Glioma is the term referring to brain tumors that arise from glial or precursor cells. This group of oncological conditions includes astrocytoma, oligodendroglioma, ependymoma, oligoastrocytoma, and a few rare histologies. In medical practice, gliomas account for nearly 25% of all primary brain tumors [1]. According to the most recent WHO guideline for the classification of central nervous system (CNS) tumors, molecular, genetic, and epigenetic markers are used [2]; however, it is also possible to subgroup gliomas into a grading scale: low-grade gliomas (LGG, WHO grades 1–2) and high-grade gliomas (HGG, WHO grades 3–4).

The most common glioma is glioblastoma (GBM, WHO grade 4 astrocytoma), accounting for 14.3% of all primary brain and central nervous system neoplasms and 49.1% of all malignant brain tumors [1], with a poor prognosis for patients: the 5-year survival rate is at 6.8% [1] and it has a median survival of 14.6 months after diagnosis ^[3][4][3,4]. The patient’s prognosis remains weak despite conventional standard-of-care therapy that includes maximal tumor resection followed by radiotherapy and adjuvant temozolomide (TMZ) administration [3].

To better understand the molecular basis of GBM, extensive genomic [5] and transcriptomic [6] analyses were performed. In 2010, these allowed the classification of glioblastomas into subtypes according to their genetic expression profiles: classical (EGFR overexpression, CDKN2A deletion, and a lack of TP53 mutations), mesenchymal (altered NF1, PTEN mutations, a high activity of MET and CD44, and MERTK genes), proneural (altered PDGFRA, mutated TP53, point mutations in IDH1 and the increased expression of OLIG2), and neural (expression of neural markers such as NEFL, GABRA1, SYT1, and SLC12A5) [5]. Median survival rates of 11.5, 14.7 and 17.0 months for the mesenchymal, classical and proneural subtypes were determined.

2. Conventional Standard of Care for Glioma Therapy

For glioblastoma patients, surgical resection of the tumor to the maximal feasible extent followed by focal tumor radiotherapy and concomitant temozolomide chemotherapy combined with additional irradiation therapy should be taken as the standard treatment (Stupp treatment) [3]. However, it is known that all glioblastomas will eventually relapse or progress ^[7][15]. For low-grade gliomas, the optimal treatment involves immediate surgical resection of the tumor to avoid further cancer progression and allow precise molecular characterization of the tumor ^[8][20], which is indeed crucial for the development of a treatment plan for any glioma ^[2][9][10][2,10,11]. For high-risk cohorts of low-grade-glioma-diagnosed patients (age > 40 years; patients who do not undergo gross total resection surgery ^[11][21]), postoperative care consists of 50–54 Gy radiotherapy followed by adjuvant therapy with DNA-alkylating or cytostatic agents. Lomustine, a DNA-alkylating agent, is usually preferred due to its tolerable toxicity and blood-brain barrier (BBB) permeation properties ^[12][22]. Taking the disease progression from low- to high-grades and the recurrent nature of gliomas ^[7][15], treatment options require constant attention and rapid development. There are several FDA-approved drugs and one medical device for glioma management aside from temozolomide (TMZ) ^[13][26]: lomustine ^[14][27], intravenous carmustine ^[15][28], carmustine wafer implants ^[16][29], bevacizumab ^[17][19], and tumor treatment fields ^[18][30]. These drugs and devices are mainly approved for the management of recurrent high-grade gliomas with only TMZ, carmustine wafer implants, and tumor treatment fields suitable for application in de novo diagnoses. Except for bevacizumab, FDA-approved drugs for gliomas represent the class of DNA-alkylating agents—their effect is not targeted on tumor cells; therefore, their application is linked with systemic adverse effects. Vascular endothelial growth factor (VEGF) was shown to be one of the key regulators of malignant angiogenesis in glioblastoma cells, while its inhibition with the help of antisense oligonucleotides correlated positively with reduced tumor growth and, indeed, diminished vasculature formation ^[19][31]. This discovery led to the FDA approval of bevacizumab, a monoclonal antibody that targets VEGF-A on the surface of endothelial cells and blocks its interaction with its receptors VEGFR1 and VEGFR2, becoming the first FDA-approved targeting therapy for use in recurrent GBM. Although bevacizumab demonstrated enhanced progression-free survival (PFS), it did not provide a benefit to overall survival (OS). This suggests that solely targeting the VEGFR axis is inadequate for suppressing tumor progression ^[20][21][32,33].

