Biomarkers and therapeutic advances in glioblastoma multiforme

3.1 O⁶-methylguanine-DNA methyltransferase

Located in chromosomal position 10q26, the MGMT gene encodes proteins which would be consumed in DNA repair whereby alkyl groups are removed from guanine at its O6.14 MGMT expression is heavily modulated by various transcription factors, including nuclear factor kappa B and specificity protein 1, which would activate MGMT promoter to induce expression of more MGMT.15 The methylation of this gene has been shown in increasing the efficacy of an alkylating agent, TMZ, upon its use with chemoradiotherapy. As highlighted by Hegi and colleagues, MGMT promoter methylation has been known to provide better outcome in patients receiving TMZ, whereby median overall survival of cases with methylation being present is 18.2 months as opposed to the significantly lower 12.2 months of cases without methylation.16 A study reported survival of unmethylated and methylated groups of MGMT-positive GBM patients to show significant differences in overall survival upon TMZ-radiotherapy treatment, being 20 and 36 months, respectively.17 Other studies such as one conducted by Kim et al. showed similar survival results despite utilizing different DNA analysis techniques.18

Epidemiologically, the MGMT promoter methylation occurs more substantially (75%) in secondary GBM as it has been shown to have a strong correlation with TP53 mutation (92%), which is abundant in secondary GBM and only 36% of primary GBM cases.19 Despite the promoter methylation, some patients might exhibit staggeringly ample expression of MGMT mRNA, which led researchers to believe that although the MGMT promoter methylation serves a valuable predictive and prognostic biomarker value, individual differences still exist. The overall combination of low promoter methylation and high mRNA expression of MGMT would eventually lead to the development of TMZ resistance. Studies involving MGMT silencing via methods such as microRNAs (miRNAs) introduction have been proven to produce favorable prognosis through bypassing TMZ resistance, which is a bane in GBM cases.20, 21

There is currently no standard method of determination of MGMT promoter methylation, but it is a general rule that false negatives of the MGMT status could be due to DNA analysis from necrotic GBM tissue samples.14, 22 The most common method in observing the methylation status of MGMT in GBM patients is via nested polymerase chain reaction (PCR) or combinatorial PCR with MS technology, SYBR Green22 or even pyrosequencing in GBM patients.18

3.2 Epidermal growth factor receptor

The amplification of EGFR and the genetic rearrangement of EGFR (EGFRvIII) are common hallmarks of GBM (40–50%), especially in the classical subtype and primary GBM.23 EGFR is encoded by a gene of the same name and codes for a tyrosine kinase receptor specific to certain growth factors. The mutation of EGFR due to histone modifications on its gene enhancer at chromosome 7p12 leads to the formation of EGFRvIII.24 This mutation truncates exons 2 and 7 which leads to the absence of extracellular ligand binding site. These results were obtained from studies of EGFR as biomarkers are traditionally performed using transfected cell lines, and conventional cell lines tend to lose the EGFRvIII mutation or have their EGFR expression dampened to normal levels.25 EGFR mutation and amplification have been classified as prognostic biomarkers as they are abundant in GBM samples.

Due to the proliferative nature of cancer which is mainly controlled by the presence of key growth factors and the abundance of their receptors, EGFR can activate pathways essential for GBM tumor cells to flourish, such as the receptor tyrosine kinase/Ras/phosphoinositide 3-kinase (RTK/RAS/PI3K) pathway.26 The activation of this pathway leads to the decreased integrity of G1 to S checkpoint in the cell cycle, enabling excessive proliferation. Often presenting itself in patients with amplified wild-type EGFR, patients with EGFRvIII mutation show significantly lower survival (0.8 years) as compared to patients without the mutation (1.4 years).24 The abundance of EGFRvIII alongside wild-type EGFR has been showcased in a study which concluded the ability of EGFR to cross-phosphorylate with EGFRvIII upon activation of a specific ligand.27, 28

Commonly present in the classical GBM subtype and primary GBM, the increased production of this genetic product would lead to a worse prognosis in most GBM cases. Despite this, some studies have yielded results which proved otherwise, giving conflicting results by concluding that the nonheterogenous of EGFR present in GBM cases may limit the translational effectiveness of drugs targeting EGFR, including immunotherapy and small molecule inhibitors,29, 30 and thus not conferring significant differences between GBM cases with and without EGFR mutations. One of the proposed mechanisms of tumorigenesis due to EGFR activation (EGFRvIII) is the remodeling of the landscape of GBM tumor cells through activation of pathways such as FOXG1 and SOX9, other than the commonly observed RTK/RAS/PI3K pathway.31

