Part one Stromal tumours David G. Huntsman BC Cancer Agency - - PowerPoint PPT Presentation

Rare tumours: some recent data and ideas Part one – Stromal tumours

David G. Huntsman

BC Cancer Agency Vancouver General Hospital University of British Columbia Canada Research Chair in Molecular and Genomic Pathology

COI –disclaimer

• Relationships with commercial entities

– Contextual Genomics – founder and CMO – Research support received from - Sanofi, Pfizer, Takeda, Novartis, GSK

• COI-Mitigating strategies

– Generic drug names only used – No products or other work from Contextual Genomics to be discussed

Cell context and mutation: The evil twins of cancer

Partners in crime from the origin of cancers through to the development of acquired resistance to targeted therapy

Outline

• GCT –time for a rethink
• SCCOHT – new treatment approaches

imminent

FOXL2 mutation 4 granulosa cell tumors

• f the ovary

Shah et al (2009) NEJM

V O A 4 9 V O A 1 5 9 V O A 1 1 9 V O A 2 9 2

FOXL2 IHC and mutational analysis as a standard diagnostic (Kommoss et al Mod Path2013)

Confirmation of FOXL2 aGCT specific c.402C>G mutation by Sanger Sequencing

TaqMan based digital mutation assay for FOXL2 aGCT specific c.402C>G mutation

7 Missense Mutations Truncating Mutations *No mutations at C134W

Cross-cancer mutation summary for FOXL2 (147 studies ) None of the cancers included in c-bioportal had the C134W mutation as none were GCT Question : what does a cancer specific mutation mean to patients

Anniina Färkkilä Mikko Anttonen Markku Heikinheimo Leila Unkila-Kallio Hugo Horlings Hannah van Meurs Maaike Bleeker Melissa McConechy Winnie Yang Blake Gilks Aline Talhouk Stefan Kommoss Sarah Brucker

3 European Centers (n =336) Finland, The Netherlands & Germany

McConechy, Färkkilä, Horlings et al JNCI 2016

• Patient’s disease

relapsed after 2.6 years and died of disease 3 years from Diagnosis

• She had been

misdiagnosed as GCT – personal

• pportunity cost

Overall survival of women with molecularly define AGCT is not distinct from population based controls

Why is the recurrent FOXL2 mutation so specific for adult type GCT

In adults expression is restricted to specialized stroma of ovary, uterus and fallopian tube Image from protein atlas

FOXL2 in Ovarian and Mullerian Stroma

Ovary Endometrium Fallopian tube Endocervix

The FOXL2 mutation is diagnostic for aGCT and its oncogenicity linked to granulosa cell pathophysiology

Granulosa cell Theca cell Oocyte

Corpus luteum Follicle

Luteinized granulosa cell FOXL2-mutant granulosa cell

Ovulation

AGCT

Other targets What else is going on in GCT genomes?

• TERT promoter mutations in 1/3
• Smattering of other cancer

mutations including targetable PIK3CA pathway mutations

• LOH of chromosome 22

Molecularly defined AGCT The power of a correct diagnosis

• Prior studies overestimated death from disease due to contaminating

misdiagnosed cases.

• Missed diagnosed cases dominate early relapses and likely past trials

studied in clinical

• Usually an indolent disease or managed well by primary and

secondary surgeries

• Treatment needed for inoperable cases of molecularly defined AGCT
• Medium time to relapse >7 years therefore current follow-up strategy

likely useless – needs testing

• Cell free DNA could be used as an adjunct to monitor patients in trials
• If no means of targeting FOXL2 derived then

analysis for other targetable mutations through inclusion in a basket type trial may be feasible

• Described by Scully

1979

• A rare disease in

young women

• Sometimes familial
• Undifferentiated

small cells

• Elevated serum

calcium level (60%)

• Very lethal 35% 2yr

survival

Small cell carcinoma of the ovary, hypercalcemic type (SCCOHT)

• Rare and rapid growth, but poorly

differentiated tumor with small rounded cells and variable numbers

• f larger cells, 60% with

hypercalcemia

• Sporadic and familial
• Mean age at diagnosis: 24 yrs (range

14 months – 43 yrs.)

• Cell of origin is unclear
• Surgical debulking followed by

aggressive chemotherapy/radiotherapy combination

Small Cell Carcinoma of the Ovary, Hypercalcemic type (SCCOHT)

Estel et al. Arch Gynecol Obstet 2011

• Highly aggressive tumor with 2-

year survival less than 35%

SMARCA4 inactivating mutation- the only recurrent somatic mutation in SCCOHT

Missense mutation 1 1,647 a.a. Germline Somatic Splice site mutation Nonsense mutation QLQ Bromo HELICc DEXDc SNF2_N BRK HSA SMARCA4 Kupryjanczyk et al. Polish J Pathol 2013. Ramos, Karnezis et al. Nature Genetics 2014. Witkowski et al. Nature Genetics 2014. Jelinic et al. Nature Genetics 2014 Ramos et al. Rare diseases 2014 Karnezis et al J Path 2016

