Introduction

Despite the development of multimodal treatments for glioma, patient prognosis remains poor. The standard treatment is surgical excision to the maximum extent without worsening symptoms, followed by postoperative radiation and concurrent chemotherapy [1]. The prognosis correlates with the removal rate; a removal rate of 78% or higher has been associated with a significant prognostic benefit [2]. However, gliomas are often located in or next to the eloquent area, and complete removal is sometimes challenging [3]. Moreover, glioma cells are highly invasive; they can be present in the surrounding brain edema, and glioma often recurs locally or at distant sites in the brain despite complete resection [4]. Tumor-treating field therapy, which suppresses mitosis by alternating electric fields, was recently shown to prolong the overall survival (OS) of glioma patients [5]; however, the effects are not satisfactory. Thus, novel curative therapies are urgently required.

Because of the invasive nature of glioma, an ideal therapeutic approach is the systemic administration of therapeutic agents that can reach a large area of the brain and are effective against unresectable lesions. Temozolomide (TMZ), which can pass through the blood–brain barrier (BBB), has been used to treat glioma [1]. Bevacizumab is effective in randomized studies, but it only prolongs progression-free survival (PFS) with no significant difference in OS [6]. A recent study showed that vorasidenib, an oral brain-penetrant inhibitor of mutant isocitrate dehydrogenase 1 (IDH1) and IDH2 enzymes, prolonged PFS and delayed the time to the next intervention in IDH1- or IDH2-mutant grade 2 gliomas [7].

Another approach for glioma treatment involves the use of vehicles for therapeutic agents [8]. Gene therapy, including viral therapy, is emerging as a novel therapy for glioma. However, the specific delivery of the therapeutic gene to target cells is critical, so the use of carriers such as cells or exosomes is necessary. Human mesenchymal stem cells (MSCs) have the capacity to migrate and accumulate in tumor cells and have shown potential as a promising systemic delivery tool for cancer therapy [9,10,11].

In this review, we summarize research on the characteristics of MSCs and the current developments of MSCs for use as vehicles of therapeutic agents for glioma treatment. We also discuss the potential of stem cell-based therapies for the treatment of glioma.

Stem cells

Stem cells exhibit multilineage differentiation and self-renewal abilities and can be classified into two major groups: embryonic stem cells and induced pluripotent stem cells (iPSC), which are pluripotent and have the ability of differentiating into a variety of cell types; and adult stem cells, which are limited in their differentiation ability and associated with the formation, maintenance, and regeneration of tissue [12], including MSCs and neural stem cells (NSCs). The research and clinical application of human embryonic stem cells is fraught with ethical and legal difficulties [13]. Therefore, autologous stem cells for therapeutic use must be obtained from organs, as stem cell harvest is easier, less invasive, and not associated with major ethical problems.

MSC characteristics

The International Society of Cellular Therapy proposed a definition of human multipotent MSCs: [1] MSCs are plastic-adherent when maintained in standard culture conditions, [2] MSCs express CD105, CD73, and CD90 and do not express CD45, CD34, CD14 or CD11b, CD79a or CD19, or HLA-DR surface molecules, and [3] MSCs can differentiate into osteoblasts, adipocytes, and chondroblasts in vitro [14]. MSCs have attracted attention in the field of regenerative medicine because of their ability to migrate to lesions such as inflammation sites, injury sites, and tumors, their anti-inflammatory property, and their ability to differentiate into target tissues [15]. MSCs have been found in and isolated from many tissues in the body, including bone marrow, adipose tissue, muscles, dental pulp, spleen, liver, and lungs; and in perinatal sources, including umbilical cord, umbilical cord blood, placenta, and peripheral blood [16]. MSCs derived from various tissues have been studied for clinical applications. Dental pulp stem cells (DPSCs), which show many common features with bone marrow MSCs, originate from neural crest cells and exhibit high proliferation and multipotency, including adipogenic, myogenic, osteogenic, chondrogenic, and neurogenic potential [17,18,19]. Among DPSCs, stem cells from human exfoliated deciduous teeth (SHED) are promising because they exhibit fast cell growth and are easy to handle. Moreover, as SHEDs are isolated from the dental pulp of deciduous teeth, their invasiveness is minimal or absent [20].

