WHO 2016 CLASSIFICATION OF CNS TUMORS - Neuro · CNS TUMORS DIAGNOSED IN THE UNITED STATES IN 2009– 2013 Distribution of All Primary Brain and Other CNS Tumors by CBTRUS Histology

WHO 2016 CLASSIFICATION OF CNS TUMORS

BNS 27th of April 2019

Alex MichotteDept of Neurology and Pathology

UZ Brussel

I

Based on the consensus of an International Working Group

Objective: establish a classification and grading of

tumors accepted and used worldwide

1979 1st edition: Zülch KJ. Morphological classification

1993: 2nd ed: Kleihues, Burger, Scheithauer.

Introduction of IHC in diagnostic pathology

2000: 3th ed: Kleihues and Cavenee. Incorporation of

genetics, epidemiology, imaging, prognosis and

predictive factors

2007: 4th ed Louis et al. Addition of new entities,

variants and patterns of differentiation

2016: 4th revised ed Louis et al. Classification of some

tumors based on molecular data

OBJECTIVES AND HISTORY OF WHO

CLASSIFICATION

2

2016 WHO classification CNS tumors BNS 27th of April 2019

I

Slide from P Wesseling, BANO, Brussels, 2 dec 2016

3


I

GLIAL CELL TYPES AND

ASSOCIATED TUMORS OF CNS

Sanai et al. NEJM 2005; 353:811-22

4


I

WHO GRADING OF BRAIN TUMOURS

LOW-GRADE : I AND II

Grade I

Low proliferative

potential (slowly

growing tumors)

Well-circumscribed

Good prognosis

after surgery (>5y or

cure)

Grade II

Low proliferative

potential (slowly

growing tumors,

nuclear atypia)

Diffuse

Tendency to

progress to higher

grades of

malignancy

WHO defines a Grading scheme that is a malignancy scale ranging

across a wide variety of neoplasms rather than a strict histological

grading system

5


I

WHO GRADING OF BRAIN TUMOURS

HIGH-GRADE : III AND IV

Grade III

Focal signs of fast-

growing activity

(anaplasia: incl nuclear

atypia and mitotic

activity) in a pre-existing

low-grade tumour or de

novo

Grade IV

Signs of generalized

anaplasia (mitotically

active, necrose, and/or

endothelial proliferation)

”de novo” or in a pre-

existing low-grade

tumour

rapid pre- and

postoperative disease

evolution and a fatal

outcome

6


I

SURVIVAL OF PATIENTS WITH ASTROCYTIC TUMORS

Ohgaki H, Kleihues P, J Neuropathol Exp Neurol (2005) 64:479-48

7


I

LIMITATIONS OF THE WHO CLASSIFICATION:

Distinction between tumoral cells and surrounding

parenchyme may be difficult

Interobserver variability due to the lack of quantification of

histological criteria

For mixed gliomas (oligoastrocytoma): less an issue with

the 2016 classification

Beween grade II and III: still an issue

Lack of representability of the surgical specimen (Sampling

error!)

Importance of pathological-radiological correlation

Lack of reproductibility

8


I

DIFFUSE GLIOMAS ARE (OFTEN) HETEROGENEOUS

CAVE SAMPLING ERROR!

Low grade diffuse glioma High-grade diffuse glioma

Low grade diffuse glioma = true

Low grade diffuse glioma = False!

Glioblastoma = true

Adapted from P Wesseling 2007

9


I

CBTRUS STATISTICAL REPORT: PRIMARY BRAIN AND OTHER

CNS TUMORS DIAGNOSED IN THE UNITED STATES IN 2009–

2013

Distribution of All Primary Brain and Other CNS Tumors by CBTRUS Histology

Groupings and Histology (N=368,117), CBTRUS Statistical Report: NPCR and

SEER, 2009-2013.

Gliomas account for 25-30% of all

tumors and 80% of malignant tumors

10


I

WHY MOLECULAR MARKERS IN ROUTINE

NEUROPATHOLOGY?