• David Huntsman
• Anthony Karnezis
• Shary Chen
• Winnie Yang
• Sarah Maines-Bandiera
• Christine Chow
• Clara Salamanca
• Ashley Wang
• Michelle Woo
• Michael Anglesio
• Dawn Cochrane
• Friedrich Kommoss
• Jeff Trent
• Holly Yin
• Will Hendricks
• Jessica Lang
• Pilar Ramos
• David Craig
• Alex Sekulic
• Patrick Pirrotte
• Jeff Kiefer
• Bernard Weissman
• Krystal Orlando
• Shujie Song
• BC cancer agency
• Marcel Bally
• Nancy Dos Santos
• Nicole Wretham
• Dana Masin
• Hong Yan
• Ada Leung
• Gregg Morin
• Shane Colborne

Re-expression of SMARCA4 induces neuronal differentiation in SCCOHT cells

BIN67 SCCOHT1 GFP SMARCA4 20 40 60 80 100 24 48 72 96 120 144 168 % cell confluence Time (hours)

pLDpuro-GFP pLDpuro-SMARCA4

MAP2

24

Cross-cancer mutation summary for SMARCA4 (147 studies )

Missense Mutations Truncating Mutations In-frame Mutations

SCCOHT can be familial: associated with germline SMARCA4 mutations in 40% of cases

Witkowski, Nature Genetics 2014 and Gyn Onc 2016

SWI/SNF chromatin remodeling complex

• Complex composition:

Mutually exclusive ATPases: SMARCA4/A2 Core subunits: BAF155, BAF170 and SNF5 Accessory subunits: i.e. ARID1A/B, BAF180, etc

Wilson BG and Roberts CW. Nature Reviews Cancer 2011

• Open chromatin structure for transcriptional

regulation of gene expression

• Regulate many biologic pathways, i.e. cell

cycle control, cell death, cell differentiation

SMARCA4/2 SMARCA4/2 SMARCA4

• Frequently mutated in cancer
• SMARCA4-deficient lung cancer cells depend on

SMARCA2 activity for survival

SWI/SNF ATP-Dependent Chromatin Remodeling Complex

Figure modified from Riccio, Nat Neurosci 2010

BAF47/INI1 ARID1A

SCCO-012 SCCO-002 SCCO-014 SCCO-015 PDX-040 PDX-065 BIN67 SCCOHT1 ACTB BCL11B DPF3 SMARCA2 PHF10 SMARCD2 DPF2 SMARCA4 BCL11A BRD9 BRD7 ACTL6A ARID1A SMARCB1 ARID1B PBRM1 SMARCC1 BCL7B DPF1 SMARCD1 SMARCE1 ACTL6B SS18 SMARCC2 SS18L1 ARID2 SMARCD3

3 2 1

• 3
• 2
• 1

Log2 ra o

Dual loss of SMARCA4/A2 in SCCOHT

1 2 3 4 Relative mRNA level

SMARCA2 SMARCA4

SMARCA4 SMARCA2 Vinculin BIN67 SCCOHT1 SVOG3e KGN

B

Karnezis, Wang, Ramos et al. J Path 2015

This cancer is defined by two features that would be lethal In most other cell types

Model of SCCOHT Pathogenesis Why is it so rare?

SCCOHT BRM+ cell BRM- cell ubiquitous rare synthetic lethality SMARCA4 mutation Immature teratoma Normal

• varian cells

Working model and Hypothesis

Progenitor cells

Neuronal differentiation Stalled differentiation

• ncogenic transformation

X

EZH2 suppression?

• Histone methylatransferase activity (EZH2)
• Primarily trimethylates histone H3K27
• Silence gene
• Maintain stem cells
• Upregulated by SMARCB1 depletion
• Required for SMARCB1-induced tumorigenesis

EZH2 is highly expressed in SCCOHT

Strong diffused staining Variable staining

19/24 (79%) 5/24 (21%)

2 4 6 8 10 12

EZH2 EZH1 EED SUZ12 Relative gene expression Normal ovary SCCOHT

EZH2 is downregulated in SMARCA4- reexpressing SCCOHT cells

H3K27Me3 Vinculin EZH2 SMARCA4 SMARCA4 - + - + BIN67 SCCOHT1

Total H3

Depletion of EZH2 suppresses the growth of SCCOHT cells

EZH2

shctrl shEZH2_1 shEZH2_2

H3K27Me3 H3

BIN67

20 40 60 80 100 120 140

BIN67 COV434 SCCOHT1 ES-2 % Growth shCtrl shEZH2_1 shEZH2_2

SCCOHT cells

EZH2 inhibitors suppress SCCOHT cell growth in vitro

A B

• 2
• 1

1 50 100

Log(conc)

% Survival

EPZ-6438

BIN67 SCCOHT1 COV434 G401 ES-2 RMG1 OVCAR-8 OVISE OVTOKO

• 2
• 1

1 50 100

Log(conc)

