Expression profilesfor prognosis and prediction

Laura J. Van ‘t VeerThe Netherlands Cancer Institute, Amsterdam

Breast Cancer - Survival premenopausal patients, lymph node negative

time (years)

surv

ival

~30% die of breast cancer

~70% survive breast cancer

Kaplan-Meier Survival Curves

1) Who to treat2) How to treat

61 y-old, fit,61 y-old, fit,postmenopausalpostmenopausalNode negativeNode negativepT = 0.9 cm pT = 0.9 cm ductal cancerductal cancerER and PR ER and PR negativenegativeHER2 negativeHER2 negativeGrade 2Grade 2

HOW SHOULD ONE TREAT HOW SHOULD ONE TREAT A SMALL (<1CM) BREAST TUMOR ?A SMALL (<1CM) BREAST TUMOR ?

FA(E)C x 6

0

10

20

30

40

50

NONE CMFx6 ACx4 TAM OTHER

%

4848

2525

151544

88

Choices of 40 experts worldwideChoices of 40 experts worldwide

Courtesy of Martine PiccartCourtesy of Martine Piccart

Gene expression profiling to improve prediction of clinical

outcome• aim

-to identify patients at risk to develop distant metastases and die of cancer

-to accurately select for adjuvant therapy

-to predict treatment response

Cy3-dUTPgreen fluorescent

reverse transcriptase,T7 RNA polymerase

Cy5-dUTPred fluorescent

cRNA

sample 2(reference)

RNAcDNA

sample 1(tumortissue)

RNAcDNA

cRNA

RNA extraction and labelingto determine expression level

sample of interestcompared tostandard reference

Scanned image of flexjet25K human oligonucleotide

microarray

Hybridized with mixture of ‘red’-labeled cRNA of a tumor sample and ‘green’-labeled reference cRNA (pool of tumor samples)

Determine:• fluorescence intensities• fluorescence ratio’s

2010: Outcome for Prognosis

Micro-array

Treatment 2

Prognosis poor

Treatment 1

Treatment 3

Classifier

• Tumor

Response for treatment 1

Micro-array

Treatment 2

Prognosispoor

Treatment 1 resistant

Treatment 3

Classifier

• Tumor

• Tumor

Micro-array

Treatment 2sensitive

Prognosispoor

Treatment 1resistant

Treatment 3

Response for treatment 2

Classifier

• Tumor

Response for treatment 3

• Tumor

Micro-array

ClassifierTreatment 2sensitive

Prognosispoor

Treatment 1resistant

Treatment 3sensitive

78 breast tumors (‘83-’94)patients < 55 yearstumor size < 5 cm

lymph node negative (LN0)no adjuvant therapy

no distant metastasesin at least 5 years (n=44)

distant metastases< 5 years (n=34)

Prognosis Reporter Genes

Who to treat?Retrospective series –

TissuebankNo adjuvant treatment

Classification Prognosis7

8 t

um

ors

goodsignature

poorsignature

threshold

threshold set with 10% false negatives91 % sensitivity, 73% specificity

metastases

(white = +)

70 significant prognosis genes

Nature 415, p530-536, 2002

Breast Cancer - Metastases riskby profiling

time (years) 151 patients, <53, LN010 year survival curve

Distinguish in: 40% good profile, 60% poor profile

meta

stase

s-fr

ee good profile:

~13% develop metastases~87% disease-free

poor profile:~56% develop metastases~44% disease-free

Implementation of Profiling

Breast cancer patients, lymph node negative:

~20-30% reduction of unnecessary treatmentand avoidance of 2-3% undertreatment

Who to treat

70 genes

Prospective NKI - Raster trial - work in progress

Good signatureLow risk

Poor signatureHigh risk

• accrual 75 patients• microarray 40 patients

24 low, 16 high risk

- biopsy of primary tumor - upfront chemotherapy

- response of primary tumor by imaging

determine therapy response profiles in biospy

How to treatNeo-adjuvant trials

Chang et al, Lancet 2003

Neo-adjuvant docetaxol response

92 differentially expressed genesdifferentiate sensitive andResistant breast tumors.Confirmed in ‘6’ sensitive tumors

• identification of predictive factors to neoadjuvant chemotherapy using gene expression profiling

• comparison of two different drug combinations:– AC (adriamycine/cyclophosphamide)– AD (adriamycine/docetaxel)

Neo-adjuvant response profile

NKI/AVL study locally advanced breast cancer

S. Rodenhuis, M. van de Vijver

Why these drugs ?

• standard drug combination for neoadjuvant chemotherapy: AC– but: resistance against anthracyclines can

exist or develop under primary therapy

• good responses observed with docetaxel containing combinations– high activity also in anthracycline-resistant

disease– but: very toxic agent

Study design (NOODCD)

• prospective phase III trial within the NKI/AvL

Patients included

• first 62 patients included into the study• enough RNA obtained from 49 patients:

50% tumour cells– 46 biopsies, 18 tumours

CT-arm biopsy/tumour

pairs biopsies

only tumours

only total amount of patients

AC 7 15 3 25

AD 8 16 0 24

Patients response

CR: complete remission

PR: partial remission

MR: minor remission

SD: stable disease

PD: progressive disease

CT-arm (almost) CR PR MR/SD/PD NA total

AC 7 (28 %) 9 (36 %) 8 (32 %) 1 (4 %)* 25

AD 5 (21 %) 11 (46 %) 8 (33 %) 0 (0 %) 24

Interim Conclusions NOODCD

• Thus far, no major differences in gene expression between tumours from patients with a CR compared to all other patients– (work in progress)

• significant differences in gene expression in tumour specimens before and after treatment

Neo-adjuvant or Metastatic response relevant for Adjuvant?

Are gene activities similar or different when comparing the primary tumor and metastasis

Expression profiling of matched pairs

Primary tumors – lymph node metastasesPrimary tumors – distant metastases

microarrays

ER- positive

ER- negative

15 primary breast tumors and lymph node metastases of the same patient

Hierarchical clustering

Primary breast tumors – lymph node metastases

Hierarchical clustering

Primary breast tumors – distant metastases

8 primary breast tumors and distant metastases of the same patient

How to treatNeo-adjuvant therapy response profiles

similarity of primary/metastasis profile implies:therapy recommendations based on expression profileof the primary tumor are a rational approach towards preventing the outgrowth of micrometastases

- biopsy of primary tumor profiled - upfront chemotherapy

- response of primary tumor by imaging

Who to treat:• Prognosis profile as diagnostic tool

-> improvement of accurate selection for adjuvant therapy (less under- and overtreatment)

• Prognosis profile implemented in clinical trials-> reduction in number of patients & costs (select only patients that are at metastases risk)

Therapeutic implications

How to treat:•Predictive profile for drug response

-> selection of patients who benefit