1. More objective/less subjective criteria

2. Use of molecular markers allowing a

1. Better diagnosis (diagnostic marker)

2. Better assessment of prognosis (prognostic marker)

3. Better assessment of response to therapy (predictive

marker)

Prognostic markers separate biological entities

Marker A-

Marker A+ Marker A+

Marker A-

Therapy

Predictive markers indicate response to therapy

11


I

1. Diagnostic entities should be defined as

narrowly as possible to optimize interobserver

reproducibility, clinicopathological predictions and

therapeutic planning

2. Diagnoses should be ‘layered’ with histologic

classification, WHO grade and molecular

information listed below an ‘integrated diagnosis’

12


I

• Integrated Diagnosis (incorporated all aspects of tissue diagnosis)

• Histological Diagnosis

• WHO Grade (histological grade)

• Molecular information

ISN-Haarlem format of “layered diagnoses”

ISN-Haarlem

layered diagnosis format

13


I

4TH REVISED 2016 WHO CLASSIFICATION OF CNS TUMORS

Louis et al. Acta Neuropathol 2016; 131:803-20


I

4TH REVISED 2016 WHO CLASSIFICATION OF CNS TUMORS


15


I



I

Major restructuring of

diffuse gliomas, with incorporation of genetically defined entities

medulloblastomas, with incorporation of genetically defined entities

other embryonal tumors, with incorporation of genetically defined

entities and removal of the term “primitive neuroectodermal tumor”

Addition of newly recognized entities, variants and

patterns

IDH-wildtype and IDH-mutant glioblastoma (entities)

Diffuse midline glioma, H3 K27M–mutant (entity)

Embryonal tumour with multilayered rosettes, C19MC-altered (entity)

Ependymoma, RELA fusion–positive (C11orf95-RELA fusion) (entity)

Diffuse leptomeningeal glioneuronal tumor (entity)

Anaplastic PXA (entity)

Epithelioid glioblastoma (variant)

Glioblastoma with primitive neuronal component (pattern)

17


I

Deletion of former entities, variants and terms

Gliomatosis cerebri

Protoplasmic and fibrillary astrocytoma variants

Cellular ependymoma variant

“Primitive neuroectodermal tumour” terminology

Addition of brain invasion as a criterion for atypical meningioma

Restructuring of solitary fibrous tumor and hemangiopericytoma

(SFT/HPC) as one entity and adapting a grading system

Expansion and clarification of entities included in nerve sheath

tumors, with addition of hybrid nerve sheath tumors

Expansion of entities included in hematopoietic/lymphoid tumors

of the CNS (lymphomas and histiocytic tumors)

18


I

WHO classification 2016WHO classification 2007

19


I

USEFUL MOLECULAR MARKERS IN DIFFUSE GLIOMAS

Required in current classification:

1p-19q codeletion

Isocitrate dehydrogenase (IDH1 and 2)

Histone H3K27M mutation

RELA-fusion

C19MC locus at 19q13.42

Useful but not required in current classification:

P53

ATRX

EGFR amplification

BRAF V600E mutation and BRAF-KIAA1549 fusion mutation

MGMT methylation status

20


I

1P-19Q CODELETION

In 1994 Reifenberger et al. reported for the first time that many

oligodendroglial tumors show loss of heterozygosity (LOH) for

chromosome arms 1p and 19q resulting from a non-balanced

translocation t (1:19) (q10:p10). The loss of the entire arms of 1p

and 19q is clinical relevant and associated with sensitivity to

chemotherapy and improved outcome.

Cairncross et al. in 1998 were the first to point out that recurrent

anaplastic oligodendrogliomas with 1p/19q codeletion were far

more responsive to PCV (procarbazine–ccnu–vincristine)

chemotherapy, with virtually all tumors responding.

In large prospective randomized studies on diffuse glioma,

1p/19q codeletion was associated with improved overall survival

but also with increased benefit of adjuvant PCV chemotherapy

given after radiotherapy (EORTC 26951,RTOG 9402.)

1p-19q codeletion is in 100% of the cases associated with an IDH

mutation

21


I

ANAPLASTIC OLIGODENDROGLIOMA

1P LOSS

Left: normal : 2 red and 2 green

signals/nucleus

Right: deletion 1p36

Red labeled target DNA on chromosome 1p36

Green labeled 1q25 reference probe

22


I

ANAPLASTIC OLIGODENDROGLIOMA

19Q LOSS

Left: normal : 2 red and 2 green

signals/nucleus

Right: deletion 19q13

Red labeled target DNA on chromosome

19q13

Green labeled 19p13 reference probe

2016 WHO classification CNS tumors Brussels 22th of April 2017

23


I

IMPORTANCE OF 1P/19Q CO-DELETION IN GLIOMAS

Cairncross et al. JCO 2013; 31: 337-43

24


I

IMPACT OF MOLECULAR BIOMARKER: 1P-19Q CODELETION

Diagnostic

PredictivePrognostic

Diagnostic biomarker:

