Stem cell-based therapies for tumors in the brain: are we there yet?

Abstract

Advances in understanding adult stem cell biology have facilitated the development of novel cell-based therapies for cancer. Recent developments in conventional therapies (eg, tumor resection techniques, chemotherapy strategies, and radiation therapy) for treating both metastatic and primary tumors in the brain, particularly glioblastoma have not resulted in a marked increase in patient survival. Preclinical studies have shown that multiple stem cell types exhibit inherent tropism and migrate to the sites of malignancy. Recent studies have validated the feasibility potential of using engineered stem cells as therapeutic agents to target and eliminate malignant tumor cells in the brain. This review will discuss the recent progress in the therapeutic potential of stem cells for tumors in the brain and also provide perspectives for future preclinical studies and clinical translation.

Tumors in the brain are categorized as either primary brain tumors, defined by their origin from cells autochthonous to the brain, and secondary brain tumors, which originate from metastatic cells from peripheral tumor sites.¹ Although the exact incidence and ratio are controversial, metastatic brain tumors are known to be far more common than primary brain tumors and have been estimated to outnumber primary brain tumors by more than 4 to 1.² Primary brain tumors are commonly further subcategorized according to their cell of origin, which can be within the brain tissue itself (eg, glioma), the membranes around the brain (eg. meningioma), the nerves (eg, nerve sheath tumors), glands (eg, pituitary tumors), and vessels (eg, choroid plexus tumors) within the cranium.³ The most frequent type of primary malignant brain tumor is glioma, of which > 50% may be categorized as WHO grade IV glioblastoma (GBM).⁴ GBM may either arise de novo (referred to as primary GBM) or much less often (about 10%) may progress from lower-grade or anaplastic astrocytomas (secondary GBM).¹ GBM is characterized as malignant, mitotically active, and predisposed to necrosis, and its very low median survival (∼14.6 months) is attributed to unique treatment limitations such as a high average age of onset, tumor location, and poor current understanding of the tumor's pathophysiology.⁵ With a low median survival, both primary and secondary GBMs are both associated with an extremely poor prognosis, with virtually all patients suffering from recurrence and <5% of patients surviving 5 years after diagnosis.^6,7

The current standard therapy for GBM is a combination of cytoreductive surgery followed by concurrent and adjuvant chemotherapy with temozolomide (TMZ) and radiation therapy.⁸ Current limitations in the treatment of GBM involve hindrances in achieving complete tumor resection, tumor resistance, challenges in penetrating the blood brain barrier (BBB), and insufficient accumulation of therapeutic agents at the tumor site (www.ncbi.nlm.nih.gov). Thus, patients with GBM are in dire need of alternative treatments that can circumvent the limitations posed by the current standard of care. Many adult stem cells (SCs) display intrinsic tumor-tropic properties, making them attractive candidates for targeted delivery of anticancer biologics by modifying them to stably express/release a variety of anticancer agents. This review aims to highlight the most recent advances in SC-based treatments for treatment of tumors in the brain, particularly GBM, and discuss the most recent SC-based therapeutic clinical trials for these tumor types.

Stem Cell Sources

Stem cells are the natural source of embriogenetic tissue generation and continuous regeneration throughout adult life. Over the last few decades, 3 major stem cell types (ie, embryonic, adult [mesenchymal and neural], and induced pluripotent) have been identified and extensively characterized. Embryonic stem cells (ESCs) are derived from the inner cell mass of blastocyst-stage embryos, and closely related embryonic germ cells were among the first cell types identified as possible sources for stem cell therapy.⁹ However, owing to numerous concerns related to the use of ESCs, adult stem cell types (namely mesenchymal stem cells [MSCs], neural stem cells [NSCs], and induced pluripotent stem cells [ipSCs]) have been explored extensively for therapeutic purposes.¹⁰ MSCs are spindle-shaped, fibroblast-like multipotent stem cells that are responsible for regeneration and cellular homeostasis in almost all tissues and have the potential of converting to tissue types of other lineages both within and across germ lines.¹¹ MSCs have been isolated from a number of organs including bone marrow, adipose tissue, fetal tissues, dental pulp, umbilical cord, Wharton's jelly, and other tissue types.¹² Most of the preclinical studies to date have been performed with bone marrow-derived MSCs; however, adipose tissue and umbilical cord blood are other MSC sources that have received considerable attention in recent years. NSCs, also referred to as neural precursor cells (NPCs), are a heterogeneous population of self-renewing and multipotent cells of both the developing and the adult CNS¹³ and are especially enriched in the subventricular zone (SVZ) and the subgranular zone in the hippocampal dentate gyrus DG. NSCs have been proven to be an invaluable source for developing cell-based therapies, especially for neurodegenerative disorders and also for other conditions, particularly certain cancer types.^10,14 Induced pluripotent stem cells (iPSCs) generated from somatic cells have emerged as popular and effective stem cell types for therapy.¹⁵ The iPSCs are generated by allowing a differentiated somatic cell to revert to embryonic stage via induced expression of transcription factors OCT4, SOX2, KLF4, and C-MYC associated with pluripotency.¹⁶ iPSCs can be derived from a wide variety of starting cells, even though easy accessibility to fibroblasts makes them the most common source for iPSC generation.¹⁷ Recent studies have shown the ability of ectopic expression of cell type-specific transcription factors to directly switch cell fates between somatic cells, thus circumventing the pluripotent state and eliminating any risk of malignant transformation.¹⁸ Similarly, MSCs derived from induced pluripotent stem cells (iPSC-MSCs) have also evolved as a promising alternative cell source for MSCs and regenerative medicine. Several studies have revealed successful derivation of functional iPSC-MSCs^19,20 that were shown to have similar characteristics as bone marrow-derived MSCs,²¹ including expression of typical mesenchymal markers and the capacity to efficiently differentiate into osteogenic and chondrogenic lineages.

