Overcoming the Blood–Brain Barrier: Successes and Challenges in Developing Nanoparticle-Mediated Drug Delivery Systems for the Treatment of Brain Tumours

Classification of Brain Tumours

The most frequent brain tumours (gliomas) originate from glial cells, and range from low infiltrating to highly aggressive. In the 2007 World Health Organization (WHO) classified gliomas within four grades, basing on histopathologic features, such as mitotic index, anaplasia, cytological atypia, microvascular proliferation, and necrosis: grade I (ie pilocytic astrocytoma), grade II (ie astrocytomas and oligodendrogliomas), grade III (ie anaplastic astrocytomas and oligodendrogliomas), and grade IV (ie glioblastoma multiforme). In 2016 WHO included in the classification also molecular diagnostic criteria for infiltrating gliomas, including mutation of isocitrate dehydrogenase, deletion of 1p/19q chromosome, and histone mutations.¹ However, malignant or high grade (III and IV) gliomas are characterized by very poor prognosis. Furthermore, 8–10% of the adult patients with cancer develop brain metastases, with considerably variable incidence among different primary cancers. Lung, breast, colon, kidney cancer or melanoma can lead to brain metastases, 70% of which originating from lung and breast cancer.²

Current Therapy of Gliomas

Surgery is the first-line treatment both in low and high-grade gliomas³ and the extent of resection has demonstrated a positive prognostic effect.⁴ Several techniques have been designed to refine tumour resection: neuronavigation, use of 5-aminolevulinic acid,⁵ and intra-operative magnetic resonance imaging (MRI). There is evidence that the combined use of these techniques improves the rate of gross total resection. The choice and the timeframe of subsequent adjuvant chemotherapy and radiation therapy (alone or as combined treatments) is still considered controversial. A survey within the European Low-Grade Glioma Network showed a relevant heterogeneity in the usage of chemotherapy. Generally, oral temozolomide (TMZ) is the first-line treatment after surgery for high-risk low-grade gliomas, or at progression, although, according to the Radiation Therapy Oncology Group, combination of radiotherapy with procarbazine, lomustine and vincristine regimen has been indicated as the gold-standard treatment.⁶ While investigations are currently underway to evaluate the potential role of chemotherapy in low-grade gliomas, combined chemotherapy/radiotherapy approaches are currently practiced after surgery in high-grade gliomas. Radiotherapy is related to important side effects, such as post-radiation leuko-encephalopathy, nerve damage, hair loss, vomiting, infertility, and skin rash. Moreover, the effectiveness of chemotherapy is limited by toxic effects on healthy cells, tumour cell chemoresistance, and poor selectivity of anticancer drugs. Finally, the blood–brain barrier (BBB) is the major limit for the delivery of chemotherapeutic agents.⁷ Thus, the chemotherapeutics currently employed for high-grade gliomas are still limited to few chemical compounds. Currently, owing to the Food and Drug Administration (FDA), oral TMZ is the standard chemotherapy for glioblastoma and anaplastic astrocytoma. Bevacizumab (Avastin^®) is a monoclonal antibody that specifically binds vascular endothelial growth factor (VEGF). Despite FDA accelerated approval for bevacizumab for brain tumours, basing on its efficacy towards recurrent glioblastoma, its use has been involved with many controversies. Indeed, this anti-angiogenic therapy failed to improve patient overall survival, despite showing efficacy in shrinking or halting tumour growth.⁸ In 1996, FDA approved biodegradable polyanhydride wafers loaded with carmustine (Gliadel^®) for chemotherapy of recurrent high-grade gliomas. Patients with recurrent tumours benefit of an 8 weeks survival increase, when wafers were placed at the second surgery. Instead, the survival increase was 2.3 months in patients with early diagnosed tumours, undergoing primary resection followed by wafer placement.⁹

Experimental Drugs for Gliomas

Apart from currently approved chemotherapy, several drugs belonging to various therapeutic categories are currently under investigation for high-grade glioma treatment: the main mechanisms underlying their activity towards glioma are summarized in Figure 1. Advantages and disadvantages of such therapeutic drugs are listed in Table 1. In the following sections, the most important attempts and findings at preclinical and clinical level concerning such drugs are briefly described.

Table 1 Advantages and Disadvantages of Experimental Drugs Against High-Grade Gliomas

Figure 1 Main mechanisms of experimental drugs used against high-grade gliomas. Abbreviations: avß3/avß5, avß3/avß5 heterodimers; EGFR, epidermal growth factor receptor; EGFRvIII, mutant EGFR; MMP, matrix metalloproteinase; PDGFR, platelet-derived growth factor receptor; PI3K/Akt/mTOR, phosphoinositide 3-kinase/protein kinase B/mammalian target of rapamycin; PKC, protein kinase C; PTEN, phosphatase and tensin homolog; Ras/MAPK, Ras mitogen-activated protein kinase; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.

