Clinical Spotlight Key Advances in Hormone Receptor Positive Breast Cancer: Updates From the 2018 Oncology Annual Meeting in Chicago

Clinical Spotlight Key Advances in Hormone Receptor Positive Breast Cancer: Updates From the 2018 Oncology Annual Meeting in Chicago Reference Slide Deck Abstract 1000 Abstract 1001 Abstract 1002 Abstract 1005 Abstract LBA1006

Ribociclib + Fulvestrant in Postmenopausal Women With Hormone Receptor-Positive, HER2-Negative Advanced Breast Cancer: Results From MONALEESA-3 Abstract 1000 Slamon DJ, Neven P, Chia S, Im S-A, Fasching PA, De Laurentis M, Petrakova K, Bianchi GV, Esteva FJ, Martin M, Pivot X, Vidam G, Wang Y, Lorenc KR, Miller M, Taran T, Jerusalem G

MONALEESA-3: Phase III Placebo-Controlled Study of Ribociclib + Fulvestrant Postmenopausal women and men with HR+/HER2- ABC No or 1 line of prior endocrine therapy for advanced disease N = 726 Randomization (2:1) Stratified by: Presence/absence of liver/lung metastases Prior endocrine therapy Ribociclib (600 mg/day orally; 3-weekson/1-week-off) + Fulvestrant (500 mg)* n = 484 Placebo + Fulvestrant (500 mg)* n = 242 Tumor assessments were performed every 8 weeks for 18 months, then every 12 weeks thereafter Primary analysis planned after ~364 PFS events Primary endpoints PFS (locally assessed per RECIST v1.1) Secondary endpoints Overall survival Overall response rate Clinical benefit rate Time to response Duration of response Time to definitive deterioration of ECOG PS Patient-reported outcomes Safety Pharmacokinetics 95% power to detect a 33% risk reduction (hazard ratio 0.67) with one-sided α = 2.5%, corresponding to an increase in median PFS to 13.4 months (median PFS of 9 months for the placebo arm), and a sample size of 660 patients Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000. ECOG PS, Eastern Cooperative Group performance status; PFS, progression-free survival; RECIST, Response Evaluation Criteria in Solid Tumors *Fulvestrant administered intramuscularly on cycle 1 day 1, cycle 1 day 15, and day 1 of every 28-day cycle thereafter

Accrual and Analysis Details 726 patients randomized between June 2015 and June 2016 Data cutoff date: November 3, 2017 (361 events) Median time from randomization to data cut-off date: 20.4 months Ribociclib + fulvestrant N = 484 Full Analysis Set Placebo + fulvestrant N = 242 PFS events N = 210 PFS events N = 151 Received treatment N = 484 Safety Set Received treatment N = 241 Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Prior Endocrine Therapy Status Criteria First line (ie, treatment-naïve for ABC) Relapse >12 months after completion of (neo)adjuvant endocrine therapy OR De novo advanced/metastatic disease (no prior exposure to endocrine therapy) Second line + early relapsers (ie, received up to 1 line of prior endocrine therapy for ABC) Early relapse on or 12 months from completion of (neo)adjuvant endocrine therapy OR Relapse >12 months from completion of (neo)adjuvant endocrine therapy with subsequent progression after 1 line of endocrine therapy (antiestrogen/ai) for ABC OR ABC at diagnosis that progressed after 1 line of endocrine therapy (antiestrogen/ai) for ABC ABC, advanced breast cancer Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Patient Disposition Characteristic* Ribociclib + Fulvestrant n = 484 Placebo + Fulvestrant n = 242 Treatment ongoing 204 (42.1) 76 (31.4) Treatment discontinued 279 (57.6) 165 (68.2) Primary reason for treatment discontinuation Disease progression 193 (39.9) 142 (58.7) Adverse events 41 (8.5) 10 (4.1) Physician decision 22 (4.5) 7 (2.9) Patient/guardian decision 21 (4.3) 5 (2.1) Death 2 (0.4) 0 Protocol deviation 1 (0.2) 1 (0.4) Technical problems 0 1 (0.4) Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Primary Endpoint: PFS (Investigator-Assessed) 100 Probability of PFS, % 80 60 40 20 0 Ribociclib + Placebo + PFS (investigator Fulvestrant Fulvestrant assessment) n = 484 n = 242 Events, n (%) 210 (43.4) 151 (62.4) Median PFS, months (95% CI) 20.5 (18.5 23.5) Hazard ratio (95% CI) 0.593 (0.480 0.732) One-sided P value.00000041 12.8 (10.9 16.3) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant 0 2 4 6 8 10 12 14 16 18 20 22 24 26 Time, Months 484 242 403 195 365 168 347 156 324 144 305 134 282 116 259 106 235 95 155 53 78 27 52 14 13 4 0 0 The hazard ratio of 0.593 corresponds to a 41% reduction in risk of progression in the ribociclib vs placebo arm Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

PFS Subgroup Analysis Subgroup Events, N/N Favors Ribociclib Favors Placebo Hazard Ribociclib + Fulvestrant Placebo + Fulvestrant Ratio 95% CI All patients 210/484 151/242 0.593 0.480 0.732 Prior endocrine First line b 76/238 66/129 0.577 0.415 0.802 therapy a Second line + early relapsers c 131/236 84/109 0.565 0.428 0.744 Liver or lung Yes 116/242 77/121 0.645 0.483 0.861 involvement No 94/242 74/120 0.563 0.415 0.764 Bone lesion only Yes 36/103 35/51 0.379 0.234 0.613 No 174/381 116/190 0.658 0.519 0.833 Age <65 years 115/258 81/129 0.607 0.454 0.810 65 years 95/226 70/113 0.597 0.436 0.818 Race Asian 22/45 7/18 1.353 0.574 3.186 Caucasian 174/406 136/213 0.562 0.448 0.704 Other 8/18 3/6 0.881 0.199 3.907 ECOG PS 0 126/310 95/158 0.559 0.427 0.733 1 83/173 56/83 0.633 0.450 0.890 Number of metastatic sites <3 126/309 92/149 0.586 0.447 0.768 3 84/175 59/92 0.621 0.441 0.874 Prior tamoxifen Yes 79/193 63/104 0.620 0.443 0.866 No 131/291 88/137 0.562 0.428 0.738 Prior AI Yes 135/257 80/118 0.670 0.507 0.886 No 75/227 71/123 0.481 0.345 0.669 0.125 0.25 0.5 1 2 4 8 Hazard ratio (95% CI) Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000. a 14 patients were not included in the prior endocrine therapy subgroup analysis due to missing data or criteria not being met; b Treatment naive for ABC; c Received up to 1 line of prior endocrine therapy for ABC. Hazard ratios were estimated based on stratified Cox proportional hazards model except in the subgroups related to the stratification factors (presence or absence of lung or liver metastases and prior endocrine therapy), where an unstratified analysis was used.