3. Delivery of Therapeutic Drugs to the Brain and Associated Difficulties

3.1. Blood–Brain Barrier as an Obstacle for Drug Delivery

The blood–brain barrier poses a significant challenge in treating conditions within the central nervous system (CNS) and brain tumors are not an exception. The components of BBB include endothelial cells with tight junctions (specific for CNS), smooth muscle cells, microglial cells, astrocytes and pericytes ^[22][34]. Tight junctions in endothelial cells create a nearly impenetrable barrier, restricting the passage of small and lipid-soluble molecules through the paracellular route. Relevant to drug delivery, the BBB allows the free passage of hydrophilic molecules with a molecular weight of up to 150 Da and hydrophobic molecules with a molecular weight of 400–600 Da ^[23][37], while the penetration of large or hydrophilic drugs is limited. In order to travel through the BBB, the drug may utilize receptor-mediated transcytosis mechanism (e.g., with transferrin and insulin receptors ^[24][38]). P-glycoprotein-1 (P-gp), which can pump out more than 60% of the marketed drugs, makes it more difficult to penetrate the BBB ^[25][39]. The transcellular route of drug transport through the BBB is favorable for molecules with a low molecular weight (<500 Da) and high lipophilicity ^[26][49]. The relationship between the polar surface area and brain permeability is proved to be inverse—drug candidates that have >80 Å of polar surface and a tendency to form multiple H-bonds require higher free energy to penetrate lipophilic membranes of the BBB cells ^[26][27][28][49,50,51]. However, the predominant usage of less polar or highly lipophilic compounds also has its drawbacks as lipophilic molecules would aggregate bound to plasma proteins, effectively decreasing the amount of available active compound ^[29][52]. For the treatment of high-grade gliomas ^[30][54], only a small number of cytotoxic drugs are being used, such as TMZ, which allows about 20% penetration through the BBB at the administered systemic dose ^[31][55]. Other well-studied drugs such as paclitaxel (PTX), doxorubicin, and cisplatin, are not included in the standard-of-care treatment for GBM due to their poor BBB permeability, even though preclinical in vitro studies have shown an inspiring decrease in the viability of cancer cells using these drugs ^[32][33][34][56,57,58]. 

3.2. Targeted Drug Delivery to the Brain

3.2.1. Passive Targeting

Passive targeting is a major strategy in drug development to achieve therapeutic effects while using systemic administration, particularly for drugs that need to reach the brain. This strategy takes advantage of the abnormal vessel formation in tumors, resulting in enlarged pore sizes in the vasculature endothelium at the tumor site. As a result, compounds that usually cannot cross the blood–brain barrier can be delivered to the tumor site with a higher probability ^[35][36][77,78]. This condition is known as the Enhanced Permeation and Retention (EPR) effect. Pore cutoff sizes differ drastically in healthy (4–25 nm) and brain tumor blood vessels (380–780 nm). Taking that into account, as well as the fact that lymphatic drainage is also impaired in brain tumors, drug delivery nanosystems that are designed to utilize the EPR effect for reaching the tumor site may be up to several hundred nanometers in diameter and could stay in the targeted area for longer periods of time ^[37][79]. The size of the nanostructures is a variable parameter in drug design and could be changed to increase the tumor permeation rate and decrease systemic toxicity ^[38][80]. Dual-modality nanosystems, designed for both the diagnostics and therapeutics of gliomas, such as QSC-LP liposomes loaded with quantum dots, supramagnetic iron oxide and cytotoxic peptide cliengitide, have shown promising results in vivo by significantly prolonging overall survival and diminishing tumor size in a xenograft mouse model ^[39][81].

3.2.2. Mechanical Targeting (Local Delivery)

Local application of the therapeutic agents to the tumor site, which scholars will refer to as ‘mechanical’ targeting, could be achieved via intracerebroventricular injection, intratumoral injection, convection-enhanced delivery, and adding anti-cancer compounds intraoperatively into the post-lesion cavity, such as carmustine wafers ^[17][40][19,70]. In this case, as there is no need to pass the BBB and no application of drugs occurs in the circulation, therefore, the systemic toxic effects are significantly lower than in the case with intravenous application, as well as there being increased therapeutic efficacy of the active compounds as they are delivered locally to the site of the tumor ^{[41][42][43][44]}[82,83,84,85].

3.2.3. Active Targeting

Active targeting is a method of drug delivery that utilizes ligands (monoclonal antibodies and their derivatives, peptides, etc.) for tumor-specific recognition. Upon recognition of tumor-specific biomarkers which are represented by overexpressed or mutated surface proteins, the drug promotes its cytotoxic effect either directly or indirectly based on a type of drug system and its corresponding specifications. As an example, contrary to passive drug delivery through the EPR effect described above, targeting BBB-overexpressed glioma markers with corresponding ligands showed successful delivery to the tumor site in a phase I trial of the peptide–drug conjugate GRN1005 ^[45][67], as well as nanoparticles in preclinical in vitro ^[46][61] and in vivo studies ^{[47][48][49][50]}[62,65,88,89].

