Current concepts in Breast Cancer- Beyond TNM

1
Current concepts in Breast Cancer- Beyond TNM

• Professor Ravi Kant
• FRCS (England), FRCS (Ireland), FRCS (Edinburgh),
  FRCS(Glasgow), MS, DNB, FAMS, FACS, FICS,
• Professor of Surgery

2
Applications of Genes Assay in CA Breast

• To subclassify breast cancer
• To estimate prognosis
• To predict response to therapy

3
Applications of Genes Assay in CA Breast

• To subclassify breast cancer
• To estimate prognosis
• To predict response to therapy

4
Gene Expression Patterns of Breast Carcinomas
Distinguish Tumor Subclasses With Clinical
Implications
PNAS 20019810869-10874
5
(No Transcript)
6
Molecular classification Prognosis

• Luminal A Best prognosis
• Luminal B
• Luminal C
• Normal breast like
• Her 2
• Basal like Worst Triple Negative

7
Subtype
Type Importance
Luminal A ER , Best overall survival, Best DFS
Luminal B ER,Her2,Intermediate

Her 2 ve ER-, Intermediate

Basal like ER-,PR-, Her2 - Worst

8
IHC profiles for the breast cancer subtypes
9
Tumor based Gene assay
Test of Genes Tissue
Mammaprint (Amsterdam) 70 Fresh
Oncotype Dx 21 Fixed
76 gene 76 Fresh
Wound response Fresh
Two gene ratio 2 Fixed
Intrinsic subtype Fresh
10
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
11
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
12
MammaPrint
70 gene classifier developed further by the
company Agendia (www.agendia.com) under the name
MammaPrint. MammaPrint was approved by the FDA
in February 2007 for node negative women under 61
years of age and with a tumor lt 5cm.
13
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
14
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
15
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
16
Tumor based Gene assay
Test Aim
Mammaprint 70 To predict risk of distant mets in N-, To identify who will benefit from Chemo
Oncotype Dx 21 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
76 gene 76 To predict DFS OS in N-, early stage
Wound response To predict risk of mets death
Two gene ratio 2 To identify N-, ER who will benefit by addition of Chemo to Tamoxifen
Intrinsic subtype To predict clinical outcome
17
Applications of Genes Assay in CA Breast

• To subclassify breast cancer
• To estimate prognosis/ prediction
• To predict response to therapy

18
Conventional classification

• Convential Classification assumes that women with
  a tumor bigger than 2 cm have a high risk to
  develop distant metastasis.

19
Size is an insufficient indiactor of metastasis
risk

• Molecular studies shows that size alone is not an
  indicator of high or low metastasis risk.
• Small and large tumors can be either low or high
  risk as determined by Molecular studies

20
Molecular studies provides additional data to
assess the risk of distant metastasis
21
MammaPrint Discovers 34 Low Risk Patients in
Adjuvant! High-Risk Group
n 209 87
N 209
n 137 66
n 72 34
Patients re-characterized as Low-Risk MammaPrint
34
Buyse et al., Journal of the National Cancer
Institute. 200698(17)1183-92.
22
Gene Assay prediction gt Adjuvant Online

• Buyse M. Validation and clinical utility of 70
  gene prognostic signatures for women with node
  negative breast cancer.
• J Natl Cancer Inst 2006981183-92.

23
Different proportion of low and high risk patients

• MammaPrint profiles accurately 40 as low risk
  compared to only 15 with St. Gallen criteria.

24
MammaPrint identifies correctly
25
MammaPrint identifies correctly

• 40 of patients with low risk in comparison to
  the 15 that are identified with conventional
  methods, thus preventing many unnecessary
  chemotherapies.
• More precise in predicting the outcome of disease
  than St. Gallen when comparing survival rates.

