Management of Triple Negative Breast Cancer. Giuseppe Curigliano MD, PhD University of Milano and European Institute of Oncology

Management of Triple Negative Breast Cancer Giuseppe Curigliano MD, PhD University of Milano and European Institute of Oncology

Outline Heterogeneity of TNBC Targeting TNBC by subtypes New antibody drug coniugates

Clinical Heterogeneity of TNBC Subtype Gene expression profile Clinical Basal-like 1 high Ki-67; DNA damage response BRCA-associated Basal-like 2 GF pathways Higher pcr Immunomodulatory Immune genes Mesenchymal Cell motility Lower DDFS Mesenchymal stem-like Cell motility; claudin-low Luminal androgen receptor Steroid pathways Apocrine features, higher LRF; PI3Kmut Lehman BD, et al. J Clin Invest 2011;121:2750-67.

Metastatic triple-negative breast cancer Subtype Gene expression profile Clinical Basal-like 1 high Ki-67; DNA damage response BRCA-associated Basal-like 2 GF pathways Higher pcr Immunomodulatory Immune genes Mesenchymal Cell motility Lower DDFS Mesenchymal stem-like Cell motility; claudin-low Luminal androgen receptor Steroid pathways Apocrine features, higher LRF; PI3Kmut Lehman BD, et al. J Clin Invest 2011;121:2750-67.

Basal like 1 TNBC Triple negative breast cancer and BRCA-mutations Clinical behavior Genomic instability Stephens et al Nature 2009 vol. 462 (7276) pp 1005

Basal like 1 TNBC HER2- MBC + gbrcamt chemo for MBC (prior A and T. No plat resistance) 2:1 N~300 ~ 150 TNBC Olaparib 300 mg tablets bid 2:1 randomization Chemo TPC Capecitabine Eribulin Vinorelbine Treat until progression Primary endpoint PFS 70% prior chemotherapy for MBC (30% prior platinum) 57% BRCA1 50% TNBC Robson M et al, NEJM 2017

Olaparib in basal like 1 Non-response Response Stable disease Partial response Complete response Olaparib 300 mg bd (n=167) 15.0 25.1 50.9 9.0 Chemotherapy TPC (n=66) 33.3 37.9 27.3 1.5 Median response onset 80 60 40 20 0 20 40 60 Olaparib: 47 days Patients, % Chemotherapy TPC: 45 days Among patients with metastatic HER2-negative BC and a germline BRCA1/2 mutation in the OlympiAD study, the objective response rate with olaparib tablet monotherapy was double that seen with standard chemotherapy TPC Stable disease was for 5 weeks, recorded 6 weeks after randomization. Chemotherapy TPC, treatment of physician s choice (including capecitabine, eribulin or vinorelbine) Robson M et al, NEJM 2017

Olaparib in basal like 1 QoL olaparib > chemotherapy Response rate % to % Robson M et al, NEJM 2017

Olaparib in basal like 1 Robson M et al, NEJM 2017

Olaparib in basal like 1 Tumour burden Olaparib 300 mg bid Chemotherapy TPC 1 metastatic site, n 46 25 Median PFS, months 8.4 4.2 HR (95% CI) 0.62 (0.35, 1.13) etastati sites, 159 72 Median PFS, months 6.5 3.0 HR (95% CI) 0.59 (0.43, 0.82) Tumour location Olaparib 300 mg bid Chemotherapy TPC Visceral, n 165 84 Median PFS, months 5.9 3.9 HR (95% CI) 0.64 (0.47,0.86) Non-visceral, n 40 13 Median PFS, months 11.0 8.4 HR (95% CI) 0.65 (0.30,1.65) Although the study was not powered to detect differences in treatment effect between subgroups, the PFS benefit for olaparib over chemotherapy TPC across tumour burden and location subgroups was consistent with that seen in the overall population 1 HR, hazard ratio. Non-visceral disease includes lymph nodes, soft tissue, cutaneous and bone only. 1. Robson M et al. N Engl J Med 2017;377:523 533. Robson M et al, NEJM 2017

