The paradox of antiviral immunity in oncolytic virotherapy

Oncolytic virotherapy (OVT) is a rapidly evolving cancer immunotherapy strategy that uses naturally occurring or genetically engineered viruses to selectively infect, replicate within, and destroy malignant cells. Although direct tumor cell lysis remains central to this approach, the therapeutic activity of oncolytic viruses (OVs) extends beyond viral replication alone. Viral infection can release tumor-associated antigens (TAAs), viral pathogen-associated molecular patterns (PAMPs), danger-associated molecular patterns (DAMPs), and inflammatory mediators that reshape the tumor microenvironment (TME) and promote antitumor immune responses [1]. This dual mechanism, tumor-selective cytolysis coupled with immune activation, has fueled renewed clinical and commercial interest in OVT.

The field of OVT has expanded substantially over the past two decades. As of 2022, more than 400 clinical trials had evaluated OVs across diverse malignancies, including melanoma, glioblastoma, lung cancer, pancreatic cancer, and others. These trials have involved multiple viral platforms, including DNA viruses such as herpes simplex virus, adenovirus, and vaccinia virus among the most widely used, and RNA viruses such as vesicular stomatitis virus, Newcastle disease virus, poliovirus, and reovirus also under active investigation [1]. The clinical feasibility of this approach is highlighted by the regulatory approval of several genetically modified OVs in different countries, including the human adenovirus H101 (Oncorine®) in combination with chemotherapy in China [2], and two herpes simplex viruses, talimogene laherparepvec (T-VEC; Imlygic®) and teserpaturev/G47Δ (Delytact®) in the United States and Japan, respectively [3,4]. Many ongoing and completed studies now evaluate OVs either as monotherapies or in combination with chemotherapy, radiation therapy, immune checkpoint blockade, and other immunotherapies [5].

Despite this progress, clinical responses to OVT remain variable, and innate antiviral immunity is a major determinant of this variability [6,7]. Many OVs exploit defects in tumor antiviral pathways, particularly type I interferon (IFN-α/β) sensing and signaling, to replicate preferentially in malignant cells. However, IFN and related innate immune pathways can also restrict viral replication, dissemination, and tumor cell killing when activated too early or too strongly [8–10]. Conversely, insufficient innate immune activation may permit viral replication and localized cytolysis but fail to convert tumor destruction into durable, systemic antitumor immunity. Mathematical models similarly support this framework by showing that OVT efficacy depends on dynamic interactions among viral infection of tumor cells, viral clearance, tumor growth, and antitumoral or antiviral immune responses rather than on viral replication alone [11,12].

Rather than viewing innate antiviral immunity as merely an obstacle to viral replication, this commentary argues that innate immune activation is a context-dependent determinant of OVT outcomes. The same pathways that restrict viral spread can also promote inflammatory remodeling, immunogenic tumor cell death, and antitumor immune priming. Understanding how viral genetics, tumor-intrinsic antiviral competence, and the TME shape this balance may help explain variable responses to OVs and guide more rational therapeutic combinations (Figure 1).

Figure 1. Innate antiviral immunity as a double-edged regulator of oncolytic virotherapy outcomes. Innate antiviral immunity can either limit or enhance the therapeutic efficacy of oncolytic virotherapy depending on its magnitude and timing. (A) Excessive antiviral restriction, driven by strong PRR sensing and IFN/ISG activation, suppresses viral replication, intratumoral spread, and tumor cell killing, resulting in weak antigen release and limited immune priming. (B) A productive therapeutic window is achieved when viral replication is sufficient to induce immunogenic tumor cell death while preserving enough innate immune activation to promote cytokine and chemokine production, antigen presentation, dendritic cell (DC) activation, and recruitment of NK cells and tumor-specific T cells, thereby converting local oncolysis into durable antitumor immunity. (C) In contrast, insufficient immune activation or excessive viral permissiveness may permit viral replication and cytolysis but fails to generate effective antigen presentation and adaptive immune priming, leading to transient tumor control without lasting systemic benefit. Additional factors may shift tumors between these states.