4. Drug Delivery Systems for Active Targeting of Brain Tumors

4.1. General Scheme for Drug Delivery Systems That Use Active Targeting

Targeted drug delivery systems that utilize active targeting strategy are generally composed of the following parts (Figure 1):

• Target—this is a moiety that is specific to the cell of interest—in this case, a tumor cell. It is required to be expressed or significantly overrepresented in the tumor over healthy cells; high-level expression is advantageous, although not necessary ^[51][90]. It could also represent a molecule within an altered biochemical pathway that is specific to a tumor cell.
• Recognition moiety—a part of the drug delivery system that recognizes the target specifically and allows address delivery to a specified tumor site with minimal toxicity and off-target effects.
• Linker—an engineered connective unit that ties the recognition part of the drug delivery system to the payload.
• Payload—a cytotoxic agent that acts upon delivery to the tumor cells either directly or indirectly through the mediation of the host’s immune response.

4.2. Targets

Targeted glioma therapies commonly aim for the recognition/inhibition of the tumor-overexpressed variants of the receptor tyrosine kinase (RTK) family of transmembrane proteins that are responsible for cell growth and proliferation signaling. RTKs represent the most abundant and clinically relevant group of surface targets for GBM drug development [6]. In glioma initiation, several subclasses of RTKs are prone to alterations such as amplification and point mutations ^[6][52][6,91]. From a structural point of view, RTKs contain an extracellular domain required for ligand binding, a transmembrane domain, and an intracellular domain with tyrosine residues required for receptor activation upon phosphorylation ^[53][93]. The binding of receptor-specific ligands triggers dimerization of the RTK and subsequent autophosphorylation at intracellular tyrosine residues (activation sites). Activated residues act as landing pads for small G-proteins whose recruitment is necessary for downstream signaling through the RAF/MEK/ERK and PI3K/AKT pathways, which control tumor formation, cell survival, proliferation, and invasion ^[52][91]. In the context of GBM, alteration in the RTK/PI3K/AKT pathway, which is responsible for constitutive RTK signaling, is considered to be the most elevated core signaling pathway ^[54][92]. Alterations (mutations/deletions) in oncosuppressor genes such as TP53, NF1, and PTEN additionally promote RTK activity, which accelerates malignant progression and plays an important role in tumor initiation ^[54][55][92,94].

4.3. Recognition Moiety

Recognition moieties in the development of targeted drug delivery systems are represented by several classes of molecules. As mentioned before, the recognition particle of the drug should be tumor-specific, with minimal off-target effects as well as no reactivity with healthy tissues ^[56][98]. Recognition moieties utilized in drug delivery systems include

• Monoclonal antibodies and their derivatives (such as antibody fragments (Fabs), single chain variable fragments (scFvs), and bispecific antibodies) ^[57][99]. The most frequently used format for the development of antibody–drug conjugates is IgG1 immunoglobulin, as it is both easy to manufacture as well as the fact that it shows a strong cytotoxic effect ^[58][100].
• Peptides as recognition moieties in the targeted drugs are presented as either cell-targeting or cell-penetrating units ^[59][108]. While the first group’s mechanism of action is similar to that of antibody–drug conjugates, the latter’s is not yet fully understood ^[60][61][95,96].
• Small molecules as inhibitors of intrinsic pathways are often used as therapeutic agents in glioma treatment ^[62][63][109,110], as the dysregulation of such pathways (e.g., signaling pathways involving RTKs, PI3K, p53, etc.) has a major role in GBM development [6].
• Aptamers could also be used as a recognition unit of a targeted drug delivery system. Aptamers are short nucleic acid sequences that could be selected based on their affinity towards tumor cells via the systematic evolution of ligands by an exponential enrichment (SELEX) process ^[64][111]. I

