2.1 Standard therapy: Surgery, chemotherapy, and radiotherapy

The standard treatment of glioma is dominated by surgical abscission, supplemented by radiotherapy and chemotherapy therapy.³⁰ Surgery is the most important step and the focus of it is to maximize the removal of the tumor while protecting normal brain tissue around the tumor. But limitations of functional areas of the brain and indistinguishable tumor boundaries are major factors affecting the extent of glioma resection, which need advanced technology to assist surgical treatment, such as intraoperative magnetic resonance imaging (iMRI)³¹ diffusion tensor imaging (DTI)³² and fluorescence guidance resection (FGR).³³ The aggressive nature of gliomas makes it impossible to completely cure them with surgical treatment alone.³⁴ To delay tumor recurrence, radiotherapy and chemotherapy are used as adjuvant therapies.³⁵ Radiation therapy is an important supplement to glioma treatment, with conventional radiation doses of 50–60 Gy.^36-38 It may be possible to conduct radiotherapy using an external source, an internal source, or radioactive monoclonal antibodies.³⁹ Temozolomide (TMZ) is a first-line agent for GBM therapy,⁴⁰ but some patients with TMZ-resistance have less than ideal therapeutic effect.⁴¹

However, even with surgical resection and chemoradiotherapy, the prognosis for gliomas is not ideal, with only a slight improvement in overall survival (OS). Therefore, more novel therapies need to be actively explored.

2.2 Monoclonal antibodies

Monoclonal antibodies include bevacizumab, cetuximab, nimotuzumab, and MAB-425. Bevacizumab can inhibit vascular endothelial growth factor (VEGF) to prevent tumor angiogenesis,⁴² which has been approved for GBM treatment by FDA since 2009.⁴³ In addition to bevacizumab (NCT00345163), it has also been used in combination with irinotecan (NCT00345163),⁴⁴ etoposide (NCT00612430),⁴⁵ or with concurrent radiotherapy (NCT00595322).⁴⁶ The evidence suggests that bevacizumab can prolong the progression-free survival (PFS) of patients with glioma within 18 months but can't prolong the OS.⁴⁷ Cetuximab and nimotuzumab target epidermal growth factor receptor (EGFR), which can block its ligand activity, induce apoptosis, antiangiogenesis, and thus achieve the effect of inhibiting GBM cells.⁴⁸ In a phase II clinical study of cetuximab treated in patients with relapsed GBM, cetuximab did not benefit patients from PFS and OS, regardless of whether accompanied by EGFR amplification.⁴⁹ The addition of nimotuzumab to standard therapy was not found to improve OS in a phase III open-label trial of newly diagnosed gliomas (NCT00753246).⁵⁰ In addition to several monoclonal antibody drugs described above, there are also monoclonal antibodies binding isotopes or toxins in the clinic, such as ABT-414 (NCT02343406),^{51, 52} and ¹²⁵I-mAb425 (NCT00589706).⁵³ But monoclonal antibody therapy has limited survival benefit and is highly susceptible to drug resistance.

2.3 Immune checkpoint inhibitors (ICIs)

An immune checkpoint plays a crucial role in the development of immune tolerance in tumors.⁵⁴ An ICI works by regulating immune cell activity through pathways such as co-inhibition or costimulation to kill tumor cells.⁵⁵ It has been shown in previous studies that combining anti-PD-1 (nivolumab) with anti-CTLA-4 blocking Abs (ipilimumab) does not improve the OS during the exploratory phase I (CheckMate 143).⁵⁶ Nivolumab and bevacizumab did not meet the primary endpoint of a Phase III trial of patients with recurrent GBM, because patients treated with nivolumab did not survive longer than patients with bevacizumab (NCT02017717).⁵⁷ The emergence of ICIs has brought new way to the treatment of GBM, but its current research is still not optimistic. At present, the main difficulty lies in its “cold tumor” characteristics, which have blood–brain barrier (BBB)⁵⁸ and strong heterogeneity,⁶ low tumor mutation burden,⁵⁹ and immunosuppressive microenvironment.^{30, 60, 61} In summary, ICIs in the treatment of GBM have multiple challenges and potential, requiring more in-depth research and clinical trials.

2.4 Chimeric antigen receptor (CAR) T cell therapy

In refractory hematological cancers, CAR T cell therapies have produced sustained therapeutic effects, but in solid tumors, they have not been as successful.^{62, 63} Three antigens are currently being targeted with CAR T cells for the treatment of glioblastoma: epidermal growth factor receptor variant III (EGFRvIII), human epidermal growth factor receptor 2 (HER2), and interleukin-13 receptor alpha 2 (IL-13Rα2).⁶⁴ The EGFRvIII is an active mutant of EGFR and is present in about 30% GBM patients.⁶⁵ EGFRvIII CAR-T cells showed excellent tumor growth inhibition in immunodeficient mouse models.⁶⁶ However, there are no meaningful effects in clinical for EGFRvIII-CAR T cells in GBM patients.⁶⁷ The HER2 is a GBM-associated antigen that expressed in 80% GBM patients.⁶⁸ HER2-specific CAR-T cells have shown potent antitumor activity in multiple tumor models.⁶⁹ Several clinical trials have recently reported that HER2-specific CARs have no serious systemic toxicity In GBM (NCT01109095, NCT03500991).^{70, 71} Interleukin-13 (IL-13) can modulate immune response and inflammation by binding to the IL-13 receptor α1 (IL-13Rα1) and IL-13Rα2.⁷² IL-13Rα2, expressed in over 75% of GBMs, is related to what tumor aggressiveness and poor prognosis.⁷³ And IL-13Rα2 was the first GBM target for CAR T cell therapy in 2004, which showed effective tumor lysis in human xenografts in vitro and in vivo studies.⁷⁴ Two-thirds of treated patients had excellent tolerance of CAR T cells and the great antitumor response in clinical trial.⁷⁵ Further improvements, IL-13Rα2 CAR-T cells are modified with 4-1BB or IL-15,^{76, 77} and 4-1BB-modified IL-13Rα2 CAR-T cell therapy is currently tested to treat medulloblastoma, ependymoma, and GBM (NCT04661384) in clinical, while intratumoral delivery is being assessed in refractory or recurrent malignant glioma (NCT02208362).