26
Mets free / Survival
27
LN or -
28
(No Transcript)
29
Gene Expression vs. Clinical
30
St Gallen vs NIH
31
(No Transcript)
32
(No Transcript)
33
TRANSBIG Validation302 Patients, Node-Neg,
T1/2, Age lt 61

• Amsterdam
• Gene expression profiling
• Agilent platform
• 70-gene prognostic custom designed chip

• Tissue Samples
• UK (Guys, Oxford)1984 1996
• France (IGR, CRH)1978 1998
• Sweden (Karolinska)1980 1990
• Node negative, untreated
• lt60 years
• gt 5 years follow-up
• T1, T2
• Tumor cell gt 50

• Brussels
• Comparison of clinicalv gene signature
  assessment ofprognostic risk
• Endpoints
• Time to Distant Metastasis
• Overall Survival
• Distant Metastasis-Free Survival, Disease-Free
  Survival

Audited clinical data
Centrally reviewed path data (Milan)
Buyse et al., Journal of the National Cancer
Institute. 200698(17)1183-92.
34
Three step validation strategy
Independent validation study on archival
material N?300
Agendia 70-gene prognostic signature N78 N15
1

• The signature is
• robust
• The technology is
• reproducible

Level 2-3
Level 5 and 4
Levels of evidence for biomarkers studies
35
MammaPrint identifies early metastases risk with
highest accuracy
Buyse et al., Validation and clinical utility of
a 70-gene prognostic signature for women with
node-negative breast cancer, Journal of the
National Cancer Institute, Vol 98, No. 17, 2006
36
FDA Clearance of MammaPrint
Study Purpose Details Comments
1 NatureDevelopment 70-gene profile 200278 patients, LN0, lt55yrs6.4 adjuvant treatment Within 5 year metastasis risk by profile multivariate OR 18
2 NEJM Validation 70-gene profile 2002151 patients5.2 adjuvant treatment Metastasis-free at 10 yrs low risk 87,high risk 445 yrs low risk 93,high risk 56
3 MammaPrintDevelopment MammaPrint 2006reproducibility of (1) and (2) on MammaPrint Highly reproducible MammaPrint as a diagnostic tool
4 TRANSBIGIndependent European validation 2006302 patientsno adjuvant treatment Metastasis-free at 10 yrs low risk 88,high risk 715 yrs low risk 96,high risk 83
37
TRANSBIG, the Translational Research Network of
the Breast International Group (BIG), conducted
an independent validation study of both the
Amsterdam and Rotterdam gene signatures in a
series of 302 patients
Although there was only a 3-gene overlap between
the two signatures, both were validated on the
same patient cohort
38
So time to learn basics again
39
(No Transcript)
40
MammaPrint interrogates critical genomic pathways
MammaPrint interrogates all of the
criticalgenomic pathways associated with
tumorprogression and the metastatic cascade

• Catabolism and tumorhypoxia related metabolism
• Cell cycle and cytoskeleton related biogenesis
• Extracellular matrixadhesion and remodeling
• General signal transductionand intracellular
  transport
• Growth factor
• Immune response
• Cellular mobility or actin filament related

41
Applications of Genes Assay in CA Breast

• To subclassify breast cancer
• To estimate prognosis/ prediction
• To predict response to therapy

41
42
(No Transcript)
43
71 Gene assay predictive value

• Van de Vijver MJ,He YD, vant Veer IJ, et al. A
  gene expression signature as predictor of
  survival in breast cancer. Amsterdam
• N Engl J Med 2002 3471999-2009

44
Oncotype Dx-21 gene predictive value

• Palk S, Tang G, Shak S et al. Gene expression and
  benefit of chemotherapy in women with node
  negative, ER breast cancer. J Clin Oncol
  2006243726-34.
• Can be done on fixed tissue.

45
21 gene analysis-Oncotype Dx

• Independent of
• Tumor size
• Age
• Park S. NEMJ 20043512817-26.
• gt Accurate than Adjuvant Online
• Goldstein RP. Abstract 63. San Antonio 2007

46
21 gene analysis can predict

• breast cancer related mortality
• Habel LA. Breast Cancer Res 2006
• NSABP- B14 trial

47
21 gene Predictive value
48
Newer prognostic Indicators

• Wound response gene risk of mets and death
• 2 gene recurrence score adding chemotherapy to
  tamoxifen in ER ve, N- ve
• Chang HY . Gene signature of fibroblast serum
  predicts cancer progressionsimilAarities between
  tumors and wound.
• PloS Biol 2004 2206-14.