100 PFS by Prior Endocrine Therapy Status First Line a Second Line + Early Relapsers b 100 Probability of PFS, % 80 60 40 20 0 No. at risk Ribociclib + fulvestrant Placebo + fulvestrant PFS (investigator assessment) Ribociclib + Fulvestrant n = 238 Placebo + Fulvestrant n = 129 Events, n (%) 76 (31.9) 66 (51.2) Median PFS, months NR 18.3 Hazard ratio (95% CI) 0.577 (0.415 0.802) 0 2 4 6 8 10 12 14 16 18 20 22 24 238 129 205 109 189 99 180 91 173 88 166 85 Time, Months 159 78 149 75 141 68 97 40 49 18 31 10 7 4 26 0 0 Probability of PFS, % 80 60 40 20 0 No. at risk Ribociclib + fulvestrant Placebo + fulvestrant PFS (investigator assessment) Ribociclib + Fulvestrant n = 236 Placebo + Fulvestrant n = 109 Events, n (%) 131 (55.5) 84 (77.1) Median PFS, months 14.6 9.1 Hazard ratio (95% CI) 0.565 (0.428 0.744) 0 2 4 6 8 10 12 14 16 18 20 22 24 236 109 188 83 167 67 159 63 143 54 132 47 Time, Months 117 36 104 29 91 25 55 12 28 8 20 4 5 0 26 0 0 Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000. a Treatment naïve for ABC; b Received up to 1 line of prior endocrine therapy for ABC

100 Secondary Endpoints: ORR and CBR All Patients 100 Patients With Measurable Disease Rate, % 80 60 40 20 32.4 P =.000912 21.5 Rate, % 80 60 40 20 40.9 P =.003 28.7 Ribociclib + fulvestrant Placebo + fulvestrant 0 ORR 0 ORR In all patients, the CBR was 70.2% for ribociclib + fulvestrant vs 62.8% for placebo + fulvestrant (P =.020) In patients with measurable disease, the CBR was 69.4% for ribociclib + fulvestrant vs 59.7% for placebo + fulvestrant (P =.015) CBR, clinical benefit rate; ORR, overall response rate Overall response rate = complete response + partial response. Clinical benefit rate = complete response + partial response + (stable disease + noncomplete response/nonprogressive disease 24 weeks) in all patients and complete response + partial response + (stable disease 24 weeks) in patients with measurable disease. Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Overall Survival Ribociclib + Fulvestrant n = 484 Placebo + Fulvestrant n = 242 Events, n (%) 70 (14.5) 50 (20.7) Overall survival data were immature at the cutoff date Results did not cross the prespecified O Brien Fleming stopping boundary Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Hematologic Adverse Events Regardless of study treatment relationship AEs 5% in Either Arm, % Ribociclib + Fulvestrant n = 483 Placebo + Fulvestrant n = 241 All Grade 3 Grade 4 All Grade 3 Grade 4 Neutropenia a 69.6 46.6 6.8 2.1 0 0 Leukopenia b 28.4 13.5 0.6 1.7 0 0 Anemia c 17.2 3.1 0 5.4 2.1 0 Thrombocytopenia d 8.5 0.8 0.2 1.7 0 0 Febrile neutropenia was observed in 5 (1.0%) patients in the ribociclib arm vs none in the placebo arm a Includes neutropenia, decreased neutrophil count, febrile neutropenia, and neutropenic sepsis b Includes leukopenia, decreased white blood cell count, lymphopenia, and decreased lymphocyte count c Includes anemia, decreased hemoglobin, and decreased red blood cell count d Includes thrombocytopenia and decreased platelet count Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Nonhematologic Adverse Events Regardless of study treatment relationship AEs 5% in Either Arm, % Ribociclib + Fulvestrant n = 483 Placebo + Fulvestrant n = 241 All Grade 3 Grade 4 All Grade 3 Nausea 45.3 1.4 0 28.2 0.8 0 Fatigue 31.5 1.7 0 33.2 0.4 0 Diarrhea 29.0 0.6 0 20.3 0.8 0 Vomiting 26.7 1.4 0 12.9 0 0 Arthralgia 24.0 0.6 0 26.6 0.4 0 Grade 4 Post-baseline QTcF >480 ms, based on ECG data, occurred in 27 patients (5.6%) in the ribociclib arm vs 6 patients (2.5%) in the placebo arm Grade 3/4 elevated ALT and AST occurred in 32 (6.6%)/9 (1.9%) and 23 (4.8%)/6 (1.2%) patients in the ribociclib arm, respectively Two patients in the ribociclib arm were confirmed Hy s Law cases; AST, ALT, and TBIL levels returned to normal following ribociclib discontinuation ALT, alanine aminotransferase; AST, aspartate aminotransferase; ECG, electrocardiogram; TBIL, total bilirubin Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Conclusions Patients receiving ribociclib + fulvestrant had a statistically significant and clinically meaningful improvement in PFS vs placebo + fulvestrant Hazard ratio: 0.593; P =.00000041; 41% reduction in risk of disease progression vs placebo Ribociclib treatment benefit was consistent across patient subgroups Prolonged PFS was observed with first-line ribociclib + fulvestrant (hazard ratio: 0.577; 95% CI: 0.415 0.802) Benefit was also observed in patients who received treatment in the second-line setting (hazard ratio: 0.565; 95% CI: 0.428 0.744) Ribociclib + fulvestrant demonstrated a manageable safety profile, consistent with previous phase III ribociclib studies Ribociclib combined with fulvestrant may be a new first- or second-line treatment option for postmenopausal women with HR+/HER2 ABC This is the first study to show that CDK4/6 inhibitor + fulvestrant combinations are efficacious in patients with de novo ABC and patients with disease that relapsed >12 months after completion of prior (neo)adjuvant endocrine therapy Slamon DJ, et al. J Clin Oncol. 2018;36(suppl): Abstract 1000.

