Local Treatment of a Pediatric Osteosarcoma Model with a 4-1BBL Armed Oncolytic Adenovirus Results in an Antitumor Effect and Leads to Immune Memory

This article is featured in Highlights of This Issue, p. 395

Osteosarcoma is the most common primary bone tumor, affecting mainly children and adolescents, with a second minor peak of incidence at approximately 50 years of age (1). This malignant neoplasm is composed of mesenchymal cells that produce aberrant osteoid and immature bone, developing predominantly near the long bone metaphysis of limbs (2). The mainstay of current osteosarcoma treatment includes surgical resection of local and metastatic disease and pre- and postsurgical combination chemotherapy. The chemotherapy regimen includes a combination of doxorubicin and cisplatin with or without subsequent treatment with high-dose methotrexate and has a patient response rate of 70% (3). However, for patients with metastasis or recurrent disease, curative therapeutic options are limited, and the overall survival rate at 5 years is approximately 20%. This life expectancy is meager, underscoring the need for effective treatments, especially for high-risk populations.

Oncolytic virotherapy is a type of cancer biotherapy that has considered as a valid treatment option. Delta-24-RGD is a conditionally replicative adenovirus engineered to selectively replicate in cancer cells with an aberrant RB pathway (4, 5). In preclinical studies, Delta-24-RGD has shown efficacy in several solid adult (4, 6) and pediatric tumors, including local and metastatic osteosarcoma models (7–9). Clinically, Delta-24-RGD showed good safety and elicited a durable complete response in 20% of patients with recurrent glioblastoma (10). Importantly, the previous clinical trial revealed that intratumoral viral injection instigated an oncolytic effect followed by an inflammatory response potentially responsible for the antitumor effect and subsequent control of the tumor.

To further potentiate the efficacy of this virus, we engineered a new oncolytic virus expressing the immune costimulatory ligand 4-1BB (4-1BBL) to enhance viral antitumor immunity (11). 4-1BB is an inducible receptor promoting the survival and expansion of activated T cells. The interaction of 4-1BBL with 4-1BB enhances the stimulation of CD4⁺ and CD8⁺ T cells, causing predominant expansion of CD8⁺ T cells (12). Signals initiated by 4-1BBL–4-1BB binding promote the CD8⁺ T-cell costimulation, playing an important role in the differentiation of effector memory responses. Therefore, we hypothesized that the oncolytic adenovirus Delta-24-ACT expressing 4-1BBL could trigger an immune response that would lead to a specific antisarcoma effect not only controlling local disease but also resulting in a therapeutic effect on distant untreated lung metastasis.

The murine osteosarcoma cell line K7M2 was obtained from the American Type Culture Collection (ATCC cat. No. CRL-2836, RRID:CVCL_V455). The murine osteosarcoma POS-1 and MOS-J cells derived from mouse spontaneous osteosarcoma were provided, respectively, by Drs. Kamijo (13) and Shultz (14) and were cultured in RPMI and DMEM, respectively with 10% FBS and 1% penicillin/streptomycin. The pediatric human osteosarcoma cell line Saos-2 was obtained from ATCC (cat. No. 300331/p651_SaOS-2, RRID:CVCL_0548). 531MII, a primary osteosarcoma cell line, was developed at the Clínica Universidad de Navarra. Cell lines obtained from ATCC were cultured following the manufacturer's instructions. 531MII cells were cultured in minimum essential medium supplemented with 10% FBS and 1% antibiotic. All cells were maintained in a humidified atmosphere with 5% CO₂ at 37°C. Cell lines were routinely tested for Mycoplasma (MycoAlert Mycoplasma Detection Kit; Lonza), and human cell lines were authenticated at the CIMA Genomic Core Facility.