Veliparib in basal like 1

Veliparib in basal like 1 1.0 Median PFS, months (95% CI) Placebo + C/P N = 98 12.3 (9.3 14.5) Veliparib + C/P N = 95 HR P value * 14.1 (11.5 16.2) 0.789 (0.536 1.162) 0.231 Probability of Progression-Free Survival 0.8 0.6 0.4 0.2 RR % to % However no pcr with veliparib in neoadjuvant setting 0.0 Placebo + C/P Veliparib + C/P 0 4 8 12 16 20 24 28 32 Months Since Randomization Phase III BROCADE 3 pending Han et al, Annals of Oncol 2018; Geyer et al, ASCO 2017

Cisplatin in basal like 1 ER-, PgR-/unknown & HER2- or known BRCA1/2 Metastatic or recurrent locally advanced Exclusions include: Adju a t ta a e i o ths Previous platinum treatment Non-anthracyclines for MBC A Priori subgroup analyses: BRCA1/2 mutation RANDOMISE Basal-like subgroups (PAM50 (1:1) and IHC) Biomarkers of HRD Tutt A et al, 2104 Carboplatin (C) AUC 6 q3w, 6 cycles On progression, crossover if appropriate n-376 BRCA1/2 = 9%/12% Docetaxel (D) 100mg/m 2 q3w, 6 cycles On progression, crossover if appropriate Docetaxel (D) 100mg/m 2 q3w, 6 cycles Carboplatin (C) AUC 6 q3w, 6 cycles 13

Cisplatin in basal like 1 Randomised treatment - all patients (N=376) % with OR at cycle 3 or 6 (95% CI) 0 20 40 60 80 100 Carboplatin Docetaxel 59/188 (31.4%) 67/188 (35.6%) Absolute difference (C-D) -4.2% (95% CI -13.7 to 5.3) Exact p = 0.44 Crossover treatment - all patients (N=182) % with OR at cycle 3 or 6 (95% CI) 0 20 40 60 80 100 Carboplatin (Crossover=Docetaxel) Docetaxel (Crossover=Carboplatin) 21/92* (22.8%) 23/90* (25.6%) Absolute difference (D-C) -2.8% (95% CI -15.2 to 9.6) Exact p = 0.73 *Denominator excludes those with no first progression and those not starting crossover treatment 14 Tutt, SABCS 2014

Cisplatin in basal like 1 % patients progression free 100 90 80 70 60 50 40 30 20 10 Median PFS: Carboplatin: 3.1 mths (95% CI = 2.5 to 4.2) Docetaxel: 4.5 mths (95% CI = 4.1 to 5.2) Restricted mean survival to 15 mths: Carboplatin: 4.8 mths Docetaxel: 5.2 mths Absolute difference: -0.4 (95% CI -1.1 to 0.3) p = 0.29 Carboplatin = 181 /188 0 Docetaxel = 0 3 6 9 12 15 182 /188 18 Months from randomisation Number of events/at risk C: 0/188 90/98 40/56 32/22 9/13 5/8 0/7 D: 0/188 57/130 60/69 48/20 7/13 6/5 152/3

Cisplatin in basal like 1 % patients alive 100 90 80 70 60 50 Median OS: Carboplatin: 12.4 mths (95% CI = 10.4 to 15.3) Docetaxel: 12.3 mths (95% CI = 10.5 to 13.6) 40 Restricted mean survival to 15 mths: 30 Carboplatin: 10.7 mths Docetaxel: 10.8 mths 20 Absolute difference: 10-0.2 (95% CI -1.1 to 0.8) p = 0.31 0 0 3 6 9 12 15 18 Months from randomisation Number of events/at risk C: 0/188 23/165 18/141 24/114 22/89 14/71 22/44 D: 0/188 11/176 20/151 35/110 19/85 23/58 16/39