This commentary is intended as a focused conceptual perspective rather than a systematic review of innate antiviral signaling in OVT. Although the antiviral and immunostimulatory functions of IFN and ISG pathways have been extensively reviewed, less attention has been given to how the timing, magnitude, and cellular context of these responses determine whether OV infection becomes abortive, cytolytic, inflammatory, or therapeutically productive. We therefore propose the “productive immune window” as a framework for understanding how viral replication kinetics, tumor-intrinsic antiviral competence, innate inflammatory signaling, and antitumor immune priming converge to shape OVT outcomes.

Tumor-intrinsic antiviral immunity as a barrier to viral oncolysis

Tumor-cell permissiveness is a critical early determinant of OVT efficacy. Many OVs exploit cancer-associated defects in antiviral sensing, IFN signaling, protein translation control, or cell-death regulation to replicate preferentially in malignant cells. However, these defects are neither uniform across tumor types nor consistent among cells within the same tumor. Some cancer cells retain robust antiviral programs that can rapidly restrict viral gene expression, replication, and spread before sufficient oncolysis occurs. Thus, tumor-intrinsic antiviral immunity acts as an early gatekeeper that determines whether infection becomes abortive, productive, or therapeutically useful.

Type I IFNs are central to this gatekeeping function. IFNs are a group of cytokines classified into three major families: type I, type II, and type III. While type I IFNs, particularly IFN-α/β, can be produced by most nucleated cells, IFN-γ is primarily secreted by activated T cells and NK cells, and type III IFNs are especially important at epithelial barriers and in selected antigen-presenting cells [8]. In the context of OVT, type I IFN signaling is particularly relevant because many tumor cells can detect viral infection and respond by inducing an antiviral state. Viral PAMPs, including viral nucleic acids and replication intermediates, are sensed by pattern-recognition receptors (PRRs) such as RIG-I, MDA5, endosomal Toll-like receptors, and, for DNA viruses, cGAS-STING and related DNA-sensing pathways [10,13]. Activation of these pathways stimulates IRF3/IRF7- and NF-κB-dependent transcriptional programs that induce IFN-α/β and inflammatory mediators. Secreted type I IFNs then signal through their heterodimeric cell-surface receptor to activate JAK/STAT pathways and induce hundreds of IFN-stimulated genes (ISGs), establishing an antiviral state in infected and neighboring cells [8,10].

ISG products restrict viral replication at multiple stages of the viral life cycle. Protein kinase R (PKR), encoded by EIF2AK2, is activated by viral double-stranded RNA and phosphorylates eIF2α, thereby suppressing cellular and viral protein translation [14]. Interestingly, PKR has been shown to play a double role in cancer cells, where it not only limits the oncolytic potential of viruses but also increases the immune resistance of cancer cells through increased TGF-β signaling [15]. The OAS/RNase L pathway provides another important antiviral mechanism induced by IFNs. The 2′-5′-oligoadenylate synthetase (OAS) family of enzymes detect viral dsRNA and generate oligoadenylates that activate RNase L, resulting in degradation of viral and cellular RNAs and, in some settings, amplification of PRR signaling [10]. Pharmacological inhibition of this pathway has been reported to enhance the oncolytic activity of vesicular stomatitis virus in resistant tumor models, illustrating how intact ISG programs can limit OV efficacy [15,16]. Additional ISGs, including Mx proteins, IFITs, ISG15, and others, can restrict viral replication, translation, assembly, or trafficking, although their direct contributions to OVT remain to be fully addressed [17–19].