Stem Cell Homing and Migration to Tumors

The intrinsic homing property of a variety of stem cell types to brain pathologies such as ischemic, neoplastic, and demyelinating lesions has been unraveled during the past decade.^22–25 A number of studies have shown tumor tropism of both NSCs and MSCs when injected via different routes in mouse tumor models of brain tumors (reviewed in¹⁰). The G-protein coupled receptor CXCR4 and its only known ligand, stromal cell-derived factor 1 (SDF1; also known as CXCL12) is one of the best-studied mediators of stem cell tropism. SDF1 and CXCR4 are expressed at high levels, particularly in the DG and SVZ where they are involved in regulating the tropism of endogenous NSCs.²⁶ The migration of exogenously modified therapeutic stem cells has been shown to proceed in a similar fashion to that of endogenous NSCs toward tumors in the brain.^27,27–30 The chemokine stem cell factor (SCF) has been shown to be upregulated by cells that reside in and around lesioned areas and induce the migration of exogenous NSCs toward pathology within the brain through interaction with the tyrosine receptor kinase c-Kit, as demonstrated in a GBM model.³¹ NSCs also express CCR2 and migrate in the direction of a MCP-1 gradient toward neoplastic lesions within the brain.³² Hypoxia is known to promote NSC tropism in vitro³³ and in vivo,³⁴ mainly due to the upregulation of vascular endothelial growth factor (VEGF) by hypoxic cells, which has been observed to result in increased expression of chemotactic factors Ang2 and GROα.³⁵ These 2 proteins promote the migration of NSCs toward regions of hypoxia within the brain. VEGF can induce NSC migration in a reactive oxygen species (ROS)- and focal adhesion kinase (FAK)-dependent manner.³⁶ Other influential signaling pathways involved in SC homing have been elucidated and include hepatocyte growth factor (HGF)/c-MET receptor,³⁷ urokinase-type plasminogen activator (uPA)/uPA receptor (uPAR),³⁸ platelet-derived growth factor (PDGF)/uPAR/β1 integrin,³⁹ and transforming growth factor β (TGFβ)/TGFβ receptor (TGFβRII).⁴⁰ Macrophage migration inhibitory factor (MIF)/CXCR4 has been recently identified as the dominant chemotactic axis for recruiting human MSCs to tumors.⁴¹ Although the tumor-homing ability of MSCs and NSCs has been extensively studied and established, understanding the ability of iPSC-derived SCs to home to tumors is still in its infancy. Recently, iPSC-derived NSCs have been shown to have potent glioma tropism.⁴² The degree of SC homing to tumors in vivo is influenced by diverse factors including the nature of the SC and the tumor microenvironment.¹⁰ In addition to homing to the primary tumor mass, both NSCs and MSCs have been shown to efficiently track malignant microinvasive deposits in the brain.^24,27 Therefore ,the potential for migration of SCs to the main tumor mass and the microinvasive deposits offers a strong rationale for the use of SCs in treating brain tumors.