Cisplatin in basal like 1 Germline BRCA 1/2 Mutation (n=43) Carboplatin Docetaxel No Germline BRCA 1/2 Mutation (n=273) Percentage with OR at cycle 3 or 6 (95% CI) 0 20 40 60 80 100 17/25 (68.0%) 6/18 (33.3%) Absolute difference (C-D) 34.7% (95% CI 6.3 to 63.1) Exact p = 0.03 Percentage with OR at cycle 3 or 6 (95% CI) 0 20 40 60 80 100 Carboplatin Docetaxel 36/128 (28.1%) 53/145 (36.6%) Absolute difference (C-D) -8.5% (95% CI -19.6 to 2.6) Exact p = 0.16 Interaction: randomised treatment & BRCA 1/2 status: p = 0.01 17

Cisplatin in basal like 1 % patients progression free 100 90 80 70 60 50 40 30 20 10 0 Median PFS: C + BRCA 1/2 mutated 6.8mnths (95% CI = 4.4 to 8.1) C + BRCA1/2 not mutated 3.1mnths (95% CI = 2.4 to 4.2) Carboplatin + BRCA1/2 mutated Carboplatin + BRCA1/2 not mutated 0 3 6 9 12 15 18 Months from randomisation Tutt, SABCS 2014

PARP inhibitors in metastatic TNBC HER2- MBC + gbrcamt Prior chemo 2:1 Talazoparib (BMN 673) 1 mg po qd 2:1 randomization Chemo TPC Capecitabine Eribulin Gemcitabine Vinorelbine Treat until progression Primary endpoint PFS HER2- MBC + gbrcamt No platinum-r 2:1 Niraparib 100 mg po qd 2:1 randomization Primary endpoint PFS Chemo TPC Menu of 4 drugs

Talazoparib

Basal like 2:Growth factor signalling Lisa A. Carey et al. JCO 2012;30:2615-2623

Basal like 2:Growth factor signalling The combination of cetuximab plus Carboplatin in metastatic TNBC produced responses in fewer than 20% of patients. EGFR pathway analysis showed that most TNBCs involved activation. Lisa A. Carey et al. JCO 2012;30:2615-2623

Subsets of triple-negative breast cancer Subtype Gene expression profile Clinical Basal-like 1 high Ki-67; DNA damage response BRCA-associated Basal-like 2 GF pathways Higher pcr Immunomodulatory Immune genes Mesenchymal Cell motility Lower DDFS Mesenchymal stem-like Cell motility; claudin-low Luminal androgen receptor Steroid pathways Apocrine features, higher LRF; PI3Kmut Lehman BD, et al. J Clin Invest 2011;121:2750-67.

Evidence from clinical trials Pembrolizumab (Merck) Humanized IgG4 anti- PD-1 antibody MPDL3280 (Genentech) engineered human IgG1 anti-pd-l1 antibody

Pembrolizumab in TNBC Recurrent or metastatic ER-/PR-/HER2- breast cancer ECOG PS 0-1 PD-L1+ tumour No systemic steroid therapy Pembro 10 mg/kg Q2W CR PR/SD Discontinuation permitted Treat for 24 mo or until PD or toxicity No autoimmune disease (active or history of) No active brain metastases Confirmed PD Discontinue PD-L1 positivity: 58% of all patients screened had PD-L1-positive tumors Treatment: 10 mg/kg IV Q2W Response assessment: Performed every 8 weeks per RECIST v1.1 a PD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody. Only patients with PD-L1 staining in the stroma or i % of tu or ells ere eligi le for e roll e t. b If clinically stable, patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan perfor ed eeks later. If progressi e disease is confirmed, pembrolizumab is discontinued. An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor.