49
Wound response gene expression profile

• Activation of fibroblasts
• Active wound healing predicts ? risk of
• metastases
• Death
• ( in Breast, Lung Gastric cancer)
• Cong HY. PLoS Biol 20042206-14

50
2 gene expression profile

• 60, ER, Early stage (MaXJ.Cancer Cell 2004)
• Expression of
• Homeobox 13
• IL-17B
• ? ratio poor outcome
• Tamoxifen alone will not do in such patients

51
Six models

• All are different
• Great concordance in five out of six gene
  expression profile models
• 21 gene (Oncotype Dx) and 70 gene (Mammaprint)
  are popular
• 81 concordance between 21 70 gene.
• Perou, Fan C. NEMJ 2006

52
Breast Cancer The Treatment Dilemma
Choices of 40 experts world-wide
61 y-old, fit, postmenopausal Node negative pT
0.9 cm ductal cancer ER and PR negative HER2
negative Grade 2
Courtesy Martine Piccart
53

Of 100 women with breast cancer
54
Only 25 will develop distant metastases

55
But we treat over 75 of all patients with
chemotherapy

56
Which means that 50 of all breast cancer
patients get a toxic chemotherapy that they did
not need!

57
Applications of Genes Assay in CA Breast

• To subclassify breast cancer
• To estimate prognosis
• To predict response to therapy

58
To predict response to therapy

• Selection of the therapy based on attributes of
  the
• Tumor
• Host

59
21 gene analysis can predict

• Responsiveness to Chemo/ Hormone
• Gianni L. J Clin Oncol.2005
• Palk S. J Clin Oncol 2006243726-34
• NSABP- B14 trial

60
TAILORx Trial Assigning Individualized Options
for Treatment
61
TAILORx trial
62
MINDACT study design
6000 patients, lt70 YRS, 1-3 POS NODES
ASSESS clinical RISK AND MammaPrint
RISK (adjuvant!online MammaPrint)
10
55
35
BOTH HIGH RISK
DISCORDANT RISK
BOTH LOW RISK
RANDOMIZE decision-making
Use MammaPrint
Use clinical risk
low
high
low
high
Chemotherapy
No chemotherapy
63
MINDACTMicroarray in Node Negative Disease may
avoid Chemo
64
Her2 positive MammaPrint low risk patients have
an excellent survival
Knauer SABCC 2008
65
Benefit of adjuvant chemotherapy in MammaPrint
high risk patients
88 Endocrine Chemo (n 265)
benefit
69 Endocrine (n184)
MammaPrint low risk (n126)
HR 0.28 (0.14- 0.56 p0lt0.001)
HR 1.99 (0.00- 6.3 pnon significant)
Distant metastasis-free survival (years)
Median Follow-up 5.2 years
Knauer et al, 2008 unpublished
66
To predict response to therapy

• Selection of the therapy based on attributes of
  the
• Tumor
• Host

67
To predict response to therapy based on
attributes of the host

• Drugs metabolized by
• CYP450 encoded enzymes
• CYP 2
• CYP 3
• CYP2D6
• CYP2C19

68
To predict response to therapy based on
attributes of the host

• Drugs metabolized
• AmpliChip CYP450 by Roche

69
Prognostic Signature and Clinical Benefit -the
chemotherapy choice-

• MammaPrint prognosis signature
• Assigns patients to risk categories with high
• specificity and sensitivity
• (low risk vs high risk for recurrence)
• Low risk sufficiently low to forego
  chemotherapy
• High risk identifies patients with early
  relapse
• and shows chemo benefit (predictiveness)

69
70
Clinical applications of microarrays

• Who to treat
• Prognosis profiles as diagnostic tool
• -gt improved selection for adjuvant therapy

• How to treat
• Predictive profiles for drug response
• -gt selection of patients who will benefit most

71
What does low risk mean?

• MammaPrint Low Risk 90 metastasis-free
  without any adjuvant treatment over the following
  10 years (NEJM 2002/JNCI 2006)
• Most of the Low Risk patients are ER
• With ER patients receiving hormonal therapy, a
  further 50 risk reduction can be achieved in the
  Low Risk group, thus MammaPrint Low Risk
  means gt95 10 year metastasis-free survival

72
Summary Poor risk

• Basal like
• Luminal B
• HER2/ER-
• Poor 70 gene profile
• High 21 gene recurrence score
• Activated wound response

73
Summary New Decision aid

• 21 gene (Oncotype Dx) 70 gene (Mammaprint) is
  better than Adjuvant Online
• Personalised treatment
• Less toxicity, less cost

74
Contrast of Appearance vs.Expression Phenotyping
Microscope
Low Grade
High Grade
Microarray
Low Risk
High Risk
75
Triple Negative

• ER-, PR-, and HER2
• Basal like on gene profiling
• Adverse prognosis
• ? new Rx

76
Thanks