Phenotypic Diversity in the Evolution of Breast Cancer Metastasis

Transcription

1 Phenotypic Diversity in the Evolution of Breast Cancer Metastasis Peter Simpson, PhD. The University of Queensland, Brisbane, Australia UQ Centre for Clinical Research Molecular Breast Pathology

2 Cancer is a Genomic Disease Stratton Nature 2009

3 Cancer is a Genomic Disease

4 Curtis Nature 2012 TCGA Nature 2012 ~3000 breast tumours mrna expression, DNA copy number, exome seq, mirna expression, DNA methylation, protein profiling. what else is there??

5 Tumour Evolution & Intratumour Heterogeneity

6

7 The Evolution of Cancer Genomes Yates & Campbell Nat. Rev. Genet MRCA Most Recent Common Ancester

8 Molecular Evolution of Breast Cancer Yates & Campbell Nat. Rev. Genet brain normal hyperplasia in situ carcinoma invasive carcinoma bone liver GI metastasis lung Lymph nodes ovaries

9 Molecular Evolution of Breast Cancer HUT CCL ADH Low grade DCIS Tubular IDC ALH LCIS PLCIS ILC High grade DCIS Mixed Ductal-Lobular

10 Intratumour Morphological Heterogeneity E-cadherin Da Silva Am J Surg Path 2008

11 Intratumour Morphological Heterogeneity DCIS ILC E-cadherin b-catenin LCIS 1p-, 3p-, 4q-, 5q+, 8q+, 10p+, 11p- DCIS normal 11q+, 11q-, 16p+ 1q+, 16q- 6q- 8p- Ecad- LCIS 17q+ 22+ ILC DCIS? 22+ IDC

12 Intratumour Heterogeneity in Metastatic Progression WLE XRT Hormonal Therapy IDC NST, grade 2 Ceased Hormonal Therapy Mastectomy IDC NST, grade 2 Subtotal gastrectomy Metastasis in stomach wall ER+, PR+, HER2- ER-, PR-, HER2- ER-, PR-, HER2- E-cadherin E-cadherin E-cadherin P120 catenin

13 Mechanisms Underlying Ductal to Lobular Transition Mixed ductal-lobular carcinomas 3-5% of all breast cancers Areas of ductal and lobular differentiation Either or both component can be capable of disseminating Genomic analysis - components clonally related rather than collision tumours Model system to study mechanism of transition related to dysfunction of E-cadherin aberrant localisation (15/17 (88%) cases) rather than complete loss Not driven by epithelial to mesenchymal transition (EMT) Genomic/epigenetic? Transition zone Ductal Lobular E-cadherin

14 Breast Cancer Autopsy Series A/Prof Margaret Cummings 197 cases from Royal Brisbane Hospital ( ) 945 metastases Clinical data ages; treatment; dates of diagnosis, surgery, post mortem; PM details Morphological, Immunophenotyping & Genomic analyses % of cases with gynaecological metastasis were <50 yrs Association between metastases to bone and brain Association between metastasis to liver following surgery

15 Phenotypic Changes with Metastatic Progression 55 cases of primary tumour and multiple matched metastases IHC for ER, PR, HER2, Ki67, p53, EGFR, c-kit Breast PR Liver Dura

16 CK14 Brain Metastasis

17 Metastasis to Gynaecological Organs Age young women Primary breast tumour occurred at median age of 52yrs Gynae metastases occurred at median age of 57yrs Ovarian metastasis Time to diagnosis median 5yrs (range 0-17yrs) Histological type enriched for ILC IDC 54% (28/52) ILC 43% (23/52) Immunophenotype ER+, HER2-35/37 (95%) primary tumours ER+ 75% primary BC and 92% matched gynae met ER+ (Bigorie 2010 Cancer) All HER2 negative Metastases: 87% ER+, 0% HER2+ ER E-cadherin

18 Metastasis to Gynaecological Organs Metastasis often widespread and extensive Distribution: 21/58 (36%) had metastases only to gynae 37/58 (64%) had other metastases peritoneum/omentum/ascitic fluid (68%), abdomen/gastrointestinal tract (43%) bone (26%) lung (12%), liver (9%) skin (7%) brain (2%) bone liver GI brain lung Lymph nodes peritoneum ovaries GM49: ILC age 36, ER+/PR+/HER2-, gynae mets age 44 ER Right ovary Right Fallopian tube Left ovary Uterus Omentum