Pembrolizumab in TNBC n =32 Confirmed complete response (nodal disease) Confirmed partial response Stable disease Progressive disease Objective response rate: 18.5% Stable disease: 25.9% Nanda, SABCS 2015

Pembrolizumab in TNBC Cohort A (N = 170): Previously Treated, Regardless of PD-L1 Expression Cohort B (N = 52): Previously Untreated, PD-L1 Positive 40 36 Complete response 40 36 ORR, % 32 28 24 20 16 Partial response ORR, % 32 28 24 20 16 23.1% 12 12 8 4.7% 4.8% 4.7% 8 4 4 0 Total PD-L1 Positive PD-L1 Negative 0 Total (All PD-L1 Positive) Adams S et al. ASCO 2017

Anti PD1 and anti PDL1 in TNBC Anti-PD-L1/PD-1 single agent in mtnbc 1L, PDL1+/- Objective Response Rate (%) 30% 20% 10% 0% 26% 1L Atezolizumab (n=115) 11% 2L+ No clear relationship with PD-L1 positivity 23% 1L Keynote-086, Cohort B Pembrolizumab (n=222) 4.7% 2L+ Keynote-086, Cohort A CR PR CR PR Schmid P, et al. AACR 2017; Adams S, et al ASCO 2017

Role of TILs in TNBC 4% Atezolizumab Pembrolizumab (Cohort A: >2nd line) Pembrolizumab (Cohort B: 1st line) 39.1% Objective Response Rate (%) 30% 20% 10% 0% 19% TIL high 2 9% TIL low Different levels by source of sample (archival vs new) and organ site sampled: LN>lung>liver Metastatic breast cancer is a low TIL disease 6.4% TIL high 2 1.9% TIL low TIL high 8.7% TIL low Schmid P, et al. AACR 2017; Adams S, et al ASCO 2017, Loi, ESMO 2017

Phase Ib Study of Atezolizumab and Nab- Paclitaxel in mtnbc Best Overall Response 1L (n = 9) Confirmed ORR (95% CI) a 66.7% (29.9, 92.5) ORR (95% CI) b 88.9% (51.7, 99.7) 2L (n = 8) 25% (3.2, 65.1) 75.0% (34.9, 96.8) 3L+ (n = 7) 28.6% (3.7, 71.0) 42.9% (9.9, 81.6) All Patients N = 24 41.7% (22.1, 63.4) 70.8% (48.9, 87.4) CR 11.1% 0 0 4.2% PR 77.8% 75.0% 42.9% 66.7% Response rates were higher for patients who received atezolizumab/nabpaclitaxel treatment as 1L therapy compared to 2L+ SD 11.1% 25.0% 28.6% 20.8% PD 0 0 28.6% 8.3% a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response. b Including investigator-assessed unconfirmed responses. Efficacy-evaluable patients were dosed by June 1, 2015, and were evaluable for response by RECIST v1.1. Mi i u effi a follo up as o ths. Adams S, et al. SABCS. 2015 [abstract 850477].

Phase Ib Study of Atezolizumab and Nab- Paclitaxel in mtnbc Including investigator-assessed unconfirmed responses. 11 of 17 responses (65%) continued on treatment at time of data cut off Adams S, et al. SABCS. 2015 [abstract 850477].

Phase III Study of Atezolizumab and Nab- Paclitaxel in mtnbc Randomized, double-blind, placebo-controlled Phase 3 trial of nab-paclitaxel ± atezolizumab as 1 st line therapy in mtnbc (NCT02425891) Study design Histologically documented locally advanced or metastatic TNBC No prior therapy for advanced disease ECOG PS 0-1 Measurable disease per RECIST v1.1 Patients with significant CV or CNS disease (except asymptomatic brain metastases), autoimmune disease or prior checkpoint inhibitor therapy are excluded Target accrual: ~350 pts Stratification factors: Presence of liver metastases Prior taxane therapy Nab-paclitaxel 100 mg/m 2 QW 3/4 + Atezolizumab 840 mg Q2W PD-L1 expression status (centrally evaluated by IHC using the SP142 assay) R 1:1 Nab-paclitaxel 100 mg/m 2 QW 3/4 + Placebo Q2W Co-primary endpoints: PFS in all patients PFS according to PD-L1 expression Secondary endpoints: OS ORR Response duration Safety/tolerability PK HR QoL Emens et al. SABCS 2015 (abstract OT1-01-06)