The advantages of MSCs in glioma treatment include their ability to migrate to tumor cells and tolerance to the immune system. Furthermore, their ability to pass through the BBB is a key advantage in treating brain tumors [21, 22]. Notably, some disadvantages and uncertainties remain regarding the effect of MSCs on tumors and tumor microenvironments (TME). The characteristics of MSCs have led to the recent interest in MSCs as a therapeutic cell vehicle for glioma. On the other hand, NSCs are one of the most studied cell vehicles in clinical trials. Immortalized human NSCs derived from the human fetal brain [23] have been converted into therapeutic NSCs, such as the FDA-approved HB1.F3-CD, which is commonly used in model experiments and several clinical trials [24]. NSCs derived from iPSCs are also being investigated as cellular vehicles [25].

In the following sections, we discuss the tumor-homing, immune tolerance, and tumor growth regulatory or promoting activities of MSCs.

Homing to tumor cells

The migratory properties of MSCs are important for an effective therapeutic approach to glioma cells invading the neural tissue. Intratumoral injection, transvenous administration, and transarterial administration of stem cells used as carriers of therapeutic agents have been shown to lead to stem cell accumulation in tumors [26,27,28,29]. Human MSCs chemotaxis to tumor-secreted growth factors such as vascular endothelial growth factor (VEGF) and platelet-derived growth factor-BB (PDGF-BB) and chemokines such as C-X-C motif chemokine ligand12 (CXCL12, also known as stromal cell-derived factor-1, SDF-1), stem cell factor (SCF), and IL-8 [30,31,32]. The main factors involved in MSC migration are shown in Fig. 1. Moreover, inflammatory mediators from the TME such as interleukin 1β (IL-1β), interleukin 6 (IL-6), tumor necrosis factor-α (TNF-α), and transforming growth factor-β (TGF-β) are markedly upregulated in glioblastomas and attract immune cells and human MSCs to the tumor [33].

Fig. 1: Chemokines and growth factors associated with MSC migration secreted by glioma cells.
figure 1

Gliomas secrete a variety of chemokines and growth factors. Chemokines attract MSCs and immune cells such as macrophages. The CXCL12/CXCR4 axis plays a key role in the migration of MSCs to the tumor cells. Several growth factors also attract MSCs to the glioma tissue. VEGFR vascular endothelial growth factor receptor, PDGFR platelet-derived growth factor receptor, NT-3 neurotrophin-3, TrkC tropomyosin receptor kinase C.

Full size image

The CXCL12/C-X-C motif chemokine receptor 4 (CXCR4) axis is one of the key factors related to MSC migration. The CXCL12 receptors, CXCR4 and CXCR7, play key roles in MSC migration and proliferation; CXCR4 is mainly responsible for MSC chemotaxis, while CXCR7 is involved in MSC viability [34, 35]. Human umbilical cord MSCs express the chemokine receptors CCR2 and CXCR4 and exhibit migratory capacity toward CD133+ glioma stem cells (GSCs) expressing C-C motif chemokine ligand 2 (CCL2) and CXCL12 [22]. Furthermore, MSCs overexpressing CXCR4 showed a stronger migration capacity toward glioma compared with naĂŻve MSCs [36, 37].

While few studies have directly compared the migration ability of MSCs and SHED to glioma, SHED exhibited a higher migration ability compared with human bone marrow mesenchymal stem cells (BMSCs) and human mesenchymal stem cells from adipose tissue (ADSCs) [32]. The expression of matrix remodeling associated 5 (MXRA5) is critical for maintaining the cell proliferation and migration of DPSCs compared with MSCs by inducing cell cycle and microtubule-related gene expression. TGF-β1 dose-dependently induces MXRA5 expression in DPSCs [38].