In addition to the cell targets above, there are some new targets such as GD2, chlorotoxin, EphA2, P32, CD133, B7-H3, and CD147.⁷⁸ The above studies have evaluated that the use of CAR T cells for brain tumors is safe and efficacious. CAR-T cell therapy is moderately effective, however, due to heterogeneous antigen expression and limited brain T cell function.⁷⁹ New strategies have been created to modify CAR-T cell function⁶⁴ such as development CAR T cells targeting multiple Tas,⁸⁰ transgenic expression of cytokines and chemokine receptors^{77, 81} and combination with radiotherapy, chemotherapy,⁸² and immune checkpoints.⁸³

2.5 Vaccines

Tumor vaccine is another kind of immunotherapy with clinical transformation prospects, based on tumor-specific antigen targets, which can instigate an immune response that is both tumor-specific and patient-specific.⁸⁴ Peptide vaccines, dendritic-cell-based (DC) vaccines, and mRNA vaccines are under investigation for glioma therapy. Peptide vaccines can induce antitumor immune responses by using small tumor-specific antigen sequences. Some peptide vaccines have achieved promising results in the study of malignant glioma, such as rindopepimut (EGFRvIII), R132H (IDH1), and Wilms tumor 1 (WT1) peptide vaccine.^85-88

Conceptually, DC-based vaccines require the patient's own DC cells, loaded with tumor lysates or peptides, and then reinfused into the patient.⁸⁴ Tumor-derived mRNA, viral antigens, and cancer stem cells also can be loaded into DCs.⁸⁹ DC vaccines are typically loaded with glioma-associated antigens (GAAs) or GAA-derived peptides, such as WT1, survivin, and IL-13Rα2.⁹⁰ More than 6 years of survival were achieved by five patients with GBM treated with a combination of chemotherapy and five synthetic peptide-pulsed DC vaccines.⁹¹ Glioma lysate-pulsed DC vaccines have shown promising results in early clinical trials. A phase III glioma clinical trial reported OS of 23.1 months in postoperative patients treated with tumor-lysate pulsed DC vaccine (NCT00045968).⁹² DC vaccines pulsed with glioma stem cells (GSCs) are able to induce interferon (IFN)-γ production to initiate T cell-mediated antitumor immunity.⁹³ Autologous DC vaccine pulsed with GBM stem-like cell lysates were used in 11 newly diagnosed patients with GBM and 25 recurrent patients, results demonstrated improved clinical benefits in OS (NCT02010606).⁹⁴

While clinical studies have demonstrated that peptide vaccines and DC vaccines can exert certain antitumor effects, the heterogeneity of gliomas is a challenge to overcome. To overcome these limitations, it may be important to combine the vaccine with surgical treatment, chemoradiotherapy, and ICIs on the basis of optimizing the DC vaccine.⁹⁵

2.6 Physical therapy

The tumor treating field (TTFieds) can inhibit tumor cell division and proliferation by interfering with the normal process of mitosis and inhibiting the normal septin localization.⁹⁶ TTFieds has been approved by the FDA for the diagnosis and recurrent GBM due to multiple great clinical trial results. A pilot clinical trial was initiated after TTFields were found to inhibit cancer growth in multiple animal tumor models. The median time to disease progression and OS are more than double historical control patients.⁹⁷ The clinical trial results for NovoTTF-100A demonstrated no significant improvement in the median OS and PFS but the less toxic and favored quality of life than chemotherapy (NCT00379470).⁹⁸ Another multicenter phase III clinical trial proved that the addition of TTFields to maintenance temozolomide chemotherapy can significantly improve the survival rate of patients with glioblastoma (NCT00916409).⁹⁹ Additionally, the synergy of the triple combination of RT, TTFields, and TMZ was explored in the clinical phase I trial（NCT03477110). Treatment with TTFields combined with scalp-sparing chemoradiation is feasible, well-tolerated, and relatively safe.¹⁰⁰ A phase III clinical trial (NCT04471844) is currently recruiting to investigate the clinical benefit of TTFields with chemoradiation treatment. Other clinical studies also found that the combination of TTFields therapy and drug therapy can improve PFS and OS in GBM, such as Bevacizumab (NCT02743078), marizomib (NCT02903069), nivolumab, and ipilimumab (NCT03430791). Although TTFields therapy is a helpful adjuvant therapy to chemotherapy in patients with GBM, its extensive application is limited by the expensive treatment cost and inconvenient equipment.

Despite extensive research, treatments that truly benefit patients have not yet been developed. Traditional therapy is the most common, but surgical methods are extremely complex, elaborate and expensive, and chemoradiotherapy has serious side effects. For high-grade gliomas, even after standard treatment with surgery and chemoradiotherapy, the median OS is between 15 and 23 months and 5-year survival rate is low.¹⁰¹ Monoclonal antibodies are guidelines recommended for GBM targeting with high targeting specificity. However, due to their large molecular size, they may not easily pass through the BBB. Multiple monoclonal antibodies have shown no advantage in prolonging survival.¹⁰² ICIs have shown promising results in preclinical experiments, but clinical trial results have not been satisfactory. Clinical trials evaluating ipilumamab, pembrolizumab, and nivolumab did not meet the primary endpoint.¹⁰³ No significant clinical benefit has been observed in clinical trials of CAR-T cell drug candidates for glioblastoma. The Immunotherapies for glioblastoma described above are still limited and need to be further optimized. The TTFieds is a new and safe method, however, the device is expensive and requires prolonged wear, which can affect the quality of life of patients. Multiple clinical trials have demonstrated no OS improvement with NovoTTF alone, but combining it with other treatment strategies can improve the treatment effect.¹⁰⁴ Oncolytic virus therapy is a promising cancer treatment with positive results from many clinical trials. Delytact is the first oncolytic virus approved for the treatment of glioma. Clinical results have shown a significant improvement in survival and safety compared to standard therapies. The 1-year survival rate of the oncolytic virus treatment group was 92.3% compared with the 15% 1-year survival of standard treatment.¹⁰⁵ Therefore, oncolytic viral therapy is likely to be a cost-effective modality to treat gliomas.

