Key points

Introduction

Among the cancers that affect the central nervous system, gliomas are those with the highest incidence. Their classification is based on morphology and the similarity between neoplastic cells and their normal glia counterparts: gliomas are tumors that have originated from the different cells of the glia, astrocytomas are histologically similar to astrocytes, oligodendrogliomas are similar to oligodendrocytes, and, in the case of mixed oligodendrogliomas, neoplastic cells similar to astrocytes and oligondendrocytes are present. Based on optical microscopy and immunohistochemistry, this classification also separates gliomas into 3 stages: stage II, stage III (anaplastic), and stage IV, which is the glioblastoma multiforme (GBM), the most malignant form of astrocytoma. In clinical practice, the current parameters used for prognosis and treatment protocol are size and stage of the tumor.

Nonetheless, this classification is considered insufficient and controversial. The responses observed among patients diagnosed with tumors within the same class are too variable. There is a high degree of subjectivity when different observers analyze the morphologic distinction between an oligodendroglioma and an astrocytoma, for example. Considering that oligondendrogliomas are chemosensitive and the survival rate of patients with oligondendrogliomas is much higher than for a patient with astrocytoma, more precision in the diagnostic is markedly relevant for the medical care. By applying the knowledge of molecular genetics, it is possible to generate more information that can help in clinical practice: deletions of the 1p and 19q regions of the short arm of chromosome 1 and long arm of chromosome 19, respectively, are more frequent in tumors that are predominantly oligodendrocytic. The loss of heterozygosity (LOH) of 1p is correlated with chemosensitivity and of 1p/19q with a prolonged therapeutic response and survival rate in the patient.

GBM correspond to 54% of all brain gliomas. In the United States in 2012, statistics indicated an incidence of 3.19 per 100,000 inhabitants, with a higher incidence in men (men 1.6 times higher than women), and a survival rate after 5 years of less than 5%. GBM are divided into primary (approximately 90%–95% and more frequent in older patients), and secondary (approximately 5%–10%, usually diagnosed in younger patients). Differently from secondary gliomas, primary GBM show no evidence of a progression from lower-stage gliomas (grades II or III).

Despite comprehensive therapy, which includes surgery, radiotherapy and chemotherapy, the prognosis of patients with GBM is very poor. Diagnosed individuals present an average of 12 to 18 months of life, the prognosis being worse for patients older than 60 years and better for younger individuals. Because of its aggressiveness, difficult treatment, and poor 5-year survival rate, gliomas have serious effects not only for the patient and immediate family but for the health system in general. Nowadays little can be offered to patients to improve their prognosis. However, many studies are being conducted to enable better understanding of the molecular biology and genetics of GBM in the hope that, as has happened with other types of cancers such as breast, melanoma, leukemia, and lung, a better molecular characterization may lead to the development of target-specific drugs and innovative approaches in the diagnosis, prognosis, and treatment of gliomas.

Cancer is a complex disease of the genome in which genes that regulate physiologic processes responsible for cellular differentiation, proliferation, and death are affected. When these genes are altered, still in the germinal cells or, more frequently, in somatic cells, and are not repaired, the mutations (the term here used to refer to any chromosomal alteration such as point mutation, deletion, insertion, amplification, and so forth), in most cases more than 1, are passed to daughter cells, which may confer characteristics that give the mutated cell advantages over normal cells. Not only chromosomal alterations but also alterations in methylation patterns and other forms of transcriptional regulations contribute to the complexity of cancer.

The mutations identified in cancer can be divided into 2 types, drivers and passengers. The former occurs in genes that belong to biological pathways essential for the survival of the malignant cell, and are frequently associated with an advantage in survival. The latter results from the genetic instability and does not have a pathologic role.

Driver mutations seem to be tissue specific; in other words, only a minority of the driver mutations associated with GBM is also identified in breast or colorectal tumors.

The comprehension of the biological pathway altered by a mutation is important for the genetic analysis of a tumor; mutations in different genes of the same biological pathway, present in one sample or in different samples, generally result in the same phenotype (exclusivity principle). For example, the analysis of GBM samples has indicated that the gene with the highest frequency of mutation is TP53 (35.8% of the samples had an altered TP53 ). However, if the biological pathways are considered, pathways of apoptosis and guanosine triphosphatases have mutations in 79.2% of the analyzed samples. The principle of exclusivity usually holds; it is rare to observe alterations in multiple genes the same biological pathway. For example, it is unlikely to find the same tumor bearing a mutation in KRAS with a mutation in BRAF , which is downstream of the same pathway.

This review aims to provide an overview of the molecular genetics of glioblastomas, probably the most extensively studied type of cancer to this date. Therefore, the amount of information published and/or available is overwhelming. Although it is by no means possible to give a thorough and complete picture on this topic, the reader is referred to other excellent reviews published, which will be able to complement the descriptions herein of an already very complex picture of the genomic alterations identified in these tumors.