Tolerance of the immune system

MSCs are better suited for cell-based therapy than other cell types because of the reduced or weakened immune response elicited by transplanted allogeneic MSCs [39]. First, MSCs are considered immune-evasive because MSCs have a low-level expression of major histocompatibility complex (MHC) class I and no class II surface proteins and co-stimulatory molecules such as CD80 and CD86, and they do not induce activation of allogeneic lymphocytes [40, 41]. Second, increasing evidence has shown that MSCs can inhibit cell proliferation and modulate the function of immune cells. MSCs inhibit the differentiation of monocytes into dendric cells (DCs) and T-cell proliferation through the downregulation of cyclin D2 [42, 43]. Moreover, MSCs can modulate immune cell functions such as cytokine secretion and the cytotoxic effects of T and natural killer (NK) cells, B cell maturation, and antibody secretion [44]. Human MSCs support the differentiation and proliferation of regulatory T cells through a process involving MSC-derived prostaglandin E2 (PGE2) and TGF-β upon cell contact [45]. MSCs exhibit immunomodulation mainly through direct cell contact with innate and adaptive immune cells and several secretomes such as indoleamine-pyrrole 2,3-dioxygenase (IDO), TGF-β, PGE2, hepatocyte growth factor (HGF), nitric oxide (NO), matrix metalloproteinases (MMPs), human leukocyte antigen-G5, interleukin-6 (IL-6), and interleukin-10 (IL-10), suppressing the activity of various immune cells [46, 47]. Therefore, for the therapeutic application of highly immunogenic substances such as viruses, the use of MSCs as vehicles is advantageous in preventing their recognition from the immune system.

Anti- and pro-tumorigenic effects

To use MSCs as vehicles of therapeutic agents for tumor therapy, it is important to consider how MSCs affect tumors. Whether native MSCs exhibit pro-tumor or anti-tumor effects remains inconclusive. Several studies have reported tumor-suppressive effects of MSCs. Human bone marrow-derived MSCs suppress tumor angiogenesis through the downregulation of PDGF-BB and IL-1β, resulting in a significant reduction in tumor volume [48]. Co-culture of human BMSCs may inhibit the invasion and migration of U251 cells by suppressing epithelial-mesenchymal transition (EMT) and promote apoptosis through downregulation of the phosphoinositide 3-kinase (PI3K)/AKT pathway [49]. Murine BMSCs expressing bone morphogenetic protein (BMP2) induced apoptosis of human GSCs and inhibited proliferation and GSC stemness through upregulation of Nanog and octamer-binding transcription factor 4 (OCT4) [50].

Exosome and extracellular vesicles are small- and medium-sized vesicles that are secreted by various cell types and play important roles in intercellular communication, signaling pathways, and molecule transfer [51]. Exosomes act as mediators of cell-to-cell communication and carriers of lipids, proteins, and nucleic acids, including DNA, mRNA, and non-coding RNAs, and they have attracted attention for their inhibitory effects on tumor cells. For example, exosomes secreted by human ADSCs inhibit tumor proliferation by reducing integrin genes in glioma cells [52].

MSCs have also been reported to have angiogenic, glial invasive, and pro-tumorigenic effects [22]. Human ADSCs promoted glioma cell proliferation and angiogenesis and enriched the expression of angiogenic factors (VEGF, angiopoietin 1, PDGF, and IGF-1) and chemokine (CXCL12). Moreover, CXCL12 inhibited tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) activation of the apoptotic pathway in glioma cells [53]. MSCs secrete TGF-β1, which is involved in immunomodulation, proliferation, migration, and EMT. Proteomic profiling of secreted products from co-culture of human MSCs and glioma cells reveals that most of the secreted proteins were in exosomal cargoes and associated with cell motility and tissue development independent of TGF-β1 [54].