Cytotoxic Agents

Different marketed cytotoxic drugs have been tested off-label in preclinical glioma models and clinical trials, including nitrosoureas (alkylating agents), platinum salts, inhibitors of topoisomerase I (etoposide) and II (camptothecin, irinotecan, topotecan), mitotic inhibitors that is taxanes derivatives (paclitaxel, docetaxel), anthracyclines such as doxorubicin (DNA intercalation and topoisomerase II inhibition), and paclitaxel–cisplatin–vincristine (PCV) combination.¹⁰

Prodrugs of Cytotoxic Agents

Lipophilic prodrugs with molecular weight (MW) lower than 500 Da, and capable of forming less than 8 hydrogen bonds, should be able to overcome the BBB:¹¹ the cytotoxic drug chlorambucil was modified accordingly, with improved brain delivery.¹² However, recent interest has been growing concerning higher MW compounds, such as fatty acid – paclitaxel (PTX) prodrugs. In particular cis-linoleic acid conjugate with PTX (CLA-PTX) resulted in much higher plasmatic half-life and brain accumulation than free PTX, with encouraging therapeutic effect on brain tumour bearing rats.¹³

Angiogenesis Inhibitors

The expression level of VEGF directly correlates with tumour grade, with a nearly 10-fold gap between high and low-grade gliomas.¹⁴ Thus, endothelial cells are a suitable target for high-grade glioma treatment. Apart from monoclonal antibody bevacizumab, that selectively binds VEGF, potential therapeutic agents include thalidomide and VEGF receptor (VEGFR) inhibitors, belonging to receptor tyrosine kinase (RTK) inhibitors category.¹⁵ Cediranib and vatalanib are orally bioavailable VEGFR inhibitors, with simultaneous inhibitory activity on tyrosine-protein kinase KIT (c-Kit) and platelet-derived growth factor receptor (PDGFR), that, currently, are undergoing clinical trials for high-grade gliomas.¹⁶

Epidermal Growth Factor Receptor (EGFR) Inhibitors

EGFR, together with downstream signalling pathways (such as phosphoinositide 3-kinase/protein kinase B/mammalian target of rapamycin - PI3K/Akt/mTOR and Ras mitogen-activated protein kinase - Ras/MAPK), regulates cell survival, proliferation, angiogenesis and migration. EGFR is overexpressed in over 40% of glioblastomas.¹⁷ Furthermore, 25% of glioblastomas are characterized by the expression of an EGFR mutant (EGFRvIII), which is defective of the extracellular ligand-binding domain.¹⁸ Erlotinib, lapatinib, and gefitinib, three RTK inhibitors acting on EGFR, have been tested at preclinical and clinical stage for malignant gliomas treatment, with limited efficacy.¹⁵ Also, cetuximab and nimotuzumab, two mouse anti-human EGFR antibodies, were employed in clinical trials for recurrent high-grade glioma patients.^9,19

PDGFR Inhibitors

PDGFR is an RTK, which is overexpressed in more than 20% of glioblastomas,²⁰ and, like EGFR, it is upstream of Ras/MAPK and PI3K/Akt/mTOR signalling pathways.²¹ Imatinib, an inhibitor of PDGFR, c-Kit, Abelson murine leukemia viral oncogene (Abl), and arginase (ARG), was tested in several clinical studies involving high-grade glioma patients.^15,22

Ras/MAPK Inhibitors

The Ras/MAPK signalling pathway is downstream to several RTK inhibitors, such as fibroblast growth factor receptor (FGFR), insulin-like growth factor (IGFR), and the aforementioned EGFR, VEGFR, PDGFR. Therefore, it is responsible for cell proliferation and migration processes, as well as for cell cycle progression, and malignant transformation. Ras/MAPK is also upstream to PI3K/Akt/mTOR signalling pathway.^15,21,23 Furthermore, malignant gliomas are characterized by Ras/MAPK up-regulation.²³ Oral tipifarnib, a farnesyl transferase inhibitor with demonstrated radio-sensitizing effect is currently undergoing several clinical trials for high-grade gliomas.²⁴

Protein Kinase C (PKC) Inhibitors

PKC enzymes belong to a serine/threonine kinases family, and it is downstream to different RTKs, including VEGFR and PDGFR.²⁵ PKC is involved in chemoresistance and radioresistance in malignant gliomas .²⁶ Furthermore, PKC is frequently overexpressed in malignant gliomas.^26–28 Enzastaurin, a lipophilic, orally administered PKC inhibitor, underwent different clinical trials as mono or combination therapy, but with limited clinical benefit.⁹

PI3K/Akt/mTOR Pathway Inhibitors

PI3K/Akt/mTOR is down-regulated by phosphatase and tensin homolog (PTEN), which, in turn, is altered (deleted, inactivated, or mutated) in nearly 40–70% glioblastomas.^29,30 Sirolimus, an orally bioavailable peptide macrolide, which is able to overcome the BBB and acts as mammalian target of rapamycin (mTOR) inhibitor, is currently employed in clinical trials for malignant gliomas.³¹ Also, everolimus and temsirolimus, two analogs of sirolimus, respectively, orally and i.v administered, have been tested in clinical trials.^32,33 However, mTOR inhibitors induce significant activation of protein kinase B (Akt) in 50% of the patients, potentially causing a reduced time to progression: simultaneous inhibition of mTOR and Akt could overcome this limitation.³⁴

Multi-Target Inhibitors

The simultaneous inhibition of different aberrant signalling pathways should result in great clinical benefit for targeted therapies. Apart from the combination of different drug therapies, multiple-target inhibition can be achieved through employment of small molecule inhibitors of newer-generation.¹⁵ Dasatinib, pazopanib, sunitinib, sorafenib, vandetanib, tandutinib are currently in evaluation in clinical trials for glioblastoma, alone or in combination therapy.⁹