4.4. Linker

For antibody- and peptide-conjugated drugs, linkers could be classified into cleavable and non-cleavable subgroups. Cleavable linkers in conjugated drugs are developed for the cytotoxic drugs to be released upon cellular uptake of the drug conjugate, which reduces off-target drug release with associated unwanted toxicity in healthy tissues ^[65][66][114,115]. In general, there are three main triggers for linker cleavage and drug release for conjugated drugs: acid-cleavable linkers undergo degradation and drug release due to the usage of the difference between pH levels in plasma (approx. pH 7.4) and endosomal/lysosomal compartments of the tumor cells ^[67][68][116,117] that could be as low as pH 4.5–4.7 ^[69][118]; enzyme-cleavable linkers are being cleaved upon internalization in the endosomal or lysosomal compartment of the cell based on the linker structure, with help of intracellular cathepsin B and beta-glucuronidase enzymes ^{[70][71][72][73]}[119,120,121,122], and redox-sensitive disulphide linkers undergo reductive cleavage based on the difference in the number of reducing equivalents in the form of glutathione between the plasma (2–20 μM/ L) and cytoplasm (0.5–10 mM) ^[74][123]. Non-cleavable linkers in the conjugated drugs require retention of the drug activity upon degradation of the recognition moiety within the endosomal compartment of the cells while the linker or its part stays intact. The advantage of using such types of drugs is associated with a decreased chance of off-target effects, as the drug is more likely to transform to its active form only upon internalization to the cells of interest ^[75][76][77][124,125,126].

4.5. Payload (Cytotoxic Agent)

Cytotoxic payloads that are commonly used in the arsenal of targeted therapeutics include

• Conventional chemotherapeutics such as doxorubicin ^[78][79][133,134] or paclitaxel ^[45][80][81][67,135,136];
• Microtubule assembly inhibitors: auristatins that inhibit tubulin polymerase and promote cell cycle arrest ^[82][137] such as monomethyl auristatin E (MMAE) ^[83][138] and monomethyl auristatin F (MMAF) ^[84][139], or maytansines and their derivatives that bind to tubulin and therefore interfere with the assembly of microtubules ^[85][140]. Common examples include mertansine (DM1) ^[86][87][88][141,142,143] or ravtansine (DM4) ^[89][90][144,145] as parts of the cytotoxic agents in targeted therapeutic drugs.
• Bacterial toxins—toxic compounds produced by Pseudomonas aeruginosa (Pseudomonas Exotoxin A) and Corynebacterium diphtheria (Diphtheria toxin) are the most commonly used toxins of bacterial origin that are utilized for cancer therapies ^[91][146]. Both toxins act upon binding and irreversibly modify the eukaryotic EF2 elongation factor, which results in impaired protein synthesis and cell death ^[92][93][147,148].
• Radioligands fused with peptides or antibodies with tumor specificity are used as a cytotoxic moiety for the targeted delivery of drugs to the tumor sites ^{[94][95][96][97]}[71,72,150,151], which allows tumor-specific distribution of the radioactivity as compared to standard beam irradiation in the standard-of-care Stupp protocol [3].
• Photodynamic therapy (PDT) is a therapeutic approach that requires the incorporation of non-toxic and inactive photosensitizer molecules into the cells of interest with subsequent light irradiation, in which the therapeutic molecules become activated. Upon irradiation, photosensitizer molecules transfer light energy and excitate molecular oxygen present in the surrounding tissues to a triplet or singlet state. In the triplet state, oxygen is capable of the generation of reactive oxygen species that react with molecules containing double bonds, leading to their damage and subsequent cell death ^[98][152]. In glioma studies and treatment, PDT is implemented via placing fiberoptics at the site of the tumor in order to irradiate cells within the brain that have obtained photosensitizer molecules ^[99][100][153,154].
• Oncolytic virotherapy is a novel approach for the targeted therapy of neoplasms that utilizes viruses for the elimination of tumor cells. While the exact mechanisms and the nature of tropism for such viruses might still be unknown, the main theories involve either direct killing of the targeted cells or indirect modulation of the host’s immune system response in order to mediate immunogenic cytotoxicity ^[101][102][158,159].

