME is dependent upon CD8⁺ T cells as well as the 2 major cytokine modulators of adaptive immunity, TNFα and IFNγ.

The immune system is now recognized as a key mediator of the efficacy of ART (4). Aside from direct cancer kills, it also produces antitumor adaptive immunity; however, repopulation of the TME by MDSCs, TAMs, and Tregs can stifle these responses (3, 35, 36). SMAC mimetic compounds, known to have radiosensitizing properties (10–12), can also promote both innate and adaptive immune responses (19, 20). Using an LLC-OVA model, we have demonstrated the SMAC mimetic Debio 1143 potentiates proinflammatory cytokine release and the tumor-specific adaptive immune response to ART (Fig. 6), which included antitumor immune memory in a subset of mice. The combination of Debio 1143 and ART resulted in the elimination of the immunosuppressive compartment from the TME in a manner dependent upon CD8⁺ T cells, TNFα, and IFNγ.

Independent of immunity, combination treatment with SMAC mimetics and radiation disinhibits apoptosis by degrading IAPs and promotes RIPK1/MLKL-mediated necroptosis via TNFα (10, 11). By increasing noncanonical NFκB signaling, SMAC mimetics convert the TNFα produced by irradiation from a prosurvival to a cell death signal (14, 37, 38). Similar mechanisms may partially account for therapeutic synergy between ART and Debio 1143 that we observed in vivo (Fig. 1B and C). Notably, in the absence of host immunity, LLC-OVA cells were markedly resistant to Debio 1143 alone in vitro (Supplementary Fig. S3A). Furthermore, Debio 1143 failed to sensitize LLC-OVA cells to doses of ionizing radiation between 2 and 6 Gy (Supplementary Fig. S3B). We performed a similar clonogenic assay using ablative doses of radiation. As shown in Supplementary Fig. S3C, the addition of Debio 1143 significantly reduced colony formation at 30 Gy, but not 10 Gy or 20 Gy (Supplementary Fig. S3C). It is unclear why radiosensitization by Debio 1143 was only observed at 30 Gy, although Debio 1143 only reduced colony survival by a modest 11.3%. Regardless, these data indicate that direct sensitization of LLC-OVA by Debio 1143 cell to ART-induced cell death could play a minor role in the synergy between ART and Debio 1143.

Beyond direct cell death, the role of the adaptive immune system in mediating the synergy between ART and Debio 1143 was clear. Combination therapy increased tumor infiltration by tumor antigen–specific, activated T cells (Fig. 2A–C), and complete responders to ART + Debio 1143 group maintained this adaptive immune response against LLC-OVA tumors upon rechallenge (Fig. 1D; refs. 16, 39). Furthermore, the therapeutic synergy between ART and Debio 1143 on LLC-OVA tumors was abrogated by CD8 depletion (Fig. 4B). Notably, we found similar trends in an LLC tumor model without OVA expression, suggesting that combination therapy induces antitumor immunity even in the absence of the artificial immunodominant OVA antigen (Supplementary Fig. S4). To our knowledge, this represents a first-time report of immunogenic synergy between ART and SMAC mimicry.

Furthermore, our data suggest that antigen presentation and T-cell priming underlies the immunogenic synergy between ART and Debio 1143. Indeed, SMAC mimetics and ART are known to activate antigen presentation by nonredundant mechanisms. SMAC mimetic–induced NFκB activation in dendritic cells and macrophages mimics CD40 engagement, driving tumor antigen presentation (16, 18). Likewise, the roles of radiation in enhancing tumor antigen release, in eliciting ATP and IFN release to recruit and activate antigen-presenting cells have been elaborated (24, 40–42). After irradiation, calreticulin, a phagocyte activator, undergoes cell membrane translocation in irradiated cancer cells, driving the production of IL1β from dendritic cells, which promotes T-cell priming and polarization at a transcriptional level (34). In keeping with the known potential of both SMAC mimetics and ART to activate antigen-presenting cells, combination therapy with ART and Debio 1143 augmented dendritic cell infiltration (Supplementary Fig. S1) and IL1β production (Fig. 3B) in the TME. This coincided with robust increases in tumor antigen–specific CD8⁺ T-cell responses (Fig. 2A) and Tc1 effector cell infiltration (Fig. 2B and C) as well as the local and systemic production of Tc1-subtype cytokines (Fig. 3A and B), the end-products of antigen priming and presentation. This suggests that the immunogenic synergy between Debio 1143 and ART may occur at the level of antigen presentation and T-cell priming.

Negative regulators of antitumor immunity, MDSCs, TAMs, and Tregs within the TME restrict the immune response (6–9). Studies have reported conflicting effects of radiation on these cell populations with some reporting their mobilization to the TME and others reporting their clearance from the TME in response to ART alone (6, 9, 24). Similar to other studies of single fraction 30 Gy ART, we found that ART alone decreased infiltration by Tregs and CD11b⁺Gr-1⁺ cells; however, contrary to that study, ART increased TAM infiltration (Fig. 2D–F; ref. 24). Despite this, ART monotherapy also increased IL10 and Arg-1 (Fig. 3A and B), yet it is unclear whether these data reflect increased TAM infiltration or, alternatively, enhanced CD11b⁺Gr-1⁺ cell immunosuppression (43). These differences may partially be explained by differing radiation doses, tumor models, host immune backgrounds, timing of data collection, and immune cell markers analyzed. For example, while the Cd11b and Gr-1 cell markers used in this study have been shown to identify cells with an MDSC phenotype in LLC syngeneic tumor models (44), they do not distinguish between monocytic and polymorphonuclear distributions, which may have differing function (45). Regardless, our data are consistent with the overarching principle that ART, in addition to promoting adaptive immunity, triggers compensatory responses among the immunosuppressive cell compartment that restrict the magnitude of the antitumor immune response. Most notably, our data indicate that the addition of Debio 1143 can reverse these immunosuppressive adaptations to ART, depleting the TME of TAMs, Tregs, and CD11b⁺Gr-1⁺ cells as well as Arg-1 and IL10 (Fig. 6D).