Therapeutic Applications for Brain Tumors

The major limitations of current GBM therapies are (1) marginal chance of curative resection due to the highly invasive nature of GBM, which allows tumor cells to infiltrate deeply into normal tissue; (2) the frequent relapses after seemingly curative resection due to the seeding of microsatellite deposits in the surrounding neuropil; (3) the pharmacokinetic problems associated with the blood-brain-barrier that impede the accumulation of therapeutic drug concentrations in the tumoral interstitium; and (4) the caution that has to be incorporated with high-dose drug treatment and radiation therapy due to the vulnerability of normal brain tissue and the serious consequences of brain damage to the patient's quality of life. Transplanted SCs, engineered to release therapeutic agents, can be used as an engraftable, highly mobile, “organic” delivery vehicle for tumor treatment. This is reliant on the biological and physiological differences between normal and malignant cells and also on the GBM tumor-homing properties of engineered SCs. The latter is a crucial factor in clinical effectiveness as engineered SCs can “home in” on microsatellite deposits of escaping GBM cells that are known to give rise to recurrent tumors in the brain. Both unmodified SCs, which possess intrinsic antitumor effects attributed to their released factors and physical interactions with tumor cells,⁴³ and modified SCs have been utilized to treat tumors in the brain. Some of the most promising are outlined in Fig. 1 and are discussed below.

Fig. 1.

Modifying stem cells for brain tumor therapy. Stem cells can be expanded and manipulated in vitro to express a variety of tumor-antagonizing transgenes to establish antitumor effect following stem cell implantation in vivo.

Modifying Stem Cells Genetically to Release Direct and Indirect Effector Proteins

SCs are typically modified by viral transduction, although nonviral methods to express transgenes encoding secretable effector proteins have been reported.^25,44 Depending on whether the therapeutic proteins act directly on malignant tumor cells in the brain or upon tumor-supporting cells in the tumor microenvironment, SC secretion of such proteins can be divided into 2 broad categories: direct and indirect effectors. The direct effectors include proapoptotic and antiproliferative proteins.

Proapoptotic Proteins

TRAIL (tumor necrosis factor-related apoptosis inducing ligand) binds to death receptors (DR)4/5 specifically expressed on tumor cells and induces caspase-mediated apoptosis.^24,45 Previous studies on intratumoral implantation of murine primary NSCs engineered with full length TRAIL in established GBM revealed a significant reduction of tumor burden.⁴⁶ However, TRAIL is a type 2 receptor membrane protein, and its release into the SC environment requires accidental cleavage from its membrane docking site. We have engineered recombinant TRAIL that consists of a fusion between the extracellular domain of TRAIL and the extracellular domain of the hFlt3 ligand, which is a ligand for Flt3 tyrosine kinase receptor and mediates both secretion through the endoplasmic reticulum.⁴⁷ Both MSCs and NSCs transduced with a lentiviral expression cassette for secretable (S)-TRAIL have shown anti-tumor effects in mouse tumor models of both nodular and invasive GBM.^24,48 Recent studies have confirmed the cytotoxic effects of TRAIL-engineered SCs on GBMs,⁴⁹ indicating the therapeutic potential of this strategy. Recently, we have developed in vivo imageable breast-to brain metastasis tumor models and shown that intra-arterial injection of SC-S-TRAIL suppressed metastatic tumor growth and prolonged the survival of mice bearing metastatic tumors in the brain.⁵⁰

Antiproliferative Proteins

Among the number of antiproliferative molecules, a limited number of molecules can be expressed and released in the extracellular milieu. The antitumor effects of SCs expressing cytokines such as interferon (INF)-β in brain tumors was shown in an early study revealing that human SCs expressing IFN-β were able to track murine GBM tumors when administered systemically.⁵¹ IFN-β, which can also act as an alternative immunogenic candidate,^52,53 and (IFN)-α⁵⁴ have both been shown to negatively regulate tumor growth in a variety of preclinical cancer models. Recent studies have further shown that the release of biologics from SCs can outcompete or sterically block the binding of endogenous ligands to their cognate receptors such as epidermal growth factor receptor (EGFR) or its tumor-specific antigen EGFRvIII variant,^55,56 resulting in the inhibition of proliferation pathways in tumor cells.

Bone morphogenetic proteins (BMPs) activate their cognate receptors (BMPRs), and in vivo delivery of BMP4 has been shown to effectively block the tumor growth post intracerebral grafting of human GBM cells.⁵⁷ In a recent study, human adipose-derived MSCs engineered to express BMP4 were shown to prolong survival in mice bearing GBM.⁵⁸ Other studies on the regulated expression of proteins that induce cell cycle arrest, such as growth-arrest specific-1 (Gas1) in SCs, have shown antitumor effects in mouse tumor models of GBM.⁵⁹

Indirect effectors of tumor cells include antiangiogenic thrombospondin-1 (aaTSP-1),⁶⁰ PEX (a fragment of metalloproteinase-2),⁶¹ and immunostimulatory molecules such as interleukins (ILs).

