University of Toronto Province-Wide Oncology Rounds May 18, 2012 The EBCTCG Overview: Is it still relevant in 2012? By Dr. Kathleen I. Pritchard Department Division Director, Medical Oncology Professor, Department of Medicine Faculty of Medicine, University of Toronto 4/19/2017

Early Breast Cancer Trialists’ Collaborative Group The Oxford Overview Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)

EBCTCG OVERVIEW K. Albain, S. Anderson, R. Arriagada, W. Barlow, J. Bergh, J. Bliss, M. Buyse, D. Cameron, M. Clarke, A. Coates, R. Collins, J. Costantino, J. Cuzick, S. Darby, N. Davidson, C. Davies, A. Di Leo, M. Dowsett, M. Ewertz, R. Gelber, C. Geyer, J. Godwin, A. Goldhirsch, R. Gray, D. Hayes, C. Hill, J. Ingle, R. Jakesz, M. Kaufmann, P. McGale, L. Norton, Y. Ohashi, S. Paik, E. Perez, R. Peto, M. Piccart, L. Pierce, G. Pruneri, K. Pritchard, V. Raina, P. Ravdin, J. Robertson, E. Rutgers, Y. F. Shao, S. Swain, C. Taylor, P. Valagussa, G. Viale, T. Whelan, E. Winer, Y. Wang, W. Wood.

EBCTCG OVERVIEW Oxford Secretariat Richard Peto Sarah Darby Mike Clarke Christina Davies Paul McGale Richard Gray Rory Collins Jon Godwin

EBCTCG OVERVIEW Steering Committee - Executive Marc Buyse Mike Clarke Rory Collins Sarah Darby Christina Davies Marianne Ewertz Martine Piccart Kathy Pritchard Eric Winer William Wood

EBCTCG OVERVIEW Past Chairs I. Craig Henderson William Wood Current Co-Chairs Kathy Pritchard Martine Piccart

EBCTCG September 2010. Preliminary results

EBCTCG OVERVIEW 1984  First overview process  data sought from all randomized trials of systemic adjuvant therapy  meta-analysis concept  collaboration sought built sustained  Trialists  Secretariat

EBCTCG OVERVIEW Methodology  Individual patient data  dates of randomization  treatment allocation  age  menopausal status  nodes  ER, PgR

EBCTCG OVERVIEW Methodology  Data checked for internal consistency  Data amended and updated by correspondence

EBCTCG OVERVIEW Methodology  Each trial analysed separately  Women in one trial are compared directly with only the women in the same trial  One log rank statistic per trial  Stratified by age and nodal status  Combined to give an overall estimate of the effect of different treatments

EBCTCG OVERVIEW Outcomes  Recurrence  first reappearance of breast cancer  includes contralateral breast cancer

EBCTCG OVERVIEW Outcomes  Deaths  unknown causes included with deaths from breast cancer unless specifically stated otherwise  problem of death without recurrence

 Breast Cancer Related Deaths  deaths of/with breast cancer EBCTCG OVERVIEW Outcomes  Breast Cancer Related Deaths  deaths of/with breast cancer

 Other Deaths  cardiac  stroke  other cancers EBCTCG OVERVIEW Outcomes  Other Deaths  cardiac  stroke  other cancers

EBCTCG OVERVIEW 1984  Tamoxifen improved survival  CMF chemotherapy improved survival  Ovarian ablation improved survival

EBCTCG OVERVIEW 1990  longer tamoxifen seemed better  tamoxifen effects greater in ER+ve women  tamoxifen reduced rate of contralateral breast cancer  chemo effective in older and younger women

EBCTCG OVERVIEW 1995  huge magnitude of effect of 5 years of tamoxifen  5 years of tamoxifen clearly better than 1 or 2  tamoxifen prevented contralateral breast cancer only in women with ER+ve disease  anthracycline containing regimens better than CMF

EBCTCG OVERVIEW 2000  15 year effects of chemo sustained in older and younger women  chemo effect appears greater in ER negative than in ER positive disease But is this really true?