Little is known about the impact of therapeutic MSCs on the TME. However, BMSCs are one of the important components of the TME in gliomas. Interactions between BMSCs and TME include angiogenesis, immunomodulation, and tumor cell invasion. BMSCs secrete various cytokines to act on endothelial cells, such as VEGF, PDGF, basic fibroblast growth factor (bFGF), angiopoietin, IL-6, IL-8, and TGF-β, which promote angiogenesis [55]. Moreover, MSCs promote EMT, causing an increased migratory capacity of C6 glioma cells without affecting the cell cycle and growth rate [56]. MSCs are associated with a variety of cells in the TME, not only the immune cells described in the previous section. MSCs interact with vascular endothelial cells and fibroblasts in the TME [55, 57]. For example, human BMSCs can acquire a cancer-associated fibroblast (CAF)-like phenotype in the B-cell acute lymphoblastic leukemia microenvironment, and TGF-β is a key factor to promote the SDF-1/CXCR4 pathway for MSCs to differentiate into CAFs [58]. Furthermore, MSCs can migrate into the tumor and promote tumor angiogenesis through the secretion of vasculogenic factors such as PDGF, VEGF, and FGF-2, and MSCs even differentiate directly into vascular cells [59].

The pro- or anti-tumor effects of MSCs may depend on the MSC type, tumor cell type, culture environment, and secreted factors driving a complex cross-talk between MSCs and glioma cells. Therapeutic agents such as suicide genes and oncolytic viruses can eradicate the vehicle MSCs, so the long-term impact of MSCs on TME may be limited. However, to further overcome these tumor-promoting aspects, it would be better to modify the MSC function. Waterman et al. reported that human MSCs polarize into two distinctly acting phenotypes regarding migration, differentiation, and immunomodulation by two different Toll-like receptor (TLR) activations [60]. The authors also reported that the two MSC phenotypes (MSC1 and MSC2) have divergent effects on tumor growth and metastasis [61]. In the co-culture of glioma cells and MSC1, inflammatory and pro-apoptotic effects were observed through the detection of elevated TRAIL, IL-17, and SCF. Therefore, modification of specific pathways such as TLR by transfected plasmids might regulate the bioactive molecules to inhibit pro-tumor behavior in MSCs.

MSCs as vehicles of therapeutic agents

The characteristics of MSCs described above have supported their use in new therapies. MSCs loaded with a variety of therapeutic agents have been developed for delivery to glioma by various administration methods (Fig. 2). The clinical studies of MSCs and NSCs used for therapeutic agent carriers in glioma therapy are listed in Tables 1 and 2, respectively. There are fewer clinical trials for MSC-associated glioma therapy compared with NSC-associated therapies.

Fig. 2: Schema of the treatment of gliomas using MSCs as a vehicle.
figure 2

Therapeutic MSCs loaded with chemotherapeutic nanoparticles, suicide genes, and oncolytic viruses are potential therapeutic agents. The therapeutic MSCs are administered locally, intravenously, or intraarterially, and MSCs accumulate in the glioma because of their ability to migrate to the glioma. After accumulating in the foci of glioma, the MSCs exert therapeutic effects by delivering the loaded agents and inducing anti-tumor effects.

Full size image
Table 1 The list of clinical trials using MSCs as vehicles of therapeutic agents in glioma therapy.
Full size table
Table 2 The list of clinical trials using NSCs as vehicles of therapeutic agents in glioma therapy.
Full size table

In the following sections, we discuss MSC-based strategies delivering therapeutic agents, including chemotherapeutic drugs, suicide/tumor suppressor genes, and oncolytic viruses.