Antiangiogenic Agents

Tumor angiogenesis represents the ability of a tumor to stimulate new blood vessel formation for self-sustained growth.⁶² Previous clinical studies have shown that intravascular delivery of antiangiogenic drugs transiently normalizes the abnormal vasculature, thus reducing tumor-associated vasogenic brain edema and providing a clinical benefit for GBM patients.⁶³ The decrease in mean vessel diameter and permeability⁶⁴ and increased pericyte coating of small vasculature⁶⁵ have been shown to be associated with vessel normalization. Previous studies have shown that MSCs localize to tumor vasculature upon intratumoral implantation,⁶⁶ and recent findings demonstrated that MSCs display pericyte markers and intratumorally grafted MSCs could possibly function as tumor pericytes.⁶⁷ Human NSCs expressing PEX, a naturally occurring fragment of human metalloproteinase-2 when injected directly into murine GBM, have shown efficacy post-treatment.⁶¹ In our previous studies, we have shown that NSCs engineered to express antiangiogenic repeats of anti-angiogeneic three type 1 repeats (3TSR) of thrombospondin (TSP)-1 markedly reduced tumor vessel-density, resulting in the inhibition of tumor progression in mice bearing highly malignant human GBM.⁶⁰ These studies suggest that SCs engineered to express antiangiogenic agents such as PEX and aaTSP-1 could be exploited for enhanced therapeutic benefit.

Immunostimulatory Molecules

The rationale for utilizing immunostimulatory molecules is based on shifting the immunosuppressive tumor milieu towards directing an immune response against the cancer.⁶⁸ Interleukins are known to regulate inflammatory and immune responses and have antitumor effects.⁶⁹ Both murine and human primary NSCs engineered to express interleukin-12 have been shown to have treatment efficacy on GBM in mice and rats, respectively.^70,71 Human MSCs secreting IL-12 or IL-18 have also shown efficacy in mice-bearing GBM.^72–74 The sustained presence of SC-delivered ILs resulted in the activation of natural killer cells and the recruitment of tumor-specific T cells.

Immunotherapy employing peripherally implanted cytokine-secreting inactivated tumor cells in syngeneic tumor models has shown promising results by potentially eradicating intracranial GBM through the induction of a local proinflammatory response in the tumor microenvironment.^75,76 Several reports show that MSCs can potentially synergize with this approach when exposed to cytokines such as interferon-gamma (IFNγ).^77,78 As such, peripheral immunotherapy using cytokine-secreting tumor cells has been shown to eradicate experimental GBM by inducing a proinflammatory tumor microenvironment.^76,79 Recent studies have shown that the intratumoral implantation of MSCs into rats bearing GBM, in addition to peripheral immunotherapy with interferon-gamma (IFNγ)-secreting inactivated tumor cells, significantly improved animal survival.⁸⁰ Although the complete mechanism of action in this context remains unclear, in vitro results suggest that MSCs in response to IFNγ act as conditional antigen-presenting cells, upregulate MHC classes I and II molecules, and secrete low amounts of immunosuppressive molecules, ultimately inducing an intratumoral antigen-specific cytotoxic CD8+ T cell response.

Modifying Stem Cells Genetically to Release Tumor-targeted Toxins

Pseudomonas exotoxin (PE) potently blocks protein synthesis by catalyzing the inactivation of elongation factor-2 (EF-2). Targeted PE-cytotoxins have been used as antitumor agents; however, their off-target delivery, systemic toxicity, and short chemotherapeutic half-life have confounded their effective clinical translation in GBM patients.⁸¹ In a recent study, we have created toxin-resistant SCs by modifying endogenous EF-2 and engineered them to secrete PE-cytotoxins that specfically target expressed interleukin (IL)-13 receptor subunit alpha-2 or overexpressed EGFR in GBM. The release of IL13-PE from SCs in a clinically relevant GBM resection model resulted in tumor cell-specific killing and led to increased long-term survival of mice compared with IL13-PE protein infusion.⁸² A recent study has shown that intratumoral injection of MSCs expressing a bispecific immunotoxin, VEGF165-ephrin A1-PE38KDEL, was effective at inhibiting tumor growth in a malignant GBM tumor model.⁸³