Delta-24-ACT was constructed by maintaining the Delta-24-RGD modifications of a 24-base pair deletion and introduction of RGD (15); however, m4-1BBL was incorporated in the E3 locus after removing this gene. Briefly, murine 4-1BBL was first cloned into a pCDNA3.1 plasmid using the Kpn I and Xho I restriction enzymes (New England Biolabs). Then, m4-1BBL flanked with the cytomegalovirus promoter and bovine growth hormone polyadenylation sequences was subcloned into the pAB26-RGD plasmid at the Cla I/BamH I site. Finally, the 4-1BBL expression cassette was introduced into pVK-500C-Δ24, and the Delta-24 plasmid was constructed by homologous recombination with pAB26-m4-1BBL in BJ5183 bacteria. For viral rescue, the obtained plasmid was linearized with Pac I and transfected into HEK293 cells (cat. No. 300192/p777_HEK293, RRID:CVCL_0045) with Lipofectamine 2000 (Invitrogen). After confirmation of genetic modifications by PCR and sequencing, Delta-24-ACT was amplified in A549 cells (cat. No. 300114/p686_A-549, RRID:CVCL_0023), purified, and stored at −80°C.

All cell lines were seeded at a density of 10³ cells per well in 96-well plates and infected with Delta-24-RGD and Delta-24-ACT at different multiplicity of infections ranging from 0 to 300. Cell viability was assessed 5 days after infection using a CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega). The dose–response curves and IC₅₀ values were analyzed using GraphPad software (RRID:SCR_002798 v8).

Cells were plated at a density of 5 × 10⁴ and were then infected at an MOI of 10 for human cell lines and 300 for murine cell lines. Seventy-two hours later, the cells and supernatant were collected and subjected to three freeze-thaw cycles. The supernatant was used to infect HEK293 cells, and viral PFUs were determined by the anti-hexon staining method.

Cell lines were plated in 96-well plates at a density of 2 × 10⁵ and infected at an MOI of 10 with a nonreplicative virus expressing GFP. After 24 hours, infected cells were harvested, and GFP expression was quantified by flow cytometry.

For immunoblotting, samples were subjected to SDS-Tris-Gly gel electrophoresis. Membranes were incubated with antibodies specific for the following molecules: E1A, (1:1,000; Santa Cruz Biotechnology cat. No. sc-430, RRID:AB_630843, Santa Cruz Biotechnology), fiber (1:1,000; Novus cat. No. NB 600–541, RRID:AB_521564, Novus Biological), GRB-2 (1:1,000; BD Biosciences cat. No. 610111, RRID:AB_397517), and LC3 A/B (1:1,000; Cell Signaling Technology cat. No. 4108, RRID:AB_213770).

Paraffin-embedded sections of mouse tibias, lungs, and livers were immunostained following conventional procedures with these antibodies: adenoviral mouse anti-hexon (Chemicon International, Inc.), adenovirus rabbit anti-E1A, (Santa Cruz Biotechnology), anti-CD3 (NeoMarkers), anti-CD4 (Abcam), anti-CD8a (Cell Signaling Technology), anti-FoxP3 (eBioscience, Thermo Fisher Scientific), and anti-vimentin clone V9 (IS30, Dako Denmark A/S). For immunohistochemical staining, Vectastain ABC Kits (Vector Laboratories Inc.) were used according to the manufacturer's instructions.

Cell lines were plated and infected with both viruses at MOIs of 25 and 300, and medium was collected 72 hours later. ATP release was assessed with an ENLITEN ATP Assay System (FF2000, Promega). HMGB1 and HSP-90 release into the medium was measured with an HMGB1 ELISA Kit (ST51011, IBL International) and an HSP-90A ELISA kit (ADI-EKS-895, Enzo Life Sciences Inc.).

Calreticulin translocation to the plasma membrane was determined by immunofluorescence staining of 2 × 10⁴ cells infected with Delta-24-ACT at a dose corresponding to the 3-day IC₅₀ (or with PBS as a negative control). Four hours postinfection, cells were fixed with 4% methanol-free formaldehyde (28906; Thermo Fisher Scientific) for 15 minutes at 37°C and were then stained with appropriate antibodies.