3.1 Modification of OVs

In the study of glioma, many viruses have been proposed including Adenovirus, Herpes simplex virus (HSV), poliovirus, vaccinia virus (VV), measles virus (MV), retroviruses, Newcastle disease virus (NDV), reovirus, parvovirus, and so on (Figure 2). Although early OVs have a degree of oncolytic activity, their specificity is not very high. Therefore, scientists have begun to improve tumor-specific selection and safety of OVs using various methods, such as deleting virulence genes, loading tumor-specific promoters, and carrying tumor suppressor gene miRNAs.¹⁰⁷ HSV-1 is often deleted by the ICP34.5¹⁰⁸ and ICP47 genes¹⁰⁹ to attenuate the pathogenicity of the virus; oncolytic adenoviruses often delete E1B-55K or E1AΔ24 to ensure that the virus selectively replicate in tumor cells^{110, 111}; oncolytic poxviruses increase its selectivity to tumor cells by deleting the TK gene.¹¹² Expression of oncolytic virus genes initiated by tumor-specific promoters, including human telomerase reverse transcriptase promoter (hTERT), E2F-1 and glial fibrillary acidic protein (GFAP) promoter can significantly improve oncolytic virus' anti-tumor specificity.^{113, 114} With the continuous development of viral biology, tumor immunology, and molecular genetics, the modification of OVs is also constantly advancing. There are currently five strategies for arming OVs as follows¹¹⁵: (1) encoding T cell co-stimulating molecules to improve APC function^116-118; (2) Arming OVs with chemokines to recruit tumor lymphocytes in TME^119-121; (3) Arming OVs with cytokines to improve the function of antitumor lymphocytes in TME^122-124; (4) Arming with antigens with OVs, which can be used as a tumor vaccine^{125, 126}; (5) Arming OVs with ICIs to relieve immunosuppression of TME.^127-129 We have compiled the main modifications of these OVs (Table 3).

Oncolytic adenovirus is a commonly used oncolytic viral vector, and its most important modifications are the deletion of viral genes and the insertion of RGD, tumor-specific promoters, and functional proteins gene.¹³⁰ ONYX-015 is an oncolytic adenovirus with a deletion of the E1B-55K gene.¹³¹ The E1B-55kD protein can bind to p53 to inhibit p53-mediated apoptosis.¹³² The lack of E1B-55K enables the virus to replicate in p53-deleted tumor cells. ONYX-015 demonstrated significant antitumour activity in xenograft models of glioblastoma.¹³³ E1A-delta24 (Delta-24), a tumor-selective adenovirus, had a 24-bp deletion in the Rb-binding region of the E1A gene, which can render it cancer-specific. Delta-24 inhibited tumor growth in nude mice with D-54 MG by 66.3% with a single dose, and by 83.8% with multiple injections.¹³⁴ Delta-24-RGD (DNX-2401) has an RGD-4C peptide motif of the fiber knob compared to Delta-24, which allows the adenovirus to enter cells by binding directly to integrins. To enhance the virus-mediated immune response, Delta-24-RGDOX (DNX-2440) was constructed to express the immune co-stimulator OX40 ligand (OX40L). Compared with its predecessor (DNX-2401), Glioma mouse models treated with DNX2440 exhibit increased T-cell responses and a prolonged survival rate.¹³⁵ This virus is currently being tested in a phase I clinical trial in patients with recurrent GBM by stereotactical injection (NCT03714334). CRAd-S-pk7 was created by inserting survivin promoter (S) and poly-lysine (pk7) inserting fiber region to target glioma.¹³⁶ CRAd-S-pk7 can inhibit xenograft tumor growth and 67% of treated mice were long-term survivors.¹³⁶ ONYX-015, DNX-2401, CRAd-S-pk7 are currently being explored in clinical trials. ICOVR-5, ICOVIR17, and VCN-01 had shown good anti-glioma activity in preclinical experiments. ICOVIR-5 encompassed E1A partial deletion in Rb protein–binding region, RGD-4C insertion and E2F1-responsive promoter to improve the selectivity and effectiveness of viral replication.¹³⁷ ICOVIR-5 significantly prolonged the median survival and 37% without relapse in xenografted mice.¹³⁸ ICOVIR-17 is a modified version of ICOVIR15 expressed human hyaluronidase PH20.¹³⁹ By enhancing the degradation of HA and viral spread, ICOVIR-17 showed significant antitumor effects and prolonged survival in mice.¹⁴⁰ Compared with ICOVIR-17, VCN-01, improved the infectivity, because its RGD was inserted in the fiber shaft instead of the fiber knob. VCN-01 displayed a pronounced cytotoxic effect on glioma cells and increased OS significantly in orthotopic glioma models in vivo.¹⁴¹ There have been no clinical trials of these viruses for gliomas.

HSV is a common candidate viral vector for OV due to its safety, large gene capacity, and ability to infect a variety of cells.¹⁴² Oncolytic herpes simplex viruses (oHSVs) are usually designed by deleting one or more virulence genes to improve their safety. Dlsptk, the HSV-1 mutant with the deletion of the TK gene, was the first genetically engineered HSV to be used in the treatment of glioblastoma.⁸ HSV-1716, G207, G47Δ, and rQnestin34.5 are members of oHSVs in which γ34.5 has been deleted or modified.¹⁴³ In addition to the loss virulence genes, they might be further modified to express different immunomodulatory or anticancer transgenes in pursuit of better treatment. A mIL-12 expressing OHSV, M002, have demonstrated the antitumor efficacy in intracranial syngeneic neuroblastoma murine model and human glioma xenograft tumor models.^{144, 145} A clinical trial is underway to determine whether IL-12-encoding virus (M032) can be effective in treating glioblastoma (NCT02062827). HSVs can also be engineered to improve infection effects based on cancer cells' specific receptor expression. The EGFR, EGFRvIII, HER2, which are highly expressed on the surface of glioblastomas, have received significant attention. EGFR-retargeted HSV (KEN, FGE-4), EGFRvIII-retargeted HSV (R-613, KMMP9), and HER2-retargeted oncolytic HSV (R-LM249, R-115) not only conferred tumor specificity but also showed promising anti-GBM efficacy.¹⁴⁶

Poliovirus is an envelope-free single-stranded RNA virus, a natural neuro-pathogen with inherent neuro-aggression.¹⁴⁷ Poliovirus invades cells via the receptor CD155 on the cell surface, which has strong antitumor potential due to its wide expression in solid tumors such as gliomas.¹⁴⁸ Since the internal ribosome entry site (IRSE) is a key neuro-causative factor for the poliovirus, it needs to be modified to reduce neurovirulence. PVSRIPO, which is the IRES is replaced by human rhinovirus type 2 (HRV2).¹⁴⁹ Preclinical experiments have proved that PVSRIPO can infect and lyse glioma cells and can effectively inhibit tumor growth and improve survival rate in animal models of glioma.^{150, 151}

VV is a double-stranded DNA virus that replicates in the cytoplasm of host cells, which has the advantages of wide range of infection, strong tumor targeting, and large gene capacity.¹⁵² Although unmodified VV has oncolytic effects, it has serious side effects. TK-knockout VV has high tumor targeting and safety due to the deletion of the TK gene. To further improve the oncolytic effect of VV, double-, triple-, and quadruple-deletion mutant VVs have been generated, such as TG6002, GL-ONC, and T601.¹⁵³ TG6002 is an attenuated VV with targeted deletions of the J2R gene and the I4L gene engineered to express the yeast FCU1 gene, which encodes the ribonucleotide reductase.¹⁵⁴ TG6002 is the only modified VV for glioma clinical trials. The combination of TG6002 and 5-fluorocytosine (5-FC) has shown significant antitumor effects in a variety of human xenograft tumor models.¹⁵⁴ The safety and efficacy of TG6002/5-FC for local chemotherapy in patients with recurrent glioblastoma are being tested in phase I/II clinical trial (NCT03294486).