→ aids in classifying a

tumor

Prognostic biomarker:

→ informs about naturalhistory of tumor

Predictive biomarker:→ informs about

probability of

response to specific

therapeutic regimen

1p/19q

Adapted from P Wesseling

25


I

Isocitrate dehydrogenase (IDH) gene mutations were reported by Parson et al (Science,

2008) in a subset of younger glioblastoma patients with prolonged survival

The 2 IDH enzymes reversibly oxidize isocitrate to a-ketoglutarate and concurrently

reduce NADP to NADPH.

The IDH1 gene is located on chr 2q33.3, its enzyme localizes to peroxisomes.

The IDH2 gene is on 15q26.1, its enzyme is active in mitochondria.

IDH1-2 Mutations result in transformation of a-ketoglutarate to D-2-hydroxyglutarate

and NADPH to NADP → D-2-hydroxyglutarate promotes cell proliferation.

IDH1 and IDH2 mutations are involved early in gliomagenesis (“driver mutation”), before

TP53 and ATRX mutations in astrocytic tumors, and before 1p/19q codeletion, CIC, and

FUBP1 mutations in oligodendroglial tumors.

The IDH mutation rate is 70-80% of diffuse grade II and III astrocytomas and

oligodendrogliomas and +/- 100% in 1p/19q codeleted tumors

IDH mutations occur mainly in young adults and middle aged patients and are rare in

children and elderly patients (> 55 y)

R132H is the most common IDH1 mutation in gliomas (90%).

Gliomas with mutations of IDH2 most often involve R172 (R172K).

ISOCITRATE DEHYDROGENASE (IDH)

Tanboon et al. JNEN jan 2016

26


I

ISOCITRATE DEHYDROGENASE (IDH)

IDH1 in peroxisomes; IDH2 in mitochondria

IDH1-2 Mutations result in transformation of a-ketoglutarate to D-2-hydroxyglutarate and NADPH to NADP → D-2-hydroxyglutarate promotes cell proliferation.

27


I

IDH1 (CODON 132) AND IDH2 (CODON 172 )

R132H mutation: +/- 90% (can be detected by IHC)

Other mutations (+/- 10%) must be identified by sequencing of the IDH1 and 2 genes (NGS)

Z. Reitman et al. J. Natl Cancer Inst. 2010 Jul 7

28


I

IDH1 AND IDH2 MUTATIONS IN HUMAN GLIOMAS

IDH mutation is a diagnostic biomarker for diffuse gliomas

Yan et al, NEJM, 2009; 360:765-73 Diffuse gliomas

29

No IDH mutin PA


I

SURVIVAL OF ADULT PATIENTS WITH MALIGNANT GLIOMAS

WITH OR WITHOUT IDH GENE MUTATIONS

IDH mutation is a prognostic marker

Yan et al, NEJM, 2009; 360:765-73

30


I

KAPLAN-MEIER SURVIVAL CURVES FOR GLIOMA CASES

CLASSIFIED INTO FIVE WHO 2016 ENTITIES (N = 1206)

IDH status has major

prognostic implications

IDH mutant tumors

without 1p/19q

codeletion have a worse

outcome compared to

IDH mutant, 1p/19q

codeleted grade III

tumors, but better than

grade II and III tumors

without IDH mutations.

Peckmezi et al. Acta Neuropathol 2017; on-line 2 march 2017

31


IDH mut

IDH wt

I

A-THALASSEMIA/MENTAL RETARDATION SYNDROME X-

LINKED (ATRX) GENE

The a-thalassemia/mental retardation syndrome X-linked (ATRX)

gene on Xq21.1 encodes the nuclear protein ATRX, a chromatin

remodeling factor

ATRX works together with histone chaperone death domain-

associated protein (DAXX) to facilitate the incorporation of

histone variant H3.3 into these regions in order to stabilize

chromatin structure and facilitates transcriptional elongation

ATRX helps maintaining telomeric integrity during DNA synthesis.