Neoadjuvant Neoadjuvant paclitaxel x 12 +/- pembrolizumab followed by AC x 4 Adaptive randomization on I-SPY 2 Signature All HER2- TNBC HR+/HER2- Estimated pcr rate (95% probabilty interval) Pembro 0.46 (0.34 0.58) 0.60 (0.43 0.78) 0.34 (0.19 0.48) Control 0.16 (0.06 0.27) 0.20 (0.06 0.33) 0.13 (0.03 0.24) Probability pembro is superior to control Predictive probability of success in phase 3 > 99% 99% >99% >99% >99% 88% Nanda et al, ASCO 2017, Abstract 506 The Bayesian model estimated pcr rates adjust to characteristics of the I-SPY 2 population. The raw pcr rates are higher than the 35 model estimate of 0.604 in TNBC.

Immunotherapy in TNBC Nivolumab (BMS) Human IgG4 anti-pd-1 antibody Pembrolizumab (Merck) Humanized IgG4 anti- PD-1 antibody MPDL3280 (Genentech) engineered human IgG1 anti-pd-l1 antibody MEDI4736 (AZ) Human IgG1 anti-pd-l1 antibody Tremelimumab (AZ) Human IgG2 Anti-CTLA-4 antibody

Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients: TONIC-trial (The Netherlands Cancer Institute) Radiation 3 x 8 Gy Biopsy Doxorubicin 15 mg x2 Biopsy Biopsy R Cyclophosphamide 50 mg daily Biopsy Nivolumab Cisplatin 40 mg/m 2 x2 Biopsy Kok M, et al. Ann Oncol. 2017;35(suppl): Abstract LBA14. No treatment 2 weeks Biopsy

Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients: TONIC-trial (The Netherlands Cancer Institute) Total (n = 50) Best ORR (CR + PR) irecist 24% CBR (CR + PR + SD) 26% CR 1 (2%) PR 11 (22%) SD weeks 1 (2%) ORR RECIST 1.1 22% Median PFS [95% CI] 3.4 months [2.5-3.7] Median time to response [range] 2.1 months [0.5-3.5] Median duration of response [95% CI] 9.0 months [5.5-NA] Kok M, et al. Ann Oncol. 2017;35(suppl): Abstract LBA14.

Challenges High risk breast cancer High TILs/immune activation signature/ PDL1+/ High TMB Good microbioma Low/Intermediate TILs/immune activation signature-/pdl1-/ Bad microbioma I-O as monotherapy or combination of I-O High TMB Low TMB Add CT to enhance immunogenicity or STING, TIGIT, RT No I-O

PARP inhibitors and IO Phase I Patients with OC or TNBC Dose level 1 Niraparib 200 mg + pembrolizumab 200 mg Phase II Patients with OC (target n = 48) or TNBC (target n = 48) RP2D Endpoint assessment Dose level 2 Niraparib 300 mg + pembrolizumab 200 mg Endpoint assessment Konstantinopoulos PA, et al. Ann Oncol. 2017;28(Suppl 5): Abstract 1143PD.

PARP inhibitors and IO Konstantinopoulos PA, et al. Ann Oncol. 2017;28(Suppl 5): Abstract 1143PD.

Notch pathway Phase 1b Study of docetaxel + PF- 03084014 in Triplenegative Breast Cancer PF-03084014

Notch pathway Characteristic Mean (range) age, years PF 100 mg BID/ PF 100 mg BID/ PF 150 mg BID/ All Dose D 75 mg/m 2 (N = 8) D 100 mg/m 2 (N = 3) D 75 mg/m 2 (N = 11) Levels (N = 22) 57 (43-76) 43 (32-64) 46 (27-69) 50 (27-76) ECOG PS, n (%) 0/1 4/4 (50/50) 1/2 (33/67) 8/3 (73/27) 13/9 (59/41) Primary Diagnosis, n 1/7 (13/87) 0/3 (0/100) 3/8 (27/73) 4/18 (18/82) (%) locally recurrent/metastatic Prior Systemic Therapies, n (%) 1st line/ 2 nd line 4/4 (50/50) 3/0 (100/0) 7/4 (64/36) 14/8 (64/36) M Locatelli et al, Oncotarget 2016