MV is an enveloped virus with unsegmented negative-stranded RNA, belonging to the paramyxoviridae family.¹⁵⁵ MV relies on CD150, CD46 and nectin-4 cell membrane receptors to enter tumor cells.¹⁵⁶ The high expression of CD46 receptor on the surface of malignant glioma cells makes MV a good oncolytic viral vector for the treatment of glioma. Engineered oncolytic MVs derived from attenuated strains, are engineered to express either human sodium–iodine transporter (NIS) or human carcinoembryonic antigen (CEA).¹⁵⁷ Compared with MV-CEA, MV-INS has more significant antitumor activity in vitro and in vivo. More importantly, MV-INS can be used in combination with local radiotherapy to treat gliomas.¹⁵⁸ MV-INS only conducted clinical studies in children and young adults with recurrent medulloblastoma or recurrent ATRT (NCT02962167), but MV-CEA is currently in clinical testing for gliomas.

In addition to the above genetically modified OVs, there are also many viruses with natural oncolytic activity. NDV is an enveloped, single-stranded RNA virus, which can cause local infections in humans with mild symptoms.¹⁵⁹ Various animal models and human clinical tests have confirmed that NDV is nonpathogenic, relatively safe, and has no side effects in humans. Thus, attenuated strains of NDV are currently used in the study of OVs.¹⁶⁰ At present, other NDV-based strategies for the treatment of malignant glioma are being explored, such as autologous tumor vaccine modified with nonlytic NDV (ATV-NDV).¹⁶¹ Reovirus are nonenveloped virus with 9–12 linear double-stranded RNA that occurs naturally in the mammalian respiratory and intestinal systems.¹⁶² Reovirus has a targeted lytic effect on cells activated by the RAS pathway, such as malignant gliomas cells and breast cancer cells.¹⁶³ Clinical trials and preclinical studies of Reovirus have shown great promise.¹⁶⁴ H-1 parvovirus (H-1PV) is a self-propagating virus with oncolytic and tumor-suppressing activity and is nonpathogenic in humans.¹⁶⁵ H-1PV interacts with galactactagglutinin-1 to enter cancer cells, and the LGALS1 gene (encoding galectin-1) is overexpressed in glioblastoma cells.¹⁶⁶ In immunocompetent rat models of rat (RG-2 cell-derived) and immunodeficiencies rat models of human (U87 cell-derived) glioma, a single stereotactic intratumor-injected H-1PV and multiple systemic (iv) application of H-1PV significantly improved survival in rat and human glioma models.¹⁶⁷

3.2 Oncolytic virus in clinical trials

The progress of OVs in preclinical research of glioma experiments has further promoted the translation into clinical trials. Among them, DNX-2401 and G207 play an important role in the research of glioma treatment. Next, we focus on OVs that have entered clinical trials. Table 4 provides a summary of clinical trials for OVs for gliomas.

ONYX-015 is one of the first OVs for cancer therapy. Upon testing ONYX-015 on 24 patients with recurrent gliomas, good safety, and immune cell infiltration were observed, but no antitumor effects were observed.¹⁶⁸ However, encouraging effects were seen in head and neck cancer patients by the combination of ONYX-15 and chemotherapy.¹⁷⁹ In 2005, H101 (Oncorine), similar to ONYX-015, was approved for head and neck cancer in China.¹⁸⁰

DNX-2401 is a recombinant oncolytic adenovirus that carries two genetic modifications. One is the absence of 24 bp in E1A gene. The second is the insertion of the RGD-4C motif in the HI ring of the fiber. Adenoviruses enter cells by binding to Coxsackievirus and adenovirus receptor (CAR) on the cell surface.¹⁸¹ However, glioma cells express lower levels of CAR than normal cells.¹⁸² Inserting the sequence encoding the RGD peptide into the HI loop of the fiber knob can improve the efficiency of oncolytic adenovirus glioma cells.¹⁸³ In U87MG xenografted mice, DNX-2401 resulted in longer survival than Delta-24.¹⁸⁴ Naiara et al. evaluated the anti-glioma effect of DNX-2401 in pediatric high-grade glioma (pHGG) and diffuse intrinsic pontine glioma (DIPG) models. Preclinical results shown that DNX-2401 has a significant antitumor effect in pHGG and DIPG orthotopic immunosuppressed and immunocompetent models. DNX-2401 administration triggers an immune response against tumors in addition to its oncolytic effects.¹⁸⁵ A clinical trial first assessed the safety and maximum tolerated dose (MTD) injected directly around the glioma tumors (NCT00805376). DNX-2401 improved long-term survival in recurrent HGG, which may be due to the direct oncolytic effect and the anti-glioma immune response triggered by increasing cytotoxic T-cell infiltration.¹⁸⁶ Preclinical trials have shown synergistic effects between DNX-2401 and TMZ.¹⁸⁷ The tolerance of DNX-2401 along with TMZ was tested in malignant glioma patients (NCT01956734). DNX-2401 and temozolomide was well tolerated and showed therapeutic activity. Interestingly, fibroblast growth factor 2 (FGF2) can be a prognostic biomarker of DNX-2401 treatment, increased FGF2 expression was correlated with OS lengthening.¹⁸⁸ DNX-2401 was well tolerated as monotherapy while the addition of IFN-γ did not provide additional benefits or improve survival rates (NCT02197169).¹⁸⁹ DNX-2401 combined with pembrolizumab has been evaluated in a phase II trial (NCT02798406), but no results have been released. A phase I clinical trial tested the safety of convection-enhanced delivery (CED) of DNX-2401 in patients with recurrent GBM. CED of DNX-2401 was proved to be safe and feasible in the tumor and in surrounding brain with a local inflammatory reaction.¹⁶⁹ A phase I clinical trial of allogeneic bone-marrow-derived human mesenchymal stem cells (BM-hMSCs) loaded with the DNX-2401 via intra-arterial injection in patients with recurrent high-grade glioma tested the ability of BM-hMSCs to deliver DNX-2401 (NCT03896568). The new technique of perfusion guided-ESIA injections (PG-ESIA) enhances the ability of BM-hMSCs-DNX-2401 to perform targeted delivery to brain tumors.¹⁹⁰

