Dr Nik Zeps

1
Apoptosis and Prediction of Response to
Chemotherapy in Breast Cancer The Role of
Molecular Markers and Digital Imaging in Clinical
Practice

• Dr Nik Zeps
• Sir Charles Gairdner Hospital

2
Breast Cancer

• Breast Cancer affects 1 in 8 Australian women1
• Greatest number of cancer-related deaths1
• Divergence in patient outcome
• Cannot accurately predict patient outcome or
  disease course

1. AIHW Breast cancer in Australia an overview,
2006
3
Polychemotherapy for early breast cancer an
overview of the randomised trials Early Breast
Cancer Trialists' Collaborative GroupLancet
1998 Volume 352 Issue 9132 Page 930-942

• Interpretation
• lt 3 women with good prognosis
• derive gain per 100 treated
• c. 7 women with node -ve
• derive gain per 100 treated
• c. 11women with node -ve
• derive gain per 100 treated
• NB these women are all pre or
• peri-menopausal
• Chemo could be simply causing
• chemical ovariectomy!
• Could just give Tamoxifen/
• LHRH agonists

4
Polychemotherapy for early breast cancer an
overview of the randomised trials Early Breast
Cancer Trialists' Collaborative GroupLancet
1998 Volume 352 Issue 9132 Page 930-942

• Interpretation
• lt 1 woman with good prognosis
• derives gain per 100 treated
• lt 3 women with node -ve
• derive gain per 100 treated
• lt 4 women with node -ve
• derive gain per 100 treated
• NB these women are all post
• menopausal
• Chemo really has no benefit at all

5
Chemotherapy in Breast Cancer

• Where is the evidence for treating EVERY woman?
• Only marginal benefit for treating 100 of women
• Imperative to discriminate which women will get
  benefit
• avoid over-treating gt90 of women who dont need
  it

6
Personalised Medicine

• Individualised therapy based on clinical and
  molecular markers
• Requires
• Biomarker discovery
• Validation

While the other explosives played during recess,
Charlie studied diligently in hopes of one day
becoming a Smart Bomb
http//www.junkertoons.com/toonbomb.htm
7
those placed in the low-risk category by the
microarray analysis had a 96 probability of
overall survival and 87 likelihood of disease
free survival for 10 years, whereas those with a
high-risk profile had a 50 overall survival and
48 probability of disease-free survival for 10
years.
MINDACT (Microarray In Node negative Disease may
Avoid ChemoTherapy)
http//www.eortc.be/services/unit/mindact/MINDACT_
websiteii.asp
8
How do you do this?

• Need Samples
• Biobanks
• Quality material
• Clinical Annotation
• Issues
• Where are they?
• How well set up are they?
• Do they have enough samples?
• Are samples processed suitably
• Clinical information?
• Treatment
• Follow Up

9
The pathologists, none of whom in the lead cases
were responsible for obtaining consent for
post-mortems, bear no responsibility for the
failure which I have found to exist. The Hon Mr
Justice Gage
10
Trust Comes on foot, leaves on horseback

• "The crowd of foolish scribblers is scarcely less
  than the swarms of flies in the height of summer,
  and threatens with their crude and flimsy
  productions to stifle us as with smoke." William
  Harvey
• Knock on effect of Alderhey is a general mistrust
  of doctors intent with human tissue

11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
WA Oncology Collection

• Carries out
• Consent
• Collection
• Processing
• Storage
• Data linkage (WAGER)

• AOCS (ovarian)-1000 (200)
• RADAR (prostate) -400
• IAB (Prostate)-150
• TARGIT IORT (Breast)-200
• Cancer Fatigue 100
• LABC 50 pa
• Colorectal 400 pa
• Sporadic breast 300 pa

15
Clinical Annotation

• Clinical Trial
• Standardised therapy
• Randomisation
• Controls
• Rich detail
• Multidisciplinary Group
• As above except not randomised

16
Locally Advanced Breast Cancer (LABC)

• 10 of breast cancers
• Large tumours, skin involvement, invasion to
  axillary lymph nodes, inflammatory
• Management neoadjuvant chemotherapy, surgery,
  adjuvant therapy (radiotherapy, chemotherapy,
  hormonal therapy)
• In vivo model of response to therapy