Genetic Landscape of Resistance to CDK4/6 Inhibition in Circulating Tumor DNA (ctdna) Analysis of the PALOMA3 Trial of Palbociclib and Fulvestrant Versus Placebo and Fulvestrant Abstract 1001 Turner NC, O Leary B, Cutts R, Liu Y, Hrebien S, Huang X, Beaney M, Fenwick K, Andre F, Loibl S, Loi S, Garcia-Murillas I, Bartlett CH, Cristofanilli M

CDK4/6 Inhibitors in Breast Cancer Treatment Palbociclib is an oral selective inhibitor of CDK4/6 Palbociclib in combination with endocrine therapy is a standard of care for patients with advanced hormone receptor positive, HER2-negative breast cancer A number of potential mechanisms of acquired resistance to CDK4/6 inhibitors have been identified in preclinical work, 1-3 although the relevance of these mechanisms to the clinic is largely unknown 1. Condorelli R, et al. Ann Oncol. 2017;29(3):640-645. 2. Herrera-Abreu MT, et al. Cancer Res. 2016;76(8):2301-2313. 3. Yang C, et al. Oncogene. 2017;36(16):2255-2264. Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

PALOMA-3 Study Design n = 347 qd; N = 521 1 wk off n = 174 Plasma samples for ctdna analysis were banked at baseline and end of treatment Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

Aims To identify resistance mutations acquired during palbociclib plus fulvestrant using paired baseline and end-of-treatment circulating tumor DNA (ctdna) analysis To identify differential mutation profiles between palbociclib plus fulvestrant vs placebo plus fulvestrant (fulvestrant alone) Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

Paired ctdna Analysis Methods Paired ctdna sequencing in 193 paired baseline and end-of-treatment plasma 125 palbociclib plus fulvestrant, 68 fulvestrant alone Amplicon error-corrected sequencing Targeted panel of 17 targetable driver and CDK4/6-related genes Calling to 0.5% allele frequency for whole genes and 0.3% for hot spots Capture exome sequencing in 14 patients Variant calling, copy number analysis, clonality analysis Targeted panel coding exons of RB1, CDK4, CDK6, CDKN1A, CDKN1B, NF1, TP53, and mutational hot spots in ERBB2, PIK3CA, AKT1, ESR1, FGFR1, FGFR2, FGFR3, KRAS, NRAS, and HRAS Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

PALOMA-3 ITT Versus Biomarker Population Progression-Free Survival Probability, % Progression-Free Survival Probability (%) ITT N = 521 Biomarker Subset N = 193 Palbociclib + Fulvestrant (n = 347) 100 Median Palbociclib+Fulvestrant PFS = 11.2 months (N=347) Median PFS=11.2 months 90 95% CI (9.5, 12.9) 95% CI (9.5, 12.9) Placebo Placebo+Fulvestrant + Fulvestrant (N=174) (n = 174) 80 Median PFS=4.6 = 4.6 months 95% 70 CI (3.5, 5.6) 60 50 40 30 20 Hazard Ratio=0.497 = 0.497 10 95% CI CI (0.398, 0.620) 2-sided p<0.000001 P<.0000001 0 0 2 4 6 8 10 12 14 16 18 20 22 Time, Months (Month) Number of patients at risk PAL+FUL 347 276 245 215 189 168 137 69 38 12 2 1 PCB+FUL174 112 83 62 51 43 29 15 11 4 1 0 2 4 6 8 10 12 14 16 18 20 Time, Months (Month) Number of patients at risk The ITT and biomarker populations had similar baseline characteristics Progression-Free Progression-Free Survival Survival Probability, Probability % (%) PAL+FUL PCB+FUL Palbociclib + Fulvestrant (n = 125) 100 Median Palbociclib+Fulvestrant PFS = 9.2 months (N=125) Median PFS=9.2 months 90 95% 95% CI CI (5.6, (5.6, 11.1) 11.1) Placebo+Fulvestrant + (N=68) (n = 68) 80 Median PFS=3.6 = 3.6 months 70 95% CI (2.0, 5.5) 60 50 40 30 20 Hazard Ratio=0.569 = 0.569 10 95% CI (0.408, 0.794) 2-sided p=0.0007 P<.0007 0 125 89 72 67 59 51 40 22 14 7 1 68 42 30 24 18 15 8 5 5 1 ITT, intent to treat Turner NC, et al. N Engl J Med. 2015;373:209-219. Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

PIK3CA and RB1 Mutations Are Enriched at EOT P =.00018 P =.00018 P values from McNemar s test, mutations increased in frequency at EOT No mutations were observed in CDK4, CDK6 (not shown) EOT, end of treatment Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

Time On Palbociclib Plus Fulvestrant Associates With Acquisition of Mutations Patients without acquired mutation at EOT Patients with acquired mutation at EOT 80 30 Patient 60 40 Patient 20 20 10 0 0 5 10 15 20 PFS Time (Month) Disease Progression 0 0 5 10 15 20 PFS Time (Month) Disease Progression Median 7.4 months P =.0066 Median 13.6 months Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

Mutations 5 0 RB1 Mutations Are Acquired On Palbociclib RB1 Q257X S565X S443X N519fs S588fs W681X DUF3452 RB_A RB_B Rb_C 0 200 400 600 800 928 aa A llele fraction 0.6 0.4 0.2 Patient 390 No RB1 mutations were detected at baseline 0% (0/193) RB1 mutations were acquired in: 4.8% (6/125) patients on palbociclib and fulvestrant 0% (0/68) on fulvestrant alone All RB1 mutations detected were truncating N ot detected Day 1 EOT RB1 N519fs PIK3CA E545K RB1 Q257X Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

PIK3CA Mutations Are Acquired in Both Treatment Arms + EOT - PIK3CA mutations a Day 1 + - 18.1% 0% (35) (0) 8.3% 73.6% (16) (142) P =.00018 PIK3CA mutations acquired in: 7.2% (9/125) patients on palbociclib plus fulvestrant and Day 1 End of treatment 10.3% (7/68) on fulvestrant alone Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001. a both treatment groups combined P values from McNemar s test