CRAd-S-pk7 is a new oncolytic virus that is being studied in clinical trials. A phase I clinical trial will determine the safety and MTD of neural stem cells loaded with CRAd-S-pk7 (NSC-CRAd-S-pk7) in newly diagnosed malignant glioma patients combined with standard chemoradiotherapy (NCT03072134). The median PFS was 9.1 months and median OS was 18.4 months. The NSC-CRAd-S-pk7 trial proved feasible and safe, and 1.50 × 10⁸ NSCs loading 1.875 × 10¹¹ viral particles were recommended for further trial.¹⁹¹ Jennifer et al. proposed a regimen by in-brain administration of multiple doses, which achieved good therapeutic results in IND studies (IND 19532).¹⁹²

HSV-1716 has a 759 bp deletion in each copy of ICP34.5 gene.¹⁹³ ICP34.5 is a neurovirulent protein of HSV, the ICP34.5 gene mutation may be useful for glioblastoma selectivity.¹⁹⁴ Phase I clinical trial have demonstrated the safety of HSV1716 when injected into the brain tumor resection cavities after surgical resection in HGG.¹⁷⁰ However, another phase I study for refractory or recurrent HGG in younger patients has been terminated without reason (NCT02031965).

G207 was derived from HSV-1 by deleting ICP34.5 gene and inserting LacZ gene into ICP6 gene, which can increase glioblastoma selectivity.¹⁹⁵ In some preclinical trials, the antitumor activity and safety of G207 were proved in different tumor models, especially glioblastoma.^196-198 Phase I/II and phase Ib studies (NCT00028158) also demonstrated the safety and effectiveness of G207 in recurrent malignant gliomas, even within a high dose of 3 × 10⁹ plaque-forming units (pfu).^171-173 Another phase I study of G207 alone or with radiation in recurrent GBM shows safety and radiographic responses (NCT00157703).¹⁹⁹ The combination of G207 and radiotherapy in patients with recurrent or progressive pHGG prolonged the median OS(NCT02457845).^{200, 201} Additionally, G207 alone or combined with a single low dose of radiation in recurrent or progressive pHGG was tested to assess the efficacy and tolerability of G207 and to survey for virologic shedding following G207 (NCT04482933). Phase I trial of G207 of refractory or recurrent cerebellar brain tumors is currently recruiting (NCT03911388).

G47Δ, originated from G207 by a lack of the nonessential α47 gene, showed a greater replication capability and a higher antitumor efficacy than G207.²⁰² G47Δ exhibited antitumor efficacy in several tumor types, including malignant glioma.^203-205 It was designated as Sakigake breakthrough therapy by the MHLW in February 2016, which allowed its expedited approval.²⁰⁶ G47∆ showed no toxicity or serious adverse events in progressive GBM patients (UMIN000002661).²⁰⁷ A phase II study tested the efficacy of G47∆ in 19 patients with residual or recurrent, supratentorial glioblastoma after radiation therapy and TMZ (UMIN000015995). This study showed that the 1-year survival rate of 84.2% and the median OS and PFS of 20.2 months and 4.7 months, which compared favorably with other treatments.²⁰⁸ The good survival benefit and safety profile led to the approval of G47∆ as the first oncolytic virus product for malignant glioma by MHLW on 11 June 2021 in Japan.

rQNestin34.5 can control the expression of ICP34.5 via the nestin-1 promoter, allowing it to replicate selectively in glioma cells.²⁰⁹ rQNestin34.5 showed anticancer efficacy and safety in mouse GBM.²¹⁰ This virus is now assessing safety in malignant glioma patients in clinical trial (NCT03152318).

The C134 virus is a replication-competent HSV that express the human cytomegalovirus (HCMV) IRS1 genes. IRS1 protein can allow viruses to evade PKR-mediated protein shutoff and maintain late viral protein synthesis, resulting in improving virus replication/production and killing ability.²¹¹ Compared with γ34.5-deleted HSV, C134 reduced tumor volumes and improved survival in two murine brain tumor models.²¹² C134 directly administered in the CNS demonstrated safety in HSV-susceptible CBA/J mouse and nonhuman primates.²¹³ The safety of C134 is testing in recurrent GBM administered intratumorally (NCT03657576).

Many clinical trials of PVSRIPO for gliomas have been carried out. Sixty-one patients with recurrent GBM treated with PVSRIPO in a phase I clinical trial. Results showed that PVSRIPO had no neurotoxicity, and the survival rate of 24 and 36 months among patients was extended (NCT01491893).²¹⁴ PVSRIPO was tested to determine the efficacy and potential toxicity in two ongoing clinical trials (NCT03043391, NCT02986178), where PVSRIPO will be delivered to the tumor by CED. A phase II open-label, single-arm study of PVSRIPO and pembrolizumab for recurrent glioblastoma is currently ongoing (NCT04479241).

Since MV-CEA showed potent antitumor activity and was shown to have synergistic effects with radiotherapy in vitro malignant glioma models.^{215, 216} MV-CEA was studied in treating glioblastoma multiforme patients (NCT00390299) to determine the side effects and dosage. The order of surgery and oncolytic virus injection was different between the two groups. The first group of patients undergo tumor resection on Day 1, followed by MV-CEA administered. In the second group, patients receive MV-CEA inside the tumor on Day 1, and the tumor mass is removed on Day 5 and then administered MV-CEA into the resection cavity. There was no DLT after intracranial MV-CEA doses of up to 107 TCID50 were used, but the final experimental results were not published.²¹⁷

Vocimagene amiretrorepvec (Toca511) is derived from a gamma retroviral replicating vector that encodes cytosine deaminase, which converts the prodrug 5-FC into the potent anticancer drug 5-fluorouracil (5-FU) in infected tumors.²¹⁸ Animal tests have shown that Toca511 replicated stably in glioblastoma xenograft models and after 5-FC administration resulted in complete regression of tumor and prolonged survival.²¹⁹ A study of recurrent HGG treated with Toca511 and 5-FC (NCT01470794) has shown durable complete responses. The median duration for responders is 35.7+ months.²²⁰ The study of Toca511/5-FC in patients with recurrent malignant glioma has ended, but the results have not yet been published (NCT01156584). A phase II/III study (NCT02414165) of comparing Toca511/5-FC with standard therapy for recurrent HGG has been terminated since neither improved OS nor other efficacy endpoints compared to the control group.²²¹

MTH-68/H and NDV-HUJ are attenuated strains of NDV currently used to treat gliomas. Four patients with high-grade glioblastoma multiforme were treated with MTH-68/H after the failure of traditional antitumor therapy, and the quality of life was improved, and the survival time was extended to 5–9 years.¹⁷⁴ A Phase I/II trial was conducted to determine the safety and tumor response of repeated intravenous administration of NDV-HUJ. Five of the 11 participants developed Grade I/II constitutional fever and one had complete remission.¹⁷⁵ In general, NDV treatment of malignant glioma has been preliminarily confirmed to be safe and reliable, and has certain efficacy, but there are few relevant clinical studies and no major breakthroughs have been made.