We also found that the clearance of CD11b⁺Gr-1⁺ cells from the TME was dependent upon these CD8⁺ T cells (Fig. 4C). Although the mechanisms by which MDSCs suppress T-cell immunity are well described (8), the reciprocal mechanisms by which T cells eliminate MDSCs are poorly defined. One possibility is that Fas-L interactions on T cells with the Fas death receptor on MDSCs promotes their demise. Studies demonstrate that MDSCs avidly express the Fas death receptor (46), and SMAC mimetics upregulate Fas-L on T cells, shifting the T-cell killing mechanism from perforin/granzyme pathways toward Fas/FasL killing (39). However, therapeutic synergy as well as the clearance of CD11b⁺Gr-1⁺ cells from the TME with combination therapy were both dependent upon TNFα (Fig. 5B and C) and IFNγ (Fig. 6B and C). Thus, it is possible that the role of T cells in clearing MDSCs in the TME is limited to their production of these cytokines. Prior studies reporting the effects of TNFα on MDSCs have been mixed. Whereas some have suggested that TNFα blocks the differentiation and enhances the suppressive activity of MDSCs (47), our results corroborate those of Deng and colleagues, who reported that TNFα produced by CD8⁺ T cells induces apoptosis in MDSCs after ART and PD-1 blockade (4). One possible explanation for these different results is that the biologic consequences TNFα signaling and NFκB activation in of CD11b⁺Gr-1⁺cells (i.e., survival vs. cell death) are contextual and shifted in favor of cell death in the presence of Debio 1143. Yet another possibility, TNFα may have differing effects on of CD11b⁺Gr-1⁺ cells at high doses generated by ART and PD-1 blockade or SMAC mimetics compared with lower amounts produced in chronic inflammatory states (4).

Interestingly, the depletion of CD8 cells, TNFα, and IFNγ failed to completely restore the CD11b⁺Gr-1⁺ cell population in the TME to levels of vehicle controls. Although we observed only minor, statistically insignificant reductions in CD11b⁺Gr-1⁺ cells after Debio 1143 alone (Fig. 2D), it is possible that the incomplete restoration of CD11b⁺Gr-1⁺ cells in the 3 depletion experiments is due to a direct cytotoxicity of SMAC mimetics upon monocytic CD11b⁺Gr-1⁺ cells (17). However, Deng and colleagues have also reported that TNFα and, to a lesser extent, IFNγ can mediate CD11b⁺Gr-1⁺ cell apoptosis in a CD8-independent manner (4). This raises the possibility that any TNFα or IFNγ-producing cell (i.e., CD4⁺ cells or even tumor cells) could play an indirect role the clearance of CD11b⁺Gr-1⁺ cells from the TME after treatment with ART and Debio 1143. Nevertheless, while we studied the primary modulators of CD8⁺ T-cell adaptive immunity, we cannot exclude the possibility of other cell types, such as natural killer cells or neutrophils, participating in CD11b⁺Gr-1⁺ cell clearance, as well.

MDSCs are well-known to upregulate programmed death ligand-1 (PD-L1) in response to radiation, which activates T-cell anergy and apoptosis upon ligation to its receptor, PD-1, on T cells (4, 48). PD-L1 expression on CD11b⁺ MDSCs also impedes the early adaptive immune response to SMAC mimetic monotherapy (16). Several groups have shown that the induction of effective adaptive antitumor immunity with either radiation or a SMAC mimetic compound requires blockade of the PD-1/PD-L1 immune checkpoint (16, 20, 49), and Deng and colleagues showed that this was dependent upon TNFα-mediated clearance of CD11b⁺Gr-1⁺cells (4). It is certainly plausible that additional therapeutic synergy could be achieved by triple therapy: ART, Debio 1143, and PD-1/PD-L1 inhibition. However, it is also possible that the TNFα-dependent clearance of MDSCs with the combination of ART and Debio 1143 suffices to reduce PD-1/PD-L1 interactions in the TME, thereby disinhibiting adaptive immunity even without PD-1 blockade.

In summary, the combination of SMAC mimetic Debio 1143 and radiotherapy resets the immunosuppressive nature of the TME in a TNFα, IFNγ-, and CD8⁺ T-cell–dependent manner and produces a systemic proinflammatory cytokine cascade. This represents a first-time report of remarkable therapeutic synergy, including durable systemic antitumor immunity, with the combination of ART and a SMAC mimetic. These data support the design of clinical trials evaluating the use of SMAC mimetics to harness the immunogenicity of ART.

B. Lu reports receiving commercial research grants from Debiopharm. No potential conflicts of interest were disclosed by the other authors.

Conception and design: Z. Tao, N.S. McCall, N. Wiedemann, G. Vuagniaux, B. Lu

Development of methodology: Z. Tao, N.S. McCall

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Z. Tao

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Z. Tao, N.S. McCall, B. Lu

Writing, review, and/or revision of the manuscript: Z. Tao, N.S. McCall, N. Wiedemann, G. Vuagniaux, B. Lu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Z. Tao, B. Lu

Study supervision: Z.-Y. Yuan, B. Lu

This work was supported by funding from Debio Pharma; National Natural Sciences Foundation of China (grant nos. 81672524, 81602678, and 81502227); and Natural Science Foundation of Tianjin (grant No. 17JCQNJC12300).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.