At first glance, studies showing that suppression of PKR or OAS/RNase L can enhance OV replication may appear to conflict with studies showing that innate immune activation contributes to antitumor efficacy [20]. These findings can be reconciled by distinguishing early tumor-intrinsic antiviral restriction from later immune-stimulatory inflammation. In models where intact PKR or RNase L activity prevents sufficient viral gene expression, transient inhibition of these pathways can increase viral replication and cytolysis [15]. However, when viral infection proceeds far enough to release tumor antigens, PAMPs, DAMPs, and inflammatory mediators, innate sensing becomes beneficial by promoting antigen presentation and immune-cell recruitment [21,22]. Thus, the same pathway may be detrimental or beneficial depending on the timing of activation, the OV platform, the baseline antiviral state of the tumor, and whether therapeutic outcome is measured as viral replication, direct cytotoxicity, or durable antitumor immunity [23].

Importantly, antiviral resistance can be present before infection occurs. Tumors with high basal IFN tone or constitutive ISG expression may enter treatment already poised to restrict viral replication [23]. Following infection, paracrine IFN signaling can further induce ISGs in surrounding tumor and stromal cells, creating a local antiviral “firebreak” that limits intratumoral viral spread. These mechanisms help explain why tumors with intact or elevated antiviral signaling, including some glioblastomas and other difficult-to-treat malignancies, can show resistance to OVT despite being biologically aggressive [24]. Therefore, tumor-intrinsic antiviral immunity can impose a major barrier to viral oncolysis by restricting viral entry-dependent replication, protein synthesis, spread, and productive cytotoxicity. However, when infection proceeds far enough to generate antigen release and inflammation, these same pathways can also provide the immune-activating signals that support the therapeutic benefits discussed below.

Innate immune activation as a driver of therapeutic benefit

Once sufficient infection and oncolysis have occurred, innate antiviral immunity can become essential for therapeutic efficacy. In this context, oncolytic viruses function not only as cytolytic agents, but also as inflammatory stimuli that convert tumor cell death into an immunologically productive event [1,9]. Viral infection and tumor cell lysis release TAAs, DAMPs, viral PAMPs, cytokines, and chemokines that reshape the TME, recruit innate immune cells, promote antigen uptake and presentation, and help convert immunogenic cell death into adaptive antitumor responses [25,26]. Thus, the therapeutic value of OVT may depend not simply on how efficiently a virus replicates within tumor cells, but on whether infection sufficiently inflames the tumor and engages the host immune system.

A key feature distinguishing OVT from many other forms of tumor cell death is that viral replication provides an intrinsic source of immune stimulation. Viral nucleic acids and replication intermediates are detected by PRRs, ultimately leading to the induction of IFNs and inflammatory mediators that amplify local immune activation [10,27]. While these responses may further limit viral replication, they also recruit dendritic cells (DCs), macrophages, natural killer (NK) cells, and T cells to sites of infection and tumor destruction [27]. In this sense, antiviral signaling can act as an endogenous adjuvant, helping the immune system interpret tumor cell death as a danger-associated process requiring coordinated innate and adaptive immune responses.

This inflammatory environment can determine whether tumor cell debris is cleared silently or converted into an antigenic source for adaptive immunity. Viral oncolysis releases TAAs and neoantigens in the presence of signals that promote the recruitment, activation, and maturation of antigen-presenting cells. Macrophages and DCs can take up dying tumor cells, process tumor-derived material, and present antigens to T cells in draining lymph nodes or within tertiary lymphoid-like structures in the TME [28]. Cross-presenting DCs are particularly important because they link extracellular tumor antigen uptake to CD8+ T-cell priming [9,25,29]. In parallel, virus-induced cytokines and chemokines support NK cell activation, enhance T-cell infiltration, and promote cytotoxic effector responses. Together, these processes can convert local tumor cell destruction into an immunologically productive event, linking antigen release and innate inflammation to adaptive antitumor immunity [30].

The cumulative effect of these immune events is the potential conversion of poorly inflamed “cold” tumors into more immunologically active “hot” tumors. Cold tumors are typically characterized by limited T-cell infiltration, reduced antigen presentation, impaired inflammatory signaling, an immunosuppressive TME enriched in regulatory T cells, myeloid-derived suppressor cells, M2-like macrophages, and inhibitory cytokines [31]. By contrast, productive OV infection can increase inflammation, induce chemokines that recruit effector immune cells, and promote the infiltration and activation of DCs, macrophages, NK cells, and tumor-specific T cells [1,26]. This “heating” provides a rationale for combining OVT with immune checkpoint blockade and other immunotherapies, particularly in tumors otherwise poorly responsive to T-cell-directed treatments [32].