Reolysin is a wild-type reovirus serotype-3 strain.²²² A Phase I clinical trial of intra-tumoral Reolysin infusion for treating recurrent malignant gliomas in adults has been completed (NCT00528684). Although dose-limiting toxicity was not identified and MTD was not reached, anti-glioma activity was shown.¹⁷⁶ Samson et al. achieved breakthrough success in the treatment of malignant glioma through intravenous injection of Reolysin, which proved for the first time that reovirus can pass through the BBB, replicate and kill tumor cells, and stimulate T cell activation to exert antitumor immunity.¹⁷⁷ Reolysin in combination with sargramostim in treating younger patients with high-grade relapsed or refractory brain tumors is currently being evaluated in phase I clinical trial (NCT02444546).

ParvOryx is a H-1 parvovirus (H-1PV). Due to good preclinical experimental data showing that H-1PV has strong cytotoxicity and tumor suppressor effects in glioma cells, it led to the phase I/IIa trial of ParvOryx01 in patients with recurrent GBM to tested the safety, tolerance, and efficacy (NCT01301430). It was found that ParvOryx can pass through the BBB and induce the establishment of immunogenic tumor microenvironment (TME).²²³ Another phase I/IIa study also demonstrated the ability of parvovirus to switch the immunosuppressed TME toward immunogenicity, even when administered systemically.¹⁷⁸

3.3 Novel OVs

A number of other novel oncolytic viral vectors have also been investigated for gliomas, but no clinical trials have been conducted yet. The safety and efficacy of different OV have been validated in preclinical trials, and these favorable results will provide valuable data for clinical translation.

The Zika virus (ZIKV) is a flavivirus of the Flaviviridae family. The virus tends to infect neural progenitor cells, which can cause severe neuropathy.²²⁴ Multiple studies have confirmed that ZIKV preferentially targeted GSCs, exerted oncolytic effects, and prolonged the survival time of tumor-bearing mice in a dose-dependent manner.^{225, 226} Kaid et al. also analyzed the safety and antitumoral effect of Zika virus intrathecal injections in three dogs bearing spontaneous CNS tumors. The results showed that the tumor regressed without adverse reactions.²²⁷ A recent study indicated that ZIKV induced a strong pro-inflammatory response and increased CD4 and CD8 T cell intra-tumoral infiltration and activation in GBM mouse models, which can improve immunotherapy efficacy.²²⁸ Taken together, Zika virus has shown good therapeutic prospects in brain tumors.

Vesicular stomatitis virus (VSV), a negative-sense single-stranded RNA virus, is extremely sensitive to the inhibitory action of interferons (IFNs).²²⁹ VSV Δ51 is a VSV mutant with a deletion of amino acid at the 51st position of its M protein.²³⁰ It could kill glioma cells efficiently and could infect and destroy the TMZ-resistant brain tumor stem cells (BTSCs) in combination with vvDD in vitro.²³¹ RVSV (GP), without natural neurotoxicity but with potent oncolytic activity, in which envelope glycoprotein (GP) is replaced by a mutant glycoprotein of lymphocytic choriomeningitis virus (LCMV-GP), can effectively eliminate brain cancer in multiple preclinical tumor models. Remarkably, rVSV(GP) escapes humoral immunity, thus, allowing repeated virus administrations.²³² VSV-EBOV virus, a chimeric VSV expressing the EBOV GP selectively targeted brain tumors and improved survival in tumor-bearing mice.²³³ Furthermore, a novel recombinant VSV: G protein less (GLESS)-fusion-associated small transmembrane (FAST)-VSV (GLESS-FAST-VSV) significantly prolonged the survival of normal-immunity animals harboring brain tumors through multiple intra-tumoral injections.²³⁴ As these promising experimental results, we can anticipate the clinical application of such viruses in the near future.

Myxoma virus (MXVY) belongs to the Poxviridae family and is a prototype for the genus Leporipoxvirus.²³⁵ MXVY showed efficacy in murine xenograft model of human glioma and models of medulloblastoma, and rapamycin can further enhance the oncolytic potential of MXVY.^{236, 237} M011L-deficient oncolytic MXVY (vMyx-M011L-KO) induced apoptosis in brain tumor-initiating cells and enhanced survival in an immunocompetent glioblastoma mouse, suggesting that it would be an ideal candidate to translate vMyx-M011L-KO to the clinic.²³⁸ More recently, IL15Rα-IL15-armed oncolytic MXVY provided potent antitumor effects, eliminating gliomas in 83% of treated mice.²³⁹

Another virus worth mentioning is the Seneca Valley virus-001 (SVV-001). Studies confirmed that SVV-001 could efficiently pass through BBB and eliminate medulloblastoma and pediatric glioma with intravenous injection.^{240, 241} Despite these promising results, the effect of SVV-001 in immunocompetent models was not well defined.

4.1 Delivery of OVs

In current experimental studies, the common route delivery of OVs is intra-tumoral injection. The intra-tumoral injection has been applied in most glioma trials, because it can overcome BBB penetration, maximize OVs concentration at the site of cancer to exert antitumor effects, protect the integrity of the OV from the immune system, and lead to robust immunological response.²⁴² However, intra-tumoral administration often requires neurosurgery, and the complexity of surgery and uncertainty about tumor location make it difficult to multiple administrations. A single high-grade dose is often associated with risks such as bleeding, infection, and tissue damage. Therefore, the safety and convenience of intra-tumoral administration need to be further optimized. CED is a micro-invasive technique that uses a catheter to establish a pressure gradient to locally deliver therapeutics into brain.²⁴³ CED also can significantly limit systemic toxicity by directly delivering drugs to the site of cancer. Multiple experiments have proven that CED can produce effective delivery of OVs to brain tumors.