Matrix Metalloproteinase (MMP) Inhibitors

MMPs are proteolytic enzymes active also in physiological conditions. In tumours, they promote basal membrane invasion, metastatization and angiogenesis, being directly secreted by tumour cells or by the surrounding stroma, under tumoural stimulus. Marimastat, the most studied MMP inhibitor, showed promising in vitro inhibition of malignant glioma, but no advantage in clinical trials with glioblastoma patients.^35,36

Integrin Inhibitors

The expression of avß3 and avß5 heterodimers is increased in malignant gliomas, and it has been hypothesized that they contribute to regulate angiogenesis and migration.³⁷ Cilengitide is a potent antagonist of both avß3 and avß5. Several clinical trials testing cilengitide in combination therapy regimens are currently ongoing.⁹

Rationale for Employment of Nanomedicines in Glioma Therapy

The main mechanisms underlying the rationale of employing nanomedicines for glioma treatment are summarized below (Figure 2).

Figure 2 Rationale of employing nanomedicines for glioma treatment. Abbreviations: BBB, blood–brain barrier; BBTB, blood–brain tumour barrier; CMT, carrier-mediated transport; EPR, enhanced permeation and retention; HIF, hypoxia-induced factors; P-gp, P-glycoprotein; RMT, receptor-mediated transport.

Passive Targeting

The permeability of adjacent brain vasculature changes during the growth of brain tumours. At their early stage, the growth of tumour cells depends only on normal brain vessels and the BBB is intact. With tumour progression, glioma cells invade the surrounding healthy tissues. When a large enough volume (>0.2 mm³) is reached by the tumour cell cluster, a structural damage will affect the BBB, and blood–brain tumour barrier (BBTB) will be formed.³⁸ Within this concern, claudin-1, a protein specifically expressed in the tight junctions of endothelial cells, is down-regulated in vessels surrounding high-grade gliomas, thus increasing the permeability of the BBB.^39,40 This “leaky” BBTB is a common feature of high-grade gliomas, because of their increased metabolic requirements. Furthermore, it has been demonstrated that VEGF, associated with the high angiogenic nature of high-grade gliomas, increases the BBB permeability, by stimulating angiogenesis in response to hypoxia.^41,42 The “leaky” BBTB can be targeted for drug delivery purposes, by exploiting the so-called enhanced permeability and retention (EPR) effect. Since the pores cut-off size at the BBTB is highly variable, the size of nanocarriers largely influences the extent and efficacy of drug delivery. In fact, the vascular leakage is significantly reduced in the tumours within the brain microenvironment compared with other districts, and it is affected by the tumour stage and/or to the tumour model employed.⁴³ For instance, the critical cut-off size of the BBTB for intracranial U87MG xenografts ranges between 7 and 100 nm.⁴⁴ Despite the reduced size of vessel fenestrations may limit the potential of passive targeting strategy to address brain cancer, drug-loaded nanocarriers free of targeting ligands showed promising efficacy in orthotopic brain glioma models, thus allowing a translation to clinical therapy.⁴⁵

Active Targeting

However, high-grade gliomas rapidly infiltrate the surrounding healthy tissue, where the BBB is not altered and the EPR effect cannot be achieved.⁴⁶ Indeed, in human patients, contrarily to animal tumour models, chemotherapy is only adjuvant, and it should mainly be addressed towards the eradication of residual tumour cells after surgery or radiotherapy. This cell population includes cells migrating from the tumour into healthy tissue or that feed the tumour from distant sites still active. Thus, the most relevant area for chemotherapy is the one surrounding the glioma, the so-called BAT (Brain Adjacent to Tumour), including cells in the invasion phase, that do not still affect the integrity of the BBB. Since BAT should be the main target of chemotherapy, the development of formulations easily crossing the intact BBB is essential.⁴⁷ In this case, suitable active targeting of nanocarriers is needed to reach the target tissue, by exploiting carrier and/or receptors overexpressed at the BBB. Carrier-mediated transporters (CMTs) are entailed in the transport of essential small molecules into the brain. Receptor-mediated transporters (RMTs) are abundantly expressed at the BBB, being exploited by large endogenous biomolecules.^2,48 Specific peptide receptors are included in RMT, eg low-density lipoprotein (LDL) receptor, transferrin (Tf) receptor, lactoferrin (Lf) receptor, insulin receptor, and receptors for insulin-like growth factors (IGF-1 and IGF-2).⁴⁹ CMT and RMT ligands have been frequently exploited for drug delivery as molecular “Trojan horses” for nanocarriers.^50–55

