There has been a vast increase in our knowledge of the molecular pathology of central nervous system (CNS) tumors, and a new WHO classification will soon be published (1). The WHO 2016 Classification already introduced for the first time molecular criteria in the classification of gliomas by IDH status and the molecular groups of medulloblastomas (2). In many advanced centers, brain tumor samples are now routinely tested with genomic techniques, such as targeted sequencing, RNAseq, and genome-wide DNA methylation profiling (3–6). The latter has become a regular genomic diagnostic tool in many centers (7–9), and many tumor entities in the upcoming WHO classification will have the methylation groups listed as desirable diagnostic criteria (1). However, genomic tests are still expensive, especially in countries with a national health system, and also typically, their results only come back several weeks later. Interpretation of genomic data also requires expertise and reporting of genomic tests, especially those conducted by commercial companies, is variable. In contrast, single-gene tests are faster for results, often done in-house, and are much cheaper. Results of many single-gene tests, e.g., 1p19q, IDH genotyping, are needed soon after the surgical resection so that planning for further treatment can be made and patients can receive appropriate counseling as soon as possible after the operation. There is therefore a need for precise molecular diagnosis and grading of brain tumors using single-gene tests.

Isocitrate dehydrogenase (IDH) genotype has been shown to be a major classifier and prognosticator for gliomas (10, 11) and WHO 2016 classifies astrocytic tumors as to whether they are IDH mutant or IDH wild type. Together with another group, we were the first to show the usefulness of mutations of TERT promoter (TERTp), when combined with IDH genotype, in the precise prognostication of low-grade gliomas (12, 13). The Cancer Genome Atlas (TCGA) showed that a combination of IDH genotype and 1p19q status stratified survivals well for the low-grade gliomas (14), but Eckel-Passow et al. in the same issue of New England Journal of Medicine, in addition to using IDH and 1p19q, introduced also TERT mutation to stratify low-grade gliomas (15). BRAF is also a gene that can be easily tested in most laboratories without too much difficulty with a mutation hotspot at V600E. We showed that in the younger adults, it is associated with good survival, and similarly, good survival with BRAF V600E mutated gliomas was shown in smaller series (16–18). Phillips et al. showed that among BRAF mutated anaplastic pleomorphic xanthoastrocytoma, the regular molecular hallmarks of glioblastomas, e.g., EGFR amplification, 10q loss, were uncommon (19). Mutations of the chromatin genes, H3.3 and H3.1, were also found to be poor prognosticators for gliomas, and the poor prognosis is not limited to tumors located in mid-line, nor were mutations restricted to pediatric high-grade gliomas (20–24).

We hypothesized that combinations of genes IDH1/2, 1p19q, TERTp, EGFR, BRAF, TERTp, 10q, and H3 status can stratify gliomas for risk and may even be superior to conventional histological grading for prognostication. We recognize that this gene list is not exhaustive as regards the vast amount of literature concerning the molecular pathology of gliomas. However, since there is a limit as to how many single-gene tests an in-house facility can possibly carry out with a reasonable turnover time, we hypothesize that such a list would adequately molecularly grade a glioma and that it could be superior to conventional histological grading in some instances.

The diagnostic protocols for brain tumors have been vastly changed in the advanced centers due to the advent of next-generation sequencing, for both DNA and RNA and also genome-wide methylation profiling. The former can evaluate gene mutations or gene fusion on a panel of a large number of genes, while the latter can map all brain tumors according to their methylation profiling into 91 methylation classes (7). Copy number variations (CNVs) can be obtained from both methods. Both have been widely used in the day-to-day fine diagnosis of brain tumors as well as research (8, 30, 38, 39). The WHO 2016 Classification of CNS Tumors for the first time introduced molecular criteria in the diagnosis of some brain tumors, and many centers in the world use an integrated, multi-layered approach in the pathological diagnoses of brain tumors, integrating histological features, WHO grades, and molecular characteristics. Adult diffuse astrocytic tumors are classified into IDH-mutant and IDH wild type, and oligodendrogliomas are defined by both IDH mutation and 1p19q codeletion by the WHO 2016 Classification. However, grading of adult diffuse gliomas remains entirely histological depending mainly on mitoses, cellular anaplasia, necrosis, and microvascular proliferation.