Modifying Stem Cells Genetically to Induce Suicide Therapy

SC-mediated suicide therapy is a strategy whereby SCs are engineered to express an enzyme that converts a nontoxic prodrug into a cytotoxic drug, resulting in efficiently killing surrounding cells by the bystander effect. Three major prodrug systems are currently being used: cytosine deaminase (CD)/5-fluorocytosine (5-FC); herpes simplex virus thymidine kinase (HSV-TK)/ganciclovir (GCV); and carboxylesterase (CE)/camptothecin-11 (CPT-11). Both MSCs and NSCs have been modified with the CD/5-FC system, tested in mouse models of GBM^84,85 and medulloblastoma,⁸⁶ and shown to result in tumor regression and prolonged survival. The HSV-TK system has been used in modified MSCs and relies on the formation of gap junctions between the SCs and surrounding target cells for an efficient bystander effect. MSCs expressing HSV-TK have shown efficacy in animal models of GBM.^87–89 Human NSCs bearing the CE/CPT-11 system have proven effective in preclinical models of medulloblastoma.⁹⁰ The feasibility of treating GBM with SCs expressing double prodrug enzymes has been explored. Human NSCs expressing CD and TK were shown to have efficient antitumor efficacy and eradication of the SCs.⁹¹ In an effort to develop efficient SC-based therapeutic strategy that simultaneously allows killing of GBM tumor cells and eradication of SCs post tumor treatment, we have engineered MSCs to co-express HSV-TK and S-TRAIL and shown that they induce caspase-mediated GBM cell death and can be selectively eradicated post tumor treatment in mouse GBM models.⁸⁸ Similar studies utilizing SCs expressing CD/IFNβ^57,92 and CE/TRAIL⁹³ have also shown efficacy in malignant GBM mouse models.

Stem Cells Carrying Nanoparticles

Nanoparticles are emerging as novel therapeutic and diagnostic tools with great potential in cancer biology. The surface of nanoparticles can be modified, thus making them ideally suited to contain high concentrations of often insoluble therapeutic reagents.⁹⁴ Although the use of nanoparticles offers considerable delivery benefits, their efficient clearance, inefficient dissemination in solid tumors, and inability to target micrometastatic lesions⁹⁴ have presented a number of challenges. SCs have the potential to overcome these barriers and can be loaded with nanoparticles and administered in vivo, where they can deposit the loaded nanoparticles in close proximity to the tumor.⁹⁵ Previous studies have shown the potential for loading cell membranes of MSCs with porous silica “nanorattles” containing doxorubicin (DOX)⁹⁶ and their efficacy in intracranial tumors.⁹⁶ Recent studies have shown that NSCs can improve the tumor-selective distribution and retention of nanoparticles within invasive brain tumors.⁹⁷ The release of nanoparticles from SCs is either due to membrane rupture or by facilitating death via photoinduction^98,99 or inducing hyperthermia.¹⁰⁰ The combination of nanoparticle-based therapeutics with SC delivery has exciting prospects for the treatment of tumors in the brain.

Stem Cells Loaded With Oncolytic Virus

Oncolytic viruses (OVs) are natural or genetically modified viruses that, upon infection, selectively replicate in and kill neoplastic cells while sparing healthy cells.¹⁰¹ However, clearance of the virus by host defense mechanisms post systemic administration and insufficient viral spread post intratumoral administration confound their therapeutic efficacy. In an effort to improve delivery of OV-based therapeutics and circumvent antiviral immunity, a number of studies have explored the possibility of using SCs as delivery vehicles.^11,102 Different SC types have been utilized for local release of intact replicating oncolytic adenoviruses administered systemically or intratumorally into mouse models of GBM.^103–105 Previous studies have shown effective homing of NSCs loaded with the oncolytic adenovirus, CRAd-Survivin-pk7, to intracranial GBM, resulting in increased animal survival.¹⁰⁶ Recently, we have shown that hyaluronan or hyaluronic acid (HA) is highly expressed in a majority of tumor xenografts established from patient-derived GBM lines that present both invasive and nodular phenotypes.¹⁰⁷ Intratumoral injection of MSCs loaded with a conditionally replicating adenovirus expressing soluble hyaluronidase (ICOVIR17) into GBM enhanced viral spread and resulted in a significant antitumor effect and mice survival.¹⁰⁷ Oncolytic herpes simplex virus (oHSV) is inherently neurotropic and has been modified for tumor-selective replication, making it a promising candidate for brain tumor therapy.¹⁰⁸ Previous clinical studies have revealed suboptimal response rates in patients after delivery of oHSV in GBM post tumor resection,¹⁰⁹ which could be attributed to “wash out” of the oHSV in the tumor resection cavity. We have utilized MSCs for local release of oHSV and its proapoptotic variant oHSV-TRAIL into mouse models of GBM and shown prolonged median survival in mice as compared with direct injection of purified oHSV.¹¹⁰