Overcoming Extrusion Mechanisms at the BBB

Additionally, in the BBB endothelial cells (on the luminal side), as well as in glioma cells, the P-glycoprotein (P-gp) is present. P-gp is a 170-kDa glycoprotein associated to the cell membrane, belonging to the superfamily of ATP binding cassette (ABC) transporters, that actively extrudes from cells chemically different substrates. P-gp can hamper BBB overcoming by chemotherapeutic drugs, and it is responsible for chemoresistance in glioma cells.^56,57 Thus, P-gp inhibition can be considered as a potential dual strategy, suitable both to enhance drug penetration in the brain, and to reduce glioma chemoresistance. Thus, specific inhibitors of ABC subfamily member 1, such as elacridar and tariquidar, and anti-estrogen tamoxifen, were investigated for glioma treatment, owing to their anti-extrusion effect at the BBB.^58,59 Apart from chemical modulators, also nanoparticulate systems can inhibit the P-gp drug efflux mechanism: indeed they can deliver the drug within the cells by endocytosis, therefore making it less susceptible to the membrane-bound drug efflux mechanisms, since it is in the form of drug-matrix aggregates. Also, some ingredients employed in the formulation of nanocarriers, such as surfactants, can contribute to the P-gp inhibition.⁶⁰ Furthermore, co-delivery in the same nanocarriers of cytotoxic drugs and P-gp inhibitors (such as ketoconazole) has also been attempted in preclinical models.⁶¹

Inhibition of Tumour Differentiation, Migration and Neo-Vascularization

Lastly, from a histological point of view, in malignant gliomas, areas of necrosis, hypoxia and microvascular hyperplasia are present, with pseudo-palisades of cells migrating from the original necrosis area.^62,63 Accordingly, it has been recently shown that cancer stem cells (CSC) are responsible for the initiation step of human glioblastoma and medulloblastoma in intracranial mouse xenografts.⁶⁴ Unfortunately, the CSC location has yet to be precisely disclosed, but it has been demonstrated that hypoxia-induced factors (HIF) are responsible for CSC phenotype and can even shift normal cells to CSC.^65,66 Since HIF are involved both in angiogenesis and in maintaining CSC, a suitable treatment for high-grade gliomas should be able to down-regulate their expression. This can occur by scavenging of reactive oxygen species (ROS). Indeed, during hypoxia, rising intracellular ROS concentration overcomes glutathione (GSH) guard levels. ROS can stabilize HIF, allowing VEGF transcription and consequent angiogenesis.^67–70 This mechanism underlies the rationale for employing anti-oxidant and anti-VEGF agents in glioma therapy. Indeed, the combination of anti-angiogenic bevacizumab and cytotoxic irinotecan is approved for recurrent glioblastoma.^71,72 Furthermore, in mice bearing intracerebral glioblastoma, pre-treatment with Tempol, a ROS scavenger, followed by TMZ chemotherapy, suppressed tumour growth and increased survival rate.⁷³ Within this concern, nanocarriers should be able to achieve co-delivery of anti-VEGF drugs and antioxidants, together to cytotoxic agents within the tumour tissue, owing to the aforementioned mechanisms. A recent experimental study showed an increase in bevacizumab activity and permeation through the BBB, when loaded within solid lipid nanoparticles (SLN).⁷⁴ Also treatment cerium oxide nanoparticles (nanoceria), with ROS scavenging properties, caused decreased expression levels of VEGF in a human astrocytoma cell line, associated with reduced motility and capacity of endothelial cells to form new capillaries.⁷⁵ Recently, camptothecin (CPT), a topoisomerase inhibitor, was employed in a co-delivery nanoparticulate system, that could stop the stabilization of HIF in brain tumours. For this, a suitable CPT prodrug was synthesized, by linkage with a tetraethylene glycol (TEG) spacer and α-lipoic acid (ALA). The obtained CPT-TEG-ALA prodrug can be cleaved by oxidation, thus acting as a ROS scavenger, and release CPT in its active form within glioblastoma cells. CPT-TEG-ALA was loaded in nanoemulsion along with α-tocopherol, an additional ROS scavenger, thus preventing HIF production.⁷⁶

Preclinical Nanomedicines

Several approaches involve the alteration of the BBB, in order to increase brain penetration of nanocarriers loaded with chemotherapeutic drugs against glioma tumours. They include both invasive methods,⁷⁷ such as convection-enhanced delivery^78,79 and post-surgical implantation,^80–82 and noninvasive BBB opening through focused ultrasound (FUS).⁸³ Indeed, ultrasound with a frequency below 1 MHz can induce reversible and temporary BBB opening with the aid of microbubbles (Figure 3). This technique can be employed to deliver theranostic agents for the detection and treatment of various brain diseases, and it has been already subject of clinical trials.⁸⁴ In particular, Cu-Se ultrasmall nanoparticles and rare earth nanoparticles, labelled with near-infrared (NIR) dyes, have been employed, at preclinical level, to monitor FUS-induced temporary opening of the BBB and following recovery,⁸⁵ as well as to detect glioma in orthotopic animal models.⁸⁶ In further studies, the aforementioned Cu-Se ultrasmall nanoparticles, loaded with doxorubicin (DOX) and labelled with NIR dyes, were used concurrently with FUS and photodynamic therapy, and demonstrated good efficacy against orthotopic tumour models.⁸⁷

Figure 3 Scheme of brain delivery of nanoparticles with focused ultrasound (FUS) technique. Abbreviation: SF6, sulfur hexafluoride.

