1 of 51

What clinicians need to know about the WHO classification 2021

H.K. Ng, Department of Pathology

Chinese University of Hong Kong

Full ppt at http://www.acp.cuhk.edu.hk/hkng/

2 of 51

Disclaimers

  • I am not the driving pathologist of the Expert Panel of 12 (10 pathologists)
  • The actual book is written by a large number of experts
  • I am not a taxonomy person
  • I sometimes see different viewpoints from the Classification
  • Appreciate most of you treat adult patients, so will be brief with pediatric, which is actually the more complex group

3 of 51

WHO is slow…….. We are probably already out of date

4 of 51

Not happy faces !!

A few not fully alert

Consensus meetings of the Twelve, August 2020 – so we are definitely out of date already

5 of 51

Some general changes

  • Pediatric and adult gliomas separate
  • Grading changed to Arabic numerals, e.g. 1, 2,3, 4 and not I, II, III,IV
  • Grading is WITHIN each tumor group
  • NOS – not otherwise specified; NEC – not elsewhere classified
  • Methylomes a desirable criteria especially in pediatric tumors

6 of 51

Adult diffuse gliomas

7 of 51

Should the diagnosis of

glioblastoma just be

HISTOLOGICAL ?

Endothelial proliferation

necrosis

8 of 51

9 of 51

TERT promoter mutations

10 of 51

From Wesseling and Verhaak

Life history of a glioblastoma

TERT is a common end point

11 of 51

CDKN2A/B homozygous deletion (FISH or methylation) : major prognosticator in IDH mutant astrocytomas

Criteria for molecular Astrocytoma Grade IV

Key references : Shirahata, von Deimling, ANP 2019; Yang R, Ng HK. Brain Pathology 2020

CIMPACT-NOW Brat et al. 2020

Li, Ng.

NOA 2019

12 of 51

Adult diffuse gliomas – WHO 2021

  • Astrocytoma, IDH mutant

(Grades 2-4, “IDH mut glioblastoma” discarded;

homozygous deletion of CDKN2A/B as Grade 4 criteria for cases not

fulfilling histology criteria)

  • Oligodendroglioma, IDH mutant and 1p19q codeleted
  • Glioblastoma, IDH wild type

(EGFR, TERT, +7/-10 for cases not fulfilling histology criteria)

13 of 51

But that does NOT mean :

  • You need to do EGFR or TERT or 7+/10- for an obvious glioblastoma which is IDHwt

  • You need to do CDKN2A/B to diagnose an obvious glioblastoma which is IDH mutant

  • These are criteria for molecular glioblastoma or Grade 4 for the Grade 2-3 lesions which do not fulfil the histological criteria

14 of 51

So don’t order molecular tests just by ticking the boxes

Not all molecular tests are necessary for diagnosis. Check

with your pathologist

Not all glioblastomas fulfill the listed molecular criteria as

of now

We are talking about diagnosis, and not about search for

potential targets etc.

15 of 51

In a cohort of adult IDH mutant glioblastomas, only 55% showed CDKN2A/B deletion. Wong, Ng. Modern Pathology 2021

In another cohort of IDHwt TERTwt glioblastomas, 65% do not exhibit EGFR amplification or +7/-10. Liu, Ng, unpublished data

16 of 51

17 of 51

Other information obtainable from methylomes

  • MGMT
  • 1p19q status
  • G-CIMP status (have to work

Out yourself)

18 of 51

Methodology

  • WHO will not stipulate
  • TERT Sanger sequencing (technically demanding)
  • MGMT – methylation specific PCR or pyrosequencing
  • EGFR amplification – FISH is best because it is histology based. In my view, NGS less accurate
  • +7/-10 – In my view, FISH is best. Methylation profiling is also good. NGS less accurate

19 of 51

Low grade adult IDHwt astrocytoma ?

If not fulfilling criteria, regarded as NEC

(Not Elsewhere Classified)

Are all histologically lower grade IDHwt gliomas just

Glioblastomas ?

Are molecular glioblastomas (IDHwt or IDH mut) really

have the same prognosis as histologically defined

glioblastoma ?

20 of 51

Shirahata, von Deimling et al.