5. Strategies in Clinical Approaches towards the Treatment of GBM

5.1. Targeting Receptor Tyrosine Kinases

5.1.1. EGFR

Epidermal growth factor receptor is a member of the ErbB family of receptors ^[103][172]. Downstream signaling pathways for EGFR regulate DNA synthesis and since inclusion criteria for clinical trials differ between the studies, it is impossible to uniformly assess the success or failure of the tested compounds (e.g., survival parameters obtained in de novo diagnosed patient cohorts are different from survival data obtained in the studies for patients who have failed in trials for the other lines of treatment). Glioblastoma somatic genomic landscaping revealed that 57% of GBM cells harbor EGFR genetic alterations [6]. EGFR amplification and overexpression were identified in 40 and 60% of primary glioblastomas, respectively, often leading to aggressive GBM tumors resistant to treatment, and enhanced proliferation, invasion, and survival ^{[104][105][106][107]}[174,175,176,177]. Point mutations and deletions are also observed among GBM patients with around 20% of diagnosed cases ^[54][92] harboring a deletion of exons 2–7, resulting in a truncated mutant variant III (EGFRvIII) ^[108][178]. Depatuximab–mafodotin, which contains an internalizing EGFR/EGFRvIII-specific antibody linked to the cytotoxic agent monomethyl auristatin F via a non-cleavable maleimidocaproyl (MCC) linker, did not show increased survival rates when tested in combination with conventional radiotherapy and temozolomide in NCT02573324 ^[109][182] or as a monotherapy versus conventional treatment in NCT02590263 and NCT02343406 ^[110][111][183,184], as well as it not having an impact on the quality of life of the patients with recurrent glioblastoma ^[112][185]; however, preclinical evaluation in vivo showed total tumor regression in mice bearing U87MGde2–7 model tumors, as well as it showing significant tumor growth reduction in patient-derived xenograft models ^[113][186]. AMG 595, composed of the maytansinoid DM1 toxin attached to a highly selective anti-EGFRvIII antibody via a non-cleavable linker, showed promising cytotoxic effects in vitro and in vivo ^[86][141] and was tested in a phase I trial (NCT01475006) on a group of 32 patients with recurrent glioma, in which 2 patients showed a partial response, 15 showed stable disease and 15 showed progressive disease outcomes ^[114][179]. The D2C7-IT dual-specific immunotoxin, comprising an EGFR wild-type and mutant-specific (EGFRvIII) monoclonal antibody (mAb) fragment linked to a genetically engineered form of the Pseudomonas exotoxin A via PCR, is currently being investigated in two active clinical trials, NCT04547777 and NCT05734560, with predicted completion in 2025 and 2028, respectively. Preliminary results summarized in a study abstract by ^[115][187] show that this drug might be promising for patients diagnosed with malignant glioma. The first proposed antibody for EGFR targeting was the chimeric human–murine monoclonal antibody Cetuximab. Unfortunately, phase II clinical trials for this drug did not show therapeutic benefit for patients with relapsed GBM, both as monotherapy, showing median PFS of 1.9 months and OS of 5.06 months ^[116][69], and in combination with bevacizumab (a monoclonal antibody targeting neoplastic angiogenesis) and irinotecan (DNA synthesis blocker), showing median PFS of 16 weeks and OS of 30 weeks ^[117][190]. The first fully human monoclonal antibody targeting EGFR, panitumumab, did not show promising results in a phase II clinical trial combined with the administration of irinotecan; the study was terminated due to not showing efficacy (NCT01017653). Nimotuzumab, a humanized anti-EGFR antibody, was tested in a phase III clinical trial (NCT00753246) comparing its efficacy in a treatment combined with a standard Stupp protocol (tumor resection followed by radiotherapy and temozolomide), in which no significant prolongation of the OS and PFS was observed for patients with both recurrent and primary glioblastomas ^[118][201].

5.1.2. PDGFR

Platelet-derived growth factor receptors are a family of six subunit homo- and heterodimers that form tyrosine kinase receptors, whose physiological function involves the regulation of hematopoiesis, embryogenesis, and glial cell development, and the protection of neuronal tissue ^[119][207]. PDGFR alteration and genetic mutations play a driving role in GBM cell proliferation and malignancy, including the dedifferentiation of glial cells and epithelial-to-mesenchymal transition, as well as the effects of intratumoral vessel formation and immunosuppression ^[120][121][208,209]. In GBM patients, amplification and genetic alterations in PDGFR are found in 13% of diagnosed cases, overexpression of this RTK is related to poor prognosis [6]. Activation of PDGFR signaling in glioma cells induces malignant vascularization, cell survival, and growth ^[122][123][112,210], and dysregulation of normal PDGFR activity in tumors leads to oncotransformation of healthy glia and GBM growth through autocrine signal transduction ^[124][125][211,212]. MEDI-575 is a PDGFRα specific monoclonal antibody that was tested in a Phase II study (NCT01268566) and is a monotherapeutic agent. It showed disappointing clinical outcomes as the median PFS for patients was only at the 1.4 months level with median overall survival at 9.7 months ^[126][213]. Olaratumab—a PDGFRα specific monoclonal antibody—was tested as a monotherapeutic agent in a phase II NCT00895180 clinical trial and showed median overall survival at 34.3 weeks level, with 7.5% of patients from the recruitment cohort reaching 6 months progression-free survival endpoint and only 2.5% of the olaratumab treated cohort achieving response from the treatment.