Anyway, the nanocarriers discussed here below include only those used to overcome the BBB/BBTB without disruption, aiming to deliver anti-cancer drugs to the brain tumours. They can be made of different matrixes, including metals (ie calcium phosphate, iron oxide), lipids (SLN), phospholipids (liposomes), proteins (bovine serum albumin – BSA), synthetic polymers (ie poly-lactide – PLA; poly-lactide-glycolide – PLGA; PEG: polyethyleneglycol), natural polymers (ie chitosan) and polymer-lipid hybrid nanoparticles (PLN).⁸⁸

Requested features of such nanomedicines are the following:⁴⁶

• They should stabilize the drug from physicochemical or biological standpoints;
• They should avoid opsonization and, therefore, clearance by the reticuloendothelial system (RES), prolonging plasma circulation time and allowing EPR effect;
• They should be endowed with selective targeting strategies to the brain;
• They should not stimulate immune response.

However, the fact that almost 98% of drugs are unable to overcome the BBB is the main reason for employment of nanocarriers in high-grade glioma treatment. Two approaches have been documented in literature: the first employs plain nanocarriers, the second exploits active targeting, in order to enhance uptake of the nanocarriers by the endothelial cells at the BBB/BBTB.⁸⁹

Within this concern, a huge number of nanocarriers aiming to glioma treatment have been engineered in the last decades. However, the following discussion will be focused only on the ones that underwent preclinical in vivo testing. Indeed, despite the presence of several in vitro investigation methods,⁹⁰ only preclinical animal models can be predictive of the real potential of administered nanomedicines, because of the simultaneous presence of biological barriers and metabolism/distribution. Pharmacokinetic and biodistribution experiments are relevant in order to foresee the metabolic fate of nanocarriers and loaded drugs, while glioma models allow to predict in vivo efficacy. Different mouse and rat glioma models are described in literature:⁹¹ ethyl-nitrosourea induced orthotopic models in allogeneic or syngeneic healthy animals, orthotopic xenografts, or genetically engineered models. Within this concern, it should be pointed out that the glioma model employed can affect in a relevant manner the measured efficacy of the therapeutic drug delivery system under investigation, with the immune system playing a key role. In fact, today the so-called “immune privilege” of the brain is no more a retained concept.^92,93 Thus, if the absence of the immune system could cause an under-estimation of the therapeutic effect in xenografts, on the other side, the presence of the immunity in immuno-competent models can hamper the reproducibility of the glioma model itself. Indeed, the superior reproducibility of syngeneic models over allogeneic has been documented, owing to a lower graft rejection.⁹⁴ Furthermore, despite surgical resection is the primary approach for high-grade gliomas, the majority of preclinical models focus solely on drug treatment of solid intracranial tumours. Within this concern, recently a resection and recurrence orthotopic model has been developed, with potential for the investigation of tumour ablation combined with local and systemic chemotherapy.⁹⁵

Plain Nanocarriers

Plain nanocarriers can improve drug delivery to gliomas by stabilizing the drug in the bloodstream, such as in the case of TMZ, which suffer from pH-dependent nonenzymatic chemo-degradation at the neutral pH, despite being the most employed drug for the treatment of brain cancers. Published reports demonstrated the therapeutic advantage of loading TMZ inside nanostructured lipid carriers (NLC), which is an optimized form of SLN with higher drug loading capacity.^96,97 Naked nucleic acid delivery is also associated with fast degradation, aspecific biodistribution and poor cell internalization. In fact, small interfering RNA (siRNA) and microRNA (miRNA) are promising tools to treat various diseases, but, due to their instability and poor delivery within target tissues, naked RNA is not robustly used. Thus, entrapment in positively charged nanocarriers can prevent from degradation/metabolism, and favour cell internalization, both at the BBTB and in glioma cells.⁹⁸

Nanocarriers can also increase drug half-life (for EPR effect), and/or favour endocytosis across BBTB endothelial cells, also owing to P-gp inhibition. In particular, this mechanism has been well documented for lipid nanocarriers, such as liposomes and SLN.⁹⁹ The physiological nature of the lipid matrix employed should improve natural brain uptake, and the employment of surfactants, such as Brij 78 or Polysorbate 80, to coat SLN can improve the nanoparticle brain uptake and drug activity, being the two molecules associated to P-gp inhibition.^100,101 However, SLN mechanism of action excludes any significant toxic effect on the BBB, as assessed by the absence of alterations to cell integrity and permeability, vessel blood flow, and choline active transport. Indeed, Western blot analyses of occludin and claudin-1 in the BBB cells, run following SLN administration, showed no modifications in protein expression.¹⁰² Interestingly, in the case of the insoluble prototype anticancer drug named edelfosine (EDF) and of ion-paired idarubicin, also oral administration of drug-loaded SLN was effective against glioma subcutaneous models¹⁰¹ and in enhancing drug concentration in the brain,¹⁰³ respectively.

Finally, plain nanocarriers can also be employed for stimuli-responsive formulations. Indeed, the combination of silver nanoparticles (AgNPs) with magnetic nanoparticle hyperthermia (MNPH) was used as a treatment in the glioma model.¹⁰⁴

The most important plain nanocarriers, employed in in vivo preclinical studies for glioma non-invasive treatment, are shown in Table 2.