Introduction

Among the cancers that affect the central nervous system, gliomas are those with the highest incidence. Their classification is based on morphology and the similarity between neoplastic cells and their normal glia counterparts: gliomas are tumors that have originated from the different cells of the glia, astrocytomas are histologically similar to astrocytes, oligodendrogliomas are similar to oligodendrocytes, and, in the case of mixed oligodendrogliomas, neoplastic cells similar to astrocytes and oligondendrocytes are present. Based on optical microscopy and immunohistochemistry, this classification also separates gliomas into 3 stages: stage II, stage III (anaplastic), and stage IV, which is the glioblastoma multiforme (GBM), the most malignant form of astrocytoma. In clinical practice, the current parameters used for prognosis and treatment protocol are size and stage of the tumor.

Nonetheless, this classification is considered insufficient and controversial. The responses observed among patients diagnosed with tumors within the same class are too variable. There is a high degree of subjectivity when different observers analyze the morphologic distinction between an oligodendroglioma and an astrocytoma, for example. Considering that oligondendrogliomas are chemosensitive and the survival rate of patients with oligondendrogliomas is much higher than for a patient with astrocytoma, more precision in the diagnostic is markedly relevant for the medical care. By applying the knowledge of molecular genetics, it is possible to generate more information that can help in clinical practice: deletions of the 1p and 19q regions of the short arm of chromosome 1 and long arm of chromosome 19, respectively, are more frequent in tumors that are predominantly oligodendrocytic. The loss of heterozygosity (LOH) of 1p is correlated with chemosensitivity and of 1p/19q with a prolonged therapeutic response and survival rate in the patient.

GBM correspond to 54% of all brain gliomas. In the United States in 2012, statistics indicated an incidence of 3.19 per 100,000 inhabitants, with a higher incidence in men (men 1.6 times higher than women), and a survival rate after 5 years of less than 5%. GBM are divided into primary (approximately 90%–95% and more frequent in older patients), and secondary (approximately 5%–10%, usually diagnosed in younger patients). Differently from secondary gliomas, primary GBM show no evidence of a progression from lower-stage gliomas (grades II or III).

Despite comprehensive therapy, which includes surgery, radiotherapy and chemotherapy, the prognosis of patients with GBM is very poor. Diagnosed individuals present an average of 12 to 18 months of life, the prognosis being worse for patients older than 60 years and better for younger individuals. Because of its aggressiveness, difficult treatment, and poor 5-year survival rate, gliomas have serious effects not only for the patient and immediate family but for the health system in general. Nowadays little can be offered to patients to improve their prognosis. However, many studies are being conducted to enable better understanding of the molecular biology and genetics of GBM in the hope that, as has happened with other types of cancers such as breast, melanoma, leukemia, and lung, a better molecular characterization may lead to the development of target-specific drugs and innovative approaches in the diagnosis, prognosis, and treatment of gliomas.

Cancer is a complex disease of the genome in which genes that regulate physiologic processes responsible for cellular differentiation, proliferation, and death are affected. When these genes are altered, still in the germinal cells or, more frequently, in somatic cells, and are not repaired, the mutations (the term here used to refer to any chromosomal alteration such as point mutation, deletion, insertion, amplification, and so forth), in most cases more than 1, are passed to daughter cells, which may confer characteristics that give the mutated cell advantages over normal cells. Not only chromosomal alterations but also alterations in methylation patterns and other forms of transcriptional regulations contribute to the complexity of cancer.

The mutations identified in cancer can be divided into 2 types, drivers and passengers. The former occurs in genes that belong to biological pathways essential for the survival of the malignant cell, and are frequently associated with an advantage in survival. The latter results from the genetic instability and does not have a pathologic role.

Driver mutations seem to be tissue specific; in other words, only a minority of the driver mutations associated with GBM is also identified in breast or colorectal tumors.

The comprehension of the biological pathway altered by a mutation is important for the genetic analysis of a tumor; mutations in different genes of the same biological pathway, present in one sample or in different samples, generally result in the same phenotype (exclusivity principle). For example, the analysis of GBM samples has indicated that the gene with the highest frequency of mutation is TP53 (35.8% of the samples had an altered TP53 ). However, if the biological pathways are considered, pathways of apoptosis and guanosine triphosphatases have mutations in 79.2% of the analyzed samples. The principle of exclusivity usually holds; it is rare to observe alterations in multiple genes the same biological pathway. For example, it is unlikely to find the same tumor bearing a mutation in KRAS with a mutation in BRAF , which is downstream of the same pathway.