5.1.3. VEGF/VEGFR

Neoplastic vessel formation is one of the hallmarks of malignant tumor growth and progression, with VEGF receptors and ligands signaling pathways being a potent driver of angiogenesis ^[127][214]. Extensive vascularization as a result of angiogenesis is one of the tumor hallmarks, indeed present in glioblastoma with VEGF being detected in 64% of diagnosed GBM cases ^[128][167] and up to 17% of diagnosed cases showing amplification of the VEGFR2 gene ^[129][168]. Several reports outline that VEGF/VEGFR axis members are strongly upregulated in glioblastoma with a direct correlation to the level of malignancy of the tumor and therefore VEGF/VEGFR amplification could serve as a marker of a poor prognosis ^[130][131][215,216]. VEGF/VEGFR signaling in the context of tumor plays a role in tumor growth ^[132][217] and proliferation of the cancer stem cells ^[133][218] as well as it is responsible for tumor immunoevasive microenvironment ^[134][219]. Since 2009, when bevacizumab (VEGF-A specific monoclonal antibody) was approved by the FDA the for treatment of recurrent glioblastoma ^[135][223], it has become a treatment of reference for combinatorial clinical trials within recurrent GBM. Two major lines of clinical investigation exist in targeting VEGF/VEGFR axis, similar to PDGFR treatment strategies—usage of monoclonal antibodies and small molecule inhibitors. The combination of bevacizumab with irinotecan (DNA topoisomerase-I inhibitor) showed a promising trend in increased progression-free survival for the patients ^[135][136][223,228] and was a basis for combinatorial trials of bevacizumab with standard-of-care therapy, radiotherapy, and conventional chemotherapeutics (outlined in the table below). Surprisingly, in the majority of the studies that involve bevacizumab as a monotherapy or a part of a combinatorial treatment, only progression-free survival is increased significantly while no therapeutic benefit is observed for the overall survival of the patients ^{[20][137][138][139][140]}[32,220,222,225,226]. CT-322, an investigational peptide drug that is comprised of a modified extracellular domain of fibronectin 10th type 3 with a specificity towards VEGFR-2 was tested in a phase II clinical trial NCT00562419 as a monotherapy or in combination with irinotecan. CT-322 acts as a high-affinity blocker of VEGFR-2 preventing its binding with VEGF ligands, most importantly, VEGF-A, therefore blocking VEGF-A induced dimerization of the receptor and subsequent MAPK signaling ^[141][229]. While showing adequate tolerability and side effects profile, this compound did not reach the trial’s prespecified efficacy measures and was terminated ^[142][227].

5.1.4. c-MET Pathway

Another example of a receptor with tyrosine kinase activity that is altered in GBM cells is c-MET. According to glioblastoma molecular landscaping, 1.6–13.1% of diagnosed cases harbor c-MET overexpression, with gains in MET found in 47% of primary and 44% of recurrent GBM cases ^[6][143][6,169] and up to 4% of cases present amplification of MET ^[54][144][92,170]. Since it was shown that inhibition of angiogenesis through blockade of the VEGF/VEGFR axis stimulates c-MET activation and leads to an invasive form of glioblastoma ^[145][235], it is important to target the c-MET pathway to effectively treat tumors. In clinical trials, c-MET pathway is targeted with monoclonal antibodies with specificity towards c-MET receptor or its ligand HGF ^[146][147][13,171].  Onartuzumab, a monoclonal antibody targeting c-MET, was investigated in a clinical study NCT01632228 and showed a trend towards reducing tumor growth, however, no significant increase in median overall survival and progression-free survival were observed in patients undergoing combinatorial therapy with onartuzumab and bevacizumab as compared to bevacizumab plus placebo ^[148][236]. Rilotumumab, an anti-HGF monoclonal antibody, was tested in a NCT01113398 clinical study in combination with bevacizumab pretreatment. Unfortunately, this investigational drug did not show a single-agent antitumor activity in the cohort of patients enrolled for the study ^[149][237].

5.1.5. FGFR

FGFR family of receptor tyrosine kinases is comprised of four transmembrane receptors named FGFR1–4 ^[150][238]. Upon activation of FGFRs, downstream signaling includes the ERK/MAPK pathway, which in turn is responsible for the regulation of cell survival, proliferation, angiogenesis, development, and differentiation ^{[151][152][153][154]}[239,240,241,242]. Even though only 3.2% of glioblastoma patients harbor amplification or point mutation in FGFR genes [6], it was shown that GBM cells could evade EGFR and MET inhibition through FGFR-mediated bypass, therefore pointing FGFR inhibitors as possible potentiators of EGFR or MET-targeted therapies ^[155][243]. Currently, the research in the field of targeted FGFR inhibitors is mostly preclinical with limited information available for clinical trials in the context of GBM. One clinical trial evaluated the effect of pan-FGFR inhibitor infigratinib in patients with recurrent glioblastoma (NCT01975701) after this drug showed potent inhibition of FGFRs and cell growth on a panel of cancer cell lines in vitro ^[156][244], resulting in progression-free survival of 1.7 months and median overall survival of 6.7 months, Unfortunately, infigratinib as a therapeutic compound was outlicenced since then and no further trials and investigations were performed ^[147][171].

