10 jaar MammaPrint Emiel Rutgers
De klinische vraag: Wat moeten/willen de breast specialist en de patiente weten? • •
De prognose: is deze zo goed dat aanvullende therapie niet/nauwelijks aan levensverwachting toevoegt. Als niet: welke therapie is het meest geschikt
Dus als je geen indicatie voor chemo ziet, dan hoef je verder ook niets van de tumor te weten (chemo sensitiviteit etc)
Prognose: hoe schatten we die in?
- NABON richtlijnen - Adjuvant!Online - St. Gallen
NABON richtlijn Laag risico: geen AST
Hoog risico
2004: - N-ve - >35 jaar - Graad 1 & 2 1-3 cm
- N+ - N -, ≤ 35 jaar behalve graad 1 en tumor ≤ 1cm - N -, > 35 jaar, graad 3 > 1cm, graad 2 > 2cm of alle tumor > 3cm
2008: - N-ve - >35 yrs - Graad 1-3 < 1cm - Graad 1 <2 cm
- N+ - N -, < 35 jaar behalve gr 1 ≤ 1cm - N -, ≥ 35 jaar, graad ≥ 2 en tumor 1,1-2cm of alle tumor > 2cm
Prognose: St Gallen International Expert Consensus 2009/11
70-genen MammaPrint • Is not just another prognostic factor • Is designed from the beginning to tell you the metastatic potential of an individual breast cancer
Met Microarrays kan je de activiteit van alle 25,000 genen in een tumor sample meten
MammaPrint: ontworpen op gearchiveerd bevroren tumorweefsel 78 78 breast breast tumors tumors
Age < 55 years, Tumor size < 5 cm Lymph node negative & No adjuvant therapy
Distant Distant metastases metastases within within 55 years years van ‘t Veer et al. (2002) Nature, Vol 415, 530-536
No No distant distant metastases metastases for for at at least least 55 years years
MammaPrint: 70 genes voorspellen de recidiefkans
78 Patients
Low Risk Signature
High Risk Signature
70 Genes
Uitgebreide klinische Validatie…. 4,879 Patienten Validation Studies Country
Reference
Years 2002
2006
2007
2008
2009
2010
First validation study
van de Vijver et al. NEJM
Independent European study
Buyse et al J NCI 17
Prospective Study - Dutch patient cohort
de Mesquito et al. Lancet Oncol. 8
Recent Diagnosis Study
de Mesquita et al Breast Cancer research tr
Core Needle biopsies
Mayordomo et al. ESMO Meeting
35
Validation in US patients
Wittner et al. Clin Cancer Res 14
100
Dutch patient cohort
Bueno de Mesquita et al. Br C Res
123
Validation in 1-3 LN+ patients
Mook et al. Breast Cancer Res Treat.
241
Validation in 4-9 LN+ patients
Saghatchian et al. St. Gallen Conf
168
Validation in HER2+ patients
Knauer et al. Br J Cancer
168
Patients treated with Tamoxifen
Kok et al. (submitted)
192
Postmenopausal patients (>61)
Mook et al. breast cancer research tr
148
German patient cohort
Kunz et al. St. Gallen Conference
140
Japanese patient cohort
Ishitobi et al. Jap Breast Cancer Symp
118
Neoadjuvant predictive study
Somlo et al. ASCO meeting
Neoadjuvant predictive study
Straver et al. Breast Cancer Res Treat
Predictiveness (Meta-analysis) study
Knauer et al. St. Gallen Conf.erence
Validation in T1 Tumors
Mook et al. Ann Oncology
295 302 427 123
68 167 1,637 427
Validatie 1: N = 151 vd Vijver et al, N Engl J Med 347: 1999-2009, 2002
Validatie 2: N = 307 Buyse M, et al.: J Natl Cancer Inst 98: 1183-92, 2006
De klinische vraag. Stap voor stap denken. Stap 1: Is de prognose zo goed dat overlevings voordeel van adjuvante chemotherapie niet opweegt tegen de nadelen ernstige late termijn bijwerkingen?