Viral 41BBL expression in osteosarcoma cell lines was determined by flow cytometry. Cells were stained first with a cell death detection antibody in the Zombie-NIR Fixable Viability Kit (423105, BioLegend) following the manufacturer's protocol and then with a PE-conjugated anti–4-1BBL antibody.

Calreticulin cell surface expression was determined by flow cytometry. Cells were stained first with a cell death detection antibody in the Zombie Green Fixable Viability Kit (423111, BioLegend) and then with a fluorophore-conjugated anti-calreticulin antibody (Abcam).

Fluorescence emission was analyzed using a FACSCanto II system with FACSDiva software (RRID:SCR_001456).

To identify immune cell populations, infiltrating immune cells were surface stained with the following antibody panel: anti–CD45-AF700 (BioLegend), anti–Ly6G-PerCP-Cy5.5 (BioLegend), anti–Ly6C-FITC(BioLegend), anti–F4/80-APC (BioLegend), anti–CD8-BV510 (BioLegend), anti–CD11b-BUV395 (BioLegend), and anti–CD4-BUV496 (BioLegend). PromoFLuor840 maleimide (PromoKine) was used as a viability marker. Samples were acquired with a CytoFlex flow cytometer (Beckman Coulter RRID:SCR_019627) and data analyses were performed using FlowJo v10 (RRID:SCR_008520).

Splenocytes and tumor-infiltrating lymphocytes (TILs) isolated from treated mice were cocultured with 10,000 K7M2 cells. K7M2 cells were left uninfected or infected with Delta-24-RGD or Delta-24-ACT at an MOI of 100 MOI. Four hours later, K7M2 cells were stimulated with recombinant mouse IFNγ protein and were collected and fixed with 4% formaldehyde 48 hours later. After 72 hours of coculture, supernatants were collected, and IFNγ ELISA (R&D Systems, Inc.) and MACS Plex Mouse Cytokine 10 kits (Miltenyi Biotec, Bergisch Gladbach) were used for analysis following the manufacturers’ instructions.

The Ethical Committee of the University of Navarra [Comite Etico de Experimentacion Animal (CEEA)] approved the animal protocols performed in this study (CEEA/022–16). All animal studies were performed at the veterinary facilities of the Center for Applied Medical Research following institutional, regional, and national laws and ethical guidelines for experimental animal care.

To establish osteosarcoma animal models, K7M2 cells were injected into female BALB/c mice through the tibial plateau into the primary spongiosa of the right tibia. PBS or viruses (10⁸ pfu/mouse) were administered intratumorally on days 10 and 18 after tumor engraftment. Tumor development and mouse weight were monitored twice weekly during the experiment. The tibia size was measured from the beginning of the experiment, and mice were sacrificed when the tumor volume reached 400 mm². Tumor volumes were determined throughout the course of the experiment by measuring the tumors along the two perpendicular diameters with a caliper and calculating the tumor volume with the following formula: volume = D × (d)² × 0.5, where D is the largest diameter and d is the smaller diameter. Animals were sacrificed when the tibial tumor volume reached 430 mm³ or when they lost more than 20% of their body weight, and we harvested tibias to examine primary bone tumors, lungs to examine metastases, and livers to evaluate safety.

Thorax tomography was performed on mice anesthetized by intraperitoneal injection of ketamine and xylazine. Then, an intratracheal cannula was placed and connected to a flexiVent rodent ventilator (Scireq) set at a rate of 200 breaths/minute and a tidal volume of 10 mL/kg. Animals were maintained on 2% isoflurane inhalation until completely relaxed.

Lung 3D tomographic images were acquired using an X-ray micro-CT system (Quantum-GX, Perkin Elmer) with the following parameters: X-ray source voltage of 90 kVp, current of 88 μA, high-speed scan protocol for a total acquisition time of 14 minutes, and gantry rotation of 360°. Breathing artifacts were denoised using respiratory gating in each acquisition. The three-dimensional tomographic images containing the whole lung had a total of 512 slices with an isotropic voxel size of 72 μm and a resolution of 512 × 512 pixels per slice. Analysis of the lung volume in each sample was carried out using Fiji/ImageJ (RRID:SCR_002285), open-source Java-based image processing software. In brief, lung images were segmented by applying a fixed threshold, and total lung volume was measured over the obtained mask (mm³). Quantified values were normalized to those of normal, healthy lung parenchyma.