Translational studies support this immune-priming model. In patients with high-grade glioma or brain metastases, intravenous oncolytic reovirus led to detectable tumor infection, increased cytotoxic T-cell infiltration, upregulation of IFN-regulated gene expression, and induction of the PD-1/PD-L1 axis, supporting combination with PD-1 blockade [33]. Similarly, in a neoadjuvant model of triple-negative breast cancer, OV treatment before surgical resection sensitized otherwise refractory tumors to immune checkpoint blockade and prevented relapse in most treated animals [34]. Together, these studies suggest that OVT can function as a TME-conditioning strategy that increases the likelihood that checkpoint blockade will encounter an inflamed, antigen-rich, and T-cell-infiltrated tumor. Thus, the therapeutic value of innate antiviral immunity lies not only in detecting viral infection, but in bridging local oncolysis with systemic antitumor immunity.

Viral genetics and cellular context as determinants of innate immune balance

If tumor-intrinsic immunity helps determine whether infection is restricted or immunologically productive, viral genetics help determine how strongly that response is triggered, evaded, or amplified. Current OVT strategies include both DNA and RNA viruses from diverse taxonomic families. As such, currently explored OVs differ in genome organization, replication strategy, tumor tropism, capacity for genetic engineering, sensitivity to innate immune restriction, and ability to encode immune-modulatory payloads. For example, T-VEC is an engineered herpes simplex virus expressing granulocyte-macrophage colony-stimulating factor (GM-CSF), illustrating how OVs can be designed to couple tumor-selective replication with enhanced immune-cell recruitment and antigen presentation [4,5,35]. This strategy has been extended across multiple platforms, including other herpes simplex viruses engineered to co-express GM-CSF and IL-12, and recombinant adenoviruses and reoviruses expressing GM-CSF [36–39]. Together, these examples demonstrate that viral genetics can influence OVT outcomes through intrinsic differences in replication and immune sensing, and through deliberate engineering to amplify antitumor immunity.

Importantly, these engineering strategies illustrate that increasing immune stimulation is not always equivalent to improving therapeutic efficacy. The effect of a viral transgene depends on how it alters viral replication, infected-cell death, innate immune sensing, and immune-cell recruitment within a specific tumor context. For example, GM-CSF can enhance antigen-presenting cell recruitment and activation, but differences in how GM-CSF is encoded or expressed may also influence viral fitness, cytotoxicity, immunogenic cell death, and the magnitude or quality of antitumor immunity [36]. Thus, viral design choices can shift the therapeutic “set point” of OVT toward greater replication, stronger inflammation, enhanced immune priming, or, in some settings, premature antiviral restriction. These examples reinforce the need to view OV engineering as a strategy for tuning the balance between direct oncolysis and immune activation rather than simply maximizing either one.

Viral tuning, however, is not limited to engineered transgenes or rationally designed deletions. Naturally occurring viral diversity can also alter how an OV interacts with tumor-intrinsic antiviral defenses, including differences in receptor engagement, replication kinetics, IFN induction, immune evasion, and cell-death programs. This is particularly relevant for viruses with segmented genomes, where reassortment can generate progeny with new combinations of viral genes and phenotypes. In this context, reovirus provides a useful case study for how natural genetic variation can reshape the balance between replication, innate immune activation, and tumor cell killing.