Validatie 3: N = 123 Bueno-de-Mesquita JM: Breast Cancer Res Treatm 2008
Het 70-gene Profiel is beter dan: -Adjuvant Online -St Gallen criteria -Nottingham PI -NABON richtlijn
Validatie 4: N = 100 Wittner et al., Clin Cancer Res 14: 2988, 2008 MGH series, Boston; Time to metastasis
St. Gallen Recommendations 2011 MammaPrint Accepted into St. Gallen’s Oncology Guidelines for Early Stage Breast Cancer Treatment “The Panel accepts the use of validated molecularly based tools if readily available as an adjunct to high-quality standard histopathologic assessment in patients with ER+ breast cancer when the doctor and patient are uncertain or ambivalent about the administration of adjunctive chemotherapy.” In addition, the Panel felt “intermediate” results were of little clinical value. Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thuerlimann B, Senn HJ; Panel members. Strategies for subtypes--dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer . Ann Oncology Aug. 2011
Kritische vragen bij het vroege mammacarcinoom & adjuvante chemotherapie • • • •
Is goede pathologie even goed? Kleine carcinomen altijd goede prognose? Her 2 overexpressie: chemo? N +ve: altijd slechte prognose waarvoor chemo is geindiceerd? • Selecteren we dan die carcinomen waarvoor chemo het meest effectief is?
Kritische vragen bij het vroege mammacarcinoom & adjuvante chemotherapie • Is goede pathologie even goed?
MammaPrint voegt toe aan optimale gradering: 1/3 discordant
764 of 1630 patients (47%) were classified as good prognosis and 866 (53%) as poor prognosis by MammaPrint Histological grading was centrally reviewed for all patients
MammaPrint high risk MammaPrint low risk
DDFS N -ve Survival Functions MammaPrint 0 1 0-censored 1-censored
DDFS N-ve Survival Functions MammaPrint 0 1 0-censored 1-censored
Kritische vragen bij het vroege mammacarcinoom & adjuvante chemotherapie • Kleine carcinomen altijd goede prognose?
Patienten inclusie criteria: Alle T1 mammacarcinomen uit MammaPrint database Dus los van: • Leeftijd • N status • ER,PgR, Her-2 => 965 patienten Mediane follow-up 7.1 jaar (0.2-25.2)
MammaPrint: Tumor T1c BCSS 11 – 22 mm Tumors
99% 92% Good signature (n=441)
88% 72%
Poor signature (n=384)
Log rank p<0.001 HR at 10 yrs: 4.42 (95% CI 2.73-7.17); p<0.001
Log rank p = 0,036 T1c tumors derived from pooled database of all MammaPrint validation studies (all, n=1696) Mook et al, Ann Surg Oncol, 2010
MammaPrint and Tumorsize T1ab BCSS 0 – 10 mm Tumors
100% 88% Good signature (n=84)
90%
Poor signature (n=55)
73%
Log rank p=0.06 HR at 10 yrs: 3.12 (95% CI 0.91-10.67); p=0.07
Log rank p = 0,036 T1ab tumors derived from pooled database of all MammaPrint validation studies (all, n=1696) Mook et al, Ann Surg Oncol 2010
En Her-2 positief Mammacarcinoom….
• Altijd slechte prognose?
BCSS: All patients
100
100
80
80
60
Good signature (n=27) Poor signature (n=142)
40
Log-rank (Mantel-Cox) Test
20
P value 0
0
2
Survival probability
Survival probability
DDFS: All patients
60
Good signature (n=27) Poor signature (n=142)
40
Log-rank (Mantel-Cox) Test
20
P value
0.0153 4
6
8
10
Time in years
Figure 1: Distant disease-free survival (LEFT) and breast cancer-specific survival
0
0
2
0.0242 4
6
8
10
Time in years
(RIGHT) According to the 70-gene signature for all 169 Her2-positive breast cancer patients.
Michael Knauer et al.Br J Cancer 2010
BCSS: without chemotherapy/trastuzumab
100
100
80
80
60
Good signature (n=21) Poor signature (n=69)
40
20
0
0
2
60
Good signature (n=21) Poor signature (n=69)
40
20
Log-rank (Mantel-Cox) Test
P value
Survival probability
Survival probability
DDFS: without chemotherapy/trastuzumab
Log-rank (Mantel-Cox) Test
0.0127 4
6
P value 8
10
Time in years
Figure 2: Distant disease-free survival (LEFT) and breast cancer-specific survival
0
0
2
0.0403 4
6
8
10
Time in years
(RIGHT) according to the 70-gene signature for 90 patients without adjuvant chemotherapy or trastuzumab. Michael Knauer et al.Br J Cancer 2010
BCSS: highly endocrine responsive
100
100
80
80
60
Good signature (n=11) Poor signature (n=31)
40
20
0
0
2
60
Good signature (n=11) Poor signature (n=32)
40
20
Log-rank (Mantel-Cox) Test
P value
Survival probability
Survival probability
DDFS: highly endocrine responsive
Log-rank (Mantel-Cox) Test
0.1383 4
6
P value 8
10
Time in years
Figure 3: Distant disease-free survival (LEFT) and breast cancer-specific survival
Michael Knauer et al.Br J Cancer 2010
0
0
2
0.1382 4
6
8
10
Time in years
(RIGHT) according to the 70-gene signature for 42 patients with highly endocrine-responsive tumors according to the St.Gallen criteria. Out of 11 low risk patients, 7 were untreated, 4 received chemotherapy and one of those received trastuzumab.