The in vitro experiments were repeated at least three times. Dose–response curves for viral cytotoxicity were obtained by nonlinear regression. Data with normal distributions were assessed by Shapiro–Wilk tests, and comparisons among groups were performed by two-tailed nonparametric tests with 95% confidence intervals for nonnormally distributed datasets or parametric tests when normality was confirmed (Student t test or one/two-way ANOVA). For comparison of groups in survival experiments, a log-rank test (Mantel–Cox) was used. GraphPad Prism 8 (Statistical Software for Sciences GraphPad Prism) was used for the statistical analyses.

Delta-24-ACT is a replicative oncolytic adenovirus based on the Delta-24-RGD (4) backbone and modified to encode the mouse 4-1BBL (m4-1BBL; Supplementary Fig. S1A). We showed that the 4-1BBL expressed by the virus was functional and effectively activated T cells in vitro and in vivo in the context of murine glioma models (15). Human pediatric osteosarcoma cells express receptors (αν3 and αν5 integrins and CAR) necessary for viral entry into cells (16), and Delta-24-ACT utilizes these receptors for recognition and subsequent entry into cells. Thus, we confirmed that Delta-24-ACT was able to infect 531MII and Saos-2 human osteosarcoma cells and K7M2 osteosarcoma cells (Supplementary Fig. S1B). Moreover, K7M2, 531 MII and Saos-2 cells infected with an increasing amount of virus showed dose-dependent expression of 4-1BBL mRNA that was efficiently translated to protein (Fig. 1A and B; Supplementary Fig. S1C). Importantly, we confirmed that after infection with Delta-24-ACT, 4-1BBL was expressed on the surface of murine (Fig. 1C and D; Supplementary Fig. S1D) and human osteosarcoma cells (Fig. 1C; Supplementary Fig. S1D).

Because viral genome modifications can interfere with effective viral replication, the replication capacity of Delta-24-ACT was quantified and compared with that of Delta-24-RGD in osteosarcoma cell lines after 72 hours of infection. No differences were found between the load of Delta-24-RGD and Delta-24-ACT progeny virions, indicating that insertion of the m4-1BBL expression cassette did not hinder viral replication (Fig. 1E; Supplementary Fig. S1E). Finally, the results of cytotoxicity tests showed that both Delta-24-RGD and Delta-24-ACT exerted a potent antitumor effect in vitro in all tested osteosarcoma cell lines (Fig. 1F; Supplementary Fig. S1F). We observed that both viruses displayed very similar IC₅₀s: Delta-24-RGD proved slightly better although not significant. In K7M2, the IC_50s were 27 ± 11.2 and 24 ± 7.9 MOIs for Delta-24-ACT and Delta-24-RGD, respectively. In 531MII were 10.1 ± 3.5 and 7.5 ± 2.4 MOIs for Delta-24-ACT and Delta-24-RGD, respectively. In addition, we confirmed the capability of Delta-24-ACT to replicate and to express m4-1BBL in osteosarcoma tumors in vivo. We detected m4-1BBL mRNA and protein expression in osteosarcoma tumors 48 hours after Delta-24-ACT treatment (Fig. 1G and H). We also observed the presence of viral proteins, specifically E1A and fiber in orthotopic osteosarcoma tumors treated with Delta-24-ACT (Fig. 1H). Importantly, the presence of the fiber protein, which is expressed in the late stages of the viral life cycle, suggested viral replication in vivo.

In summary, these results demonstrated that the newly generated virus Delta-24-ACT efficiently expressed m4-1BBL in osteosarcoma cells both in vitro and in vivo, indicating preservation of its replication capacity.