Acta Neuropathologica 2018

Showed a worsened survival

for IDH mutant Grade III

gliomas with homozygous deletion

of CDKN2A/B and when compared

with IDH mutant glioblastoma

21 of 51

IDH mutant LGG – Yang, Ng. Brain Pathology 2020

22 of 51

Survival Analysis of IDH-mutant lower-grade astrocytomas – combinations of CDKN2A, CDK4 and PDGFRA

Low risk group: Non-alterations

Intermediate risk group: CDKN2A deletion and CDK4 amplification

High risk group: PDGFRA amplification

Low

(n=73)

High

(n=27)

p<0.001

A

Intermediate

(n=49)

Probability

100

80

60

40

20

0

60

0

100

80

40

20

Progression-free survival (months)

Low

(n=74)

High

(n=30)

p<0.0001

B

Intermediate

(n=51)

Probability

Overall survival (months)

100

80

60

40

20

0

60

0

100

80

40

20

Yang, Ng. Brain Pathology 2020

23 of 51

OS (years)

Survival %

Molecularly lower-grade (n=68)

Molecularly high grade (n=51)

p <0.001

IDH wild-type lower-grade gliomas have good or bad survivals depending on molecular features

Aibaidula, Ng. Neuro-oncology 2017

24 of 51

The triple negative lower grade gliomas

(IDH, TERT, 1p19q) do not

behave like glioblastomas in spite of being

IDHwt

25 of 51

Among Low Grade Gliomas which are IDH-, 1p19q non-deleted, TERT- (n=80)

Chan, Ng. New England Journal of Medicine 2016

26 of 51

26

27 of 51

27

Aoki, Natsume

Neuro-oncology 2018

Similar message : some IDHwt LGG have better and some have worse prognosis depending on molecular

features – not all are IDHwt glioblastomas

28 of 51

Berzero, Sanson. NO 2021

- Grading is still important

In IDHwt LGG

-TERTmut only IDHwt LGG has

better prognosis

29 of 51

WHO 2021 – separate out the pediatric-type gliomas

2.1.2:     Pediatric-type diffuse low-grade gliomas�                                2.1.2.1: Diffuse astrocytoma, MYB- or MYBL1-altered�                                2.1.2.2: Angiocentric glioma �                                2.1.2.3: Diffuse low-grade glioma, MAPK-altered�                                2.1.2.4: Polymorphous low-grade neuroepithelial tumour of the young�                      �2.1.3:     Pediatric-type diffuse high-grade gliomas�                                2.1.3.1: Diffuse midline glioma, H3 K27M-mutant�                                2.1.3.2: Diffuse hemispheric glioma, H3 G34-mutant�                                2.1.3.3: Diffuse paediatric-type high-grade glioma, H3-wildtype and IDH-wildtype�                               2.1.3.4: Infant-type hemispheric glioma, H3-wildtype�2.1.4:     Circumscribed astrocytic gliomas�                                2.1.4.1: Pilocytic astrocytoma�                                2.1.4.2: High-grade astrocytoma with piloid features�                                2.1.4.3: Pleomorphic xanthoastrocytoma�                 2.1.4.6: Subependymal giant cell astrocytoma�                                2.1.4.7: Chordoid glioma�   2.1.4.8: Astroblastoma-MN1�2.1.5:     Glioneuronal and neuronal tumours                                    �                2.1.5.1: Ganglioglioma�                2.1.5.2: Desmoplastic infantile astrocytoma / ganglioglioma �                2.1.5.3: Dysembryoplastic neuroepithelial tumour             �                2.1.5.4: Diffuse glioneuronal tumor with oligodendroglioma-like features and nuclear clusters �                2.1.5.5: Papillary glioneuronal tumour

Many others………………………………

30 of 51

Bandopadhayay P Pediatric Blood & Cancer 2014

N=4,400

Grades 1 and 2 gliomas in children have similar prognosis

31 of 51

Grades 3 and 4 gliomas in children have similar prognosis in children

32 of 51

Whole genome methylation profiling – more useful in pediatric tumors

Binary

IDAT files

(1 for red,

1 for green)

33 of 51

Typical report from DKFZ Molecular Classifier

34 of 51

Capper D et al. Nature 2018

Methylation profiling listed as

DESIRABLE criteria in WHO 2021

for many entities

“Histological diagnosis without

Histology”