3.3 Platelet-derived growth factor alpha receptor

Much like EGFR, PDGFRA is a receptor for specific growth factors which if overexpressed would lead to abnormal and uncontrolled cellular growth. Gliomas, including GBM, present different PGFR ligand types and various types of receptors discovered, the first ones being PDGFRA and PDGFRB. Presenting alterations within its expression and abundance, PDGFRA serves as a prognostic biomarker of GBM, especially for the proneural subtype.28, 32 PDGFRA gain is almost always present with the presence of IDH mutations (93%) which is coincidentally present at a much higher rate in secondary GBM.33

Based on an mRNA in situ hybridization techniques, PDGFRA has been found to selectively localize its protein structures within the tumor cells, while PDGFRB was mostly found in the vasculature surrounding the tumor cells.34 PDGFR proteins have been known to express themselves as early as low-grade diffuse astrocytoma and are expressed at a highly enhanced rate in high-grade secondary GBM tumors. Based on this early presentation of PDGFRA as biomarker, the overall survival of PDGFRA-positive patients is also the longest, some reaching over 10 years with a median of 8.8 years in total samples studied.33 A study observed a marked reduction of cell proliferation in PDGFRA tumor cells after conductance of RNA interference,35 suggesting that its expression levels could be altered and is a potential drug target (e.g. crenolanib and imatinib).36

3.4 Isocitrate dehydrogenase

IDH is a protein enzyme encoded by the IDH genes in chromosome 2 whose primary function is to catalyze the oxidative decarboxylation process within the Krebs cycle. In contrast to the wild-type enzymatic function, IDH mutants attain an alternative responsibility to catalyze the production of 2-hydroxyglutarate (2-HG), which is an oncometabolite. 2-HG would eventually compete with alpha-ketoglutarate in activating enzymes which support DNA demethylation, yielding a hypermethylation in the tumor cells and eventual tumorigenesis.37 This application of mutant IDH expression has been shown to intercalate itself with methylated MGMT to give a significantly favorable prognosis for GBM patients undergoing TMZ and radiation therapy.38 This also correlates with another study done in vivo and in vitro, whereby Western blot analysis yielded overexpression of wild-type IDH which does not affect the cell cycle, but resulted in more chemotherapy resistance, forming a link between the two.39 Wang's study also presented possible mechanisms of IDH mutation, IDHR132H, which led to cell cycle arrest and reduced proliferation of GBM samples.

Posing a strong correlation with PDGFRA alterations aside from MGMT, IDH1 and IDH2 mutations are heterozygous and the distinguishing factor between the two mutations is the codons where the mutations are; codon R132 for IDH1 mutation and codon R172 for IDH2 mutation.37 IDH mutations are commonly found in the proneural subtype of GBM and 85% of secondary GBM, while only present in 7% of primary GBM.40, 41 Due to it being prominent in secondary GBM, it is thought that the IDH mutations have a significant role in gliomagenesis. Within a study of 1010 individuals with diffuse gliomas, 70.9% possess the IDH1 mutation, while only 3.1% possess IDH2 mutation.42 Detection of such biomarker is available via immunohistochemistry or even spectroscopy. The major limitation of utilizing this biomarker is determining the fate of IDH mutation in progression of diffuse gliomas. Aside from acting on its own, the mutated IDH1 would cause eventual mutations in other genes, including the ATM gene and mTOR gene, the latter being a kinase closely associated with the GBM pathogenesis.40

3.5 Loss of heterozygosity of chromosome 10

LOH is commonly occurring in malignant tumor cells, which mostly affects the tumor suppressor genes and subsequently leading to decreased protection of bodily systems toward tumorigenesis.43 The analysis of LOH in GBM patients is done using microsatellites and lastly, PCR to amplify the gene products.