ESR1 Mutations Are Both Lost and Acquired During Treatment + EOT - ESR1 mutations a Day 1 + - 16.1% 6.7% (31) (13) 13.0% 64.2% (25) (124) p = 0.07 ESR1 mutations acquired in: 15.2% (19/125) patients on palbociclib and fulvestrant Day 1 End of treatment 8.8% (6/68) on fulvestrant alone Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001. a both treatment groups combined P values from McNemar s test

ESR1 Y537S Is Likely Selected by Fulvestrant ESR1 mutations* Day 1 End of treatment Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001. a both treatment groups combined P values from McNemar s test

Conclusions First assessment of resistance mechanisms to palbociclib plus fulvestrant in a randomized study Driver mutations acquired in both treatment arms PIK3CA mutations ESR1 Y537S mutation Other oncogenic mutations at low frequency (ERBB2 and FGFR1/2) Mutations acquired in the palbociclib arm RB1 mutations, which were infrequent These findings suggest parallel evolution of resistance to fulvestrant and CDK4/6 inhibitors and may inform future treatment strategies Turner NC, et al. J Clin Oncol. 2018;36(suppl): Abstract 1001.

Abemaciclib for Pre/Peri-Menopausal Women With HR+, HER2- Advanced Breast Cancer Abstract 1002 Neven P, Rugo HS, Tolaney SM, Iwata H, Toi M, Goetz MP, Kaufman PA, Barriga S, Lin Y, Sledge Jr GW

HR+, HER2- ABC Pre/peri- a or postmenopausal ET resistant: Relapsed on neoadjuvant or on/within 1 yr of adjuvant ET Progressed on first-line ET No chemo for MBC No more than 1 ET for MBC ECOG PS 1 MONARCH 2 Study Design N = 669 Patients were enrolled in 142 centers in 19 countries Abemaciclib: 150 mg b BID (continuous schedule) Fulvestrant: 500 mg c Placebo: bid (continuous schedule) Fulvestrant: 500 mg c Statistics: 378 events for 90% power at one-sided α of.025 assuming a true HR of 0.703 114 pre/perimenopausal patients were randomized in the study 2:1 Randomization Primary endpoint: Investigator-assessed PFS Secondary endpoints: Overall survival, response, Clinical benefit rate, Safety Stratification factors: Metastatic site (visceral, bone only, or other) ET resistance (primary vs secondary) 1,2 bid, twice daily dose; ET, endocrine therapy; HER2, human epidermal growth factor receptor 2; HR, hazard ratio; HR, hormone receptor; MBC, metastatic breast cancer 1. Cardoso F, et al. Breast. 2017;31:244-259. 2. Cardoso F, et al. Ann Oncol. 2017;28(1):16-33. Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002. a Aged <60 years and have natural menstrual bleeding. Patients were required to receive gonadotropin-releasing hormone (GnRH) agonist b Dose postamendment c Fulvestrant administered per label