Theoretically, the systemic administration is a desirable delivery method because it is easy to operate and can be administered repeatedly, and the virus can be widely distributed to treat primary and metastatic tumors. However, the BBB is the biggest obstacle to the systematic administration of OVs for glioma. BBB plays an important role in controlling the microenvironment of the CNS to allow for proper neuronal function.²⁴⁴ At the same time, the BBB is a restrictive vascular barrier that limits the effectiveness of drugs delivered to brain tumors, including OVs. Even so, there are several OVs that can cross BBB to reach brain tumors, such as VV, H-1PV, SVV-001, SFV,²⁴⁵ and M1 virus.²⁴⁶ Experiments have shown that the system-injected oncolytic reovirus can be carried by immune cells to cross the BBB.¹⁷⁷ This provides new ideas for OVs to cross the blood-brain by being loaded on brain-targeted cells, such as neural and mesenchymal stem cells.²⁴⁷ Currently, BM-hMSCs-DNX-2401 (NCT03896568) and NSCs-CRAd-S-pk7 (NCT03072134) are validating the feasibility of this strategy. Endovascular selective intra-arterial (ESIA) infusion can deliver drugs into glioma tumors when injected systemically, enhancing drug delivery and targeting tumor vasculature selectively.²⁴⁸ ESIA infusion of mesenchymal stem cells loaded with delta-24 in a canine model showed feasibility and safety and supported the clinical trial of ESIA infusion of NSCs-CRAd-S-pk7.²⁴⁹ Furthermore, focused ultrasound-mediated intranasal delivery (FUSIN) can circumvent the BBB and enhance the transport of intranasally administered drugs at the FUS-targeted brain location.²⁵⁰ Experiments have shown that FUSIN can safely and effectively deliver AAV to spatial target brain locations, showing the potential to deliver OVs to glioma tumors.²⁵¹

4.2 Antitumor immunity and antiviral response

Besides the direct oncolysis, OVs can enhance innate or adaptive immune responses to exert antitumor. However, the immune response can also inhibit viral replication due to the antigenicity of the oncolytic virus itself, allowing the body to produce antibodies that neutralize the virus, thereby culminating in the clearance of viruses and reducing the therapeutic effect.²⁵² Therefore, how to balance the immune response and the antiviral response to support the virus to exert its antitumor function is a major challenge.

Viral infection of peri-tumoral cells can trigger an innate antiviral inflammatory response that releases a variety of stimulating cytokines, such as IFN that can trigger an antiviral response in TME. Several studies have shown that upregulation of endogenous IFN signaling levels is detrimental to the treatment of OVs.²⁵³ Therefore, combination therapy with OV and interferon signaling modulators/inhibitors is a strategy to reduce antiviral response immune. Ruxolitinib, JAK/STAT pathway inhibitors that can modulate the type I IFN response, increases the oncolytic efficiency of MV in patient-derived GBM xenografts.²⁵⁴ In addition, rapamycin and cyclophosphamide (CPA), which can inhibit or delay antiviral activity in the body, have been used to improve therapeutic outcomes of oncolytic virus in malignant glioma tumor models.²⁵⁵

To protect OV from nAbs that are produced by self-immunogenicity, encapsulation of OVs has been extensively explored. Masking viral surface proteins with biomaterials such as polymers, nanoparticles, liposomes or copolymers enhances protection against nAbs during delivery and targets tumors to some extent. Polyethylene glycol (PEG), polysaccharides, cationic polymers, and cationic lipids are the most commonly used biomaterials.²⁵⁶ However, the strategies still have limitations. Chemical modification approaches are not usually heritable and viral bioactivity might be impaired by excessive shielding. Another method of shielding is the use of extracellular vesicles (EVs), which are lipid-bilayer membrane structures secreted by most cells. EVs can cross the BBB and placental barrier more efficiently, and have great potential as a carrier of natural medicine.²⁵⁷ Tumor cell-derived microparticles have proven to be effective vectors for oncolytic adenoviruses.²⁵⁸ Cell-based delivery vectors are among the most promising delivery vectors for OVs. OVs loaded into cells can circumvent the recognition of OVs by the main immune system of the host. Mesenchymal cells and NSCs have been shown to have tumor-homing properties so they can be used to deliver OVs to tumors.^{259, 260} In clinical trials, BM-hMSCs and NSCs are being applied to deliver oncolytic adenovirus to treat gliomas. Methods for genetically modifying oncolytic virus surface proteins have been developed to make OVs more resistant to the host's immune system. Surface antigens of OVs can be replaced by those of other viruses so that they escape from pre-existing immune responses. VSV-GP enhanced infectivity for glioma cells and demonstrated resistance to complement-mediated inactivation.^{261, 262} Besides, Ana Mato-Berciano et al generated an oncolytic adenovirus with the albumin-binding domain (ABD) on the hexon that also can evade neutralizing antibodies.²⁶³

4.3 Immunosuppressive TME

Malignant gliomas have a highly immunosuppressive TME that becomes resistant to immunotherapy. To improve antitumor immune responses, we need a thorough understanding of the TME of gliomas. Among the components of the TME are extracellular matrix (ECM), signaling molecules, stromal cells, and immune cells.²⁶⁴ The composition of the ECM and hardness of the tumor-associated ECM change in GBM patients.²⁶⁵ The spread of OVs is hindered. Tumor-associated myeloid cells (TAMCs) are key drivers of immunosuppression in the TME, constituting up to as much as 50% of the cellular mass in tumors.²⁶⁶ TAMs, one of the most numerous subtypes of TAMCs, can release immunosuppressive factors to exert immunosuppressive effects in the TME and negatively correlate with OS in malignant glioma patients.²⁶⁷ Myeloid-derived suppressor cells (MDSCs) are other cells with potent immunosuppressive effects in the TME. In addition to secreting immunosuppressive cytokines, MDSCs can inactivate CD8 cells and disable TCR.²⁶⁸ T cells play a vital role in the adaptive immune response to malignancies. However, cytotoxic T cells are downregulated in the TME of gliomas due to T cell aging, depletion, tolerance, and incompetence.³⁵ NK cells are affected by immunosuppressive factors within TME. The unique MHC-I molecule expressed by glioma cells can also bind to receptors on the surface of NK cells, thereby inhibiting their function.²⁶⁹ DCs are also affected by immunosuppressive TME to induce the regulatory of phenotype. These regulatory DCs can in turn activate Tregs and downregulate CD8 T cells recruitment.²⁷⁰ To summarize, inhibition of immune effector cells and activation of immunosuppressive cells result in a highly immunosuppressive TME for gliomas.