Stem Cell-released Extracellular Vesicles

Extracellular vesicles (EVs) (exosomes and shedding microvesicles) are released by various cell types and contain mRNA, miRNA, proteins, and lipids.¹¹¹ Accumulating evidence supports the notion that MSCs act in a paracrine manner¹¹² by transferring the biological material to adjacent or distant cells to facilitate communication between different cell types.¹¹³ Previous studies have shown that MSCs secrete and transfer microvesicles containing specific miRNAs that can serve in intercellular communication.¹¹⁴ The ability of adult MSCs to transfer miRNA (miR-124 and miR-145) mimics to GBM cells, which express very low levels of these miRNAs, has been explored and shown to have antitumor effects in a mouse model of GBM.¹¹⁵ Similarly, intratumoral injection of exosomes derived from miR-146-expressing MSCs significantly reduced tumor growth in a rat model of primary GBM.¹¹⁶ Recent studies have reported that MSCs are able to package and deliver active drugs such as paclitaxel (PTX) through their exosomes and increase survival of mice bearing GBM.^117,118 Additional, recent studies have shown that exosomes derived from MSCs engineered to express different miRs have antitumor effects in mouse tumor models.¹¹⁹

Synergistic Approaches Enhancing Therapeutic Stem Cell Efficacy in Combination With Other Antitumor Agents

Brain tumors comprise a heterogeneous population of cells that are genetically and epigenetically unstable.¹²⁰ Given the GBM heterogeneity, it is unlikely that any one effective strategy will provide a satisfactory treatment regimen for such tumors. Therefore, using a combination approach would be more effective for treating GBM. Because brain tumor cells and the cells in the tumor microenvironment either specifically express or overexpress cell surface receptors, one strategy is to create SCs that simultaneously express/secrete different therapeutic agents targeting these receptor types. This will allow targeting multiple pathways in tumor cells and associated cells in the microenvironment and thus increase the efficacy of targeted therapies. As such, there are a growing number of studies using bimodal SCs or SCs that secrete bifunctional molecules. We have shown that 3TSR of TSP-1 upregulates death receptor (DR) 4/5 expression in a CD36-dependent manner and primes resistant GBMs to TRAIL-induced apoptosis. A single administration of MSCs expressing 3TSR and S-TRAIL targets both tumor cells and vascular component of GBMs and extends survival of mice bearing intracranial TRAIL-resistant and highly vascularized GBM.¹²¹ Recent studies have shown that SCs coexpressing immunostimulatory cytokines IL-18 and (IFN)-β significantly prolonged the survival and inhibited tumor growth in a rat intracranial GBM model by promoting cell apoptosis, antitumor cytokine production, and CD4+ and CD8+ T-cell infiltration.¹²²

An emerging branch of cancer therapies is the use of smaller antibody fragments such as single-chain variable fragments (scFv) and nanobodies (consisting of just the V_HH domain) that bind to epitopes overexpressed on tumor cells¹²³ and create bifunctional proteins that are able to perturb multiple signaling pathways specifically in tumor cells (Fig. 2). In one study, we fused S-TRAIL to an EGFR-specific nanobody to make a proapoptotic immunoconjugate (ENb-TRAIL) that concurrently targets cell proliferation and death pathways.⁵⁶ SCs engineered to express ENb-TRAIL were shown to efficiently inhibit EGFR signaling and induce caspase-mediated cell death in a panel of GBM expressing varying levels of EGFR and DR4/5.⁵⁶ In vivo, when mice bearing intracranial GBM were treated with stem cell-delivered ENb-TRAIL, tumor burden was significantly decreased and mice survival was prolonged.⁵⁶ Taken together, bifunctional antibody fragments possess many attributes, making this a novel and potentially effective means for treating malignancies.

Fig. 2.

Enhancing stem-cell (SC) efficacy by engineering bifunctional molecules. The efficacy of SC-delivered therapeutic payload can be enhanced by engineering bispecific molecules that target multiple cell surface receptor mediated signaling pathways in glioblastoma cells.