Chemotherapeutic drugs and nanoparticles

The systemic administration of chemotherapeutic agents is a fundamental approach to cancer treatment. However, some types of drugs have a narrow therapeutic range, and side effects on several organs are common issues. Because of the BBB, tumor cells in the central nervous system are protected from many chemotherapeutic agents. MSCs have the ability to pass through the BBB and accumulate in glioma tissue, and thus strategies using therapeutic drugs carried by MSCs would be an important strategy for glioma treatment.

Nanoparticles as drug carriers enhance drug reachability in the tumor tissue and increase drug efficacy, which can reduce the chances of toxicity [62, 63]. Doxorubicin delivered to gliomas by silica nanorattle-doxorubicin-anchored MSCs showed anti-tumor ability against U251 glioma with a wider distribution and longer retention lifetime than that of free doxorubicin and silica nanorattle-encapsulated doxorubicin [64]. Drug-containing nanoparticles such as poly (d, l-lactic-co-glycolic acid) (PLGA) and liposomes are used to obtain harmless retention of chemotherapeutic drugs loaded on murine MSCs for tumor-homing capacities [65, 66]. Paclitaxel (PTX)-encapsulating hyaluronic acid PLGA polymeric micelle (PTX/HA-PLGA micelle)-loaded murine MSCs showed favorable tumor tropism and drug delivery in a glioma therapeutic model [67]. Overexpression of CD44 in both MSCs and C6 glioma cells allows for effective drug loading, drug penetration, and accumulation. Moreover, nanoparticle-loaded MSCs have been used in gadolinium neutron capture therapy (Gd-NCT). Lai et al. demonstrated that magnetic nanoparticles encapsulated with gadodiamide using human umbilical cord-derived MSCs as vehicle resulted in higher Gd drug content at glioblastoma sites compared with free gadodiamide alone [68].

Gene therapy

Gene therapy can be classified into three categories in accordance with the therapeutic approach [1, 69] delivery of suicide genes: these genes act as a prodrug converter and cause cell apoptosis; this is also known as gene-directed enzyme prodrug therapy (GDEPT); [2] corrective gene therapy: therapeutic genes are introduced into cancer cells to regulate abnormalities in their genomic profiles and inhibit cell growth; and [3] toxin/apoptosis-inducing gene therapy: the delivered transgene results in the production of a toxic protein such as diphtheria toxin or TNF-α. By introducing exogenous genes or manipulating target genes, gene therapy can prevent cancer progression and enhance anti-tumor immune responses [70, 71]. This section mainly focuses on suicide gene therapy.

The herpes simplex virus thymidine kinase (HSV-TK)/ganciclovir (GCV) system and cytosine deaminase (CD)/5-fluorocytosine (5-FC) system are the two major GDEPTs. HSV-TK phosphorylates GCV, which cannot be phosphorylated by human thymidine kinase, resulting in mono-phosphorylated GCV that is further phosphorylated to triphosphorylated GCV by cellular kinases. Phosphorylated GCV is transferred to adjacent tumor cells through gap junction intercellular communication and acts as a nucleoside analog that causes cell apoptosis [72]. Several phase I–II clinical trials have demonstrated the safety and efficacy of the HSV-TK/GCV system [73]. However, in a phase III clinical study of HSV-TK suicide gene therapy, manual injection of vector-producing cells did not improve PFS or OS compared with standard therapy (surgical resection and radiotherapy). One reason for this result was the use of fibroblasts as vector-producing cells, which led to the poor spread of suicide genes [74]. Following these results, strategies for suicide gene therapy using stem cells that can migrate to tumors as carriers have been reported [27, 72]. This approach induces tumoricidal effects and activates the anti-tumor immune response. TK-cAd-MSCs showed a high secretory profile of several active anti-tumor immune response cytokines such as interferon-γ and IL-2 and chemokines such as monocyte chemoattractant protein-1 (MCP-1) and IL-12p40. IL-2 is a key molecule with pleiotropic effects on the immune system and potent anti-tumor effects in the TME [75]. However, to use MSCs as gene carriers in GDEPT, suicide genes should not be harmful to MSCs without prodrugs. The expression of HSV-TK results in excessive thymidine metabolism, which is toxic to some stem cell types [76], and the use of TET-inducible systems or optimized mutant HSV-TK are strategies to reduce toxicity to stem cells [25, 77, 78]. The first-in-human phase I clinical trial of GDEPT with local injection of ADSCs expressing TK for recurrent glioblastoma reported that it was safe and well tolerated. The median OS was 16.0 months, and contrast and fluid-attenuated inversion recovery (FLAIR) volume on magnetic resonance imaging (MRI) was significantly decreased at 6 and 12 months after injection compared with the baseline [79].