Within GBM, many chromosomes are affected with LOH, namely 9p, 10, 17p, 19q and 22, while LOH 1p and LOH 19q are molecular predictors of oligodendrocyte neoplasms that provide no prognostic or predictive significance in GBM biomarkers.44 LOH 10q (more specifically LOH 10q23) is the common hallmark of all GBM, having presented in 70% of total GBM cases, although mostly in primary GBM. An age correlation has also been made toward LOH 10q, whereby GBM subjects of ages 40 and above showcased a much more frequent LOH 10q (84.2%), while those of ages 40 and below only show 16.7%.45 Due to the uniformity of results observed in various studies, it is believed that LOH chromosome 10 is of a high value as a prognostic biomarker for both primary and secondary GBM, whereby a more specific LOH at 10q25qter is accurate in diagnosing secondary GBM.46

One of the tumor suppressor genes heavily affected by LOH 10q is PTEN, whereby it regulates the PI3K pathway and thus modulates the cellular proliferation.47 PTEN functions by inhibiting PIP3 and thus resulting in no proliferation and apoptosis. Other tumor suppressor genes which might affect include TP53 and NF1, each serving different functions in the tumorigenesis of GBM.

3.6 Tumor protein 53

TP53 gene encodes for a widely known tumor suppressor protein, p53. Dubbed as the guardian of the genome, p53 serves various roles in suppressing tumorigenesis. TP53 point mutations have been observed at a much higher rate in secondary GBM (90%) when compared to cases in primary GBM (30%), and in some cases, are absent in primary GBM.47, 48 This finding correlates with the fact that mutations leading up to GBM might occur early in the development of gliomas and accumulate as the tumor progresses. Based on its molecular subtype, proneural GBM has a high incidence of TP53 mutation, while the classical subtype seems to lack this altogether.7, 10 One proposed mechanism of TP53 mutation in aiding the progression of GBM is via the mevalonate (MVA) pathway regulation. Elevated activity of MVA pathway has been observed in R273H cells (mutant TP53) as compared to U343 cells (wild-type TP53).49 Based on qRT-PCR methods, the TP53 mutation was verified to be upregulated and is correlated with the MVA pathway activation due to the upregulation of enzymes known to promote tumorigenesis, namely MVA kinase and 3’-hydroxy-3’-methylglutaryl-coenzyme A reductase.

Mouse double minute 2 homolog (MDM2) is a strong negative regulator of the TP53 gene in regulating tumorigenesis. The use of inhibitors of MDM2 is efficacious in patients harboring TP53 mutations. RG7112, a type of MDM2 inhibitor, has shown some promise for clinical testing based on researches done in TP53-MDM2 amplified cell lines.50 Some pathways involving mTOR and Bcl-2 are interestingly conferring resistance to certain cancer drugs and thus this opens interesting avenues to delve into in TP53 mutation GBM.51

3.7 Circulating tumor cells

GBM, much like other cancers, portrays circulating tumor cells (CTCs) which might lead to eventual metastasis and spreading of the disease. CTCs of GBM may harbor prognostic values which could also serve monitoring roles in patients. Easily obtainable from bodily fluid such as blood samples and analyzed via telomerase assays or amplification of EGFR, CTC has been reported to occur at high prevalence in GBM, up to >75%.50-54 Despite no significant correlation being reported, CTC has been noted to be phenotypically similar to the subtypes of proneural and mesenchymal subtypes.55 The most interesting clinical value for GBM CTC would be its correlation to tumor progression52 and recurrence54 aside from distinguishing the different molecular subtypes of GBM. Intriguingly, CTC levels detected after chemo-radiation are also significantly lower compared to the levels prior to treatment, which might offer an invaluable insight in differentiating tumor progression from radionecrosis.

5.1 Temozolomide chemotherapy

TMZ is an alkylating agent consumed orally by GBM and astrocytoma patients. The main mechanism of action of TMZ is by arresting the cell cycle at G2/M checkpoint which would eventually lead to apoptosis of GBM cancer cells.72 Various studies have assessed the effectiveness of TMZ as a drug choice for primary GBM which yielded a median overall survival of patients treated with TMZ in conjunction with radiotherapy to be 12 months, while the untreated group which only underwent radiotherapy had a median overall survival of 8 months.72 Addition of TMZ in radiotherapy yielded as much as 18% of test subjects surviving for at least 2 years, as compared to the 4% which survived for 2 years without the administration of TMZ.73

Despite the widespread use of TMZ as a drug of choice in the case of GBM, individual differences still exist and this takes form in TMZ resistance. The mechanism of resistance is observed to be tied to the overexpression of MGMT protein within GBM cells due to the lack of promoter methylation16-22 or wild-type IDH expression.39 This would in turn lead to a worse prognosis of GBM cases. Various other mechanisms have been proposed as to how the phenomenon of TMZ resistance came about, including the mismatch repair pathway74 and p53 pathway.75 Despite most findings indicating the vast correlation of MGMT expression to worsening prognosis of GBM upon TMZ treatment, a study conducted by Bocangel et al. has shown otherwise, which led various researches to focus on other molecular determinant of TMZ resistance aside from MGMT.76 To mitigate the effects of TMZ resistance, the usage of MGMT inhibitors such as O⁶-benzylguanine or interferon beta could activate the p53 cascade and thus sensitizing TMZ-resistant cells.77