A number of different studies have combined an additional agent that synergizes with SC therapy or sensitizes an otherwise resistant population to SCs delivering a single therapy, thus increasing the effectiveness of locally delivered biological agent to the tumor site. As such, different drugs that can be delivered systemically have been shown to synergize with SC-delivered TRAIL to potentiate its p53-independent proapoptotic mode of action in brain tumors. These include proteasome inhibitors such as bortezomib, HDAC inhibitors, genotoxic drugs such as cisplatin, the PI3-kinase/mTOR inhibitor PI-103, shRNA, and micro-RNA inhibitors.^48,124 We have also designed additional supplementary treatments such as microRNA-21 (miR-21) inhibitors⁵² and novel PI3-kinase/mTOR inhibitor PI-103¹²⁵ to augment the antitumor effect of different SC-mediated S-TRAIL therapy in mouse models of GBM in vivo. The combination of valproic acid (VPA) with MSCs expressing HSV-TK was shown to enhance the bystander effect, allowing enhanced cellular transmission of GCV-monophosphate in culture and increased survival in preclinical models of GBM.⁸⁹

GBM patients are often treated by maximal surgical resection followed by ionizing radiation and TMZ.⁶ Therefore, it is imperative to initially test SC therapies with preclinical models and study designs that incorporate current standard of care in an effort to be translated into the clinic. Previous studies have shown that tumor irradiation enhances the tumor tropism of human umbilical cord blood-derived MSCs mediated by increased IL-8 expression on glioma cells.¹²⁶ Similarly, MSCs expressing TRAIL¹²⁷ or interferon-β¹²⁸ were shown to be more effective at killing malignant GBM in the presence of TMZ. In another study, NSCs carrying OV used in combination with radiation and TMZ¹²⁹ were shown to prolong the survival of mice bearing patient-derived GBM xenografts.

Promising Preclinical Studies and Clinical Trials

The CNS lesion site(s) (focal vs multifocal) very much define(s) the route of SC transplantation.¹⁴ Direct local transplantation is suggested as a preferred route of NSC transplantation for focal CNS disorders, whereas the multifocality of certain CNS disorders represents a major limitation for intralesional cell-transplantation approaches. In mouse models of GBM, our earlier studies have shown that the intraventricular route of delivery is much more efficacious than the intravenous or intraperitoneal route.³³ Recent studies suggest that intranasal delivery of SCs is an emerging noninvasive means for the treatment of brain tumors.^130,131 Given that about 75% of GBM patients undergo tumor debulking,^31,132 another innovative strategy is the encapsulation and implantation of SCs in the resected GBM tumors in intracranial mouse GBM models. Biodegradable hydrogels and synthetic extracellular matrix (sECM) materials composed of hyaluronic acid, alginate, agarose, and other polymers¹³³ have been utilized to encapsulate SCs in a variety of rodent cancer models^133,134 and thus offer an important prospect for harnessing on-site delivery of tumor-specific agents. We have investigated a new approach to GBM treatment by encapsulating therapeutic SCs in GBM resection xenograft mouse models that recapitulate the clinical scenario of surgical debulking of GBM³¹ (Fig. 3). Human MSCs and NSCs encapsulated in sECM were retained in the resection cavity and permitted tumor-selective migration. sECM-encapsulated SCs engineered to express proapoptotic TRAIL protein or loaded with oHSV-TRAIL were placed into the mouse GBM tumor resection cavity and shown to increase survival significantly.^31,110 Recent studies have further shown significant therapeutic efficacy of MSCs loaded with conditionally replicating adenovirus expressing soluble hyaluronidase (ICOVIR17) compared with direct injection in a clinically relevant mouse model of GBM resection.¹⁰⁷ Similarly, we have shown that a low dose of cisplatin in combination with sECM-encapsulated SC-TRAIL transplanted in the GBM resection cavity significantly decreases tumor regrowth and increases survival in mice bearing GBM.¹³⁵ Together, these studies provide a platform for clinical translation of engineered SCs to target residual brain tumor cells post surgery.

Fig. 3.

Encapsulating engineered therapeutic stem cells in clinically relevant mouse tumor models of GBM. Stem cells expressing bifunctional molecules or loaded with oncolytic viruses can be encapsulated in biodegradable synthetic extracellular matrix and placed in the tumor resection cavity created after GBM tumor debulking in mouse tumor models of GBM.

Currently, a few trials are testing if modified SCs have active antitumor activity in brain tumor patients (Table 1). Genexine, Inc is sponsoring NCT02079324, which is intended to establish the safety and efficacy of MSCs expressing IL-12 (GX-051) administered intratumorally in patients with advanced head and neck cancer. In another trial, allogeneic NSCs modified to express CD are being tested in recurrent GBM (NCT01172964). Two more complementary trials have been approved that combine NSC-CD therapy with either leucovorin calcium (NCT02015819) or irinotecan hydrochloride (NCT02055196) to further sensitize GBM cells.

Table 1.