Baseline Characteristics of Pre/Perimenopausal Patients Abemaciclib + Fulvestrant n = 72 Placebo + Fulvestrant n = 42 Median age (range) 46 (32-57) 47 (32-66) Race, n (%) Most recent ET, n (%) a Prior AI, n (%) Sensitivity to ET, n (%) a PgR status, n (%) Metastatic site, n (%) Measurable disease, n (%) Asian Caucasian Other Neoadjuvant or adjuvant Metastatic Yes No 51 (70.8) 14 (19.4) 7 (9.7) 44 (61.1) 26 (36.1) 10 (13.9) 62 (86.1) Primary resistance b 28 (38.9) Secondary resistance c 42 (58.3) Positive Negative Visceral Bone only Other Yes No 54 (75.0) 18 (25.0) 43 (59.7) 19 (26.4) 10 (13.9) 51 (70.8) 21 (29.2) 24 (57.1) 16 (38.1) 2 (4.7) 21 (50.0) 20 (47.6) 12 (28.6) 30 (71.4) 15 (35.7) 26 (61.9) 38 (90.5) 4 (9.5) 17 (40.5) 15 (35.7) 10 (23.8) 28 (66.7) 14 (33.3) AI, aromatase inhibitor; PgR, progesterone receptor a 2 patients in the abemaciclib arm and 1 patient in the placebo arm received no prior ET b Patients whose disease relapsed 2 years while receiving (neo)adjuvant ET or progressed 6 months of receiving ET for ABC [Cardoso F, et al. Breast. 2017;31:244-259. Cardoso F, et al. Ann Oncol. 2017;28(1):16-33.] c Patients receiving prior ET who do not meet the definition of primary resistance were considered to have secondary resistance Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Progression-Free Survival, % 100 80 60 40 20 ITT Population Patients at risk Abemaciclib 446 367 314 281 234 171 101 65 32 2 0 Placebo 223 165 123 103 80 61 32 13 4 1 0 Progression-Free Survival Median PFS abemaciclib + fulvestrant: 16.4 months placebo + fulvestrant: 9.3 months HR (95% CI):.553 (.449,.681); P<.00000001 Progression-Free Survival, % Pre-/Perimenopausal Population placebo + fulvestrant: 10.5 months 0 0 0 3 6 9 12 15 18 21 24 27 30 0 4 8 12 16 20 24 28 Time, Months Time, Months 100 80 60 40 20 Median PFS abemaciclib + fulvestrant: not reached HR (95% CI):.446 (.264,.754); P =.002 Patients at risk Abemaciclib 72 59 52 47 36 20 10 0 Placebo 42 34 24 15 11 4 2 0 PFS benefit confirmed by BICR HR: 0.460; 95% CI: 0.363-0.584, P<.00001 PFS benefit confirmed by BICR HR: 0.423; 95% CI: 0.236-0.793, P<.005 BICR, blinded independent central review Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Change in Tumor Size and Response Rate Change in Tumor Size, % abemaciclib + fulvestrant Pre/perimeopausal ORR CR CBR ITT [N = 72] 43.1% 2.8% 77.8% Measurable (n = 51) 60.8% 3.9% 74.5% Change in Tumor Size, % placebo + fulvestrant Pre/perimeopausal ORR CR CBR ITT [N = 42] 19.0% 0% 69.0% Measurable (n = 28) 28.6% 0% 71.4% CR, complete response a 4 patients without post-baseline tumor assessment are not depicted Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Time to Chemotherapy ITT Population Pre-/Perimenopausal Population Chemotherapy-Free Survival, % 100 80 60 40 20 Median PFS abemaciclib + fulvestrant: not reached placebo + fulvestrant: 26.3 months HR (95% CI):.65 (.50,.85); P<.002 0 0 3 6 9 12 15 18 21 24 27 30 Time, Months Patients at risk Abemaciclib 446 410 377 350 331 283 195 124 67 13 0 Placebo 223 193 171 154 142 124 91 47 24 4 0 Progression-Free Survival, % 100 80 60 40 20 Median PFS abemaciclib + fulvestrant: not reached placebo + fulvestrant: 19.2 months HR (95% CI):.61 (.32, 1.15); P<.123 0 0 4 8 12 16 20 24 28 32 Time, Months Patients at risk Abemaciclib 72 65 58 55 42 29 17 0 0 Placebo 42 39 35 30 26 12 5 0 0 Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Treatment-Emergent Adverse Events Abemaciclib + Fulvestrant n = 71 Placebo + Fulvestrant n = 42 20 in Abemaciclib Arm All G2 G3 G4 All G2 G3 G4 Any, n(%) 70 (98.6) 20 (28.2) 40 (56.3) 4 (5.6) 40 (95.2) 17 (40.5) 7 (16.7) 0 Diarrhea 62 (87.3) 22 (31.0) 8 (11.3) 0 10 (23.8) 1 (2.4) 0 0 Neutropenia a 42 (59.2) 9 (12.7) 28 (39.4) 2 (2.8) 3 (7.1) 1 (2.4) 1 (2.4) 0 Leukopenia 31 (43.7) 15 (21.1) 12 (16.9) 0 2 (4.8) 1 (2.4) 0 0 Infections and infestations 31 (43.7) 26 (36.6) 1 (1.4) 0 11 (26.2) 7 (16.7) 2 (4.8) 0 Abdominal pain 25 (35.2) 2 (2.8) 0-5 (11.9) 2 (4.8) 0 - Anemia 24 (33.8) 15 (21.1) 7 (9.9) 0 1 (2.4) 0 0 0 Headache 24 (33.8) 7 (9.9) 0-13 (31.0) 2 (4.8) 0 - Vomiting 23 (32.4) 5 (7.0) 1 (1.4) 0 3 (7.1) 0 1 (2.4) 0 Nausea 20 (28.2) 6 (8.5) 1 (1.4) - 10 (23.8) 3 (7.1) 1 (2.4) - a 1 patient (1.4%) experienced grade 3 febrile neutropenia in the abemaciclib arm. Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Grade 2 & 3 Diarrhea: Pre- and Post-Dose Amendment Pre Amendment (200 mg starting dose) n = 20 Post Amendment (150 mg starting dose) n = 51 50 50 Patients, % 40 30 20 Grade 2 Grade 3 Patients, % 40 30 20 Grade 2 Grade 3 10 10 0 0 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 17 19 Cycles Cycles Events Due to Diarrhea Pre/Post Amendment Discontinuation 5% / 0% Grade 2 55% / 21.6% Grade 3 15% / 9.8% Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Conclusions Abemaciclib plus fulvestrant and a GnRH agonist significantly improved PFS (median NR vs 10.5 months; HR: 0.446) and ORR (60.8% vs 28.6% in patients with measurable disease) in pre/peri-menopausal women with ABC. These results are consistent with the ITT population Abemaciclib plus fulvestrant and a GnRH agonist led to substantial tumor regression Abemaciclib and fulvestrant delayed the time to initiation of subsequent chemotherapy in the ITT population. Similar results were observed in pre/peri-menopausal patients Diarrhea associated with abemaciclib was generally predicable (occurred early), manageable (with conventional doses of antidiarrheal medication and dose reduction), and reversible Abemaciclib dosed on a continuous schedule plus fulvestrant demonstrated a manageable safety profile consistent with other studies of abemaciclib. No additional toxicities were observed by adding a GnRH agonist to abemaciclib plus fulvestrant Neven P, et al. J Clin Oncol. 2018;36(suppl): Abstract 1002.

Everolimus + Exemestane vs Everolimus Alone or Capecitabine for Estrogen Receptor-Positive, HER2 Advanced Breast Cancer Abstract 1005 Jerusalem G, Kovalenko E, Yardley DA, de Boer R, Hurvitz S, Ejlertsen B, Blau S, Özgüroğlu M, Landherr L, Ewertz M, Taran T, Fan J, Noel-Baron F, Louveau A-L, Burris H