19 Genomics of Metastasis Progression Autopsy Case 11 (2007): IDC, grade 3; ER-, PR-, HER2+ Breast Lymph node Lung Adrenal Breast Breast Adrenal Lung

20 Clonal Evolution and Intratumour Heterogeneity in Metastatic Progression Autopsy Case 7 (1989): mixed Ductal-lobular, grade 3; ER-, PR-, HER2+ Breast Breast Lung Lymph node Chromosome 17 Lymph node pm Breast Lung Lung Liver Chromosome 2 Adrenal Breast Lung Chromosome 10

21 Axillary lymph lymph node node Intratumour Heterogeneity - identifying the lethal clone? Non-axillary lymph lymph node node Chromosome 2 Breast Lymph Node Axillary lymph node Lung Lung Lung Liver Liver A Lymph Non-axilla Node lymph node PM B Chromosome 10 Breast Adrenal Lung B B Breast Breast C Breast Breast Liver Liver D Lung Lung Breast C C Breast Breast E E E Breast Breast Adrenal Adrenal F F F E Breast F D D Lung Lung Breast Breast G G G Lung H H H G H Lung Lung I I J J

22 Identifying Driver Genes of the Lethal Clone Autopsy Case 33 (Primary tumour 1966; PM 1969): IDC, grade 3; ER-, PR-, HER2- Chromosome 5 Breast Liver Breast Chromosome 20 Liver

23 Summary Morphological and molecular features of breast cancer quite well characterised Huge molecular data resources now available from consortia such as ICGC, TCGA, METABRIC Intratumour heterogeneity Limitations of single biopsy/single time point analysis Evident throughout tumour development & progression to metastasis Remains a considerable problem for patient management Which is the lethal clone? Which clone to treat? We are investigating intratumour heterogeneity to unravel mechanisms of disease progression Pre-invasive disease Mixed ductal-lobular carcinomas Metastatic progression Autopsy studies end stage, but reveal important clinical and biological insight into the natural history of dissemination Contemporary longitudinal series support findings from autopsy series and effects of treatment selection on lethal clone.

24 Acknowledgements UQ Centre for Clinical Research Molecular Breast Pathology Group UQ Centre for Clinical Research & QLD Health Pathology Prof Sunil Lakhani A/Prof Margaret Cummings Amy Reed, Jodi Saunus, Chanel Smart Leo Da Silva, Jamie Kutasovic, Janani Jayanthan Queensland Institute of Medical Research Georgia Chenvix-Trench Kum Kum Khanna Peter O Rouke, stats team UQ Institute of Molecular Biosciences Sarah Song, Nic Waddell, Sean Grimmond Queensland Centre for Gynaecological Cancer Andreas Obermair, Karen Sanday Brian Fritz, Gabe Kolle Michael Sandery (Illumina) Helen Spiers (Ramaciotti), Dan Belluoccio (Agilent) UQ Centre for Clinical Research

25

26 Clinical/pathology characteristics of gynaecological metast All mets to gynae n=53 Breast primary Gynae metastasis Other metastatic sites Average age diagnosis Peritoneum/omentum/a scitic fluid 28/41 (68%) Range Abdomen/GI 18/41 (43%) Type Bone/bone marrow 11/41 (26%) ILC 23 (43%) Lung 5/41 (12%) IDC 28 (54%) Pelvis/iliac LN 5/41 (12%) Other (micropapillary/medullary) 2 (3%) Liver 4/41 (9%) Grade Skin 3/41 (7%) 1 5 (18%) Kidney 1/41 (2%) 2 10 (37%) Gallbladder 1/41 (2%) 3 12 (44%) Bladder 1/41 (2%) Size (mm) average n=38 Range Thyroid 1/41 (2%) Cerebrospinal fluid 1/41 (2%) LN n=43 30/43 positive (69%) Brachial plexus 1/41 (2%) ER+ 34/39 (87%) 23/24 (95%) 8/9 (88%) HER2+ 1/23 (4%) 4/18 (22%) 0/2 (0%) * * * Brain 1/41 (2%) Time to diagnosis: 0-17 years, Median 5 years