Stem cell therapies for cancer in clinical trials

Stem-cell Type	Therapeutics Loaded	Phase (study start)	National Clinical Trial Number	Targeted Cancer Types
NSC	Cytosine deaminase	I (12/2013)	NCT02015819	Recurrent high-grade glioma
NSC	Carboxylesterase	I (01/2016)	NCT02192359	Recurrent high-grade glioma
MSC	IL-12	I (03.2014)	NCT02079324	Head and neck cancer
MSC	MV-NIS	I/II (03/2014)	NCT02068794	Recurrent ovarian cancer
MSC	IFNβ	I (03/2016)	NCT02530047	Ovarian cancer
MSC	ICOVIR5	I/II (01/2013)	NCT01844661	Metastatic and refractory solid tumor
HSPC	rHIV7-shl-TAR-CCR5RZ	N/L (08/2015)	NCT02337985	AIDS-related non-Hodgkin's lymphoma
HSPC	LVsh5/C46 (CAL-1)	I (11/2015)	NCT02378922	Relapsed or refractory acute myeloid leukemia
HSPC	LV-C46/CCR5/P140K	I (03/2016)	NCT02343666	Lymphoma with HIV infection

Abbreviations: HSPC, hematopoietic stem/progenitor cell; ICOVIR5, oncolytic adenovirus; IFNβ, Interferon-β; L-12, interleukin-12; LV-C46/CCR5/P140K, lentiviral vector encoding multiple inhibitors for HIV replication; LVsh5/C46 (CAL-1), dual anti-HIV lentiviral vector; MSC, mesenchymal stem cell; MV-NIS, oncolytic measles virus encoding thyroidal sodium iodide symporter; N/L, not listed in clinical trial database; NSC, neuronal stem cell; rHIV7-shl-TAR-CCR5RZ, lentiviral vector encoding multiple anti-HIV RNAs.

Prospects and Caveats on the way to the Clinics

The adult SC represents a potentially powerful tool and offers great potential for use in the treatment of tumors in the brain. Transplantation experiments have revealed that both NSCs and MSCs migrate extensively towards both primary and metastatic tumors in the brain. These intrinsic abilities have been harnessed to deliver apoptosis-inducing proteins, immunostimulatory signals, antiangiogenic factors, cell cycle modulators, inducers of differentiation, and OVs. However, the integration of nascent therapeutic technologies such as genetically modified SCs would require considerable understanding of fundamental mechanisms before they can be efficiently translated into clinical settings. The use of a particular SC type in clinical trials will depend on the ability SCs to preferentially home to tumors in the brain, their easy availability, and their relative ease of manipulation in vitro without requiring immortalization. A major advantage of using autologous stem cells is their immunological compatibility, which has a profound effect on cell survival post transplantation. MSCs with multilineage potential show a broad migratory capacity for different brain tumor types including glioblastoma, medulloblastoma, ependymoma, and astrocytoma.^136,137 Hence, they have been studied as a better alternative to NSCs that, despite their strong tumor-tropic properties, have limited availability and ethical issues.¹³⁸ Due to their abundant availability, easy isolation and expansions, fewer ethical concerns and, most importantly, eligibility for autologous transplantation, human adipose tissue-derived MSCs are one of the most attractive vehicles for the delivery of therapeutic agents.¹³⁸ Although iPS-derived NSCs have been considered a feasible, effective, and autologous source for clinical applications, their therapeutic ability has not yet been fully addressed.¹⁹ The selection of the SC type, its purity, thorough understanding of its biology, and the effect of any modifications on its function should be established in preclinical studies before its consideration for clinical translation.

In recent years, patient-derived primary GBM lines have been created from isolated human brain tumor tissue and utilized for preclinical studies. A number of studies have shown that xenografts of these primary patient-derived cell lines often recapitulate clinical GBM more faithfully, thus providing additional insights and more stringent testing of promising new anti-GBM therapies.¹³⁹ A thorough biological understanding of such GBM tumor types will improve the chances of therapeutic success by yielding new therapeutic targets and enabling mechanism-based SC therapies to be developed, tested, and refined in response to a specific cancer profile.¹⁴⁰ The use of genetically engineered mouse brain tumor models in which tumors arise in situ or incorporation of surgical resection into GBM tumor studies might hold more validity and need exploration.

In clinical settings, the safety of the grafted SC is a major concern as there are still scientifically contentious issues regarding basic SC biology. Of particular concern is whether SCs promote the growth of certain tumors^141–143 or indeed form tumors themselves.^144,145 Importantly, non-immortalized adult stem cells provide fewer safety concerns than immortalized adult stem cells and may be used without posing a risk to the patient. The incorporation of suicide genes such as HSV-TK into therapeutic SCs would enable their controlled eradication post treatment and alleviate this concern. Because patient safety is vital, these issues should be resolved before starting the next generation of clinical trials.

Funding

This work was supported by R01CA138922 and R01CA173077.