Randomized, Open-Label, Phase II Study BOLERO-6 randomized 309 patients to receive EVE + EXE (n = 104), EVE alone (n = 103), or CAP (n = 102) Eligibility Criteria Postmenopausal women with ER+ HER2 metastatic or recurrent BC, or locally advanced BC not amenable to curative surgery or radiotherapy Recurrence or progression on ANA or LET Measurable disease per RECIST v1.1 or bone lesions (lytic or mixed), and ECOG PS 0 2 N = 309 Randomization (1:1:1) a EVE 10 mg PO qd + EXE 25 mg PO qd (n = 104) EVE 10 mg PO qd (n = 103) CAP 1250 mg/m 2 PO bid (2 weeks on, 1 week off) (n = 102) Primary Objective Estimate HR of investigatorassessed PFS for EVE + EXE vs EVE alone b Key Secondary Objective Estimate HR of PFS for EVE + EXE vs CAP b Other Secondary Endpoints OS, ORR, CBR, and safety BOLERO-6 was not powered to perform statistical comparisons between arms a Stratified by presence or absence of visceral disease (lung, liver, heart, ovary, spleen, kidney, adrenal gland, malignant pleural or pericardial effusion, or malignant ascites; b Stratified multivariate Cox regression models were adjusted on treatment and the following prognostic and baseline covariates where imbalances between arms were observed: bone-only lesions (yes vs no); prior chemotherapy (yes vs no); ECOG PS (0 vs 1 2); organs involved (2 vs 1, and 3 vs 1); race (Caucasian vs non-caucasian); age (<65 vs 65 years). ANA, anastrozole; bid, twice daily; LET, letrozole; NSAI, nonsteroidal aromatase inhibitor; OS, overall survival; PO, oral administration; qd, once daily Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Characteristic, % Median age (range), years <65 years, % Race, % Caucasian Other ECOG PS, % a 0 1 2 Metastatic site of cancer, % b Visceral c (excluding CNS) Bone Bone only Other Metastatic sites, % 1 2 3 Baseline Characteristics EVE + EXE n = 104 61.0 (32 86) 63 75 25 52 45 66 85 13 64 50 50 EVE Alone n = 103 61.0 (38 88) 62 83 17 47 51 CAP n = 102 59.5 (35 84) 68 89 11 56 42 More patients in CAP- vs EVE-containing arms: <65 years of age Caucasian a Data missing for 7 patients in the EVE + EXE (n = 3), EVE alone (n = 2), and CAP (n = 2) arms; b CNS in 2 patients in the EVE alone (n = 1) and CAP (n = 1) arms; c Lung, liver, heart, ovary, spleen, kidney, adrenal gland, malignant pleural or pericardial effusion, or malignant ascites. CNS, central nervous system 64 77 16 59 54 46 62 83 24 59 56 44 Fully active (ECOG PS 0) Had bone-only metastases Fewer patients in CAP- vs EVE-containing arms had 3 metastatic sites Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Primary Objective: Estimated HR of PFS for EVE + EXE vs EVE Alone EVE + EXE offers a PFS benefit vs EVE alone PFS, % 100 90 80 70 60 50 40 30 20 10 0 Patients still at risk EVE + EXE EVE alone n/n mpfs, HR* (90% CI) Months Censoring EVE + EXE 80/104 8.4 0.74 (0.57 0.97) EVE alone 74/103 6.8 0 3 6 9 12 15 18 21 24 27 30 33 36 Time, Months 104 73 52 39 26 19 11 10 10 10 9 5 1 0 103 66 40 26 14 9 7 4 4 4 2 1 0 0 39 Estimated HR of PFS for EVE + EXE vs EVE alone was 0.74 (90% CI 0.57 0.97) Censored for initiating new antineoplastic therapies: EVE + EXE arm, 9% EVE alone arm, 18% A stratified multivariate Cox regression model accounting for baseline imbalances and known prognostic factors gave a consistent HR (0.73; 90% CI 0.56 0.97) for EVE + EXE vs EVE alone Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Key Secondary Objective: Estimated HR of PFS for EVE + EXE vs CAP CAP may have been favored by baseline imbalances and potential informative censoring PFS, % 100 90 80 70 60 50 40 30 20 10 0 Patients still at risk EVE + EXE CAP n/n mpfs, HR* (90% CI) months Censoring EVE + EXE 80/104 8.4 1.26 (0.96 1.66) CAP 68/102 9.6 0 3 6 9 12 15 18 21 24 27 30 33 36 Time, Months 39 42 104 73 52 39 26 19 11 10 10 10 9 5 1 0 0 102 68 48 38 33 26 19 14 10 9 6 3 2 1 0 Estimated HR of PFS for EVE + EXE vs CAP was 1.26 (90% CI 0.96 1.66) Censored for initiating new antineoplastic therapies: EVE + EXE arm, 9% CAP arm, 20% A stratified multivariate Cox regression model accounting for baseline imbalances and known prognostic factors gave a HR of 1.15 (90% CI 0.86 1.52) for EVE + EXE vs CAP Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005. a EVE + EXE vs CAP (obtained from a stratified Cox model)

OS, % 100 90 80 70 60 50 40 30 20 10 0 Patients still at risk EVE + EXE EVE alone CAP Overall Survival: EVE + EXE vs EVE Alone or CAP n/n 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 Time, Months 104 101 92 81 74 67 63 53 48 43 39 22 13 8 3 1 0 0 103 96 86 81 72 69 66 57 55 49 43 27 21 11 4 2 0 0 102 94 88 83 78 70 64 61 54 43 38 31 21 16 7 3 1 0 mos, months HR* (90% CI) Censoring EVE + EXE 71/104 23.1 EVE alone 59/103 29.3 1.27 (0.95 1.70) CAP 58/102 25.6 1.33 (0.99 1.79) New antineoplastic therapies initiated at EOT: EVE + EXE arm, 78% EVE alone arm, 81% CAP arm, 79% A stratified multivariate Cox regression model accounting for baseline imbalances and known prognostic factors gave a HR of 1.27 (90% CI 0.94 1.70) for EVE + EXE vs EVE alone and a HR of 1.19 (90% CI 0.88 1.62) for EVE + EXE vs CAP mos, overall survival Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Adverse Events AE, a % EVE + EXE n = 104 EVE alone n = 103 CAP n = 102 All Grades Grade 3 4 All Grades Grade 3 4 All Grades Grade 3 4 Total 100 70 98 59 100 74 Stomatitis b 49 9 46 5 25 7 Fatigue 38 8 31 3 35 8 Diarrhea 35 5 33 3 54 8 Anemia 32 13 25 10 22 7 Most frequent all-grade AEs: Stomatitis in EVE-containing arms PPE syndrome and diarrhea in CAP arm Elevated GGT 15 9 16 12 2 2 Elevated AST 15 7 14 8 9 1 Hypertension 14 6 8 2 5 3 Hyperglycemia 13 4 17 8 8 1 Pneumonia 11 7 9 3 3 2 Neutropenia 4 0 4 2 15 6 PPE syndrome 3 1 3 0 61 27 Grade 3-4 AEs more frequent in EVE + EXE arm vs EVE alone arm, and comparable between EVE + EXE and CAP arms a 5% grade 3 4 events in any arm; b BOLERO-6 was not designed to use the SWISH 1 protocol for stomatitis prevention. AE, adverse event; AST, aspartate aminotransferase; GGT, gamma-glutamyl transferase; PPE, palmar-plantar erythrodysesthesia. 1. Rugo HS, et al. Lancet Oncol 2017;18(5):654-662. Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Other Safety All-grade serious AEs Regardless of causality 50 All-grade AEs leading to discontinuation Regardless of causality 50 Patients, % 40 30 20 10 37 (36%) 30 (29%) 30 (29%) Patients, % 40 30 20 10 18 (17%) 20 (19%) 21 (21%) 0 EVE + EXE n = 104 EVE Alone n = 103 CAP n = 102 0 EVE + EXE n = 104 EVE Alone n = 103 CAP n = 102 Serious AEs more frequent with EVE + EXE vs EVE alone or CAP Incidence of AEs leading to discontinuation comparable in each arm Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Deaths Deaths, n All study deaths EVE + EXE n = 104 71 EVE Alone n = 103 59 CAP n = 102 58 Disease progression 67 53 53 AE On-treatment deaths a 4 9 6 5 5 2 Disease progression 6 2 0 AE 3 3 2 188 patients died during the study Disease progression most frequent reason in each arm Few deaths occurred due to AEs 16 of 188 deaths occurred on-treatment a Disease progression most frequent reason in EVE + EXE arm Incidence of death due to AEs comparable in each arm Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005. a Occurring up to 30 days after treatment discontinuation.

