la nuova classificazione who dei tumori cerebrali: aspetti ...€¦ · classifications of cns...

La nuova classificazione WHO dei tumori cerebrali: aspetti istopatologici e molecolari

Department of (Neuro)Pathology, AMC, University of Amsterdam

Eleonora Aronica

Outline

Classifications of CNS Tumors

WHO classifications 1979 – 2016

Old concept of CNS tumor diagnoses

Novel approach of CNS tumor diagnoses in the

molecular era

Genetically defined entities….

Future challenges

Classification of CNS Tumors

• Individual patient care (estimating prognosis and guiding therapy)

• Conduct and interpretation of clinica trials

• Analysis and understanding of basic scientific experiments

• Elucidation of population-based disease trends that may implicate environmental or other etiologies

• Allocation of resources by governments and health insurers tosupport health care

Accurate classification

Brain Pathology ISSN 1015-6305

Periodic revisions of tumor classifications therefore have diverse and important effects on many aspects of individual and population health

the care of individual patientsthe care of individual patients

Technical issues and challengesClassification of CNS Tumors

Periodic revisions

Balance between incorporating the latest molecular findings and the practical issues of clinical diagnosis and current patient management

Incorporate the latest molecular findings

Utilize the most accurate techniques

Do not disrupt the current clinicaldiagnosis and patient management

Availability and cost of the new diagnostic techniques

In 1865, Virchow was the first to describe gliomas

pathologically

Classifications of CNS Tumors

Rudolf Virchow (1821 –1902)

In 1926, Bailey and Cushing introduced the

concepts of histogenesis and brain tumor grading

1979 the first WHO classification of CNS tumors

was published

The classification cannot solve all the unresolved

problems of interpretation…

It must be taken into account that tumours very often

consist of a mixture of cells…Therefore, when

classification is possible only with difficulty, some

mixed groups are foreseen.

The process of malignant dedifferentiation is

accounted for, in all groups where such changes

occur, by the introduction of a higher grade which is

OutlineWHO classifications of CNS Tumors 1979 – 2007

WHO classifications of CNS Tumors 2000-2007

Increasing evidence supports the genetic basis of tumorigenesis

In 2007 CNS WHO the diagnostic categories were established by conventional

histology, genetic alterations are used to provide prognostic or predictive data

• FISH

• 1p/19q

• BRAF:KIAA1549

• (EGFR) • DNA

• IDH1/2

• H3F3A K27M & G34R/V

• BRAF V600E

• MGMT

• TERT

• BRAF:KIAA1549

Molecular Diagnostics

12.10.2016

Euro-CNS Amsterdam

Discussion and challenges 2007-2016

Tools

• Molecular Antibodies• IDH1 R132H• ATRX• H3.3 K27M• BRAF V600E• P53• INI1

High throughput DNA

High throughput epiDNA

450K methylation analysis

NGS• Panel-Sequencing

• Exom-SequencingHigh throughput RNA

RNAseq

Integrated diagnosis

Multiple “tiers”

Integration of genotype, phenotype and outcome data

Patient-specific therapeutic approach

IDH-1

WHO 2016

• Formulating concept of how CNS tumor diagnoses are structured in the molecular era

• Major restructuring of diffuse gliomas, with incorporation of genetically defined entities

• Incorporation of a genetically defined ependymoma variant

• Major restructuring of medulloblastomas, with incorporation of genetically definedentities

• Major restructuring of other embryonal tumors, with incorporation of genetically defined entities and removal of the term “primitive neuroectodermal tumor”

• Novel approach distinguishing pediatric look-alikes, including designation of novel, genetically defined entity

“integrated” phenotypic and genotypic parameters for CNS tumor classification

WHO 2016: nomenclature, new entitiesMajor restructuring of diffuse gliomas, with incorporation of genetically defined entities

WHO 2016

“integrated” phenotypic and genotypic parameters for CNS tumor classification

Histology and molecular genetic features: simplified algorithm

Acta Neuropathol (2016) 131:803–820

WHO 2016

“integrated” phenotypic and genotypic parameters for CNS tumor classification

Possibility of discordant results:

1. Diffuse glioma with astroglial features:

IDH mutation and 1p/19q codeletion: oligodendroglioma IDH-mutant

2. Diffuse glioma with oligodendroglial features

IDH, ATRX and TP53 mutations: diffuse astrocytoma, IDH-mutant

Histology and molecular genetic features

Diffuse gliomas

The diagnostic use of both histology and molecular genetic features also raises the possibility of discordant results

WHO 2016

“integrated” phenotypic and genotypic parameters for CNS tumor classification

Histology and molecular genetic features

Diffuse gliomas

WHO grade determinations are still made on the basis of histologic criteria

Casus 2008

- 29 years, female

- Headache, vision reduction..