5.2 Gene therapy

The technique to incorporate tumoricidal genes within viral vectors has been around in the field of oncology for a few years. One such example is Toca 511 and Toca FC which is still in development to deploy transgene cytosine deaminase via a retroviral replicating vector which catalyzes the production of 5-fluorocytosine to the active 5-fluorouracil.78 Toca 511 and Toca FC have been utilized in phases I and II clinical trials which yielded positive results in terms of lifespan of high-grade glioma patients.79, 80 The most interesting aspect of this retroviral system is that it does not trigger host's immune system and thus allow an oncolytic agent to be deployed and spread rapidly across functional tumor cells.

Another approach in combating GBM with viral gene transfer technique is the use of adeno-associated virus (AAV) vectors injected directly into GBM cells within the brain to express genes which are tumoricidal in nature. The usage of AAV9 vector to produce an anticancer agent, sTRAIL that killed as much as 60% of GBM cells in murine models and transfected cell lines as compared to the control, has been demonstrated by Crommentujin et al.81 AAV9 viral vector is a brilliant choice of a viral vector due to its serotype ability to pass blood–brain barrier in IV administration.82 The statistics obtained from the use of AAV vectors provide promising approaches to therapeutic advancements in GBM.

Other approaches for gene therapy include the transfer of suicide genes (thymidine kinase) within herpesvirus vectors which allowed cytotoxic nucleoside analogues to be phosphorylated. This phosphorylation would in turn cause the analogues to combine with the DNA of proliferating cells and thus halt their cell division.83 The introduction of exogenous tumor suppressor genes (i.e. p53, PTEN and p16INK4A) via nonreplicating adenovirus serotype 5 into GBM cells also showed a significant decrease in cellular proliferation, tumorigenesis and tumor progression in vivo.84 Despite the excellent opportunity for a therapy which combines tumor suppression and gene therapy, the gene transfer technique has limitations which include potential resistance and poor gene transfer. To mediate the downside of gene therapy, efforts such as inhibition of histone deacetylase 6 in tumor cells have shown promising results in increasing the oncolytic properties of viral gene transfer.85

5.3 Immunotherapy

The use of our body's own immune system as opposed to utilizing exogenous agents to ward off cancers such as GBM has been an interesting field in oncology. One such effort is the use of vaccines to trigger the body's immune response. Peptide vaccines such as the EGFRvIII found in primary GBM and classical GBM subtype could be administered to trigger immunity toward EGFRvIII expressed in GBM tumor cells.86 A phase II trial with 18 patients enrolled yielded 26 months of overall survival of patients treated with EGFRvIII vaccine, while the control group only showed 15 months of overall survival.87 This result allowed the trial to proceed and given the safety of this vaccine's toxicity profile, this vaccine harbors a future in the immunotherapeutic advancements in GBM. Other available vaccine options under clinical trials are tumor lysate vaccine and the dendritic cell-based vaccine (DCVax-L).88

The development of chimeric antigen receptor (CAR) T cells via T cell engineering (i.e. adoptive cell transfer) is another futuristic method in the combating GBM.89 By constructing T cells displaying CARs to target glioblastoma specific antigens (e.g. EGFRvIII and GBM-specific interleukins), these engineered CAR T cells have been observed to initiate potent tumoricidal activity.90, 91 Such effects portrayed by CAR T cells have been proven clinically in various clinical trials which specifically recruited GBM patients.91, 92 The development of CAR T cells provided results which are consistent with its usefulness compared to antibody-based immunotherapy as the engineered T cells can penetrate solid GBM tumors and recruit supporting molecules into the cancer microenvironment for a better immune response.93 Despite its benefits, one of the main concerns with the use of CAR T cells is the cytokine storm which might occur due to fast-paced T-cell proliferation, triggering various other health problems in GBM test patients.94

Immunotherapy has been refined throughout the decades to meet a bigger niche and fields of diseases by combining results of biomarker discovery or even cell engineering. Developing an effective, immunotherapeutic treatment which is nontoxic would be the end-goal of the immunotherapy advancements of GBM therapy.