The clinically developed oncolytic reovirus pelareorep, previously known as Reolysin®, is derived from the naturally occurring serotype 3 Dearing (T3D) strain, which has been widely used because of its benign clinical profile and preferential cytotoxicity toward transformed cells [40]. However, reliance on T3D also raises an important question: whether other reovirus genotypes may be better suited for specific tumor contexts. T3D is a potent inducer of IFNs and IFN-stimulated gene (ISG) expression in several systems, in part because it lacks the M1 gene-associated repression of IFN signaling described for the reovirus type 1 Lang strain [41–43]. More recent work demonstrates that even closely related T3D laboratory strains can differ in replication kinetics, IFN induction, host gene expression, and oncolytic potency [44]. These observations support the idea that subtle viral genetic differences can substantially alter the innate immune landscape of infected tumor cells. Consistent with this model, hybrid reassortant reoviruses generated from genetically distinct parental strains have shown enhanced infectivity and cytotoxicity in triple-negative breast cancer cells [45] and, in our recent work, in human fibrosarcoma and diverse epithelial cancer cell lines [46,47]. Thus, for viruses with segmented genomes, genetic reassortment provides a natural strategy for exploring how viral strain, tumor cell type, infection kinetics, and cell-death pathways interact to determine whether infection is primarily restricted, cytolytic, inflammatory, or therapeutically productive.

At the same time, reassortment should be viewed as a discovery and optimization strategy rather than an immediately straightforward path to clinical translation. Reassortant OVs would require rigorous evaluation of genetic stability, lot-to-lot reproducibility, scalable manufacturing, viral fitness, biodistribution, shedding, and safety in immunocompetent systems [48,49]. These considerations are especially important for live, replication-competent biological products, for which product identity, purity, potency, stability, and transmission risk must be carefully controlled. Thus, while reassortment can reveal viral gene combinations that improve infection, cytotoxicity, or immune activation in preclinical models, clinical development would require demonstrating that these phenotypes are stable, manufacturable, and safe across relevant tumor and host contexts.

Importantly, strain-dependent effects are unlikely to be universal across tumor types, and our recent work further supports this hypothesis [47]. The same viral genotype may produce distinct outcomes depending on tumor-cell receptor availability, entry efficiency, viral protein synthesis, basal IFN tone, inducibility of antiviral genes, and the cell-death pathways available in a given cancer cell. In this sense, viral genetics and cellular context should be viewed as interacting determinants of OVT outcome rather than independent variables. A reovirus variant that replicates rapidly and induces robust cytotoxicity in one tumor setting may be restricted by early IFN activation in another, whereas a more inflammatory but less cytolytic infection may be valuable if it enhances antigen presentation and immune-cell recruitment. Defining this balance will require comparing viral strain performance across tumor models while measuring not only viral replication and cell viability, but also IFN induction, inflammatory gene expression, immunogenic cell death, and immune activation. Transcriptomic and immunologic comparisons of host responses induced by distinct oncolytic reoviruses across cancer cell types may help identify virus–tumor pairings optimized for direct oncolysis, immune priming, or rational combination therapy.

Toward rational tuning of innate immunity in oncolytic virotherapy

Operationally, the productive immune window could be defined by paired measurements of viral replication and immune activation over time (Table 1). On the viral side, useful parameters include the fraction of infected tumor cells, intracellular viral RNA or protein accumulation, infectious viral yield, intratumoral spread, and the rate of viral clearance [12]. On the host-response side, relevant parameters include the magnitude and timing of IFN-α/β production, ISG induction, inflammatory cytokine and chemokine expression, antigen-presentation markers, and recruitment or activation of DCs, NK cells, and tumor-specific T cells [11,50]. Importantly, these variables should be interpreted kinetically rather than as static endpoints. Early and high-magnitude IFN/ISG induction may suppress viral amplification before sufficient tumor cell killing occurs, whereas delayed or intermediate innate activation may permit productive infection while preserving the inflammatory cues needed for antigen presentation and adaptive immune priming [12,50].