- MRI: Frontal (right)

T2 FLAIRLahl et al., 2003

T1 +ctrT2

2008: oligodendroglioma III

Revision 2017

2017: progression

Second resection and revision

T2

FLAIR

Lahl et al., 2003

PCR NGS, neuro-oncologie panel: mutation in exon 4 of IDH1 gene (c.395G>A, p.R132H)

Neuro-PA Genen

AKT1 FUBP1 NPRL3

AKT3 H3F3A PIK3CA

ATRX HIST1H3B PIK3R1

BRAF HIST1H3C PIK3R2

CDK6 IDH1 PTCH1

CIC IDH2 PTEN

CTNNB1 KDM6A SMARCA4

DDX3X MDM2 SMARCB1

DEPDC5 MTOR SMO

EGFR MYB SUFU

EZH2 MYBL1 TP53

FGFR1 NPRL2

No 1p/19q codeletion

Anaplastic astrocytoma IDH-mutant

• CpG methylation of over 850.000 individual CpG

• more than 99% of RefSeq‐annotated genes covered

• Material: 100-500ng DNA from formalin fixed paraffin embedded (FFPE)

tissue

• approximately 5 x 10µm slides (tumor area >1cm)

Illumina Infinium Human Methylation 850k Array

Integrated diagnosis

DNA methylation-based classification

Integrated diagnosis

DNA methylation-based classification

Online classifier (www.molecularneuropathology.org)

Machine learning algorithms are capable of reliably and robustly differentiating between 82 CNS tumour DNA methylation classes (represented by ~2800 reference samples)

• CpG methylation of over 850.000 individual CpG

• more than 99% of RefSeq‐annotated genes covered

• Material: 100-500ng DNA from formalin fixed paraffin embedded (FFPE)

tissue

• approximately 5 x 10µm slides (tumor area >1cm)

Illumina Infinium Human Methylation 450 or 850k Array

Integrated diagnosis

DNA methylation-based classification

Human Methylation 850k Array

Integrated diagnosis

Anaplastic astrocytomaIDH-mutant

DNA methylation-based classification of human central nervous system tumorsCapper et al., in preparation

Online classifier tool (www.molecularneuropathology.org)

Machine learning algorithms are capable of reliably and robustly differentiating between 82 CNS tumour DNA methylation classes (represented by ~2800 reference samples).

Diffuse glioma IDH-mutant WHO II

Casus 2018

IDH1= negative

FLAIR

PCR NGS, neuro-oncology panel: mutation in exon 4 of IDH2 gene (c.515G>T, p.R172M) mutation in the promotor region of TERT gene (c.-124C>T).

-25 years, male

- Epilepsy

- MRI: fronto-parietal tumor

GFAP

1p/19q codeletion

Diffuse oligodendrogliomaIDH-mutant WHO II

WHO 2016: nomenclature, new entitiesMajor restructuring of diffuse gliomas, with incorporation of genetically defined entities

Casus 2017

IDH1= negative

FLAIR

PCR NGS, neuro-oncologie panel: hotspotmutation in exon 2 van het H3F3A gen gevonden (c.83A>T, p.K28M).

H3K27-M= positive

Glioblastoma (IDH -wildtype, WHO IV)

-23 years, male

- Headache, vomiting....

- MRI: thalamus tumor

Diffuse midline glioma H3 K27-mutant

LEAT Task Force 2012-2016…..

Brain tumors associated with epilepsy

G

no unequivocalclusters DN… ? GNT

dGNT

dGNT

GG GG/DNT ?

GG GNT-NOS

GNT ..do not meet strict definitions

GG

IT’S A MESS

GG DNTGG/DNT

2015

• Low-grade glio-neuronal tumors associated with early seizure onset (LEAT) are

difficult to classify at the microscopical level (only 40% interrater agreement).

• Integration of molecular data was helpful to increase the diagnostic yield (79 %)

• Difficult-to-classify glio-neuronal tumors with diffuse growth patterns separated in

only two molecular classes

LEAT Task Force 2012-2018…..

Roadmap to a better understanding of LEAT

A proposal towards international consensus and integrated genotype-phenotype classification of LEAT is in progress

2016 WHO classification

Substantial step forward ?

Change in our current diagnostic work-up of brain tumors:

molecular parameters are used to establish brain tumor diagnoses

More precisely defined entities will facilitate clinical, experimental

and epidemiological studies

Intermediate stage: requiring further

incorporation of molecular data

Computational Pathology

Digital Pathology

Computational Pathology

James