Autophagy mediates the release of damage-associated molecular patterns (DAMPs) and tumor antigens; therefore, its stimulation has been proposed to be required for immunogenic cell death (ICD; refs. 17–20). Consequently, we assessed whether Delta-24-ACT treatment induces autophagy. We observed an increase in the conversion of LC3I to LC3II and a decrease in the P62/GRB2 ratio in Delta-24-ACT–infected cells compared with Delta-24-RGD–infected cells (Supplementary Fig. S2A). Next, we evaluated whether Delta-24-ACT–induced autophagy led to the release of several DAMPs by evaluating ATP, HMGB1, and HSP-90α release and calreticulin (CRT) exposure in the membrane as hallmarks of ICD. We observed an increase in ATP release in K7M2 (P = 0.04) and 531MII (P = 0.03) cells 72 hours after treatment with Delta-24-RGD or Delta-24-ACT (Fig. 1A). The level of HMGB1 release tended to increase, although nonsignificantly, in both cell lines after viral treatment (Supplementary Fig. S2B). In addition, HSP-90α release, a marker of ICD in human cells, was significantly increased in 531MII cells infected with either Delta-24-ACT or Delta-24-RGD compared with control cells (mock-infected; Fig. 2B). Moreover, calreticulin translocation from the endoplasmic reticulum to the cell membrane, forming clusters in K7M2 and 531MII cells after Delta-24-ACT infection, was observed (Fig. 2C; Supplementary Fig. S2C). Quantification of calreticulin expression in the membrane of K7MII and 531MII cells showed a significant increase in both cell lines infected with the either virus (P = 0.002 and P = 0.02 for K7MII cells and 531MII cells, respectively; Fig. 2D). In summary, Delta-24-ACT treatment of osteosarcoma cells triggered the production of several DAMPs in vitro.

Previously, we showed that Delta-24-RGD treatment triggered an immune response in glioma and DIPG models (7, 10, 21). In fact, viral administration induces the production of DAMPs that attract immune cell populations to the tumor site, eliciting a specific immune response against tumor cells (17). Therefore, BALB/c mice bearing K7M2 murine osteosarcoma cells orthotopically were treated twice with PBS, Delta-24-RGD, or Delta-24-ACT (days 10 and 18). Anatomopathological studies revealed a significant increase in infiltrating CD3⁺ and CD8⁺ cells in tumors treated with Delta-24-ACT compared with those treated with Delta-24-RGD (P < 0.05) or PBS (P < 0.001). The infiltration density was increased almost twofold in Delta-24-ACT–treated mice compared with Delta-24-RGD–treated mice (Fig. 3A and B). On the other hand, CD4⁺ cells were increased in treated tumors, although the difference was not significant (Supplementary Fig. S3A). Quantification of several immune cell populations by flow cytometry at day 7 posttreatment (Supplementary Fig. S3B) confirmed that both viruses significantly increased the number of CD45⁺, CD3, CD8, CD4, Trg, NK cells, macrophages, and DCs in the tumor (all P < 0.05). However, we did not find significant differences among both viruses (Fig. 3C). A second experiment confirmed that Delta-24-ACT increased the numbers of CD3⁺ cells, CD4⁺ cells, CD8⁺ cells, and macrophages compared with those in PBS-treated mice (P = 0.002, P = 0.03, P = 0.04, and P = 0.001, respectively; Fig. 3D). Moreover, CD4⁺ and CD8⁺ cells tended to express higher levels of CD137⁺ (Fig. 3D), suggesting that T-cell activation was increased. Quantification of IFNg production in splenocytes isolated from mice in the different treatment groups showed that compared with PBS, both viruses increased significantly the level of this cytokine measured by ELISA (P = 0.01 and P < 0.001 for Delta-24-RGD and Delta-24-ACT, respectively; Fig. 3E). Moreover, splenocytes from Delta-24-ACT–treated mice displayed a more active phenotype than those from Delta-24-RGD (P = 0.02). This finding was further confirmed by ELISPOT, which showed that the coculture of K7M2 cells with splenocytes from Delta-24-ACT–treated mice led to a nearly threefold increase in the production of IFNg as compared with the PBS group (P < 0.0001; Fig. 3F). Delta-24-ACT was also superior to Delta-24-RGD (P = 0.006). Moreover, intratumoral injection of Delta-24-ACT led to upregulation of proinflammatory cytokines such as IFNg, GM-CSF, IL5, and IL4 in the tumor. Interestingly, we did not observe any difference in the expression of IL2 or IL12. In addition, virus injection did not induce any changes in protumoral cytokines (Supplementary Fig. S3C).