With these parameters in mind, combination strategies should be viewed as tools for shaping the timing, magnitude, and consequences of innate antiviral immunity. Chemotherapy and radiation therapy can increase tumor stress, antigen release, vascular permeability, and immunogenic cell death, potentially creating a tumor environment that is both more permissive to OV activity and more inflammatory [1,51]. Similarly, immune checkpoint inhibitors, adoptive cell therapies, and cytokine-based approaches may capitalize on OV-induced antigen release, inflammatory signaling, and immune-cell recruitment. In this framework, combination therapy is not simply intended to increase tumor cell killing through additive cytotoxicity, but to coordinate viral replication, innate immune sensing, tumor-cell death, and adaptive immune priming. The therapeutic goal is therefore to identify combinations that move tumors toward a productive immune state in which viral infection is sufficient to inflame and reduce tumor burden while preserving the immune signals required for durable antitumor immunity.

The effectiveness of these combinations is likely to depend strongly on therapeutic timing and sequence. Administering chemotherapy or radiation before OVT may increase tumor-cell stress, antigen availability, vascular access, or susceptibility to viral replication [52,53], whereas delivering an OV before immune checkpoint blockade may first convert a poorly inflamed tumor into a more antigen-rich and T-cell-infiltrated lesion [9,33,34]. Conversely, suppressing antiviral signaling too early may enhance viral replication but blunt the inflammatory cues required for antigen presentation and immune priming. The same intervention may therefore produce different outcomes depending on whether it is used to prepare the tumor for infection, extend viral replication, amplify immune activation, or sustain adaptive antitumor responses after viral clearance. Rational OVT combinations should therefore be evaluated not only by which agents are paired, but by when each component is delivered relative to viral infection and the evolving innate immune response [9,54].

Rational tuning of innate immunity will also require biomarkers that define the antiviral and immune state of each tumor before treatment. Baseline IFN signaling, ISG expression, PRR expression or activity, antigen-presentation capacity, and defects in JAK/STAT signaling may influence whether a tumor is permissive to viral replication or rapidly restricts infection [8,23]. At the same time, the immune contexture of the TME, including T-cell infiltration, DC abundance, PD-L1 expression, suppressive myeloid populations, and cytokine profiles, may determine whether viral infection is likely to generate productive antitumor immunity (Table 1). For example, tumors with high IFN activity may be relatively resistant to viral spread but poised for inflammatory immune activation, whereas tumors with impaired antiviral signaling may support greater viral replication but require additional strategies to promote antigen presentation and T-cell priming. Incorporating such biomarkers into preclinical studies and clinical trial design could help match viral platforms and therapeutic combinations to tumor-specific vulnerabilities rather than treating OVT as a one-size-fits-all intervention.

Clinically, defining this window will require distinguishing biological resistance from delivery failure. Lack of response may reflect insufficient viral access to tumor tissue, rapid neutralization or clearance, poor intratumoral spread, intact tumor-intrinsic antiviral defenses, immunosuppressive myeloid or stromal barriers, or failure to generate adaptive immune priming [55,56]. These mechanisms have different therapeutic implications: delivery failure may require alternative routes of administration or carrier strategies, whereas antiviral restriction may require transient modulation of IFN/ISG pathways, and immune-excluded tumors may require combinations that enhance antigen presentation, T-cell recruitment, or checkpoint responsiveness [55]. Clinical experience with pelareorep illustrates this challenge, as pelareorep plus carboplatin and paclitaxel did not improve progression-free survival in metastatic pancreatic adenocarcinoma despite evidence of immunomodulatory activity and biomarker associations [57]. Clinical trial designs should therefore incorporate paired pre-treatment and on-treatment biopsies, viral detection assays, blood-based cytokine and immune profiling, toxicity monitoring, and response endpoints that capture both local oncolysis and systemic immune activation [58,59]. Evidence supporting these candidate biomarker categories and clinical endpoints is drawn from recent systematic reviews and meta-analyses of OVT efficacy, safety, and predictive biomarkers [59–61], and is summarized in Table 1.