The CD is an enzyme of bacterial or fungal origin. CD metabolizes 5-fluorocytosine (5-FC), an antifungal drug, into 5-fluorourasil (5-FU), which exhibits anti-tumor effects. The 5-FU then induces apoptotic cell death via caspase 3 and 9 [80]. The bystander effect caused by the CD/5-FC system does not require direct cell contact, as 5-FU can disperse to surrounding target cells by passive diffusion [81]. Transplantation of CD-expressing human MSCs and sequential treatment with 5-FC and TMZ suppressed tumor progression in a U87 glioma model. Administration of the CD/5-FC system followed by TMZ-induced cell cycle arrest and DNA breakage [82]. In a phase I first-in-human clinical trial, the CD/5-FC system using NSCs was well tolerated in the patients, and autopsy confirmed non-tumorigenicity and migration of NSCs to distant tumor sites [24], while phase I and II clinical trials using MSCs expressing CD are underway and results are pending (Table 1). Some studies have demonstrated enhanced efficacy by introducing both CD and TK into stem cells as a double prodrug therapy [83, 84]. In a colorectal cancer model, coadministration of human MSCs expressing cytosine deaminase (MSC/CD) and 5-fluorocytosine (5-FC) with α-galactosylceramide (α-GalCer) promoted the infiltration of immune cells such as NK cells, macrophages, DCs, and T cells into the TME [85].

Oncolytic virus

Oncolytic viruses (OVs) represent a promising strategy for glioblastoma treatment. OVs are naturally occurring or genetically modified viruses that specifically replicate in cancer cells and enhance tumor lysis by viral replication and antiviral immunity while sparing normal cells [86, 87]. Several studies have examined various wild-type and altered OVs for glioma therapy, including adenovirus, herpes simplex virus [88], measles virus [89], reovirus [90], vesicular stomatitis [91], Newcastle disease virus [92], vaccinia virus [93], and poliovirus [94]. In addition to the direct tumor lysis of the virus, both innate and adaptive immune responses against cancer are enhanced by OV infection and oncolysis [87]. OVs also modulate the TME and enhance anti-tumor immune responses. Tumor-associated antigens are released by tumor lysis, and tumor-associated antigens are internalized by antigen-presenting cells and presented to CD8+ T cells or cytotoxic T lymphocytes, enhancing their cytolytic granules containing perforin and granzyme to induce anti-tumor effects [95]. To improve the anti-tumor effects of OVs, combinations with immune checkpoint inhibitors have been explored as a strategy to enhance indirect oncolytic effects by the immune system [93]. OVs affect the immune system and the TME, including influencing vascular permeability and signaling to adjacent tumor cells [88, 96].