EBCTCG OVERVIEW 2000  15 year effects of 5 years of tamoxifen sustained and of great magnitude  door opened to question of 5 years versus longer tamoxifen  ovarian suppression/ablation effective but not significantly so when added to chemotherapy

EBCTG OVERVIEW 2005  2000 Overview: Lancet, 2005  Trialists meet: new Steering Committee formed  Many new trials added  More women-years of follow-up for all major questions  But major trials still missing

EBCTCG OVERVIEW 2006  Trialists met: new questions  type of anthracycline-based regimen  taxane trial status  aromatase inhibitors  trastuzumab  chemoendocrine therapy (only in ER+, pre- and postmenopausal subsets)  Subcommittees of the SC formed

 Locoregional therapy EBCTCG OVERVIEW 2010  Tamoxifen  AI’s  Chemotherapy  Locoregional therapy

2010 EBCTCG OVERVIEW TAMOXIFEN TAMOXIFEN VS NOT LONGER VS SHORTER TAM No of women No of women 1 yr vs not 2 yr vs not 5 yr vs not 9126 23940 21457 2 – 4 vs 1 – 2 y 5 vs 1 - 2 y 10 vs 5 y 3200 20000 22000 54523 45200 Median follow-up = 15y 22% are ER- PR- Median follow-up = 5y 50% are ER ?

2010 EBCTCG OVERVIEW Tamoxifen for 5y vs same management but No Tam Risks (all) Benefits (ER+) Death w/o recurrence * RR 1.05 (+ 0.07) 2p>0.1 Endometrial incidence RR 2.33 (+ 0.25) 2p<0.00001 Absolute gain at 15y Proportional risk reductions Recurrence 38% (2p<0.00001) BC mortality 30% (2p<0.00001) All deaths 22% (2p<0.00001) Contralateral BC 39% (2p<0.00001) 13% 9% * Numerical excess of deaths due to stroke, pulmonary embolus, uterine cancer (15 vs 13 ; 6 vs 0; 8 vs 1)

2010 EBCTCG OVERVIEW Tamoxifen for 5y vs same management but no Tam Learning more about Tam benefits  On types of B.C. events…  In subgroups  In relation to chemotherapy administration  Over time…

2010 EBCTCG OVERVIEW Tamoxifen for 5y : Impact on BC events

2010 EBCTCG OVERVIEW Tamoxifen for 5y: Benefits in subgroups

2010 EBCTCG OVERVIEW Tamoxifen for 5y vs same management but no Tam

2010 EBCTCG OVERVIEW Tamoxifen for 5y : Benefits in subgroups Nodal status AGE TAM for 5y : BENEFITS for whom ? Tumor diameter Tumor grade All do benefit !!

2010 EBCTCG OVERVIEW

2010 EBCTCG OVERVIEW

2010 EBCTCG OVERVIEW Tamoxifen for 5y: Benefits in subgroups Yes No ER levels (fmol/mg prot) ER+ PR+ ER- PR- TAM for 5y : BENEFITS for whom ? ER- PR+ ER+ PR- Uncertain Yes

2010 EBCTCG OVERVIEW Tamoxifen for 5y : Benefit over time

Duration of adjuvant Tam and outcome 2010 EBCTCG OVERVIEW Duration of adjuvant Tam and outcome

2010 EBCTCG OVERVIEW Impact of TAM duration Even 1y only provides significant benefit 10y provide small benefit which could ↑ over time

2010 EBCTCG OVERVIEW Tamoxifen for 10y : Benefits vs risks at 10 y Mean follow-up only 5y Absolute Xcess Absolute gain Proportional risk reductions Recurrence 8% (2p=0.03) BC mortality 10% (2p>0.1) Contralateral BC Death w/o recurrence * + 1.5% (2p=0.59) Endometrial cancers + 0.7% (2p=0.00004) 1% (2p 0.03) 2.9% (2p 0.55) 1.3% (2p 0.03) *Numerical excess of deaths due to cerebrovascular events (42 vs 38 in y0-4; 27 vs 24 in y5-10), thrombo-embolic events (10 vs 56 in y0-4), end. cancers (8 vs 6 in y0-4, 4 vs 2 in y5-10)

2010 EBCTCG OVERVIEW TAMOXIFEN  5y in ER+ disease  reduces  recurrence by 38%,  BC death by 30%  all deaths by 22%  contralateral BC by 40%  benefits all women with ER+ disease  unclear benefits in ER-PgR+ disease  benefits women with ER very rich tumors more  increases endometrial cancer by 2.3 fold