En N +ve Mammacarcinoom?
• Altijd indicatie voor chemotherapie?
70-gene Profile and Prognosis in Breast Cancer with 1-3 Axillary Lymph Node Metasases S. Mook et al., Breast Cancer Res Treatm 2008. Good profile (n=99) Poor profile (n=142)
Distant metastases as first event
Overall survival
95%
95%
77%
73%
HR 4.1
HR 5.4
(95%CI 1.7 – 10.0)
(95%CI 2.1 – 13.8)
p=0.002
p<0.001
MammaPrint laag risico: voldoende goed?
Background
Objective
Methods
Results
Conclusions
De klinische vraag: Wat moeten/willen de breast specialist en de patiente weten?
Stap 2 • selecteer je de ‘goede’ carcinomen voor chemotherapie?
Benefit of neo-adjuvant chemotherapy for MammaPrint high risk patients • Netherlands Cancer Institute • 2 clinical trials
Eligible patients N=167
• T-stage >3 cm and/or LNplus (SNB/FNA)
• Antracycline-like • Antracyclin-Taxane • Taxane
• ultrasound guided 14 gauge biopsies • MRI imaging • Pathology
Good prognosissignature
Poor prognosissignature
N=23 (14%)
N=144 (86%)
pCR(axilla+breast)
pCR(axilla+breast)
n=0
n=29 (20%) P=0.015
Straver et al, BCRT 2009
• pCR: • pathological • complete remission
Neo-adjuvant Standard Chemotherapy and MammaPrint Clinical Benefit • 70 gene MammaPrint High Risk Signature patients show significantly higher chemosensitivity • All pCR are found in the High Risk Signature group
High Risk Signature Patients show Clinical Benefit of Chemotherapy Straver et al, BCRT 2009
Tot nu geen recidieven in de laag risico groep
36
Straver et al 2009, Br Cancer Res and Treatment
Chemotherapie effect MammaPrint LOW RISK patients (n=252) DDFS: MammaPrint LOW RISK (n=252)
BCSS: MammaPrint LOW RISK (n=252)
100
99% 93% ET (n=174, 69%) ET+CT (n=78, 31%)
60
40
HR 0.26 (0.03‐2.02) p=0.20
20
0
0
1
2
3
4
99% 97% ET (n=174, 69%) ET+CT (n=78, 31%)
80
Percent survival
Percent survival
80
100
60
40
HR 0.58 (0.07‐4.98) p=0.62
20
5
Time in years
Knauer et al, ASCO 2009, BCRT 2010 Albain et al 2009
0
0
1
2
3
Time in years
4
5
Chemotherapie effect MammaPrint HIGH RISK patients (n=289) DDFS: MammaPrint HIGH RISK (n=289)
BCSS: MammaPrint HIGH RISK (n=289)
100
100
94%
88% 80
76% 60
ET (n=141, 49%) ET+CT (n=148, 51%) 40
HR 0.35 (0.17‐0.71) p<0.01
20
0
0
1
2
3
4
in years 12% Time absolute benefit 50% relative benefit Knauer et al, ASCO 2009, BCRT 2010 Albain et al 2009
Percent survival
Percent survival
80
81% ET (n=141, 49%) ET+CT (n=148, 51%)
60
40
HR 0.21 (0.07‐0.59) p<0.01
20
5
0
0
1
2
3
4
13% absolute Time in years benefit 68% relative benefit
5
MINDACT trial • Is niet ontworpen om de prognostische waarde van MammaPrint te valideren • Zal ons vertellen of chemotherapie terecht is onthouden aan patienten met een MP laag risico test die ‘anders’ wel chemo gehad zouden hebben met een strake limiet van een 5 jaars breast cancer specific survival van 93-95%.