Collectively, these results showed that Delta-24-ACT increased immune infiltration, inducing a proinflammatory microenvironment.

Because 4-1BB agonists have shown hepatic toxicity in clinical settings (reviewed in refs. 8, 22), we first evaluated the safety of intratumoral administration of Delta-24-ACT. Administration of one or three doses of virus did not affect the weight of the treated animals (Fig. 4A). Moreover, quantification of murine hepatic transaminases 48 hours after a single dose and after three doses of Delta-24-ACT revealed no differences in blood alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels (Fig. 4B), suggesting normal hepatic function. Histopathologic analysis confirmed that the livers from virus-treated animals were morphologically normal (Fig. 4C), indicating the safety of the virus.

Next, we evaluated the therapeutic effect of the virus. For this study, two local injections were administered on days 10 and 18 (Supplementary Fig. S4A). Mice were weighed until day 40 to rule out toxicity (Supplementary Fig. S4B). Efficacy studies showed that control mice developed visible bone tumors faster than treated mice, and by day 25, all but two control mice exhibited measurable tumors (Fig. 5A). Moreover, tumors developed in only four of the 12 tibias in the treated group. The tumor diameters in control mice ranged from 90 to 500 mm³, whereas the four tumors that developed in the Delta-24-ACT–treated group had volumes between 250 and 450 mm³ (Fig. 5A). Treatment with Delta-24-RGD also delayed tumor growth (Supplementary Fig. S4C). Kaplan–Meier survival curves showed that local administration of Delta-24-ACT led to significantly extended overall survival times, with median survival times of 30 days for PBS-treated mice and 47 days for Delta-24-ACT–treated mice (P < 0.0001; Fig. 5B). In a second survival study, we compared the efficacy of Delta-24-ACT with Delta-24-RGD and PBS treated mice. We observed that efficacy of Delta-24-ACT was superior to the one exerted by Delta-24-RGD. Kaplan–Meier survival curves showed that local administration of Delta-24-ACT or Delta-24-RGD led to significantly extended overall survival times, with median survival times of 33.5 days for PBS-treated mice, 38 days for Delta-24-RGD (P < 0.05) and 60 days for Delta-24-ACT–treated mice (P = 0.0006). Both viruses led to the generation of long-term survivors: one (1/10) and four (4/10) for Delta-24-RGD and Delta-24-ACT, respectively (Supplementary Fig. S4D). K7M2 cells are a highly lung metastatic cell line; therefore, we evaluated the potential for lung metastasis. Micro-CT analysis showed that the integrity of the normal lung parenchyma was significantly maintained in mice treated with Delta-24-ACT compared with control mice (P = 0.05; Fig. 5C), indicating that the viral treatment can control spontaneous lung metastasis. Delta-24-RGD treatment also significantly reduced the number of tumor nodules in the lungs of treated mice (Supplementary Fig. E). Rechallenge of the long-term survivors with K7M2 cells in the contralateral leg showed that 100% of long-term survivors (3/3) were protected against rechallenge through the development of protective immune memory. Moreover, anatomopathological studies showed normal tibias in the animal previously treated with Delta-24-ACT in stark contrast with the tibias of the naïve treated mice (Fig. 5G). Lung evaluation also revealed that long-term survivors showed healthier lung parenchyma than naïve counterparts (Fig. 5H).