Most general pathology laboratories in the world are not in a position to carry out next-generation sequencing or genome-wide methylation profiling routinely on all their neurosurgical specimens, as these tests remain expensive and demanding on expertise for interpretation, e.g., t-distributed stochastic neighbor embedding (t-SNE) plots. Most national types of health finance systems are unable to support such extensive genomic tests regularly. However, most pathology departments in the more advanced societies will have the capacity for single-gene studies and molecular tests like Sanger sequencing, for example, for BRAF V600E, IDH1 R132, and IDH2 R172H. Similarly, FISH is also widely available in pathology departments, especially in the context of Her-2 testing for breast cancers. We therefore believe that there is a need to develop a relatively simple molecular grading algorithm for the gliomas based on single-gene testing based on current literature on the molecular pathology of gliomas (15, 32, 33, 40–42). There are many excellent studies on the grading of special groups of gliomas by a panel of single genes. For example, TCGA concluded that IDH and 1p19q status satisfactorily grades diffuse lower-grade gliomas, while Eckel-Passow et al. further fine-graded lower-grade gliomas with the addition of TERT status (14, 15).

Mutations of the promoter region of TERT have been suggested to be a key terminal event in the evolution of glioblastomas, but it has an intriguing property of being mutated also in oligodendrogliomas, which are mostly low-grade gliomas (43). The reason for its dual and seemingly contradictory properties is unknown. The dual impact of TERT can be seen in that other than being present in IDH mutant, 1p19q codeleted tumors, it was also a prognosticating factor among IDH mutant 1p19q non-deleted tumors (44, 45) and, in this series, the difference between molecular grade 2 (TERT mutant, IDH mutant, 1p19q non-codeleted) and molecular grade 3 (TERT wild type, IDH mutant, 1p19q non-codeleted). However, molecular group 2 remains a small group in our cohort (n = 59).

Other than lower-grade gliomas, there have also been a few studies in the literature devoted to molecular-grade special groups of gliomas by single-gene studies (32, 33), but we are only aware of one previous attempt to use a panel of single-gene status to grade holistically all adult gliomas (46). In this study, Jiao et al. showed that gliomas can be satisfactorily graded for prognosis by complete sequencing of IDH1/2, ATRX, CIC, and FUBP1. However, the latter genes are large, and sequencing of entire genes is laborious and expensive.

The biomarkers used in this study were known in the literature on their own. Group 1, IDH mutant, 1p19q codeleted is the regular oligodendroglioma as currently defined by WHO Classification, and Group 6 constitutes the bulk of the regular IDH wild-type glioblastomas as defined by cIMPACT-NOW update 3 (47). Group 5 would constitute most of the IDH wild-type lower-grade gliomas that do not fulfill the molecular criteria of glioblastoma (33, 48, 49). It is also interesting that in our molecular Group 4, BRAF V600E was able to impart a better prognosis among the IDH wild-type gliomas. This is consistent with the few papers concerning the potentially better prognosis of BRAF V600E in adult gliomas (16–18). The overall number in this IDH wild type, BRAF mutant group is small (27/1,028 cases), consistent with the overall low percentage of BRAF V600E in adult gliomas (50). These tumors were not epithelioid, PXA-like, or gangliogliomas on histological review; similarly in the three references quoted, there was also no mention of such histological features.