Table 2 Plain Nanocarriers Aiming to Glioma Therapy Employed in Preclinical Studies

Functionalized Nanocarriers

Nanocarriers can be functionalized on their surface with suitable ligands, in order to exploit RMT and CMT at the BBB. As previously mentioned, several specific receptors, in particular for endogenous proteins and peptides, were employed as molecular “Trojan horse” for nanocarrier systems.⁴⁹ LDL receptor is overexpressed both at the BBB and in glioma cells and, therefore, it is a potential target to be exploited for the delivery of therapeutic agents. Apolipoproteins (Apo) are usually employed to target LDL receptors, but Apo being a high MW protein, usually shorter chimera peptides including the receptor-binding domain are employed.^50,105 Angiopep-2 (TFFYGGSRGKRNNFKTEEY, molecular weight 2.4 kDa), a peptide belonging to the Kunitz domain-derived family, is a potent ligand of LDL on the BBB employed for nanocarriers functionalization.¹⁰⁶ Within this concern, it should be pointed out that Polysorbate 80 is frequently employed as suspending agent for plain nanocarriers aiming to drug brain delivery. Indeed, it has been hypothesized that this nonionic surfactant can adsorb endogenous Apo E present in serum on the surface of nanoparticles.^107,108 Mammalian Lf is a cationic iron transporting glycoprotein (80 kDa), whose receptor is expressed on the endothelial side of the BBB. Studies with membrane preparations of mouse brain have shown that the Lf receptor at the BBB has two classes of binding sites: a high-affinity, with a dissociation constant K_d of 10.61 nM, and a low-affinity with a K_d of 2228 nM. Plasmatic concentration of endogenous Lf (~5 nM) is lower than the K_d of Lf receptors at the BBB. Thus, the competitive inhibition with endogenous Lf is avoided, allowing employment of Lf functionalized nanocarriers for drug delivery in glioma models.¹⁰⁹ Finally, apart from protein ligands, nanocarriers can be functionalized also with small molecules, recognized at the BBB from specific receptors/carriers, such as GSH receptor,¹¹⁰ folate receptor and monocarboxylic acid transporter (MCT-1).^61,111

However, different critical issues are associated with active targeting to the BBB. First of all, most of the RMT are not present exclusively at the BBB. For example, Tf receptor is expressed in monocytes, red blood cells, lungs, hepatocytes, and in the gut, along with the BBB. Within this concern, it is reported that active targeting of nanocarriers should benefit from employment of a “spacer” between the nanoparticles’ surface and the grafted protein, and that grafting a low amount of protein should enhance selectivity for the BBB rather than nontarget tissues.⁵¹ Moreover, nanocarriers grafted with physiological ligands (ie Tf) undergo binding competition with the corresponding endogenous protein, which can decrease targeting efficacy;⁴⁶ then, most of the available targeting proteins can cause immunogenic reactions; finally, some receptors on the BBB, such as insulin and Tf, control homeostasis of iron and glucose within the brain, and nanocarriers grafted with monoclonal antibodies against these receptors may down-regulate their activity and raise safety concerns.¹¹²

Functionalization of nanocarriers can be exploited also to target specific receptors overexpressed in glioma and/or at the BBTB. Previously mentioned overexpression of EGFR and EGFRvIII mutant at the BBTB allows a selective targeting.¹¹³ Also, integrins play important roles in tumour invasion and angiogenesis: in glioblastoma avß3 and avß5 integrin receptors are overexpressed on brain tumour cells and neo-vessels, favouring interaction with the extracellular matrix. Integrins can be specifically targeted by employing arginine-glycine-aspartic acid (RGD) peptides, such as cilengitide (cyclo [RGDfV]), a cyclic RGD peptide.¹¹⁴ Nucleolin, instead, exists only in the nucleus of cells, thus offering an attractive target for cells characterized from a high internalization rate. F3 peptide, that specifically binds to nucleolin, was utilized to decorate nanocarriers, in order to realize glioma cell and neo-vasculature dual-cellular targeting.¹¹⁵ Furthermore, as previously mentioned, MMPs are a group of zinc-dependent proteins acting as key modulators of tumour invasion and metastatization, due to their degrading capacity of the extracellular matrix. In addition, MMP-2/9 is also required for the tumoural angiogenic switch. Thus, MMP-2/9 conjugated low-molecular-weight protamine (ALMWP) was employed for glioma targeting of nanocarriers.¹¹⁶ Also chlorotoxin (CTX), a small peptide (36-amino acid) derived from Leiurus quinquestriatus (scorpion) venom, can bind to MMP-2, and it was employed for nanocarriers functionalization.¹¹⁷ Finally, IL-13Ra2 receptor is overexpressed in pilocytic astrocytomas and glioblastoma. Therefore, interleukin-13 (IL-13) can be used to target nanocarriers to glioblastoma tissue.

A summary of the most important functionalized nanocarriers employed in in vivo preclinical studies is shown in Table 3, and a scheme of the main active targeting strategies is displayed in Figure 4.

Table 3 Functionalized Nanocarriers Aiming to Glioma Therapy Employed in Preclinical Studies

Figure 4 Scheme of the main mechanism used for active targeting of nanocarriers in glioma therapy. Blue: endogenous ligands; red: exogenous ligands. Abbreviations: avß3/avß5, avß3/avß5 heterodimers; CTX, chlorotoxin; EGFR, epidermal growth factor receptor; GSH, glutathione; HFE, homeostatic iron regulator protein; IL 13, interleukin 13; LDL, low-density lipoprotein; Lf, lactoferrin; MCT-1, monocarboxylic acid transporter 1; MMP, matrix metalloproteinase; Tf, transferrin.