2010 EBCTCG OVERVIEW TAMOXIFEN 10 yrs vs 5 yrs of adjuvant TAMOXIFEN in ER+/? Disease  absolute reduction in recurrence by 8% (2p=0.03)  reduces contralateral BC by 10% (2p=0.03)  increases endometrial cancer by 4 fold  reduces BC mortality by 3% (2p=0.55)  increases death without recurrence by 1.5% (2p=0.59)

2010 EBCTCG OVERVIEW TAMOXIFEN Messages for clinical practice in 2010  PgR does not predict for benefit of adjuvant TAM  For ER-PgR+ patients, the tumor should be retested and if doubt remains, TAM could be offered

2010 EBCTCG OVERVIEW TAMOXIFEN Messages for clinical practice in 2010  There is presently little incentive to prescribe more than 5y of TAM, in postmenopausal women  More than 5y of TAM may be useful at least for DFS in premenopausal women especially those without a uterus

Aromatase inhibitors EBCTCG SEPTEMBER 2010

Data from 1st analysis No unplanned cross-over Cut-off 30 Sept 2006 Cohort 1: 5yrs AI vs 5yrs tam Cohort 2: 2-3 yrs of AI vs 2-3 yrs of tam after 2-3 yrs tam

5 years AI vs tamoxifen: life table curve, recurrence JCO, 2010, 28, 509-518

5 years AI vs tamoxifen: subgroup analysis, recurrence JCO, 2010, 28, 509-518

5 years AI vs tamoxifen: life table curves, br ca mortality JCO, 2010, 28, 509-518

2-3yr AI vs tam after 2-3 yrs tam: life table curve, recurrence JCO, 2010, 28, 509-518

2-3yr AI vs tam after 2-3 yrs tam: subgroup analysis, recurrence JCO, 2010, 28, 509-518

2-3yr AI vs tam after 2-3 yrs tam: life table curve, br ca mortality JCO, 2010, 28, 509-518

Message for Clinical Practice in 2010 2010 EBCTCG OVERVIEW Aromatase Inhibitors Message for Clinical Practice in 2010  AIs > tamoxifen  recurrence  survival  good given  early  after 2 yrs

Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Comparisons between different polychemo- therapy regimens for early breast cancer: meta-analyses of long-term outcome among 100,000 women in 123 randomised trials Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Published online December 6, 2011 in The Lancet DOI:10.1016/S0140-6736(11)61625-5 EBCTCG, Lancet 2011

Direct and indirect comparisons between different polychemotherapy regimens, based on ~100,000 randomised women 45,000 taxane vs no taxane* (44,000 with anthracycline in both arms) 22,000 anthracycline vs CMF (18,000 vs “standard” CMF) 5,000 more vs less anthracycline (2000 comparing currently relevant doses) 31,000 polychemotherapy vs no adjuvant chemo (13,000 CMF vs Nil; 10,000 anthr.-based regimen vs Nil) * Excludes trials of one taxane regimen vs another EBCTCG, Lancet 2011

Trials of chemotherapy vs no adjuvant chemotherapy - Any anthracycline-based regimen (eg, standard 4AC) vs nil - Standard CMF vs nil EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Chemotherapy vs no adjuvant chemotherapy L: anthracycline-based regimen (eg, standard 4AC), R: standard CMF EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Chemotherapy vs no adjuvant chemotherapy L: anthracycline-based regimen (eg, standard 4AC), R: standard CMF EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Chemotherapy vs no adjuvant chemotherapy L: anthracycline-based regimen (eg, standard 4AC), R: standard CMF EBCTCG, Lancet 2011

by TYPE of treatment comparison Breast cancer mortality ratio: anthracycline-based regimen (eg, standard 4AC) or standard CMF vs no chemotherapy, by TYPE of treatment comparison EBCTCG, Lancet 2011

ER+ disease only: by ENTRY AGE Chemotherapy (anthracycline-based regimen or standard CMF) + 5 year endocrine therapy vs 5 year endocrine therapy only, ER+ disease only: by ENTRY AGE EBCTCG, Lancet 2011

Trials of any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy: Subgroup analyses by age, stage and ER status, and by subsets of ER+ disease EBCTCG, Lancet 2011