For glioblastomas, cIMPACT and the new WHO (2021) Classification recommended the use of EGFR amplification, TERT, and chromosome 7+/10− as criteria for diagnosis of molecular glioblastoma in IDH wild-type adult gliomas, which do not show diagnostic histological features like necrosis and microvascular proliferation (1, 47). For IDH mutant histologically lower-grade gliomas, the molecular diagnostic criterion for glioblastoma is the homozygous deletion of CDKN2A/B (51). In this study, we have used extensively EGFR amplification and 10q loss in addition to TERT, as 10q loss has been shown to be associated with glioblastomas for many years (52, 53). The WHO (2021) and cIMPACT criteria actually recommend combined whole chromosome 7 gain and 10 losses (1, 47), with molecular events that are usually not tested in a general pathology laboratory, and the critical reference for this recommendation used methylation profiling to test for whole chromosomes (7).

The molecular grading scheme we propose can easily be made available in general pathology departments. Adoption of this system provides a fine prognostication for gliomas and in, some instances, is superior to the histological grading, as shown in Figure 3 . The use of this scheme also fulfills many of the recommended markers of the WHO Classification. While most of the biomarkers used in this grading scheme including 1p19q codeletion, IDH2 mutations, 10q loss, and TERT mutation do not have any immunohistochemical tests, IDH1 and BRAF V600E mutations are readily tested by immunohistochemistry, and these can be used in busy diagnostic laboratories. We recognize that not all single-gene biomarkers that were described in the new WHO Classification and cIMPACT (e.g., CDKN2A/B) have been included in the grading scheme. In our cohort of molecular Grade 2 (IDH mutant, 1p19q non-codeleted, TERT mutant) and Grade 3 (IDH mutant, 1p19q non-codeleted, TERT wild type) tumors, we were unable to demonstrate a statistical significance between cases with or without homozygous CDKN2A deletion. Marker and Pearce also showed that CDKN2A deletion was not associated with prognostication in Grades II and III IDH-mutant astrocytomas (54). Methylated MGMT has been known for a long time to be a favorable prognostic marker in glioblastomas (55) and predicts their response to alkylating chemotherapy (56, 57). However, it is not in regular use for the lower-grade glioma as a prognosticator, and most of the low histological grade IDH mutant gliomas are likely to be MGMT methylated.

We also recognize the limitations of our study, as we used retrospective analyses of existing data, and it would be impossible for us to re-do a new panel of single-gene markers for 1,275 cases according to the recommendation of the new WHO Classification. Ours is not a proposal to replace the new WHO Classification (1) but merely a proof of principle that a set of well-placed single-gene biomarkers can fully cover risk stratification in adult gliomas. Even in the new WHO Classification (1), there are cases that do not fit and are classified as NEC (Not Elsewhere Classified), e.g., IDH wt gliomas that do not exhibit EGFR amplification or TERTp or combined whole chromosomes +7/−10q. It is possible that other algorithms may have developed that use some biomarkers might be shown to be equally useful in molecular grading of gliomas in the future. But our study showed that a simple molecular grading scheme such as the one proposed can satisfactorily fine-grade adult gliomas, and these tests can be performed relatively quickly and are highly achievable in practically all pathology laboratories. This grading scheme will provide critical and timely information for immediate clinical management of glioma patients.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

The studies involving human participants were reviewed and approved by The Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee; the Ethics Committees of Huashan Hospital, Shanghai; and The First Affiliated Hospital of Zhengzhou University, Zhengzhou. The patients/participants provided their written informed consent to participate in this study.

The authors confirm contribution to the paper, as follows: study conception and design: AC, W-SP, HL, and H-KN. Data acquisition: Z-FS, KL, W-WW, HC, NC, DC, X-ZL, Z-YZ, and YM. Analysis and interpretation of data: AC, Z-FS, KL, NC, and H-KN. Drafting of the manuscript: AC and H-KN. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

This study was supported by the Health and Medical Research Fund (HMRF), the Food and Health Bureau of Hong Kong (07180736); the National Natural Science Foundation of China (82072020, 81702465, U1904148 and U1804172); and Shanghai Municipal Science and Technology Major Project (2018SHZDZX01).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.