17
Locally advanced breast Cancer

• Neoadjuvant therapy (TAC)
• T3 or T4, or N2 or N3 and M0 breast cancer
• TAC (Docetaxel 75mg/m2, Doxorubicin 50mg/m2,
  Cyclophosphamide 500mg/m2 given d1 q 3 weeks
• Pts w clinical SD/PR or CR after 6 cycles TAC,
• breast conserving surgery based on
• patient preference, residual tumour size lt 4cm,
  acceptable cosmesis,
• absence of extensive microcalcification, absence
  of multifocal disease, non-inflammatory cancer
  and ability to achieve minimal of 5mm margin
  clearance

18
LABC tissue collection

• Multi-centre collection
• Mount, RPH, SCGH
• Collected in parallel with clinical trial
• 44 patients gave consent

19
Apoptosis Predictive Markers

• Chemotherapy induces apoptosis
• Many chemo agents induce cell death through DNA
  damage
• Many chemo agents induce mitochondrial-dependent
  apoptosis

20
ATP synthase apoptosis
Mt matrix

• Cytochrome C release
• Acidification of cytosol early event in apoptosis
• pH regulated by ATP synthase
• Apoptotic stimuli causes reverse operation of the
  ATP synthase pump

IMS
BAX
Cytosol
APOPTOSIS
21
How important is the ATP synthase pump in
apoptosis?
Mt matrix

• ATP synthase blockers inhibited Bax-induced cell
  death and cyt C release in yeast models.
• Restoration of mutated ATP synthase subunits
  restored sensitivity to Bax
• Does not completely abrogate apoptotic signaling
  affects efficiency

IMS
BAX
Cytosol
22
ß-F1-ATP synthase and its role in predicting
response to chemotherapy in locally advanced
breast cancer

• Otto Warburg 1931 Nobel prize
• Glycolysis higher in cancer cells
• ß-F1-ATP synthase rate limiting step for making
  ATP
• Bax can induce apoptosis
• requires functional ATP synthase to do so
• loss of the proton pump may prevent Bax from
  integrating into the outer mitochondrial membrane
  to form a channel and facilitate cytochrome c
  release

23
Hypothesis

• Patients with impaired mitochondrial function
• Mitochondrial number
• Mitochondrial DNA mutations
• ATPase expression
• will have poorer response rates to chemotherapy
  than patients with normal mitochondrial function

24
Scientific Approach

• Immunohistochemistry
• Specific mitochondrial proteins
• PCR
• Mitochondrial number
• Integrity of mitochondrial DNA

25
Tissue Micro-arrays

• 500 patients ? 10 slides
• 400 with normal tissue
• Clinico-pathological features of all cases
• Treatment and survival data

26
Results to date

• Staining breast TMAs complete
• Collaboration established with Molecular
  Discovery Systems for staining analysis
• ß software
• Volume
• Image size
• Data set size
• 6.6GB Xcel Data from 19 TMA cores

4x
40x
27
Results to date image de-convolution and
visualisation
28
Automated cell finding.
Nuclei de-clumped
Some nuclei still clumped
Nuclei touching
However, if you require the nuclei to be fully
separated the segmentation can be reduced to a
seed , this is of particular use if the cytoplasm
is what is required for measurements and nucleus
just for identification of object.
29
(No Transcript)
30
Refine the Map for Higher Magnification
10X 750 tiles
20X 3000 tiles
40X 15000 tiles
31
Output Changes
32
Issues

• Most systems have proprietary file formats eg
  Aperio .svs
• The images of full slides can be 13GB
• Compression of images as a way of controlling
  image size leads to loss of information and image
  artefact

33
Analysis Measurements, which ones, and how many??
34
Link to Health information
35
Summary

• Need samples for marker development
• Have to be good quality
• Good clinical annotation
• Need projects
• Image analysis complex
• Must be linked to clinical info for outcomes

36
Acknowledgements

• Sally Mclaren, PhD student
• Technical
• Sam Gallagher (MDS)
• Lisa Spalding
• Anne-Marie Shearwood
• Funding
• Lions Cancer Institute
• National Health and Medical Research Council
• Radiation Oncology, SCGH
• University of Western Australia
• Cancer Council of WA
• Tissue Collection
• SCGH breast clinic/ RPH breast assessment clinic
  (Prof Christobel Saunders)
• Mount Hospital breast surgeons Dr Hastrich, Mr
  Ingram, Mr Wilsher
• Dr Arlene Chan
• PathWest