ATRX mutations are associated with the alternative lengthening of

telomeres (ALT) phenotype, IDH and TP53 mutations, but not with

1p/19q loss

ATRX mutations are thus a marker of astrocytic tumors and may

be associated with a better prognosis (Wiestler et al)

ATRX mutations are detected in 33% of Gr II astrocytomas, 46% of

gr III astrocytomas, 80% of IDH mut GBM and 7 % of IDHwt GBM

(Xiao-Yang 2012)

32


I 33


I

ALGORITHM FOR CLASSIFICATION OF THE DIFFUSE GLIOMAS

BASED ON HISTOLOGICAL AND GENETIC FEATURES


34


I

MGMT (O6-METHYLGUANINE DNA METHYLTRANSFERASE )

MGMT is a DNA repair

enzyme encoded by the

MGMT gene on10q26

and involved in the

repair of the damage

caused by alkylating

agents

Promotor methylation

of MGMT(+/-40% of

GBM) leads to a

decreased expression of

the enzyme → better

response to radio-

chemotherapy and

prolonged survival

Potential use as

predictive biomarker in

the treatment of elderly

GBM patients

Hegi, NEJM, 2005

35


I

IMPACT OF MOLECULAR BIOMARKER

Diagnostic

PredictivePrognostic

Diagnostic biomarker:

→ aids in classifying a

tumor

Prognostic biomarker:

→ informs about naturalhistory of tumor

Predictive biomarker:→ informs about

probability of

response to specific

therapeutic regimen

MGMT

Adapted from P Wesseling

36


I

“INTEGRATED” DIAGNOSTIC APPROACH FOR ADULT ASTROCYTOMA,

OLIGODENDROGLIOMA AND GLIOBLASTOMA

Reuss et al. Acta Neuropathol 2015;129:133-46

Changes from initial to integrated diagnosis in 405 adult patients with supratentorial glioma. Width of bars indicates relative proportions of the initial tumor groups. A astrocytoma, OA oligoastrocytoma, Ooligodendroglioma, GBM glioblastoma, GBMo glioblastoma with oligodendroglial component, GBMs

secondary glioblastoma, gcGBM giant cell glioblastoma, GS gliosarcoma

37


I

Low grade part with calcifications

Anaplastic part: high cell densitywith mitoses

H&E

H&E

Clinical info

M, 54 y-old

First epileptic seizure

Left frontal tumor with focal enhancement

Immunohistochemistry

Olig 2 +

IDH1+ (IDH1 R132H mut)

ATRX: normal expression

P53 < 10%

Ki67 focally 25%

Molecular biology

1p/19q codeletion

HISTOLOGICAL DIAGNOSIS

Anaplastic oligodendroglioma Gr III


I

Anaplastic oligodendroglioma Gr III


Olig2 Ki67

39


I

INTEGRATED DIAGNOSIS WHO 2016

Anaplastic oligodendroglioma, IDH mutant and 1p/19q codeleted

GrIII

IDH1 ATRX

40


I


Anaplastic oligo-astocytoma Gr III

Predominant astrocytic population Predominant oligodendroglial population

H&EH&E

41


I



Clinical info

F, 31 y-old

2014: right frontotemporal

astrocytoma Gr II

Jan 2017 enhancing area

on imaging: malignant

transformation?

Immunohistochemistry

Olig2 +

P53 + (>50%)

IDH1 – (wild type)

ATRX loss of nuclear

expression (mutation)

Ki67 ++

Molecular Biology

No 1p/19q codeletion

Anaplasia: cell density ++; mitoses (arrow); endothelial proliferation

H&E

42


I



olig2 P53

43


I

INTEGRATED DIAGNOSIS WHO 2016

Glioblastoma IDH wild type, ATRX mutated, Gr IV

Prognostic implic: better prognosis than IDHwt-ATRX wt GBM

(Peckmezi et al)

ATRX Ki67

44


I

DIFFUSE MIDLINE GLIOMA H3 K27M MUTANT

45

H3.3 is the major histone involved in chromatin changes and gene

expression profiles and associated with active chromatin and translation

Mutations will interfere with chromatin function, inducing defects in

chromatin remodelling and tumorigenesis.