Finally, efforts to modulate innate immunity must account for the risk of suppressing protective antitumor responses. Transient inhibition of IFN signaling or downstream antiviral pathways may enhance viral replication in tumors with strong antiviral defenses, but broad or prolonged suppression could impair antigen presentation, DC maturation, NK-cell activation, T-cell priming, and the inflammatory remodeling needed for systemic immunity [8,62]. Conversely, strategies that intensify innate immune activation may improve immune recruitment and checkpoint responsiveness but also risk premature viral clearance or inflammatory toxicity. The central challenge is therefore not to eliminate innate antiviral immunity, but to tune its timing, magnitude, and cellular targets.

Oncolytic virotherapy occupies a unique position at the intersection of virology, tumor biology, and cancer immunology. Its therapeutic activity depends not only on whether a virus can infect and kill malignant cells, but also on how the host interprets that infection. Innate antiviral immunity is central to this process. The same pathways that detect viral infection and restrict viral spread can also provide the inflammatory signals needed to recruit immune cells, promote antigen presentation, and convert local tumor destruction into systemic antitumor immunity [8]. Thus, innate immunity should not be viewed simply as a barrier to be overcome, nor as a response to be maximized without restraint.

A useful model is that successful OVT requires a productive balance between viral replication and innate immune activation (Figure 1). Excessive early antiviral restriction may prevent viral dissemination, reduce tumor cell killing, and limit antigen release. Conversely, insufficient innate immune activation may allow local viral replication but fail to generate the inflammatory context needed for durable immune priming [8]. The optimal therapeutic window likely lies between these extremes: sufficient viral replication to infect, inflame, and reduce tumor burden, coupled with sufficient innate immune sensing to recruit and educate antitumor immunity before viral clearance or immunosuppressive feedback dominates.

Future OVT strategies should therefore move beyond a one-size-fits-all approach and toward rational matching of viral platforms, tumor contexts, and combination therapies. Viral genetics, whether through natural strain variation, reassortment, or deliberate engineering, can shape replication kinetics, IFN sensitivity, cell-death pathways, and immune activation. At the same time, tumor-intrinsic variables such as receptor availability, basal IFN and ISG profiles, antigen-presentation capacity, and cell-death competence can determine whether infection becomes abortive, cytolytic, inflammatory, or immunologically productive. Integrating these variables into preclinical testing and clinical trial design may help identify virus–tumor pairings that are optimized for direct oncolysis, immune priming, or combination therapy.

Future studies should therefore pair viral replication measurements with time-resolved immune profiling rather than evaluating either endpoint in isolation. Preclinical studies should compare OV platforms and viral variants across tumor models with distinct basal IFN/ISG states, while measuring viral growth, IFN and ISG induction, antigen presentation, immunogenic cell death, and immune-cell recruitment (Table 1). Clinically, trials should incorporate biomarker-guided stratification, paired biopsies, viral detection assays, blood-based immune profiling, and rational sequencing of combination therapies. Such studies would allow the productive immune window to be tested as a measurable therapeutic state rather than treated only as a conceptual model.

Combination approaches will be especially important for placing tumors within this productive immune window, but they must be guided by timing and mechanism. Transient modulation of antiviral signaling may improve viral replication in some settings, but excessive suppression could compromise antigen presentation, DC activation, NK-cell function, and T-cell priming [62]. Conversely, strategies that intensify inflammation may improve immune recruitment and checkpoint responsiveness but risk premature viral clearance or toxicity.

Ultimately, the goal of OVT should not be maximal viral replication or maximal innate immune activation in isolation. Rather, the field should aim to define and manipulate the conditions under which viral infection, tumor cell death, innate immune sensing, and adaptive immune priming can optimally occur. Understanding this balance will be essential for improving response consistency, designing more effective combinations, and realizing the full potential of OVs as cancer immunotherapies.

This work was supported by institutional faculty support from Elon University to E.E.R.-S. A.T.R. received financial support from the Elon Undergraduate Research Program, the North Carolina Independent Colleges and Universities, and the North Carolina Academy of Science.

The authors declare that no conflicts of interest exist.