The 2016 WHO Classification of paediatric CNS tumours – the essentials
Posted in Neurosurgery on 2nd Aug 2018
Harsha Narayanamurthy MBBS, MRCS, is a Specialty Registrar (ST7) in Neurosurgery on the Severn and Peninsula Deanery training programme, and is interested in Neuro-oncology, tumour genetics and novel approaches to brain tumour treatment.
Peter Whitfield BM (Distinction in Clin Med), PhD, FRCS Eng., FRCS, SN, FHEA is a Consultant and Associate Professor in Neurosurgery. He is Chairman of the SAC in Neurosurgery and the Fellowship of the European Board of Neurosurgery Examination. His clinical practice is wide, encompassing traumatic brain injury, subarachnoid haemorrhage, brain tumours, hydrocephalus and degenerative spine disease.
Kathreena Kurian BSc (Exp Pathol), MBBS, MD, FRCPath (Neuro) is a Reader in Brain Tumour Research at the University of Bristol and Honorary Consultant Neuropathologist, North Bristol Trust. She runs the Brain Tumour Research Centre; Institute of Clinical Neurosciences, Bristol Medical school which aims to reduce the burden of brain cancer by identifying risk factors and biomarkers that will inform prevention, personalised diagnosis and treatment.
Correspondence to: Mr H Narayanamurthy, Department of Neurosurgery, Southmead Hospital, North Bristol NHS Trust, Bristol.
Conflict of interest statement: None declared.
Provenance and peer review: Submitted and externally reviewed.
Date first submitted: 20/5/17
Date resubmitted after peer review: 29/4/18
Acceptance date: 22/8/18
To cite: Narayanamurthy H, Whitfield P, Kurian K. ACNR 2018;17(4):5-8
Published under a Creative Commons license
List of abbreviations:
IDH: Isocitrate Dehydrogenase TP53: Tumour protein (gene) 53 BRAF: proto-oncogene B-raf/ v-Raf murine sarcoma viral oncogene homolog B H3 K27M: Lysine 27 to methionine substitution in histone variant H3.3 HIST1H3B, HIST1H3C: Histone cluster 1 H3 Family member B and C WNT: Wingless-related integration site SHH: Sonic hedgehog signalling pathway INI1: Integrase Interactor 1 SMARCB1: SWI/ SNF related, matrix associated, actin dependent regulator of Chromatin, subfamily B, member 1 SMARCA4: SWI/ SNF related, matrix associated, actin dependent regulator of Chromatin, subfamily C, member 4 C19MC: Chromosome 19 microRNA cluster RELA: V-Rel Avian Reticuloendotheliosis Viral Oncogene Homolog A L1CAM: L1 cell adhesion molecule.
Brain tumour classification is critical for understanding the behaviour of tumours. The WHO classification of CNS tumours has been among the best resources for the latest knowledge about tumours of the nervous system. New to the 2016 classification are concepts of ‘layered diagnosis’ and ‘integrated diagnosis’ which are a combination of traditional histological grading with molecular genetics. It is important that clinicians understand the basic concepts and important differences from the earlier versions for their daily practice.
Classification of the different brain tumours has been one of the primary means of understanding them by organising them according to their various characteristics. Rudolf Virchow’s report on brain tumour classification in 18631 was the first known attempt at classifying brain tumours. From then, through the seminal papers of Bailey and Cushing in 1926 explaining the potential relations between the developing brain and brain tumours,2 to the concept of tumour grading introduced in 1949 by James Kernohan and colleagues,3 there have been many significant contributions in this field. Zulch et. al., published the first WHO based classification in 1979, followed by the second (1993), third (2000) and fourth (2007) editions. A briefly popular alternative tumour grading system called the St Anne – Mayo grading system was also published in 1988.4 More recently, The Cancer Genome Atlas (TCGA) has enhanced our understanding of the molecular basis of brain tumours exponentially. More precise molecular classification of brain tumours may improve the success rate of clinical trials by comparing similar molecular subtypes and lead to personalised therapeutic options for patients. The 2016 WHO classification introduces the concept of a ‘layered diagnosis’5 combining histology and molecular genetics, with molecular genetics, grading, histology and final integrated diagnosis forming layers 4 to 1. This combined phenotypic and genotypic grouping puts tumours with similar prognostic markers together and guides treatment for biologically and genetically similar tumours. However, some disadvantages are unavoidable, like a delay in getting the final result due to the wait for molecular genetics testing, a potential change in the grade of the tumour with the final integrated diagnosis, and discrepancies in access to molecular analysis facilities and expertise in some parts of the world. Relevant information regarding some common paediatric tumours are presented here, but the list is not exhaustive. It aims to provide clinicians a list of the conditions usually encountered in a normal paediatric neurosurgical practice.