Oncolytic viral therapy disrupts immunosuppressive TME to some extent by causing tumor cells to lysis, releasing TAAs, and inducing innate immunity. Several OVs have been shown to induce immune cell responses, such as DNX-2401, ONYX-015, oncolytic reovirus, and oncolytic parvovirus. Engineering OVs to express immunostimulatory factors is a potential approach to achieve higher levels of pro-inflammatory factors in the TME to effectively turn tumors from “cold” to “hot.” Some recombinant OVs expressing cytokines or chemokines have shown good antitumor activity in glioma models. DNX-2440 stimulated T-cell responses in tumors and prolonged survival rate in glioma mouse models, which is currently undergoing in clinical (NCT03714334). G47Δ-mIL-12 has significantly enhanced survival of mouse in orthotopic U87 and human GSC-derived GBM xenograft models.²⁷¹ M032, IL-12-encoding HSV, is currently being testing in a phase I trial in glioblastoma patients (NCT02062827). FLT3L is a vital growth factor for DCs that can induce of antitumor immune responses by increasing DC numbers, and FLT3L-expressing Ad or HSV improved survival in mouse glioma models.^{272, 273} JX-594, VV with expressing GM-CSF, has shown potential anti-glioma activities in both in vitro and immunocompetent rodent models.²⁷⁴

4.4 Improve oncolytic efficiency

Although OVs as a single drug has shown great promise for brain tumors, the therapeutic efficiency of OVs still needs to be improved. So far, many researchers have focused on the engineered oncolytic virus itself to improve its antitumor activity. In addition to the small space for modification and high technical difficulty, excessive modification of the virus skeleton will reduce the infectious capacity and oncolytic function of OVs. More and more evidence suggested that the infection and replication of OVs in tumors can activate antitumor immunity, turning “cold” tumors into “hot” tumors, thereby enhancing the effectiveness of treatment.¹² Therefore, combining OVS with chemotherapy drugs or immunotherapy is an effective way to improve the efficacy of patients with brain tumors.

Chemoradiotherapy is the most traditional method for glioma, and their combination with oncolytic virus has shown certain application prospects. OVs combined with radiotherapy have the synergy. In pHGG and DIPG models, DNX-2401 in combination with radiotherapy produced a synergistic anti-glioma effect.²⁷⁵ Chemotherapy can improve the treatment effectiveness of oncolytic virotherapy by helping OVs to evade antiviral immune responses. Many studies have shown that combination therapy between OV and TMZ shows synergistic antitumor activity. TMZ in combination with DNX-2401 can improve CD8+ T cell infiltration and significantly improve survival in C57BL/6 mice carrying intracranial GL261 tumors.¹⁸⁷ Good results have also been shown in 31 patients with recurrent Glioblastoma (NCT01956734). However, experiments have shown that TMZ does not improve the survival rate of G47Δ-IL12-treated tumor-in-situ mice, but reduces the therapeutic effect of G47Δ-IL12.²⁷⁶ Cyclophosphamide, an immunosuppressive drug, can improve the antitumor efficacy of Ad, HSV, MV, and reovirus by inducing apoptotic cell death of tumor cells and affecting the innate and adaptive immune responses.²⁷⁷ The prodrug 5-FC is also in combination with Toca511 and TG6002 to treat gliomas, which are under clinical evaluation.²⁷⁸ Overall, the combination of OVs with chemoradiotherapy is additive and synergistic. However, to reach its full potential, the optimal conditions for the use of the clinical combination should be carefully considered.

Oncolytic virotherapy is usually combined with ICIs to increase its potency. It is known that OVs can promote immune cell infiltration and increase PD-L1 expression in tumors. Positive results have been demonstrated in a number of preclinical and clinical trials. Belcaid et al demonstrated that Delta24-RGD combined with PD-1 blockade resulted in increased OS in both the GL261 and CT2A glioma model, with a low dose of the Delta24-RGD upregulated PD-1 expression on CD8 T cells and significantly altered the immune microenvironment in murine glioma.²⁷⁹ Furthermore, an ongoing phase II clinical trial is investigating DNX-2401 combined with pembrolizumab in GBM patients (NCT02798406). An increase in PD-1/PD-L1 axis expression in tumors is induced by reovirus, and PD-1 blockade enhanced systemic therapy in the GL261 glioma model by reovirus.¹⁷⁷ Highly effective antitumor responses and improved long-term survival were demonstrated when α-PD1 therapy was combined with oncolytic VV expressing IL-21 treatment.²⁸⁰ MV in combination with anti-PD-1 antibody blockade overcome immunosuppression and enhance antitumor activity against GBM, thus improving therapeutic outcome.²⁸¹ Saha et al. used a triple-combination approach in combination with anti-CTLA-4, anti-PD-1, and intra-tumoral G47Δ-mIL12 treatment due to the two independent immune inhibitory pathways. Indeed, triple combination therapy cured 77% of treated mice with 005-derived tumors.²⁸² Therefore, the combined application of multiple immunotherapies may be the focus of future research to achieve improved antitumor effects.

Oncolytic virotherapy in combination with CAR T cell therapy is another strategy to improve the oncolytic potency. The viral infection-induced local inflammatory reaction may reverse the immunosuppressive TME and enhance more efficient homing and penetration of CAR-T cells into the tumor stroma.²⁸³ IL-7-loaded oncolytic adenovirus and B7H3-CAR-T combination prolongs survival and reduces tumor burden because of the enhanced T cell proliferation and reduced T cell apoptosis.²⁸⁴ Lp2-CAR-transduced T cells (Lp2-CAR-T) specifically targeted podoplanin-expressing glioma cells and were also able to kill patient-derived GSCs. Combining Lp2-CAR-T with G47Δ can further inhibit tumor growth and improve survival.²⁸⁵ The synergy between OVs and CAR-NK cells has also been assessed. Ma et al. generated an oHSV-1 expressing human IL15/IL15Rα sushi domain fusion protein (OV-IL15C). OV-IL15C plus EGFR-CAR NK cells synergistically suppressed tumor growth and significantly improved survival in an immunocompetent model by increasing intracranial infiltration and activation of NK and T cells and elevated persistence of CAR NK cells.²⁸⁶ Overall, combination therapy with oncolytic virus and CAR T cells may improve antitumor efficacy. However, no clinical trials have been carried out in patients with gliomas.