Nanocarrier-Mediated Cell Therapy

In addition to the selective BBB, glioma chemotherapy can be hampered from the fact that it is difficult to obtain a sufficiently high drug concentration in the tumour, in order to kill infiltrative malignant cells, without hampering healthy brain tissue. Thus, even if a relatively high drug accumulation within brain tissue is obtained, chemotherapeutics may undergo unwanted distribution in the extracellular matrix, or within intratumoral necrotic pockets, but without exerting the desired effects on target glioma cells. These limitations may be overcome by using stem cells as carriers of nanoparticulate delivery systems.¹¹⁸ Indeed, cell-based therapies represent an innovative and valuable tool for cancer treatment, and recent interest is growing also for high-grade glioma therapy. A number of cells have been recently studied, either in preclinical or clinical trials for brain tumour targeting, and can be exploited also in combination with nanocarrier systems.

In particular, adult mesenchymal stem cells (MSC) show a tumour tropic ability, that can be exploited simultaneously to target tumour and deliver therapeutic agents.¹¹⁹ With genetic modification, MSC are able to home cancer tissues and affect the tumor growth by the secretion of cytotoxic molecules. Moreover, MSC are able to trespass the BBB in physiological and pathological conditions, and preclinical studies showed that MSC engineered to express suicide gene enhance the antitumor response in glioblastoma animal models. Within this concern, the combinatorial approach of cell therapy, small-molecule chemotherapy and nanomedicine strategies can open new opportunities for glioma treatment (Figure 5). Internalization/binding of drug-loaded nanoparticles into MSC can be exploited to increase the antitumor efficacy by targeted delivery to the tumour microenvironment.¹²⁰ Indeed, PTX loaded PLGA nanoparticles showed an enhanced activity in brain tumor, which is ascribed to the sustained release of the drug.¹²¹ In a recent work, a hybrid system consisting of MSC spheroids and methotrexate (MTX)-loaded nanoparticles was engineered, in order to increase retention at tumour site: this system improved tumour inhibition in a heterotopic glioblastoma murine model.¹²² Even though drug-loaded MSC are a promising strategy, some major issues should be considered, such as the fact that the conjugation of nanodrugs to MSC surface could affect the tumour homing ability, as demonstrated by a lot of studies carried out on brain tumour xenografts.

Figure 5 Scheme of nanocarrier mesenchymal stem cells (MSC) combined therapy of gliomas.

Interestingly, also adipose-derived stem cells (ADSC) show tumour homing ability. This behaviour can be exploited to design drug delivery systems for brain tumours, such as cell-based carriers for nanoparticulate systems. Moreover, it is possible to endow such nanoparticles with stimuli-responsive properties for targeted drug delivery. Indeed, superparamagnetic iron oxide nanoparticles (SPION) were loaded with PTX and subsequently taken up by ADSC via endocytosis.¹²³ Drug release was then activated by high-frequency magnetic field in a glioblastoma murine model. The dual-modality therapeutic strategy including ADSC and smart nanoparticles may be further investigated in the near future for clinical translation.

Given that the “immune privilege” of the brain is no more a retained concept, also the recent advances in immunotherapy strategies offer new opportunities for a synergistic combination of cell therapy with chemotherapeutic drug delivery systems, in order to target brain tumours.¹²⁴ For example, the combination of dendritic cell-targeted vaccines with nanoparticles loaded with anticancer drugs should lead to a superior anticancer activity.¹²⁵ Combination with nanocarrier loaded with checkpoint inhibitors could be another valuable therapeutic approach.¹²⁶ Furthermore, neutrophils (NE) have a native ability to traverse BBB/BBTB and penetrate the glioma site, where the tumour associated NE favour the continuous recruitment of circulating NE; in addition, local brain inflammation, following surgical tumour removal, activates NE migration towards to the inflamed brain. This amplification of inflammatory signals supports an enhanced brain tumour targeting of nanocarriers loaded in NE, such as PTX loaded liposomes¹²⁷ and DOX loaded magnetic mesoporous silica nanoparticles, which is evident also in resection and recurrence orthotopic models.¹²⁸ Finally, chimeric antigen receptor (CAR) T cells should be considered, too. They are T cells, generally autologous, ex vivo artificially engineered on their surfaces in order to recognize tumour-associated antigens. The CAR shows both an antigen-binding and a T cell activating function. These modified CAR T cells can produce the lysis of the cells presenting the associated tumor antigen when administered intravenously in patients. Currently, CAR-T cell therapy has been approved in B-cell lymphoma and leukemia, but clinical trials are ongoing for glioblastoma treatment.^129,130 Very recently CAR T cell administration showed to improve anti-glioma response.¹³¹ Also in this case, CAR-T cells could be associated with nanocarriers in order to improve the therapeutic outcome, and CAR-T cell surface can also be modified to targeting purposes. Indeed, human epidermal growth factor receptor 2 (HER-2) functionalized CAR T cells demonstrated increased persistence over time.

