In additional efforts to construct clinically-relevant and objective tumor subclasses, two recent landmark studies have proposed classification schemes defining subgroups of lower-grade glioma (grades II and III) with distinct patterns of molecular and clinical features. Using unsupervised clustering of genomic, transcriptomic, epigenomic, and proteomic data from 293 lower-grade gliomas, an analysis from the TCGA network defined three disease subgroups according to IDH, TP53, and 1p/19q status [15]. Two groups were defined by those with IDH mutation with or without 1p/19q codeletion; the first included patients who also tended to have CIC, FUBP1, NOTCH1, PI3 kinase pathway, and TERT promoter alterations, while the second group was characterized by patients who also tended to have TP53 and ATRX mutations. The authors suggested that development of these two types of lower-grade gliomas is initiated by IDH mutation and G-CIMP establishment, then driven forward by either TP53 mutation or 1p/19q codeletion. Tumors without an IDH mutation marked a third group that was molecularly and clinically comparable to that seen in primary grade IV glioma (glioblastoma), displaying frequent PTEN, EGFR, NF1, TP53, or PIK3CA alterations.

In a second, larger study by Eckel-Passow et al. including TCGA cases as well as cases from two independent sets from the UCSF Adult Glioma Study and the Mayo Clinic, 615 lower-grade gliomas were classified into five groups according to IDH mutation, 1p/19q codeletion, and TERT promoter mutation status [16]. The classification scheme proposed by Eckel-Passow et al. is similar to that proposed by the TCGA study but incorporates TERT promoter rather than TP53 mutation status to further refine three groups into five, generating a comprehensive classification system that also incorporates higher glioma grades; a comparison of the molecular alterations that define the two classification schemes is illustrated in Fig. 3.2. The lower-grade glioma groups included tumors without any of the three alterations (7%, “triple-negative”), with IDH mutation only (45%), TERT mutation only (10%), TERT and IDH mutations (5%), and with all three alterations (29%, “triple-positive”). As in the TCGA study, subgroups were associated with characteristic acquired alterations: CIC, FUBP1, and NOTCH1 mutations were seen in triple-positive tumors, PI3 kinase pathway mutations in triple-positive and TERT mutation only tumors, TP53 and ATRX mutations in tumors with IDH mutation with and without TERT promoter mutation, and EGFR, PTEN, and NF1 mutations in triple-negative and TERT mutation only tumors. Subgroups also had characteristic patterns of copy-number changes and TCGA expression subtypes [17]; interestingly, tumors harboring IDH mutations with or without TERT and 1p/19q alterations were most closely related to the proneural expression pattern, while triple-negative or TERT mutation only tumors were more likely to display classical or mesenchymal expression patterns, indicating greater molecular similarity to primary glioblastoma.

Additionally, the five subgroups had notable differences in the distribution of histologic subtypes. These data, along with population-level incidence data, can be used to construct population-based estimates of the proportion of grade II astrocytomas and oligodendrogliomas by histology and molecular subtype [16, 18] (Fig. 3.3). According to these estimates, nearly three-quarters of grade II astrocytomas and oligodendrogliomas combined are triple-positive or have IDH mutation only. Interestingly, among grade II astrocytomas and oligodendrogliomas, triple-positive tumors tend to be predominantly oligodendrogliomas, while all other subtypes tend to be astrocytomas. It is important to note that these estimates do not include oligoastrocytomas due to the lack of available population data by grade for this histologic type, and that histological subtypes are based on traditional definitions before the 2016 updated WHO guidelines.

While DLGGs are most often sporadic, ~5% of glioma cases are familial in nature, and the risk of glioma development is approximately twice that in first-degree relatives of glioma patients than for the general population [19–23]. Although such familial aggregation can be a reflection of shared environmental exposures, a growing number of constitutive genetic risk factors have been identified that cause glioma or contribute to glioma susceptibility.

Epidemiological studies have explored many exposures for association with glioma risk with inconsistent results, particularly for low-grade glioma development [25, 52]. Due to the rare nature of these tumors and frequent pooling of glioma grades in many previous epidemiologic studies, the evidence for associations with many factors (e.g., occupational exposures, dietary factors, anti-inflammatory agents) is limited and inconclusive, and will require larger studies with detailed exposure histories and well-annotated biospecimens for increased study validity. A summary of the established, probable, and unlikely risk factors for low-grade glioma is presented in Table 3.3. The risk factors with the best evidence for contributing to low-grade glioma risk, namely exposure to ionizing radiation and history of allergies or atopic disease, are discussed here.

While survival after diagnosis with DLGG is better than that for high-grade glioma patients, it has not improved over the last several decades and DLGG is still ultimately a fatal disease with median survival between 7 and 12 years [3, 4]. Tumor histology and certain clinical factors have been linked to DLGG survival; however, these features are limited in their clinical value and ability to predict tumor progression and/or survival. In recent years, novel molecular biomarkers and tumor classifications have emerged that are remarkably informative for outcome prediction.