27 ER and HER2 results ER+ Ovary (23) Uterus- endometrium,cervix,vuvla (14) Tubes (12) Omentum (10) Large bowel, rectum (6) Peritoneum (5) Small bowel (3) Pouch of douglas (1) Liver (1) Bladder (1) Diaphragm (1) Pelvis (2) Appendix (4) Pleura (1) Iliac vein (1) Stomach, duodenum (2) sites: 66 ER+ (75%) 49 gynae sites on TMA: 43 ER+ (87%) Her2 negative 87/87 (100%)

28 Linear model of tumour evolution K Polyak Clin Can Res 2008 initial genetic and epigenetic alterations dominantly drive tumour phenotype

29 Non-linear development Intra-tumoural heterogeneity Clonal diversity arises as a consequence of genomic instability The dominant clone may change depending on the selection conditions Micro-environmental factors treatment K Polyak Clin Can Res 2008 more aggressive and harder to treat?

30 Klein Science 2008

31 % of cases Immunophenotype primaries metastases ER+ PR+ HER2+ Triple negative Basal-like p53+ Ki-67 (>10%)

32 54% of DCIS of uniform grade 1 (29.2%) 2 (22.5%) 3 (2.5%) 46% showed diversity in grade 30% grade % grade % grade 1-3 Diversity in grade correlated with mutation of p53

33 Columnar Cell Lesions (CCL) spectrum of pre-invasive lesions of breast tdlu high frequency in biopsy due to microcalcification precursor to low grade DCIS and invasive carcinoma? co-existence with low grade DCIS & invasive carcinoma morphological & IHC overlaps molecular data of clinical importance

34 Case 3 CCC + atypia (class 6) DCIS (class 8) CCH + cytological and architectural Atypia (class 5)

35 Right mastectomy for wide spread DCIS 20 samples analysed: normal tdlu CCL DCIS Case q, 11q 15q, 16p 16q, 17q 19, p 6q 7q 8q 10q 16q 11q 15q 16p 16q 17q p 12q Xq 16q 5q 8q 12q 15q 16p 19 7p 15q 16p 16q 11q 11q 15q 16p 16q 19p slice 10 (lateral) slice 5 (medial) q 15q 16p 16q 19 9p 16q 18q 3p 15q 16p 19 1q 6 12p+q 13q 18q 15q 16q 11q 15q 16p 16q 17q 18q 19 16q 15q 16p 16q 19p 7q 14q 1 2p 15q 19 20p 19

36 wide local excision CCL spectrum within TDLU 15 samples analysed: normal acini CCL ALH DCIS Case 3

37 Case 3 normal CCH (class 2) CCC (class 1) CCH (class 2)

38 Right mastectomy for invasive carcinoma 24 samples analysed: CCL HUT ALH DCIS IDC Case q 16q none none 3p+q 1q 3p+q 9q 16q 19p 22 none none none slice 7 (lateral) none none slice 2 (medial) q 19 7q 8p+q 9q 18p none 11q 12q Xq X 1p, 1q 5, 6, 8 10, 13q 14q, 16p 16q 1q 9q 16q 18p none Xq X none none none

39 Distribution of metastases by patient age UQCCR UQ Centre for Clinical Research Molecular Breast Pathology

40 Supp Fig 7 - Case 33 A Chromosome 5 Breast B Chromosome 6 Breast Node non-axillary Node nonaxillary Liver Liver C Chromosome 8 D Chromosome 17 Breast Brea st Node non-axillary Node non-axillary Liver Liver

41 A Chromosome B Chromosome 11 Supp Fig 5 case 13 Breast 8 Brea st Node axilla Node axilla Lung Lung Kidney Kidney C Chromosome 16 D Chromosome 21 Breast Breast Node axilla Node axilla Lung Lung Kidney Kidney

42 Introduce what will be covering from molecular pathology point of view Intra-tumour heterogeneity Path IHC phenotype Genomics Maybe just speak about this on previous slide as a contents page? Pre-invasive disease how frequently heterogeneous? See Abdel fatah paper % of DCIS heterogeneous according to grade MDLs 3-5% of tumours share D and L features Heterogeneity within types lobular carcinomas Heterogeneity during metastatic progression HOW TO INTRODUCE THIS? HOW DOES FLOW WORK 2 parts to my talk: part 1 morphological heterogeneity during progression from preinv-mets; part 2 use of a large collection of primary and mets samples from autopsy series and contempoary surgical series to examine Biology