Any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, by ENTRY AGE EBCTCG, Lancet 2011

Any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, by NODAL STATUS EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by AGE and STAGE Breast cancer mortality ratio: any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, by AGE and STAGE EBCTCG, Lancet 2011

Any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, by ER STATUS EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by ER STATUS and subsets of ER+ Breast cancer mortality ratio: any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, by ER STATUS and subsets of ER+ EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 ER+ disease only: by ENTRY AGE Any anthracycline-based regimen (eg, standard 4AC) vs no adjuvant chemotherapy, ER+ disease only: by ENTRY AGE EBCTCG, Lancet 2011

Trials of any anthracycline-based regimen. vs standard CMF Trials of any anthracycline-based regimen* vs standard CMF *Standard 4AC, standard 4EC, or higher-cumulative-dosage regimens (eg, CAF or CEF) EBCTCG, Lancet 2011

Approximate equivalence: Definitions of “standard” CMF and 4AC (mg/m2 x frequency/cycle) Standard CMF: Six 4-weekly cycles of C100x14 oral M40x2 iv F600x2 iv Standard 4AC: Four 3-weekly cycles of A60 iv C600 iv Approximate equivalence: in the trials of standard AC vs standard CMF, both appeared to be of comparable efficacy EBCTCG, Lancet 2011

Standard 4AC vs standard CMF: approximate equivalence EBCTCG, Lancet 2011

Examples of higher-cumulative-dosage* anthracycline-based regimens (mg/m2 x frequency/cycle) CAF: Six 4-weekly cycles of C100x14 oral A40x2 iv F500x2 iv CEF: Six 4-weekly cycles of C75x14 oral E60x2 iv F500x2 iv * Higher dosage than standard 4AC not only of anthracycline but also of other cytotoxic drugs; scheduled dosages could be reduced for toxicity EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Anthracycline-based regimens with higher cumulative dosage (eg CAF/CEF) vs standard CMF EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by TYPE of treatment comparison Breast cancer mortality ratio: anthracycline-based regimen vs standard CMF, by TYPE of treatment comparison EBCTCG, Lancet 2011

by age, stage and ER status Trials of any anthracycline-based regimen vs standard CMF: subgroup analyses by age, stage and ER status EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by AGE and STAGE Breast cancer mortality ratio: anthracycline-based regimen vs standard CMF, by AGE and STAGE EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by ER STATUS and subsets of ER+ Breast cancer mortality ratio: anthracycline-based regimen vs standard CMF, by ER STATUS and subsets of ER+ EBCTCG, Lancet 2011

Taxane trials Data on 44,000 women in randomised trials of a taxane-plus-anthracycline-based regimen vs the SAME, or MORE, non-taxane chemotherapy 11,000 in trials where the non-taxane regimen was the SAME, and 33,000 in trials where it was MORE [15% node-negative; mean follow-up only 5 years; mean recurrence rate about 5% per year] EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Taxane-plus-anthracycline-based regimens vs (L) the SAME, or (R) MORE, non-taxane chemo.  EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Taxane-plus-anthracycline-based regimens vs (L) the SAME, or (R) MORE, non-taxane chemo.  EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 Taxane-plus-anthracycline-based regimens vs (L) the SAME, or (R) MORE, non-taxane chemo.  EBCTCG, Lancet 2011

or whether taxanes are given alone (†). Taxane comparisons, subdivided according to: (a) how the non-taxane treatments compare (active = control, active = ½ control, or an intermediate ratio), and (b) whether the cycles of taxane are given concurrently (©) with the anthracycline, or whether taxanes are given alone (†). EBCTCG, Lancet 2011

by TYPE of treatment comparison Breast cancer mortality ratio in taxane trials, by TYPE of treatment comparison EBCTCG, Lancet 2011

Taxane trials: subgroup analyses by age, stage and ER status Taxane-plus-anthracycline-based regimen vs an anthracycline-based control regimen with the SAME, or MORE, of each non-taxane cytotoxic drug EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by ENTRY AGE Taxane-plus-anthracycline-based regimen vs the SAME, or MORE, non-taxane chemo, by ENTRY AGE EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by NODAL STATUS before chemotherapy Taxane-plus-anthracycline-based regimen vs the SAME, or MORE, non-taxane chemo, by NODAL STATUS before chemotherapy EBCTCG, Lancet 2011