Diffuse midline gliomas
The new entity in this category of the classification is denoted as diffuse midline gliomas, WHO grade IV H3 K27M mutant. These tumours were historically termed diffuse intrinsic pontine gliomas (DIPG), occurring primarily in children in midline locations like thalamus, brainstem and spinal cord. They show K27M mutations in the Histone H3 gene H3F3A (80%), less commonly in the HIST1H3B gene (~20%) and very rarely in the HIST1H3C gene7 (Histones are responsible for the nucleosome structure of chromosomal fibres and play a central role in transcription regulation, DNA repair and replication, and chromosomal stability. The mutant H3.3 histone disrupts epigenetic post-translational modifications near genes involved in cancer.8 Histologically, they are diffusely infiltrative gliomas with predominantly astrocytic differentiation. Certain juvenile gliomas in the telencephalic hemispheres (but still midline) present with morphological appearances of a glioblastoma and genetic testing reveals an H3F3A mutation at G34 (H3F3A G34R/V). These tumours have a clinical prognosis lying between GBM IDHwt and GBM IDHmut.9
The classification of medulloblastomas (MB) was one of the greatest challenges during the creation of the WHO 2016 classification due to the usefulness of both its histological classification as well as the genetic classification. Hence, medulloblastomas have both genetically defined and histologically defined classifications. Since medulloblastomas are embryonal tumours, they are automatically assigned a WHO grade IV. The histologically defined classification consists of classic, desmoplastic/ nodular and large cell/anaplastic MB. The genetically defined classification, which is new, consists of WNT activated, SHH activated and non WNT/non SHH MB. The WNT (wingless-related integration site) pathway regulates crucial aspects of cell fate determination, cell migration and polarity, neural patterning and organogenesis, whereas the SHH (sonic hedgehog) pathway has a role in the induction of the floor plate of the developing embryo and plays an important role in regulation of organogenesis later. The non WNT/ non SHH group consists of Groups 3 and 4, where group 4 tumours with chromosome 11 losses are low-risk tumours and group 3 and 4 tumours without chromosome 11 losses are standard-risk tumours. The WNT activated MBs are low-risk tumours, whereas the SHH activated MBs with TP53 mutations are high-risk and without TP53 mutations are standard-risk tumours. The non WNT/ non SHH group is the most common form of MB, accounting for nearly 60% of all MBs.10 In case of difficulty establishing a histological or genetic diagnosis, medulloblastomas can be termed NOS. Since expression profiling or methylation analyses for molecular classification of MBs are not easily available in all institutions, the WHO have recommended a set of three antibodies to help classify these tumours. They are represented in Figure 19 below.
Other Embryonal tumours
This category includes entities like atypical teratoid/ rhabdoid tumours (AT/RT), CNS embryonal tumours with rhabdoid features, embryonal tumours with multi-layered rosettes (C19MC – altered and NOS), and CNS embryonal tumours (consisting of former CNS PNET entities – medulloepitheliomas, CNS neuroblastomas, ganglioneuroblastomas and a new entity, CNS embryonal tumours NOS), all of which are given a WHO grade IV due to their embryonal origin. AT/ RTs are characterised by loss of expression of either INI1 (involving mutation/ loss of locus of SMARCB1 gene) or BRG1 (encoded by SMARCA4 gene). Tumours which are histologically akin to AT/RTs but express INI1 and BRG1 are termed CNS embryonal tumours with rhabdoid features.12 A new entity introduced in this category in the 2016 classification is ‘Embryonal tumour with multi-layered rosettes’ (ETMR), and, based on amplification of C19MC expression, they are termed ETMR, C19MC altered or ETMR, NOS (if no alteration of C19MC or if the test cannot be performed).12
This category has been changed little in the new classification due to the continuing (albeit diminishing) value of the existing histological classification and grading system. Histologically, WHO Grades 1 to 3 ependymomas are described:
- WHO grade I tumours – myxopapillary ependymoma and subependymoma,
- WHO grade II tumours – Ependymoma, papillary ependymoma, tanycytic ependymoma, and clear cell ependymoma, and
- WHO grade III tumours – Anaplastic ependymoma.
There has been in research use, a molecular classification of paediatric posterior fossa ependymomas which divides them into Group A, with poorer outcomes and Group B with better outcomes.13 The new genetics based entity in the 2016 classification is the supratentorial ependymoma with a RELA fusion (majority of paediatric and small minority of adult tumours in this location) which is graded into WHO grade II and anaplastic WHO grade III tumours based on the same criteria as the non RELA-fused ependymomas. These tumours can be tested using the antibody L1CAM which only binds to RELA-fused ependymomas.14 An entity called cellular ependymoma has been eliminated from the new classification due to its resemblance to a standard WHO grade II ependymoma.