% Survival

GSK126

BIN67 SCCOHT1 COV434 G401 ES-2 RMG1 OVCAR-8

C D

0 0.1 0.5 0 0.1 0.5 μM

BIN67 SCCOHT1 H3K27Me3 H3 EPZ-6438 BIN67 SCCOHT1 H3K27Me3 0 1 5 0 0.1 1 μM H3 GSK126

EPZ-6438 induces neuronal differentiation in SCCOHT cells

MAP2 DAPI Merged

DMSO EPZ-6438 day 7 EPZ-6438 day 12 BIN67 cells

EPZ-6438 suppressed SCCOHT xenograft tumor growth in mouse

200 400 600 800 1000 1200 13 18 23 28 33 38 43 48

Tumor volume Days after cell inoculation vechicle 100mg/Kg EPZ-6438 200mg/Kg EPZ-6438

A

20 40 60 80 100 120 10 20 30 40 50

Percentage Survival Days after cell inoculation

vechicle 100mg/Kg EPZ-6438 200mg/Kg EPZ-6438

B

BIN67 xenograft model BIN67 xenograft model

H3K27Me3 H3

1 2 3 1 2 3

Control EPZ-6438 day 28

BID QD

SCCOHT cells are hypersensitive to HDAC inhibitors

6 day

HDAC inhibitors

SCCOHT lines immortalized granulosa cell line HGS lines

20 40 60 80 100 120 140 10 20 30 40 50 60

% cell viability Drugs (1 uM)

BIN67 SCCOHT1 SVOG3e ES-2 OVCAR4

Epigenetic drug library (1 mM) 6 days Cells 1 μM Fix and stain with crystal violet Dissolve, quantitation and calculation of cell viability change

SCCOHT non-SCCOHT SC 5 10 15 20

IC50 (nM)

Quisinostat

Synergism between EPZ-6438 and Quisinostat in SCCOHT cells

20 40 60 80 100 120 % Viability

BIN67

20 40 60 80 100 120 0.1 0.2 0.3 0.4 0.5 % Survival EPZ-6438 (uM)

BIN67

ctrl 1.5nM Qstat 3nM Qstat 6nM Qstat

Combination of EPZ-6438 and Quisinostat robustly suppressed the growth of SCCOHT xenograft

100 200 300 400 500 600 700 800 900 5 10 15 20 25 30 35 40 45

Tumor volume (mm3) Days post innoculation

BIN67 mouse xenograft model

Vehicle 200mg/kg EPZ-6438 5mg/kg Quisinostat EPZ-6438+Quisinostat

QD

• EZH2 alone or in particular in combination with HDAC

inhibitor strongly suppress the growth of SCCOHT in vitro and in vivo

EZH2/HDAC inhibitors

Progenitor cells Neuronal differentiation

HDAC

X X

Stalled differentiation and

• ncogenic transformation

X

HDAC

SCCOHT -conclusions

• Loss of BRG1 in an ovarian mass is diagnostic
• Cell of origin unknown
• Response to EZH2 and pan HDAC inhibitors seen in-

vitro and in-vivo – both drug tested in man but not in combination

• An international effort will be required to establish

much needed new treatments for this cancer

Thanks

• My lab: Melissa McConechy, Winnie

Yang, Yemin Wang, Tony Karnezis,, Madlen Maassen and Janine Senz

• Local collaborators : Sohrab Shah –

Bioinformatics, Blake Gilks

• GCT Collaborators: Stefan Kommoss, Anniina

Färkkilä, Mikko Anttonen, Markku Heikinheim, Leila Unkila-Kallio, Hugo Horlings, Hannah van Meurs, Stefan Kommoss, Sarah Brucker, Maaike Bleeker

• SCCOHT collaborators: Jeff Trent, William

Foulkes, Bernard Weissman , Pilar Ramos

Is SCCOHT a rhabdoid tumor of the ovary

Fahminiya et al. Oncotarget 2016

H&E

Rhabdoid-like cells in SCCOHT

• Shared clinicopathological features
• Common genetic features (mutation in SWI/SNF

complex)

Similarity between SCCOHT and ATRT, MRT

Small cells Large cells – rhabdoid

Useful to Conceptualize SCCOHT as a Malignant Rhabdoid Tumour of the Ovary (MRTO)? Yes

• Similar morphology

– Tumours with variable numbers of small, large and rhabdoid cells

• Similar genetics

– Mutations & silencing of core SWI/SNF members

• Similar treatments?

– TBD

Should We Rename SCCOHT as MRTO? No

• SCCOHT is a much better name than MRTO based on

clinicopathologic criteria

– Half of tumors composed exclusively of small cells – Large cells are minority of cells, if present – Rhabdoid cells are usually minority of large cells, if present – Most patients have hypercalcemia

• Neither name suggests cell of origin
• Similar but distinct genetics
• Little evidence that SCCOHT and AT/RT or other rhabdoid

tumours respond to similar therapies

• A more useful/effective nomenclature is needed for this

group of tumours that encapsulates cell context, genetics, histology and treatment similarities