Breast cancer mortality ratio in taxane trials, by AGE and STAGE EBCTCG, Lancet 2011

EBCTCG, Lancet 2011 by ER STATUS Taxane-plus-anthracycline-based regimen vs the SAME, or MORE, non-taxane chemo, by ER STATUS EBCTCG, Lancet 2011

by ER STATUS and subsets of ER+ Breast cancer mortality ratio in taxane trials, by ER STATUS and subsets of ER+ EBCTCG, Lancet 2011

Halving big risks and halving small risks by chemotherapy Proportional risk reduction does not depend much on age, ER status or nodal status (or on tumour grade or tumour diameter) Absolute risk reduction, however, depends on the prognosis – and, for ER+ disease, this is the prognosis with endocrine therapy Information lacking on tumour gene expression and on quantitative immunohistochemistry EBCTCG, Lancet 2011

EBCTCG September 2010. Preliminary results Effect of Radiotherapy after Breast-conserving Surgery on 10-year Recurrence and 15-year Mortality in Women with Early Breast Cancer EBCTCG September 2010. Preliminary results

EBCTCG September 2010. Preliminary results Proportional effect of radiotherapy after breast-conserving surgery (BCS ± RT) 11 000 women, pN0/pN+/pN? Any recurrence Breast cancer mortality EBCTCG September 2010. Preliminary results

EBCTCG September 2010. Preliminary results Absolute effect of radiotherapy after breast conserving surgery (BCS ± RT): 11 000 women pN0/pN+/pN? Any recurrence Breast cancer mortality Any death “One-in-four rule” one breast cancer death avoided for every 4 recurrences avoided EBCTCG September 2010. Preliminary results

EBCTCG September 2010. Preliminary results Effect of radiotherapy after breast-conserving surgery (BCS ± RT): 1100 pN+ women Any recurrence Breast cancer mortality “One-in-four rule” one breast cancer death avoided for every 4 recurrences avoided EBCTCG September 2010. Preliminary results

EBCTCG September 2010. Preliminary results Absolute effect of radiotherapy after breast-conserving surgery (BCS ± RT): 7300 pN0 women Any recurrence Breast cancer mortality “One-in-four rule” one breast cancer death avoided for every 4 recurrences avoided EBCTCG September 2010. Preliminary results

EBCTCG September 2010. Preliminary results Conclusions Radiotherapy highly effective in reducing recurrence in both pN0 and pN+ women Radiotherapy also reduces 15-year breast cancer “One-in-four” rule applies for pN0 and pN1 women Benefits not substantially reduced by fatal side-effects EBCTCG September 2010. Preliminary results

The Oxford Overview: Is it Still Relevant in 2010 ?

YES

EBCTCG OVERVIEW Tamoxifen  5 +/- 5 years  30% - 40% in recurrence  25 in deaths

EBCTCG OVERVIEW AIs  Better than Tam  for all subgroups  25% in recurrence  0 – 25% in BC mortality

 ? Stronger effect after two years of tamoxifen EBCTCG OVERVIEW AIs  ? Stronger effect after two years of tamoxifen

EBCTCG OVERVIEW Chemotherapy vs None  CMF/AC  20 – 30% recurrence  10 – 30% BC mortality A vs CMF

EBCTCG OVERVIEW  Adriamycin vs Standard CMF  10 – 20% recurrence  10 – 20% mortality

EBCTCG OVERVIEW Taxanes vs Non-Taxanes  10 – 20% recurrence  10% BC mortality

EBCTCG OVERVIEW  Natural History of Breast Cancer  ER/PgR +ve  ER and PgR -ve

EBCTCG OVERVIEW  By having all data  avoids publication bias  gives average effect size  clarifies time frames of effects  process / outcomes both useful

EBCTCG OVERVIEW  Future – Yes  Publications  2 on radiation results 2010 - 2011  one on chemotherapy 2011  one on tamoxifen  one on AIs  Meet Again September 19-22, 2012

EBCTCG OVERVIEW  Future – Yes  Publications  2 on radiation results 2010 - 2011  one on chemotherapy 2011  one on tamoxifen  one on AIs  Meet Again September 19-22, 2012

EBCTCG September 2010. Preliminary results