Luminal Androgen Receptor: Bicalutamide ER/PR(-) IHC % LABC/MBC AR+ DAKO Ab > 10% Bicalutamide 150mg daily Primary endpoint = CBR24 (CR + PR + SD > 24 weeks) Screened patients 12% AR+ (mostly TNBC) Clinical Benefit Rate = 19% (95% CI 7-39%) All SD Gucalp et al CCR 2013

Luminal Androgen Receptor: Abiraterone MBC ER/PR % 138 screened 38% AR+ % Primary Endpoint = CBR24 Median PFS 2.8m N = 30 evaluable patients ~ 2.5 prior lines Rx ~ 50% visceral mets Most common, related AEs: fatigue (18%) HTN (12%) hypokalemia (9%) nausea (6%) CBR24 = 20% (95%CI: 8-39%) 1 confirmed CR Bonnefoi et al, Ann Onc 2016

Luminal Androgen Receptor: Enzalutamide ER/PR/HER2 (-) LABC/MBC AR+ Ventana > 0% Enzalutamide 160 mg daily Primary endpoint = CBR16 (CR + PR + SD > 16) Screened weeks) patients 79% AR+ (55% by 10% cutoff) Median 1 prior Rx Evaluable (n=75 AR > 10%) CBR16 35% (24-46%) CBR24 29% (20-41%) RR 8% SAE 29% 100 80 PFS (%) 60 40 20 0 PFS 14.7 weeks 95% CI: 8.1, 19.3 0 8 16 24 33 41 49 61 64

Luminal Androgen Receptor Gene expression classifier created = PredictAR (Basal-, apocrine+, etc to identify LAR) PREDICT AR PREDICT AR+ PREDICT AR PREDICT AR+ Total, n (%) 62 (53%) 56 (47%) CBR16, % (95% CI) n 11% (5, 21) n = 7 39% (27, 53) n = 22 CBR24, % (95% CI) n CR or PR, % n 6% (2, 16) n = 4 3% n = 2 36% (24, 49) n = 20 9% n = 5 0 1 Prior Lines 2+ Prior Lines Active Confirmed CR or PR 0 8 16 24 32 40 52 Time (weeks) 64 0 8 16 24 32 40 52 64 Time (weeks) Traina et al, ASCO 2015

Luminal Androgen Receptor ITT Population 100 n = 118 Overall Survival (%) 80 60 40 20 PREDICT AR mos 32.3 weeks (95% CI: 20.7, 48.3) PREDICT AR+ mos 75.6 weeks (95% CI: 51.6, 91.4) 0 Patients at risk PREDICT AR+ PREDICT AR 0 8 16 24 33 41 49 61 64 56 62 53 55 49 46 45 37 Data cutoff 1Jul2015 ITT = intent to treat; mos = median survival; CI = confidence interval;. 42 27 Weeks 40 24 32 13 PREDICT AR+ mos 18.0 months PREDICT AR mos 7.5 months 15 6 11 6 85 3 2 50 NCT01889238 Courtesy of J. Cortes, ECCO 2015

Metastatic triple negative BC Glembatumumab Vedotin in GPNMB+ TNBC GPNMB, glycoprotein NMB Yardley DA, et al. J Clin Oncol. 2015;33(14):1609-1619.

Metastatic triple negative BC Phase II Trial Sacituzumab Govitecan Met TNBC 3/4/5 th -Line Phase II Median DoR 7.6 months Med PFS 5.5 months >90% TNBCs express Trop-2 CT, computed tomography; MRI, magnetic resonance imaging Bardia A, et al. Presented at: San Antonio Breast Cancer Symposium; December 5-9, 2017: San Antonio, Texas.

Challenges AR-targeting maybe PARPi Yes! My bet will work in somatically inactivated? I-O?

Conclusions Select the right partner and validate studies with the same backbone Demonstrate bioactivity and not MTD Metastatic breast cancer is not always the right setting Neoadjuvant Post-neoadjuvant can be more informative