ECCO16-ESMO36 - Stockholm, 23-27 September 2011
Baseline characteristics and logistics aspects after a successful accrual
Emiel Rutgers On behalf of co-PIs Fatima Cardoso and Martine Piccart & the whole team
MINDACT TRIAL DESIGN
N= 6600 Clinical-Pathological (C) risk (Adjuvant! Online) C-HIGH / G-HIGH
Genomic (G) risk (70-gene signature)
Discordant Discordant cases cases C-HIGH C-HIGH // G-LOW G-LOW or or C-LOW C-LOW // G-HIGH G-HIGH
C-LOW / G-LOW
1st randomization to treatment use Clinical vs. Genomic risk
Chemotherapy HR+
2nd randomization
Anthracycline –based vs. CapecitabineCapecitabine-Docetaxel
Endocrine therapy 3rd randomization
Tamoxifen 2y / Letrozole 5y vs. Letrozole 7y
No Chemotherapy HR+
EORTC-BIG MINDACT TRIAL DESIGN The discordant cases Evaluate Clinical-Pathological risk and 70-gene signature risk Discordant cases 32% (as per protocol estimative) Clin-Path HIGH and 70-gene LOW
Clin-Path LOW and 70-gene HIGH
1st randomization treatment decision Use Clin-Path risk to decide CT Clin-Path HIGH and 70-gene LOW chemotherapy Clin-Path LOW and 70-gene HIGH no chemotherapy
Use 70-gene risk to decide CT Clin-Path HIGH and 70-gene LOW no chemotherapy Clin-Path LOW and 70-gene HIGH chemotherapy
Potential CT rightfully foregone
Accrual Country
Pilot phase1 First 800 enrolled patients
Enrolled pts
Netherlands (NKI)
2055
France (FNCLCC)
1992
Germany (WSG)
825
Belgium (EORTC)
820
Spain (SOLTI)
523
Italy (GOIRC)
186
UK (NCRIBCG )
64
Slovenia (IOL)
37
Switzerland (EORTC)
25
Total
6527
1. Rutgers et al. Eur J Cancer 8 (3): 188, 2010 (suppl; abstr 444)
Patient characteristics • Age
•
Local hormone receptor status HR+: 88% (ER+ or PR+ or both) HR-: 12%
•
Local HER2 status 86.8% negative 10.6% positive 2.5% not done/unknown
≤ 50: 33% > 50: 67%
• Tumor size < 20 mm: 72% > 20 mm: 18%
• Tumor grade
1: 21.7% 2: 48.9% 3: 28.8% Undefined: 0.6%
• Triple negative: 9.5% •
Nodal status 79.9% LN negative 13.7% LN +1 4.4% LN +2 2.3% LN +3
Patient risk allocation Clin-path risk and 70gene risk at enrollment 70-gene risk Total
Clinical-pathological risk LOW N(%)
HIGH N(%)
Total
LOW
2586 (40)
1436 (22)
4022 (62)
HIGH
678 (10)
1827 (28)
2505 (38)
3264 (50)*
3263 (50)*
N=6527
Discordant cases (10 + 22 = 32%) match protocol hypothesis
The absolute difference between C-HIGH / G-LOW and C-LOW / G-HIGH is 11.6%
*The 50-50 split is coincidental
Conclusions Accrual has been successfully completed Its complex logistics, including real-time collection of frozen tumor tissue, were proven to be feasible in a multinational, multicentric setting A crucial fact to test the primary endpoint, the absolute difference between Clin-Path HIGH / 70-gene LOW and Clin-Path LOW / 70-gene HIGH cases, has been achieved (>10%) Compliance to treatment allocations is high
Met dank aan
MammaPrint: en nu? • Het lijkt veilig geen chemo te geven aan patienten met een vroeg mammacarcinoom en een laag risico MammaPrint test • Patienten met een MammaPrint hoog risico carcinoom profiteren het meest van adj. chemotherapie
MammaPrint: wie dan? • CBO 2011: als er twijfel is…. • Praktijk: vooral patienten van 40-70 jaar met een carcinoom tussen de 1 en 3 cm, ER +ve.
Dank De patienten, de onderzoekers, ons Mammateam, Het Mindact team. Voor uw deelname en aandacht