Several studies have reported that the use of stem cells as carriers can enhance the anti-tumor effects of OVs [28, 97,98,99]. Various factors, such as the efficacy of infection against the tumor, viral replication, distribution of viral particles, and inactivation of the OV by the immune system, can influence the anti-tumor effects. The administration method also affects the initial viral distribution; however, direct administration has diffusivity problems, and intravenous administration has reachability problems because of innate and adaptive immune responses such as antibodies induced to neutralize naked OV [100]. Therefore, a vehicle to efficiently deliver the OVs to the cancerous tissue is essential to enhance the therapeutic effect. The use of MSCs to deliver oncolytic adenovirus CRAd-CXCR4-5/3 into the mouse brain increased adenovirus infection of remote glioma cells; when adenoviral vectors were delivered by human MSCs, the number of viral copies in remote glioma tissue increased 46-fold compared with delivery of virus alone [101]. Systemic administration of hepatocellular carcinoma–targeting oncolytic adenovirus loaded in human MSCs resulted in accumulation in cancer cells, and a high level of viral concentration was observed at tumor sites [98].

The advantages of using MSCs include the ability to avoid an immune response to the virus even when administered systemically, the possibility of reducing the initial virus dose because of their ability to accumulate in tumors, and high virus replication in the MSCs. Finally, MSCs are lysed by OV replication, which can avoid any negative side effects related to surviving stem cells. Even after TMZ or PCV administration for glioma treatment, the collected BM-MSCs have the same characteristics as BM-MSCs derived from healthy individuals and are feasible carriers of OVs by carotid administration, showing anti-tumor effects [28].

Chimeric antigen receptor T (CAR-T) cell therapy has shown efficacy in relapsed acute lymphoblastic leukemia [102] and is promising for glioma therapy [103, 104]. Several clinical trials are underway, and CAR-T cell therapy is in the developing stages as a therapeutic modality for glioma. IL-12 can stimulate T cells to produce interferon to effectively promote antigenic cell activation and cause anti-tumor immunity. Thus, systemic administration of MSCs with oncolytic adenovirus expressing IL-12 and PD-L1 blockers has been attempted to enhance CAR-T cell function by allowing the adenovirus to infect tumor cells and destroy the TME [105].

Challenges and future prospects

MSCs are very useful vehicles for administering therapeutic genes and viruses, but there are some challenges that need to be overcome to establish MSCs as an efficient and safe vehicle for patient treatment. Depending on the treatment modality, the therapeutic agent can show toxicity against MSCs; if the viability of MSCs is impaired before reaching the tumor site, the migration capacity and anti-tumor effect will also be impaired. Therefore, it is important to optimize the therapeutic agents for MSCs to maintain the viability of suicide gene–expressing or OV-loaded MSCs without compromising their therapeutic efficacy. For example, there are TKs that are optimized for SHED and effective in reducing the toxicity of SHED before it reaches the tumor [78]. Moreover, for treatment using OVs, it is necessary for MSCs to reach the tumor site before the OV can proliferate and release progeny virions. However, the optimal timing of administration of MSCs carrying each kind of OV has not been well established. After infection of human MSCs with oncolytic Delta-24-RGD, the highest number of infectious units was detected after 48 h of infection, meaning MSCs would need to reach the tumor site within 48 h [21]. However, little is known regarding other OVs.

The initial distribution of MSCs after intratumoral injection is limited, as MSCs are localized primarily around the injection needle. Given the need to deliver the therapeutic agent to the entire lesion, it is ideal for MSCs to be distributed widely in the tumor tissue by intraarterial or intravenous injection. Intraarterial administration could be reasonably effective for the distribution of MSCs. However, the primary issue with intraarterial administration of cells is the risk of cerebral infarction. Studies of selective arterial administration of Delta-24-RGD-loaded MSCs from the distal internal cerebral artery or posterior cerebral artery of canines reported that injection of 2 × 106 cells/20 ml to 1 × 108 cells/10 ml can be safely achieved using commercially available microcatheters such as Echelon 14 (inner diameter 0.017”) [106]. However, MSCs may become trapped in other organs when administered intravenously.

Many uncertainties regarding the effects of MSCs on TME have not been resolved and require further investigation. To enhance the anti-tumor effects of these novel strategies, both the therapeutic agent and MSCs need to be better optimized to improve the retention of the therapeutic agent and the distribution and migration capacity of MSCs.