Conclusions Median PFS with EVE + EXE (8.4 months) consistent with BOLERO-2 (7.8 months), 1 and vs EVE alone here (6.8 months) corresponded to estimated 26% reduction of risk of disease progression or death (HR 0.74) Median PFS with EVE alone numerically longer than previously reported in a small phase II study (3.5 months) 2 No new safety signals observed with EVE + EXE A numerical median PFS difference was observed for CAP over EVE + EXE (9.6 vs 8.4 months), which may be attributed to various baseline characteristics favoring CAP and potential informative censoring Median PFS with CAP also inconsistent with previous studies (4.1 7.9 months) 3-7 1. Yardley DA, et al. Adv Ther. 2013;30(10):870-884. 2. Ellard SL, et al. J Clin Oncol. 2009;27(27):4536-4541. 3. Robert NJ, et al. J Clin Oncol. 2011;29(10):1252-1260. 4. O'Shaughnessy JA, et al. Oncologist. 2012;17(4):476-484. 5. Stockler MR, et al. J Clin Oncol. 2011;29(34):4498-4504. 6. Kaufmann M, et al. Eur J Cancer. 2010;46(19):3184-3191. 7. Harbeck N, et al. Breast Cancer Res Treat. 2017;161(1):63-72. Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Conclusions Interpretation of the results of BOLERO-6 must consider the limited sample size and open-label design Nonetheless, the results suggest that EVE + EXE offers a PFS benefit vs EVE alone, supporting the continued use of this combination in this treatment setting The benefit risk profile of EVE + EXE is unchanged from BOLERO-2 Jerusalem G, et al. J Clin Oncol. 2018;36(suppl): Abstract 1005.

Phase III Study of Taselisib (GDC-0032) + Fulvestrant (FULV) in Patients (pts) With Estrogen Receptor (ER)- Positive, PIK3CA-Mutant (MUT), Locally Advanced or Metastatic Breast Cancer (MBC): Primary Analysis From SANDPIPER Abstract LBA1006 Baselga J, Dent S, Cortés J, Im Y-H, Diéras V, Harbeck N, Krop IE, Verma S, Wilson TR, Jin H, Wang L, Schimmoller F, Hsu JY, He J, De Laurentiis M, Drullinsky P, Jacot W

PI3K Signaling Is Frequently Dysregulated 1 and PI3K Inhibition Augments ER Function and Dependence in Hormone Receptor Positive Breast Cancer 2,3 PI3K signaling is involved in tumor growth, proliferation, and survival, and is frequently activated in solid tumors 1 The PI3K pathway may be activated by gain-offunction mutations and/or amplification of the PIK3CA gene 1,4-7 PIK3CA encodes the α-isoform of the catalytic subunit of PI3K (PI3Kα) Mutations in PIK3CA are detected in ~40% of ER-positive, HER2-negative breast cancer 8 There is significant crosstalk between the ER and PI3K signaling pathways; inhibition of PI3K results in an adaptive upregulation of ER signaling 2,3 AKT, protein kinase B; ER, estrogen receptor; mtor, mammalian target of rapamycin; PI3K, phosphatidylinositol 3-kinase; PIK3CA, phosphatidylinositol-4,5-bisphosphate 3 kinase, catalytic subunit alpha 1. Fruman DA, et al. Cell. 2017;170(4):605-635. 2. Bosch A, et al. Sci Transl Med. 2015;7(283):283ra51. 3. Toska E, et al. Science. 2017;355(6331):1324-1330. 4. Samuels Y, et al. Science. 2004;304(5670):554. 5. Zhang Y, et al. Cancer Cell. 2017;31(6):820-832. 6. Zehir A, et al. Nat Med. 2017;23(6):703-713. 7. Janku F, et al. Nat Rev Clin Oncol. 2018;15(5):273-291. 8. Arthur LM, et al. Breast Cancer Res Treat. 2014;147(1):211-219. Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