Evaluation of the efficacy of Delta-24-RGD and Delta-24-ACT in immunodeficient model lacking lymphocytes T and B bearing orthotopically K7M2 murine cell line, showed no therapeutic benefit, indicating that the mechanism of action of these viruses is mediated by an efficient antitumoral immune response (Fig. 5I; Supplementary Fig. S4F). In summary, these data revealed the efficacy and safety of Delta-24-ACT in a pediatric osteosarcoma murine model.

The observation that osteosarcoma tumors exhibit genomic instability and a high mutational burden, associated with an increased presence of neoantigens, suggests that these tumors are amenable to immunotherapy (23). In fact, several immunotherapy-based clinical trials, including trials of immune checkpoint inhibitors, chimeric antigen receptor (CAR) T cells and other immunotherapies (NCT03006848, NCT03628209, NCT04539366, NCT04483778, and NCT02470091), are ongoing. However, the published results are modest regarding outcomes in patients with metastatic or refractory osteosarcoma (24, 25). Furthermore, the toxicities induced by some of these agents have hampered their clinical development (26, 27). In this work, we showed that Delta-24-ACT is safe and efficacious in murine osteosarcoma models. Our results are consistent with a plethora of trials that support the inclusion of oncolytic viruses in routine oncology and, most importantly, pediatric oncology (NCT03178032, NTC02444546, NTC03043391, NTC03330197, NTC02031965). Oncolytic viruses have shown good safety profiles and a degree of efficacy (28). These results encourage the combination of these viruses with drugs specifically geared toward targeting aberrancies in pediatric solid tumors. One interesting feature of the virus presented in this work is that the overall survival of the long-term responders (50% of treated mice) developed an effective antiosteosarcoma memory immune response. This capacity of the virus to achieve a “vaccine” effect is of relevance for osteosarcoma therapy because 20% of new cases present with metastasis at diagnosis (29). In this line of thinking, we know that treatment with oncolytic viruses leads to the release of DAMPs and pathogen-associated molecular patterns and ultimately to ICD, which in turn results in tumor antigen release (29). In our system in vitro, both viruses led to the release of DAMPs compatible with ICD, and this phenomenon was higher in Delta-24-ACT–treated osteosarcoma cells. Supporting this notion, in an elegant work by Hinterberger and colleagues, where they evaluated the efficacy of a modified Ankara virus (MVA-TAA-4-1BBL) armed with 4-1BBL and administered intratumorally in different tumor models, they observed that the virus triggered profound changes in the tumor microenvironment, including the induction of multiple proinflammatory molecules and ICD (30). However, to clarify the specific contribution of 4-1BBL to the enhancement of this type of death, further experiments would be needed.

One of the main hurdles of current immunotherapies is the immunosuppressive tumor microenvironment (TME) and the inability to induce a T-cell trafficking to tumor site. Specifically relevant to osteosarcomas is that the presence of malignant osteoid limits immune infiltration into tumor (23). In addition, single-cell RNA sequencing studies of osteosarcoma patient samples have shown that the TME in this type of cancer is highly heterogeneous, with an ample diversity of immune populations with different phenotypes, revealing the complexity of osteosarcoma tumors (30) and suggesting that one strategy does not fit all cancer. We observed that Delta-24-ACT induced T-cell trafficking to osteosarcoma tumors overcoming the immunosuppressive TME. Therefore, the virus could shake the TME and render the tumor more susceptible to other types of immunotherapies such as immune checkpoints. In that line of thinking, an elegant study by Ligon and colleagues uncovered an immune molecular and cellular signature in osteosarcoma pulmonary metastasis that expressed PD-L1, LAG3, and CSF1R and was associated with worse progression-free survival (31). It would be interesting to combine the virus with one or several inhibitors of these molecules and assess the effect on survival.