5.1.6. Multikinase Inhibitors

Imatinib is the first multikinase inhibitor whose targets are PDGFRα/β, BCR-Abl, and c-Kit. The initial trial of this drug in combination with hydroxyurea did not show an increase in survival for patients ^[157][248]; however, it was tested in trials after that, returning unsatisfactory or clinically insignificant results both in combination with hydroxyurea ^[158][249] and as monotherapy ^[159][250]. Dasatinib, an inhibitor of PDGFRβ, EPHA2, BCR-Abl, c-Kit, and SRC, was tested in clinical trials on patients with recurrent glioblastoma as a monotherapeutic agent and in combination with chemotherapy, bevacizumab or the conventional standard-of-care Stupp protocol. The monotherapeutic application of dasatinib failed in a clinical trial NCT00423735 as the study did not pass the efficacy preset and was terminated ^[160][252]. The combination of dasatinib with lomustine proved to be ineffective due to the efficacy endpoints not reaching the measures of historical controls ^[161][253]. Sunitinib, an inhibitor of PDGFRα/β, c-Kit, VEGFR1/2/3, FLT3, and RET, also showed upsetting results in clinical trials with no significant antitumor activity for patients with recurrent glioblastomas ^[162][163][256,258], while in one study a sub-cohort of patients with high c-Kit expression undergoing monotherapy with sunitinib showed median PFS of 16.0 months and median OS of 46.9 months, with the rest of the tested cohorts showing PFS and OS endpoints at 2.2 and 9.4 months, respectively ^[164][257]. This observation might prove valuable that treatment with sunitinib needs to be fine-tuned to a specific cohort of glioma patients. Ponatinib, a multikinase inhibitor that targets BCR-Abl, PDGFRα, VEGFR2, FGFR1, and Src ^[165][260] but also RET, c-Kit, and FLT1, is under investigation in an NCT02478164 clinical trial as a monotherapy for patients with bevacizumab-refractory GBM. Preliminary results show a median OS of 98 days and PFS of 28 days, with minimal activity of the investigational drug in the selected group of patients ^[166][259].

5.2. Targeting Immune Checkpoints

5.2.1. PD-1/PD-L1

Clinical trials that have investigated the effect of blockade of PD-1 on host T-lymphocytes with the PD-L1 ligand through the use of monoclonal antibodies in glioblastoma were performed for a variety of developed investigational drugs. Pembrolizumab, an anti-PD-1 monoclonal antibody, was tested in the NCT02337686 phase II trial for patients with recurrent glioblastoma. The investigation tested the effect of the adjuvant or neoadjuvant application of a checkpoint inhibitor for patients undergoing tumor resection. When compared between adjuvant and neoadjuvant cohorts, the addition of pembrolizumab significantly prolongs overall survival (13.7 vs. 7.5 months for neoadjuvant vs. adjuvant therapy) as well as a progression-free survival rate (3.3 vs. 2.4 months for neoadjuvant vs. adjuvant application). In the monotherapy cohort for patients treated with pembrolizumab (NCT02054806), a durable anti-tumor trend was observed with a median overall survival rate of 13.1 months and median PFS rate of 2.8 months, suggesting that single-agent application of this drug might be beneficial for a certain subpopulation of GBM patients. However, as the authors suggest, the increase in OS might be unclear as the baseline characteristic of patient records used as a historical cohort in that study might be non-uniformed and, therefore, this result should be taken with caution ^[167][267]. Another example of an anti-PD-1 monoclonal antibody investigated in the clinical setting for GBM patients is nivolumab. It was tested in patients with recurrent glioblastoma in a phase III NCT02017717 ^[168][268] trial as a monotherapy vs. bevacizumab monotherapy, as well as in two phases III trials in newly diagnosed glioblastoma patients with methylated (^[169][269], NCT02667587) and unmethylated MGMT promoter (^[170][270], NCT02617589). MGMT-related studies compared combinatorial treatment with nivolumab and concomitant radiotherapy with temozolomide compared to RT/TMZ alone.