Within this context, suitable attention should be devoted also to cell membrane-coated nanoparticulate systems. Wrapping nanoparticles with cell-derived membranes provide nanocarriers of a natural surface coating with complex biological entities, which are nearly impossible to synthetically replicate via ligand attachment. This allows to overcome some of the previously mentioned shortcomings of actively targeted nanocarriers, such as opsonization and incorrect ligand recognition. Cell membrane coating technology was firstly introduced by employing red blood cell membranes, in order to provide “stealth” properties to synthetic nanoparticles. Currently, it has been applied to many cell types, including cancer cells.¹³² Indeed, cancer cells possess the unique ability to self-target homologous cells (the so-called homotypic targeting), which can be translated to wrapped nanocarriers, showing also reduced immune clearance compared to uncoated ones (Figure 6). Recently, in a study involving PLGA nanoparticles coated with U87 glioma cell membrane fractions, an induction of cancer cell-specific immune response was demonstrated.¹³³

Figure 6 Scheme of homotypic targeting of glioma cell membrane-coated nanoparticles.

Clinical Nanomedicines

Despite relevant evidences at the preclinical level, and several clinical studies addressing the employment of nanocarriers for brain tumours chemotherapy, particularly for glioblastoma, till now no new nano-drug has yet approved for brain tumour therapy. In general, the approval rate for novel nanomedicines is below 10%, mainly because of safety and efficacy profile failures during preclinical and clinical studies.¹³⁴ Indeed, regulatory agencies require manufacturers to perform accurate preauthorization studies to assess the quality, safety, and efficacy profiles of a new nanomedicine.^135,136 Furthermore, the difficulty to find out suitable preclinical models that truly represent what happens in the humans is a major flaw, that hampers clinical translation of nanomedicines.⁹⁵

Within this concern, according to the European Commission Recommendation (2011/696/EU), 100 nm is the demarcating upper size limit where the properties of materials can change significantly from conventional equivalents, and the European Medicines Agency (EMA) also included an official definition of nanomedicine as being up to a size of 100 nm.¹³⁷ However, frequently nanomedicines have broader size range than the proposed definition, inducing the EMA to include all “structures” with sizes of less than 1000 nm, that are designed to have specific properties,¹³⁸ can improve site-specific drug delivery and significantly alter toxicological profiles, thus allowing to perform a case-by-case evaluation.¹³⁹ Nonetheless, despite the numerous guidelines existing for the validation methods of chemical parameters in such matrixes, suitable implementation of this legislation should require validated analytical methods for nanoparticles’ characterization, that, so far, do not exist.¹⁴⁰

Some of the most important clinical trials with new emerging nanomedicines are here reported. Few of them involve active targeting mechanisms. In fact, despite the existence of good manufacturing practices for nanodrug delivery systems, that can be employed for translation to clinical trials,¹⁴¹ actively targeted nanomedicines are associated with high costs and scale-up issues. A Phase I clinical study involving patients with recurrent glioma (grade 2–4) has been carried out by administering PTX-Angiopep-2 peptide–drug conjugate (GRN1005). Even if GRN1005 improved PTX permeation into tumour tissue, Phase II trial interim analysis did not show therapeutic response.^142,143 Tf conjugated diphtheria toxin (Tf-CRM107) showed in vitro toxicity towards glioma cells and it was effective in xenografts after local administration. Moreover, in phase I and II clinical trials, local administration of Tf-CRM107 resulted in low toxicity, encouraging response rate (35%) and promising overall survival (74 weeks) in patients with recurrent high-grade brain tumours. However, early Phase III clinical trials employing this approach were terminated due to disappointing preliminary results.^9,144 In another interesting phase I study, EnGeneIC delivery vehicle loaded with DOX was tested in patients with recurrent glioblastoma. The anti-EGFR monoclonal antibody Vectibix was used to target EGFR on cancer cells, thus leading to DOX release.¹⁴⁵

However, the most important clinical evidences were obtained with off-label employment of already marketed liposomal DOX, acting mainly owing to the EPR effect. In a clinical study liposomal DOX administered in patients with high-grade gliomas improved overall survival.¹⁴⁶ Moreover, DOX loaded in pegylated liposomes (PEG-DOX) was efficacious and well tolerated in patients with recurrent high-grade glioma.¹⁴⁷ Indeed, encouraging results were obtained with PEG-DOX alone (mOS 26 weeks) or in combination with TMZ (median overall survival 32 weeks, 6-months progression-free survival 32%). Owing to the available studies, PEG-DOX, alone or in combination regimens, can actually be considered as a treatment option for recurrent high-grade gliomas, but only if no further chemotherapy is available; however, it should be further evaluated in larger clinical trials.¹⁴⁸

Conclusions

Malignant gliomas are still associated with a poor prognosis, despite recent advances in surgical treatment. Despite the large number of potential drug candidates, the efficacy of adjuvant chemotherapy remains unsatisfactory, primarily because of the BBB. Nanocarriers can favour delivery of chemotherapeutics to malignant gliomas owing to different mechanisms, including chemical stabilization of the drug in the bloodstream, EPR effect (because of the leaky BBTB), P-gp inhibition, active targeting through CMT and RMT, inhibition of cell differentiation and angiogenesis, cell-mediated targeting, or stimuli-responsive delivery. In particular, different and efficient active targeting approaches have been attempted in preclinical studies on animal models, mainly by employing protein targeting moieties. Nevertheless, a few number of nanomedicines reached the clinical trials, and most of them include drug-loaded nanocarriers free of targeting ligands, probably because of safety and scalability concerns.