SANDPIPER Study Design ER-positive/HER2-negative locally advanced or metastatic BC Postmenopausal women Recurrence or progression during or after aromatase inhibitor No more than one chemotherapy regimen for MBC No prior fulvestrant, PI3K, or mtor inhibitor PIK3CA-mutant tumors: N = 480 (planned) Without PIK3CA-mutant tumors: N = 120 (planned) Taselisib 4 mg qd + fulvestrant 2:1 randomization Placebo + fulvestrant Taselisib 4 mg qd + fulvestrant 2:1 randomization Placebo + fulvestrant Treat until PD or unacceptable toxicity No crossover Survival data Primary Endpoint: INV-PFS in patients with PIK3CA-mutant tumors (central cobas test a ) Secondary Endpoints: ORR, OS, CBR, DoR, BICR-PFS in patients with PIK3CA-mutant tumors Safety Exploratory Endpoint: Efficacy in patients without PIK3CA-mutant tumors Stratification: 1. Visceral disease 2. Endocrine sensitivity 3. Geographic region Endocrine sensitivity: 1) If no endocrine treatment in advanced or MBC, 24 months of adjuvant endocrine treatment prior to recurrence; 2) Documented clinical benefit (CR, PR, or SD 24 weeks) to most recent endocrine treatment in advanced or MBC. a cobas test detects the following PIK3CA mutations: R88Q, N345K, C420R, E542K, E545A/G/K/D, Q546K/R/E/L, M1043I, H1047L/R/Y, and G1049R. BICR, blinded independent central radiology; CR, complete response; DoR, duration of response; INV, investigator-assessed; MBC, metastatic breast cancer; ORR, objective response rate; OS, overall survival; PD, progressive disease; PFS, progression-free survival; SD, stable disease; qd, daily Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Patients With PIK3CA-Mutant Tumors Baseline characteristics were well balanced between treatment arms Placebo + Fulvestrant n = 176 Taselisib + Fulvestrant n = 340 Age, median (range) 61 (39, 85) 60 (32, 84) ECOG PS 0 / ECOG PS 1, % 52.8 / 47.2 54.4 / 45.6 Visceral disease*, % 58.5 59.1 Bone-only disease, % 18.2 20.6 Bone metastasis, % 72.2 78.5 Measurable disease, % 76.1 77.6 Endocrine sensitivity*, % 73.3 73.8 Prior endocrine therapy Prior adjuvant endocrine therapy 68.2 59.7 Prior endocrine therapy for MBC 68.8 74.7 Prior tamoxifen (regardless of setting), % 48.9 49.4 Prior CDK4/6 inhibitor, % 1.7 3.5 Prior chemotherapy Prior chemotherapy in MBC, % 27.8 32.1 Most recent therapy in MBC, % 72.7 77.9 Number of regimens in MBC, median (25%, 75%) 1 (0, 2) 1 (1, 2) Range 0, 6 0, 5 *Stratification factor. CDK, cyclin-dependent kinases; ECOG PS, Eastern Cooperative Oncology Group Performance Status Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Primary Endpoint: INV-PFS in Patients With PIK3CA-Mutant Tumors PFS was defined as the time from randomization to first disease progression as determined by investigator using RECIST v1.1, or death from any cause. Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Forest Plot of INV-PFS PIK3CA Mutant (1) PFS was defined as the time from randomization to first disease progression as determined by investigator using RECIST v1.1, or death from any cause. Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Forest Plot of INV-PFS PIK3CA Mutant (2) PFS was defined as the time from randomization to first disease progression as determined by investigator using RECIST v1.1, or death from any cause. Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Median DoR: 7.2 months 8.7 months Secondary Endpoints ORR and CBR in Patients With Measurable Disease Placebo + Fulvestrant n = 134 Taselisib + Fulvestrant n = 264 Responders, n (%) 16 (11.9) 74 (28.0) Difference in response rates (95% CI) 16.1 (8.4, 23.8) P value (Cochran-Mantel-Haenszel).0002 CBR in patients with and without measurable disease: 41.5% 54.4% OS CR, n (%) 0 2 (0.8) PR, n (%) 16 (11.9) 72 (27.3) CBR, n (%) 50 (37.3) 136 (51.5) Placebo + Fulvestrant n = 176 Taselisib + Fulvestrant n = 340 Events, n (%) 43 (24.4) 73 (21.5) CBR is defined as an objective response or no disease progression for 24 weeks since randomization; confirmation not needed for CR and PR Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Exploratory Endpoint: Efficacy in Patients Without PIK3CA-Mutant Tumors Patients With Measurable Disease At Baseline Placebo + Fulvestrant n = 35 Taselisib + Fulvestrant n = 61 Responders, % 14.3 19.7 P value difference in response rate P =.37 *Stratified log-rank Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Summary of AEs (Safety-Evaluable Patients; Regardless of Causality) MedDRA-Preferred Term (Unless Specified), n (%) Placebo + Fulvestrant n = 213 Taselisib + Fulvestrant n = 416 Diarrhea a 19.7 60.1 Hyperglycemia a 9.4 40.4 Nausea 24.4 34.1 Decreased appetite 10.3 26.4 Fatigue 17.5 24.3 Headache 11.7 20.2 Stomatitis a 8.5 33.2 Vomiting 11.3 18.8 Asthenia 18.3 18.5 Rash a 11.3 15.2 Cough 13.1 13.0 Back pain 11.3 13.0 Abdominal pain 8.9 12.3 Dry mouth 7.5 12.3 Arthralgia 12.7 115 Alopecia 2.8 11.3 Pruritis 7.5 11.1 Pyrexia 3.3 13.6 Dyspnea 8.0 10.3 MedDRA-Preferred Term (Unless Specified), n (%) Placebo + Fulvestrant n = 213 Taselisib + Fulvestrant n = 416 Diarrhea a 0.9 11.5 Hyperglycemia a 0.5 10.8 Rash* -- 3.8 Stomatitis a -- 3.6 ALT/AST increase 0.5 3.3 Colitis a -- 3.1 Hypertension 3.3 2.4 Dehydration 0.5 1.9 Lipase increased 0.9 1.7 Neutropenia 0.9 1.7 Vomiting 0.9 1.7 Pneumonitis a 0.5 1.7 Pneumonia -- 1.7 Sepsis 0.5 1.2 Diarrhea infectious -- 1.2 Hypokalemia -- 1.2 a Frequencies of selected AEs are based on group terms of relevant events associated with taselisib, no preferred terms. Selected AEs were grade 3, except for grade 5 pneumonitis in the placebo arm and two grade 4 hyperglycemia events in the taselisib arm. ALT, alanine aminotransferase; AST, aspartate aminotransferase Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

Conclusions SANDPIPER met its primary endpoint; the addition of taselisib to fulvestrant showed a statistically significant improvement in INV-PFS in patients with PIK3CA-mutant tumors Median INV-PFS 5.4 mos 7.4 mons, HR 0.70, P value.0037 Secondary endpoints, including ORR, CBR, DoR, and BICR-PFS, showed consistent improvement with the combination; OS data remain immature Exploratory analysis of PFS suggests that a treatment effect of taselisib + fulvestrant in patients without detectable PIK3CA mutation cannot be completely ruled out The combination of taselisib + fulvestrant had an expected safety profile, with gastrointestinal toxicities and hyperglycemia being the most frequent AEs The challenging tolerability of this combination led to frequent discontinuations and may have limited the clinical benefit in this disease setting Baselga J, et al. J Clin Oncol. 2018;36(suppl): Abstract LBA1006.

prime Points The combination of ribociclib and fulvestrant was clearly superior to fulvestrant alone and provides an alternative first-line strategy to the use of a CDK4/6 inhibitor in combination with an aromatase inhibitor Resistance mechanisms to CDK 4/6 remain unknown in majority of cases RB1 mutations occur in around 5% of patients with progression on palbociclib New ESR1 and PI3 kinase mutations may partly explain acquired fulvestrant resistance Abemaciclib, in combination with an aromatase inhibitor and GnRH analogue, is effective for pre/perimenopausal women with MBC Everolimus has clinical activity as a single agent but should continue to be administered in combination with exemestane The mutant-specific PI3 kinase inhibitor taselisib significantly improved PFS in patients with mutations in the PI3 kinase pathway