35 of 51

WHO 2021 – separate out the pediatric-type gliomas

2.1.2:     Pediatric-type diffuse low-grade gliomas�                                2.1.2.1: Diffuse astrocytoma, MYB- or MYBL1-altered�                                2.1.2.2: Angiocentric glioma �                                2.1.2.3: Diffuse low-grade glioma, MAPK-altered�                                2.1.2.4: Polymorphous low-grade neuroepithelial tumour of the young�                      �2.1.3:     Pediatric-type diffuse high-grade gliomas�                                2.1.3.1: Diffuse midline glioma, H3 K27M-mutant�                                2.1.3.2: Diffuse hemispheric glioma, H3 G34-mutant�                                2.1.3.3: Diffuse paediatric-type high-grade glioma, H3-wildtype and IDH-wildtype�                               2.1.3.4: Infant-type hemispheric glioma, H3-wildtype�2.1.4:     Circumscribed astrocytic gliomas�                                2.1.4.1: Pilocytic astrocytoma�                                2.1.4.2: High-grade astrocytoma with piloid features�                                2.1.4.3: Pleomorphic xanthoastrocytoma�                 2.1.4.6: Subependymal giant cell astrocytoma�                                2.1.4.7: Chordoid glioma�   2.1.4.8: Astroblastoma-MN1�2.1.5:     Glioneuronal and neuronal tumours                                    �                2.1.5.1: Ganglioglioma�                2.1.5.2: Desmoplastic infantile astrocytoma / ganglioglioma �                2.1.5.3: Dysembryoplastic neuroepithelial tumour             �                2.1.5.4: Diffuse glioneuronal tumor with oligodendroglioma-like features and nuclear clusters �                2.1.5.5: Papillary glioneuronal tumour

Many others………………………………

36 of 51

H3K27M

  • Midline H3K27M mutant gliomas (DIPG)
  • Pitfalls : may occur outside midline; usually poor prognosis

37 of 51

6 years old female

Thalamic GBM

K27M-H3.3 mutations

(AAG → ATG, lysine → methionine)

38 of 51

“DIPG” – H3K27M mutant mid-line glioma

39 of 51

Cynthia Hawkins, AANP

40 of 51

Cynthia Hawkins, AANP web

Pediatric high grade gliomas IDHwt, H3wt are still poorly characterized

41 of 51

2019

Infant high grade gliomas comprise multiple subgroups

Characterised by novel target fusions and better survivals

Clarke M, Mackay A…….Ng HK……Jones C

Cancer Discovery 2020

Infantile gliomas are a separate group in WHO 2021

Note : NTRK inhibitors in clinical trials

42 of 51

2.1.4:     Circumscribed astrocytic gliomas�                                2.1.4.1: Pilocytic astrocytoma�                                2.1.4.2: High-grade astrocytoma with piloid features�                                2.1.4.3: Pleomorphic xanthoastrocytoma�                 2.1.4.6: Subependymal giant cell astrocytoma�                                2.1.4.7: Chordoid glioma�   2.1.4.8: Astroblastoma-MN1

43 of 51

pilocytic

Prototype circumscribed glioma : pilocytic astrocytoma

44 of 51

BRAF Gene Rearrangement (Fusion) - a key test that your lab should have – adult patients too

Two normal signals (orange) plus a smaller third signal near one of the large signals

Other methodology : Nanostring, RNAseq, NGS

45 of 51

Ependymoma

  • Supratentorial ependymoma
  • Supratentorial ependymoma ZFTA (RELA) fusion-positive
  • Supratentorial ependymoma YAP1 fusion-positive
  • Posterior fossa ependymoma
  • Posterior fossa ependymoma Group PFA
  • Posterior fossa ependymoma Group PFB
  • Spinal ependymoma
  • Spinal ependymoma, MYCN-amplified
  • Myxopapillary ependymoma
  • Subependymoma

46 of 51

Medulloblastoma

`

Medulloblastoma

47 of 51

Medulloblastoma, molecularly defined

  • Medulloblastoma, Wnt activated
  • Medulloblastoma, SHH activated and p53 wt
  • Medulloblastoma, SHH activated and p53 mt
  • Medulloblastoma, non-Wnt, non-SHH

  • Medulloblastoma, histologically defined

48 of 51

McLendon, Miller, Ng.

Hematology Oncology

Clinics 2021

As per WHO 2021

49 of 51

M/5. Anaplastic medulloblastoma with drop metastasis and MYC amplification

50 of 51

Molecular grouping of medulloblastoma

  • IHC (cannot distinguish Group 3 and Group 4)
  • Nanostring transcriptomes
  • Methylation profiling
  • ?? NGS

  • In addition
  • FISH for example for MYC or MYCN or other cytogenetics required by your clinicians if not doing methylation profiling

51 of 51

Robin O

Barnard

Peter Burger

F Stephen Vogel