K27M mutation in histone H3.3 gene (H3F3A) is a frequent event in

pediatric DIPG (+/- 70-80%)

Other Mutations in H3.3 gene (HIST1H3B; G34V/R-H3.3) are rare

Broad spectrum of histopathology (A gr II-III, O gr II-III, GBM)

K27M-H3.3 mutation is associated with a worse overall survival when

compared to patients who are wild-type for H3.3. independent of patient

age and histologic diagnosis

H3 K27M mut midline glioma is therefore a WHO grade IV tumor

irrespective of the underlying pathology

The K27M mutation can be identified by IHC and sequencing

The G34V/R-H3.3 mutation was not found in pontine gliomas whereas it

represents 13 % of H3.3 mutations in supratentorial GBM


I 46


I

DIFFUSE MIDLINE GLIOMA H3 K27M MUTANT

F 7 y old, axial FLAIR M 9y old, axial T2

47


I

F, 7 Y OLD DIFFUSE GLIOMA OF THE PONS K27M MUT

Histopathological features of a glioblastoma WHO gr IV

2016 WHO classification CNS tumors Brussels 22th of April 2017

48


I

F, 7 Y OLD DIFFUSE GLIOMA OF THE PONS K27M MUT

Ki67 +/- 20% Strong nuclear expression of the mutant H3K27M

49


I

BRAF MUTATIONS

In 10 % of pilocytic astrocytoma

(mainly extracerebellar (diencephalic)

location

In 40-60% of ganglioglioma

In 70-80% of Pleomorphic

xanthoastrocytoma (PXA)

In 50% of epithelioid glioblastoma

In 86% of papillary craniopharyngioma

(Larkin et al, Acta Neuropathol 2014)

In 70 % of pilocytic astrocytoma

In 30% of ganglioglioma

In 0 % of PXA and diffuse

astrocytomas

BRAF V600E mutation Fusion BRAF-KIAA1549

50


I

BRAF-KIAA FUSION IN PILOCYTIC ASTROCYTOMA

Collins Acta Neuropathol jun 2015512016 WHO classification CNS tumors BNS 27th of April 2019

I

CHALLENGE: BALANCING DESIRES AND NEEDS

Provide the best possible

diagnosis

Utilize the most

accurate, cuttingedge

techniques

Incorporate the latest

molecular signatures

Do not disrupt current

clinical diagnosis and

patient management

Weigh the availability

and cost of novel

diagnostic techniques

Preserve the ability for

longterm clinical,

experimental and

etiological correlations

Most accurate

diagnosis ‘Stay connected’

52

Adapted from P Wesseling BANO 20162016 WHO classification CNS tumors BNS 27th of April 2019

I

PROBLEMS OF WHO CLASSIFICATION

Intervals between WHO editions is much too long and do not keep

pace with rapid changes in the field of molecular pathology

There is a need for a platform to address gaps in

neuropathological classification and to apply the new findings in

molecular pathology more rapidly in the clinical practice

Platform has been created cIMPACT-NOW: Consortium to inform

molecular and practical approaches to CNS tumor taxonomy-not

official WHO

Members are leading neuropathologists involved in the WHO

2016 classification

This platform provides yearly updates and guidelines between

WHO editions


53

I

CIMPACT-NOW

Diffuse IDH-wt astrocytomas (grade II or III) with molecular profile of a glioblastoma

have the same aggressive behavior as a glioblastoma WHO grade 4.

• Combination of whole chromosome 7 gain and whole chromosome 10 loss (+7/-

10)

• High-level EGFR amplification

• TERT promotor mutation

Assessment by large-scale DNA methylation profiling (microarray) warrants further

study


54

I

THE FUTURE OF (NEURO)PATHOLOGY

SMART SYNTHESIS OF MORPHOLOGICAL AND MOLECULAR INFORMATION!

TCGA Research Network et al.

Nature Genetics 2013;45:1113

Adapted from P Wesseling BANO 2016

55


I

TAKE HOME MESSAGES

In spite of major advances of WHO 2016 for diffuse astrocytic and

oligodendroglial tumors, some groups remain poorly classified

cIMPACT-NOW has published major advances:

Clarification of terminology and entities (H3K27M mut gliomas should be “diffuse”

and “midline”)

The usefulness of ATRX (ATRX mutations in diffuse gliomas are only found in

astrocytomas)

Diffuse astrocytic glioma, IDH-wild type with molecular features of glioblastoma,

WHO grade 4

Large scale DNA methylation techniques are powerfull for

CNS tumor classification

The refinement and discovery of new entities

Remain expensive and not widely used

Histomolecular classification of gliomas is still ongoing


56

THANK YOU

Documents

WHO 2016 CLASSIFICATION OF CNS TUMORS - Neuro · CNS TUMORS DIAGNOSED IN THE UNITED STATES IN 2009– 2013 Distribution of All Primary Brain and Other CNS Tumors by CBTRUS Histology