5.2.2. CTLA-4

Cytotoxic T-lymphocyte associated protein 4 (CTLA-4) is an immune checkpoint receptor that acts upon binding with ligands CD80 and CD86 present on antigen-presenting cells leading to inhibition of T-lymphocyte stimulatory signaling ^[171][272]. CTLA-4 expressing T-lymphocytes are usually present in the lymph nodes unlike PD-1-expressing cells in the tumor microenvironment ^[172][273]. Tremelimumab, an anti-CTLA-4 monoclonal antibody, was recently tested in a phase II NCT02794883 clinical trial as a monotherapeutic agent or combined with anti-PD-L1 monoclonal antibody durvalumab. A combination of these drugs showed contradictory results as compared with durvalumab alone; however, the combination of durvalumab with tremelimumab might be beneficial for increasing the PFS rate of the patients.

5.2.3. TIM-3

TIM-3 is an immune checkpoint that is widely expressed in glioblastoma and is an important regulator of the inflammatory response related to anti-PD-1 inhibition by therapeutic agents ^[173][174][277,278]. Thus, the combined anti-TIM3 inhibitor and other immunotherapies are getting promising. In preclinical investigation, the combination of anti-TIM3 therapy, anti-PD-1 therapy, and radiotherapy in animal models has reported promising efficacy—Kim et al. showed, in a murine orthotopic GL261 glioma model, that the addition of anti-TIM-3 therapy to stereotactic radiosurgery (SRS), as well as triple therapy with anti-TIM3, anti-PD-1, and stereotactic radiosurgery, significantly increases overall survival and therefore needs to be studied in detail for clinical implementation.

5.2.4. LAG-3

LAG-3 is a marker of active immune cells ^[175][176][280,281] with a similar mode of immune evasion in cancers such as PD-1 ^[177][282]. In tumors, LAG-3 is usually found within T cells that have lost their functions and, therefore, they could be a potential target for immune checkpoint inhibitors in oncology treatment ^[178][179][283,284]. In GBM, the LAG-3 expression profile correlates with the expression of the CD8+ T-cells CD8A ^[180][285]; therefore, targeting LAG-3 in glioblastomas with high levels of CD8+ cells in the tumor environment might be a beneficial clinical strategy. Currently, two phases I studies for the anti-LAG-3 monoclonal antibody are in the process of investigation (NCT02658981 and NCT03493932).

5.3. Targeting Extracellular Matrix (ECM) Components in Glioma Microenvironment

The extracellular matrix in brain tumors is a promising target for therapeutic treatment, as components of the brain ECM in pathologic conditions show tumor-specific behavior and ECM-dependent metabolic activity ^[181][286], which makes it possible to differentiate between targeting healthy tissue and the tumor environment in therapeutic development. The extracellular matrix in the brain tumor environment overexpresses several markers that are possible targets for glioma-specific therapies, such as tenascin-C, fibulin-3, fibronectin, and hyaluronan; as well as this it shows increased activity of metalloproteinase MMP9 ^[182][287]. Preclinical trials of the tenascin-C A1-targeting scFv IgE based antibody showed good accumulation in a tumor site in xenograft mice bearing U87 glioma cells, with comparable results to fibronectin-targeting antibody that is undergoing clinical trials at the moment ^[183][288]. Fibronectin is a promising target for delivering drugs to the extracellular compartment of solid tumors as it is a good marker of neovascularization and therefore could be targeted for eliminating the spread of the tumor tissue ^[184][290]. Preliminary in vivo studies showed that a radiolabeled scFv fragment of the L19 antibody that targets an ED-B domain of fibronectin significantly prolongs survival in mice with intracranially placed C6 glioma cells (22 vs. 16 days in mice bearing tumors of which the volume exceeded 150 mm^{3 [185]}[291]), as well as it showing preferential binding in the neovasculature region of the tumor, which is increased when antivasculature drugs are coadministered [290

6. Conclusions

Gliomas, as representatives of CNS tumors, are characterized by a vast variety of altered markers within tumor cells and the surrounding microenvironment [2], with glioblastoma diagnosis holding poor survival rates and prognosis ^[1][3][4][1,3,4]. Many efforts in preclinical and clinical research are being invested into developing anti-glioma and GBM drugs; however, in many cases, these attempts are proving to be ineffective due to factors such as tumor location in the brain that is protected from main circulation due to presence of the BBB ^[186][293], as well as inter- and intratumoral heterogeneity ^{[5][6][187][188][189]}[5,6,294,295,296], often resulting in acquired therapy resistance. Nonetheless, accumulating data on unsuccessful trials plays a great role in enlarging the understanding of gliomas from a fundamental science point of view and helps to elucidate the mechanisms of drug resistance for future improvements in therapeutic strategies.