Neuronal and mixed neuronal – glial tumours
This category comprises of some WHO grades I – III tumours which have not been changed very much from the previous classification system. They include dysembryoplastic neuroepithelial tumours, gangliogliomas, gangliocytomas, dysplastic cerebellar gangliocytomas, desmoplastic infantile gangliogliomas, desmoplastic infantile astrocytomas and central neurocytomas (WHO grade II).11 To this has been added one new entity called diffuse leptomeningeal glioneuronal tumour, WHO grade I (DLGNT). These were previously termed diffuse oligodendroglial leptomeningeal tumours, and can show frequent BRAF: KIAA 1549 duplications and deletions, 1p (frequent), and, not that frequently, 19q deletions14 and no IDH mutations. Though they are predominantly Grade I tumours,12 there have been reports of malignant tumours as well. The other change in this category is that neurocytomas are now considered to be characterised by the absence of IDH mutations.11
The 2016 molecular WHO classification reflects a quantum leap in the application of genetics to the description of brain tumours. However, the 2016 WHO classification has also highlighted the amount of work that is still necessary to achieve better genetically based classifications for many tumours such as meningiomas and ependymomas. Initiatives such as the 100,000 genomes project will further our understanding in this regard. It is vital that neurosurgeons stay abreast of these developments and work in a multidisciplinary team including neuropathologists, oncologists and neuroradiologists to conduct an informed, up to date practice.
|1. Virchow R. Die Krankhaften Geschwulste. Hirschwald 1863-67, Berlin. 1863.|
|2. Bailey P CH. A classification of tumors of the glioma group on a histogenetic basis with a correlation study of prognosis. Lippincott, Philadelphia. 1926.|
|3. Kernohan JW, Mabon RF, Svien HJ AA. A simplified classification of the gliomas. Symposium on a new simplified concept of gliomas. Mayo Clin Proc 1949;71-75.|
|4. Daumas-Duport C, Scheithauer B, O’Fallon J, Kelly P. Grading of astrocytomas. A simple and reproducible method. Cancer. 1988 Nov 15;62(10):2152-65.
|5. Louis DN. The next step in brain tumor classification: “Let us now praise famous men”… or molecules? Acta Neuropathol. 2012 Dec 15;124(6):761-2.
|6. Louis DN, Ohgaki H, Wiestler OD CW. World Health Organization Histological Classification of Tumours of the Central Nervous System. International Agency for Research on Cancer, France. 2016.|
|7. Castel D, Philippe C, Calmon R, Le Dret L, Truffaux N, Boddaert N, et al. Histone H3F3A and HIST1H3B K27M mutations define two subgroups of diffuse intrinsic pontine gliomas with different prognosis and phenotypes. Acta Neuropathol. 2015 Dec 23;130(6):815-27.
|8. Rekhi B, Kosemehmetoglu K, Tezel GG, Dervisoglu S. Clinicopathologic features and immunohistochemical spectrum of 11 cases of epithelioid malignant peripheral nerve sheath tumors, including INI1/SMARCB1 results and BRAF V600E analysis. APMIS [Internet]. 2017 Apr 27.
|9. Korshunov A, Capper D, Reuss D, Schrimpf D, Ryzhova M, Hovestadt V, et al. Histologically distinct neuroepithelial tumors with histone 3 G34 mutation are molecularly similar and comprise a single nosologic entity. Acta Neuropathol. 2016 Jan 19;131(1):137-46.
|10. Ellison DW, Dalton J, Kocak M, Nicholson SL, Fraga C, Neale G, et al. Medulloblastoma: clinicopathological correlates of SHH, WNT, and non-SHH/ WNT molecular subgroups. Acta Neuropathol. 2011 Mar;121(3):381-96.
|11. Banan R, Hartmann C. The new WHO 2016 classification of brain tumors-what neurosurgeons need to know. Acta Neurochir (Wien). 2017;1-16.
|12. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK ED, Figarella-Branger D, Reifenberger G von DA. WHO classification and grading of tumours of the central nervous system. IARC Press Int Agency Res Cancer, Lyon. 2016.|
|13. Witt H, Mack SC, Ryzhova M, Bender S, Sill M, Isserlin R, et al. Delineation of two clinically and molecularly distinct subgroups of posterior fossa ependymoma. Cancer Cell. 2011 Aug 16;20(2):143-57.
|14. Parker M, Mohankumar KM, Punchihewa C, Weinlich R, Dalton JD, Li Y, et al. C11orf95-RELA fusions drive oncogenic NF-κB signalling in ependymoma. Nature. 2014 Feb 27;506(7489):451-5.