Besides lymphocytes, we observed a significant upregulation of macrophages in the tumors treated with Delta-24-ACT. Interestingly, this population is prevalent in patients, and their role is complex and context-dependent (32–35). In our work, we did not delve into the phenotype of these cells. In addition, we observed a significant increase in dendritic cells. It is well known that activated APC express high amounts of 4-1BBL. Moreover, it has been shown that immunization with a 4-1BBL lentivirus activates significant numbers of bystander DCs in the draining lymph nodes and leads to more efficacious CD8. In that work, the authors showed that transactivation by 4-1BBL/4-1BB interaction following DC–DC contact could participate in the immune response to infection or vaccination (36). Understanding the potential contribution of APCs to the antisarcoma effect shown by Delta-24-ACT could provide clues to improve this therapy

One limitation of our study is that adenovirus replication is hindered in murine cell lines. The murine osteosarcoma cells used in this work are nonpermissive for adenoviral replication (Fig. 1), reducing the adenoviral load and limiting 4-1BBL expression to only a first wave. This constraint probably resulted in milder stimulation of the immune response than that potentially occurring in human counterparts. However, although our study was not ideal, our data showed the efficacy of this virus.

The encouraging results achieved by local administration of Delta-24-ACT, indicating that it can trigger distant tumor remission, underscore the therapeutic potential of this virus and the need for its clinical translation.

The data that support the findings of this study are available within the article or Supplementary Information or available from the corresponding author upon request.

N. Martinez-Velez reports grants from AACR Foundation during the conduct of the study. J. Fueyo reports a patent for Oncolytic adenoviruses pending and licensed to DNAtrix and is a shareholder of DNATrix. C. Gomez-Manzano reports a patent for Oncolytic adenoviruses pending and licensed to DNAtrix and is a shareholder of DNAtrix. M.M. Alonso reports grants from DNAtrix outside the submitted work. No disclosures were reported by the other authors.

N. Martinez-Velez: Conceptualization, methodology, writing–review and editing. V. Laspidea: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. M. Zalacain: Conceptualization, investigation, methodology, writing–original draft, writing–review and editing. S. Labiano: Investigation, writing–review and editing. M. Garcia-Moure: Investigation, writing–review and editing. M. Puigdelloses: Investigation, writing–review and editing. L. Marrodan: Investigation, writing–review and editing. M. Gonzalez-Huarriz: Investigation, writing–review and editing. G. Herrador: Investigation, writing–review and editing. D. de la Nava: Investigation, writing–review and editing. I. Ausejo-Mauleon: Investigation, writing–review and editing. J. Fueyo: Formal analysis, methodology, writing–review and editing. C. Gomez-Manzano: Formal analysis, methodology, writing–review and editing. A. Patiño-García: Conceptualization, resources, funding acquisition, investigation, writing–original draft, writing–review and editing. M.M. Alonso: Conceptualization, resources, funding acquisition, investigation, writing–original draft, writing–review and editing.

The performed work was supported through the Departamento de Salud del Gobierno de Navarra (54/2018-APG); AACR-AstraZeneca Immuno-oncology Research Fellowship (18-40-12-MART; N. Martinez-Velez); Predoctoral Fellowship from Gobierno de Navarra (V. Laspidea and G. Herrador), Instituto de Salud Carlos III y Fondos Feder (PI19/01896 M.M. Alonso, PI18/00164A. Patiño-García); Amigos de la Universidad de Navarra (to M. Puigdelloses); Fundación La Caixa/Caja Navarra (A. Patiño-García and M.M. Alonso); Fundación El Sueño de Vicky; Asociación Pablo Ugarte-Fuerza Julen, Fundación ADEY, Fundación ACS, (A. Patiño-García and M.M. Alonso); and Department of Defense (DOD) Team Science Award under grant (CA 160525 M.M. Alonso, C. Gomez-Manzano, and J. Fueyo). La Marató (M.M. Alonso). This project also received funding from the European Research Council (ERC) under the European Union's Horizon 2020 Research and Innovation Programme (817884 ViroPedTher to M.M. Alonso).

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.