Breast pathology directs systemic treatment Dr. Sabine Linn Medical oncologist The Netherlands Cancer Institute Amsterdam The Netherlands
Recent decline breast cancer mortality UK, US, NL, F; 35-69 years UK 1950-2002 US 1950-2001 NL Adj CT Adj HT Screening F Peto, Lancet 2005
Dutch guideline adjuvant systemic therapy Adjuvant systemic therapy only when the absolute 10-years survival benefit is at least 5%
Current adjuvant treatment options ERanyHER2any No systemic treatment 10% ERnegHER2neg chemo 15% ERnegHER2pos Surgery + radiotherapy chemo Trastuzumab 1 year Trastuzumab 1 year 7.5% ERposHER2pos chemo Endocrine Rx 5 yrs 7.5% ERposHER2neg chemo Endocrine Rx 5 yrs 70%
N=710 node-neg ER+HER2- ER-HER2- ER+HER2+ ER-HER2+ Parker, JCO 2009
How to further improve breast cancer survival? Better patient selection Currently 90% receives adjuvant syst therapy Only 50% needs it Reduce over- and undertreatment prognostic factors tailored therapy; target molecular defect Currently one-size-fits-all chemotherapy Currently one-size-fits-all hormonal therapy predictive factors
Prognostic and predictive markers Prognostic factor % Survival No treatment - Factor + Factor Years Predictive factor % Survival No treatment Years + Factor - Factor % Survival Treatment A Years + Factor - Factor
Caveat of better patient selection The best survival for the whole group of breast cancer patients is obtained by treating everybody with adjuvant systemic therapy. because none of our prognostic and predictive tests have an accuracy of 100%!!! Retel, Eur J Cancer 2010
Breast cancer metastasis-free survival premenopausal patients, lymph node negative Metastatic-free survival ~30% metastasize & die of breast cancer ~70% survive breast cancer time (years)
Breast cancer metastasis-free survival according to the 70-gene profile Discerns: 40% favorable profile, 60% unfavorable profile So approximately 1 out of 30 women will die because of a false low risk Mammaprint result.. Metastatic-free survival favorable profile: ~13% breast cancer mets ~87% metastasis-free as adjuvant chemotherapy will be withheld based on the low risk MP, while this woman would only have been cured WITH adjuvant chemotherapy unfavorable profile: ~56% breast cancer mets ~44% metastasis-free time (years) Van de Vijver et al. NEJM 2002 151 patients, <53 yrs, N0 10-years survival data
Trade-off of better patient selection Mammaprint as an example So it costs 1 avoidable death To spare 30 women the side-effects of chemotherapy that they don t need because they were cured already Retel, Eur J Cancer 2010
Discordant risk estimates MammaPrint vs clinical guidelines 27% to 39% discordant risk estimates ~30 % discordant risk estimates resulted in changed therapy advice in ~20% of patients (mammaprint should be combined with clinico-pathological factors)
Clinical risk indexes and Mammaprint Low-low-low Good risk MP 20% 60% 100% 10-yrs OS Poor risk MP Discordant clinical risk indexes High AO / high St Gallen / high NPI; MP low risk Good risk MP should still receive chemotherapy + endocrine Poor risk MP therapy High-high-high Good risk MP Poor risk MP Bueno-de- Mesquita, Ann Oncol 2009
Differences between Mammaprint and Oncotype DX For whom? Mammaprint ER pos N0 Untreated Oncotype DX ER pos N0 TAM-treated Tissue fixation Frozen FFPE Histology Ductal Ductal Predictive for tamoxifen benefit?? Yes? Predictive for chemotherapy benefit? No Yes Prognostic in HER2 pos? Yes No
Does Mammaprint add prognostic information? AUC Clinical 0.66 Clinical + MP 0.75 MP only 0.69 Bueno-de-Mesquita, Breast Cancer Res Treat, 2008
Does Oncotype DX add prognostic information? IHC4=ER / PR / HER2 /Ki67
Does Oncotype DX add prognostic information? IHC4=ER / PR / HER2 /Ki67
Mammaprint predictive for chemotherapy benefit? Test for interaction P=NS Knauer et al. Breast Cancer Res Treat 2010
Oncotype DX predictive for chemotherapy benefit? Low recurrence score High recurrence score Test for interaction P< 0.05 Paik, JCO 2006
Tumor size measurement
Assessment of differentiation grade, ER, PR, HER2 Interobserver variation grade ~ 30% Different AOL risk group in 8% Bueno-de-Mesquita et al. Ann Oncol 2009
Advantage gene expression classifiers More reproducible than pathologists No interobserver variation
Mammaprint test: for whom? Hormone receptor positive disease Node negative Invasive ductal carcinoma NOS Age > 45 years Tumour size < 1.0 cm and grade III Tumour size 1.1 2.0 cm and grade II or III Tumour size 2.0-3.0 cm and grade I
Mammaprint test: for whom? 50 years 9% 10-yrs risk ER pos Mammaprint high risk: multiply 10-years Grade I 2-3 cm N0 2.6% risk mortality risk with factor 2 to 3 (HR 2 to 3) Benefit ET 3%, CT 2%
Conclusions Mammaprint is prognostic, not predictive Mammaprint adds prognostic information to standard prognostic factors Mammaprint has independent prognostic value within the subgroup of HER2 positive patients Better patient selection comes at the expense of a few avoidable deaths as our prognostic tests are not 100% accurate
Summary; what is needed from the pathologist? Histologic subtype, grade Tumor size, nodal status Frozen tumor tissue ER and PR in % HER2 status Ki67 acgh profile PIK3CA mutation status FGF1R amplification status IGF1R expression levels
Acknowledgements Molecular Biology Piet Borst Michael Hauptmann Sven Rottenberg Lodewyk Wessels Jos Jonkers group Karin Beelen Eelke Gort Rutger Koornstra Mark Opdam Philip Schouten Tesa Severson Marieke Vollebergh Radiation Oncology Harry Bartelink Biometrics Otilia Dalesio Harm van Tinteren Andrew Vincent Tumor Biology Wilbert Zwart Desiree Verwoerd Sjaak Neefjes Rob Michalides Medical Oncology Jos Beijnen Sjoerd Rodenhuis Jan Schellens Gabe Sonke Molecular Genetics Rene Bernards Lund University Caroline Holm Olle Stal Goran Landberg Pathology Jolien Bueno-de- Mesquita Renske Fles Renate de Groot Simon Joosse Marleen Kok Donne Majoor Petra Nederlof Marjanka Schmidt Laura van t Veer Marc vd Vijver Jelle Wesseling PSOE Frits van Dam Wim van Harten Sanne Schagen Valesca Retel Surgical Oncology Emiel Rutgers Marie-Jeanne Vrancken Peeters