This review aims to provide an overview of the molecular genetics of glioblastomas, probably the most extensively studied type of cancer to this date. Therefore, the amount of information published and/or available is overwhelming. Although it is by no means possible to give a thorough and complete picture on this topic, the reader is referred to other excellent reviews published, which will be able to complement the descriptions herein of an already very complex picture of the genomic alterations identified in these tumors.

Molecular characterization of glioblastoma multiforme: genetic alterations

In 2008, 2 seminal studies were published delineating a complex landscape of genetic alterations present in GBM. The Tumor Cancer Genome Atlas (TCGA) analyzed fewer genes in more samples than the group of Vogelstein, which analyzed more genes in fewer samples.

Both studies pointed to the importance of 3 major biological pathways, responsible for cell survival, DNA repair, progression of G1/S, blockade of G2/M, cell migration, apoptosis, and progression of cell cycle: pathways of (1) RTK (tyrosine kinase receptors) signaling, and pathways of the tumor suppressor genes (2) TP53 and (3) RB . The articles agreed on most of the genes altered and, in accordance with the principle of exclusivity, in general the alterations identified were in genes of different biological pathways. The TCGA study identified simultaneous alterations in 3 pathways in 74% of the samples of the study performed by the TCGA network.

Both studies identified samples with a hypermutated phenotype after treatment with the alkylating agent temozolomide (TMZ). The TCGA network identified 7 samples with this profile. Of these, 6 had mutations in at least one DNA-repair gene of the MMR (mismatch repair) family that seem to have an influence in the methylation pattern of MGMT . The hypermethylation of the promoter of the gene MGMT results in the lack of expression of the protein O ⁶ -methylguanine-DNA methyltransferase, which repairs DNA. The methylation of MGMT interferes with the susceptibility of cells to treatment with TMZ. Results indicate that the administration of TMZ in combination with radiotherapy can increase the survival of patients diagnosed with GBM. Therefore, these observations can have a direct impact on the clinical practice and protocols administered today. In addition, mutations on MMR and the methylation pattern of MGMT can have an important role in the frequency of mutations observed in samples from treated patients.

The most frequent alterations identified were: gene amplifications on EGFR , CDK4 , PDGFRA , MDM2 , MDM4 , MET CDK6 , MYCN , CCND2 , P1K3CA and AKT3 ; homozygotic deletions in the genes CDKN2A/B , PTEN , CDKN2C , RB1 , PARK2 , and NF1 ; and higher-frequency mutations in the genes TP53 , PTEN , NF1 , EGFR , ERBB2 , RB1 , P1K3R1 , and P1K3CA. Considering only alterations in copy number, 66%, 70%, and 59% of the 206 samples analyzed presented somatic alterations in the main pathways of RB , TP53 and RTK, respectively. If sequencing results were included, this frequency would increase to 87%, 78%, and 88% in the respective pathways.

In the genes identified as frequently mutated, Parsons and colleagues indicated candidates that had more chances of being responsible for the process of tumor formation, in other words, driver mutations: CDKN2A , TP53 , EGFR , PTEN , NF1 , CDK4 , RB1 , 1DH1 , P1K3CA , and P1K3R1 . It is worth pointing out that both studies identified mutations in the gene NF1 , possibly a relevant tumor suppressor gene in GBM; 23% and 15% of the samples in the TCGA network study and of Parsons and colleagues, respectively, had some type of alteration in this gene.

Mutations in the IDH1 gene were identified in almost all secondary GBM and correlated with a better prognosis, indicating their relevance in the refinement of the classification of these tumors. In primary GBM, these mutations were rarely detected. A mutation in IDH1 correlates with a hypermethylation status of the DNA, denominated Glioma CpG Island Methylator Phenotype (G-CIMP). Mutations in IDH2 have also been identified in GBM samples, but are less recurrent than in IDH1 . IDH1 and IDH2 are NADP ⁺ -dependent enzymes that catalyze the oxidative decarboxylation of isocitrate to α-ketoglutarate, an important biological compound in the biochemical pathway of carbohydrates, lipids, and amino acids. The identification of mutations on IDH1 and IDH2 were important not only because of their predominance but as possible biomarkers of progressive gliomas. They highlight an important characteristic of GBM, of a disease of altered cellular metabolism and correlated epigenetics.

A previous report demonstrated the amplification of tyrosine kinase receptor genes in the GBM samples from TCGA. Approximately 50% of the 206 samples had amplification in at least 1 of the 51 RTKs analyzed. The most frequent was on the EGFR gene (41%), followed by PDGFRA (10%), which was frequently coamplified with KIT (7%) and KDR (4%). MET was the third most frequent (2%). The amplification of EGFR was associated with a worse prognosis.

Lack of heterozygosity of chromosome 10 is also a frequent alteration in GBMs.

Genomic alterations such as the ones described here produce altered patterns of gene expression. Several articles have been published reporting analysis of patterns of expression (transcriptomes) in gliomas.