Zoneddy Dayao, Rachel Rabinovitch, Stephen H. Dyar, San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 (NRG Oncology) Evaluating Pathologic Complete Response Rates in Patients with Hormone Receptor-Positive, HER2-Positive Breast Cancer treated with Neoadjuvant Therapy of Docetaxel, Carboplatin, Trastuzumab, and Pertuzumab (TCHP) with or without Concurrent Estrogen Deprivation Therapy Mothaffar F. Rimawi, Reena S. Cecchini, Priya Rastogi, Charles E. Geyer, Jr, Louis Fehrenbacher, Philip J. Stella, Zoneddy Dayao, Rachel Rabinovitch, Stephen H. Dyar, Patrick J. Flynn, Luis Baez-Diaz, Soonmyung Paik, Sandra M. Swain, Eleftherios P. Mamounas, C. Kent Osborne, Norman Wolmark Thank you. It is my privilege to present to you, on behalf of my colleagues, the primary analysis of NSABP B52 conducted by NRG Oncology

San Antonio Breast Cancer Symposium, December 6-10, 2016 Rationale ER+/HER2+ tumors are less likely than ER-/HER2+ tumors to respond to dual anti-HER2 therapy. ER may act as a pathway of resistance to anti-HER2 treatment. Older trials suggested antagonistic effects of chemotherapy and endocrine therapy. The rationale for our trial is based on the fact that ER+/HER2+ tumors are less likely than ER-/HER2+ tumors to respond to dual HER2 inhibition, and that ER may act as a pathway of resistance to anti-HER2 treatment. Earlier trials that studied older chemotherapy agents combined with tamoxifen suggested antagonistic effects, but no data was available on combinations of more modern agents. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

San Antonio Breast Cancer Symposium, December 6-10, 2016 Hypothesis We hypothesized that concurrent inhibition of ER and HER2, plus chemotherapy, will not be antagonistic, and will overcome resistance to treatment thus improving pCR rates in pts with ER+/HER2+ breast cancer. We therefore hypothesized that concurrent inhibition of ER and HER2, plus chemotherapy, will not be antagonistic, and will overcome resistance to treatment thus improving pathCR rates in pts with ER+/HER2+ breast cancer. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

SABCS, December 6-10, 2016 NRG Oncology/NSABP B-52 HER2-Positive, ER and/or PgR-Positive Invasive Breast Cancer Diagnosed by Core Needle Biopsy REQUIRED BLOOD AND TISSUE STRATIFICATION RANDOMIZATION Arm 1 TCH every 21 days x 6 cycles + Pertuzumab Arm 2 TCH every 21 days x 6 cycles + Pertuzumab Estrogen Deprivation REQUIRED TISSUE Core biopsy of primary tumor before Cycle 3 of TCHP* *Obtained core biopsy in 103 pts. NSABP B52 is a randomized trial for patients with histologically confirmed ER-positive, HER2-positive breast cancer. After undergoing study required blood and tissue collection, eligible patients were stratified according to their tumor size, nodal involvement, and menopausal status, and randomized to Arm1: docetaxel, carboplatin, trastuzumab, and pertuzumab every 3 weeks, or Arm2: with that same treatment plus endocrine therapy with estrogen deprivation. After receiving 6 cycles, patients proceeded to surgery and axillary staging. SURGERY (lumpectomy or mastectomy) and axillary staging

San Antonio Breast Cancer Symposium, December 6-10, 2016 Eligibility Criteria Invasive adenocarcinoma of the breast diagnosed by core needle biopsy Clinical tumor ≥2.0 cm if clinically node negative. Any size if node positive. Tumors must be hormone receptor positive and HER2+ by ASCO/CAP The LVEF must be ≥50% regardless of the testing facility's lower limit of normal. Adequate organ function To be eligible, patients had to have: Invasive adenocarcinoma of the breast diagnosed by core needle biopsy Clinical tumor ≥2.0 cm if clinically node negative. Any size if node positive. Tumors must be hormone receptor positive and HER2+ by ASCO/CAP The LVEF must be ≥50% regardless of the testing facility's lower limit of normal. And, adequate organ function. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

} ~ 8 yrs after start of trial San Antonio Breast Cancer Symposium, December 6-10, 2016 Endpoints Primary pCR rate in the breast and nodes (ypT0-is ypN0) Secondary pCR rate in the breast Clinical complete response Toxicity Recurrence-free interval OS } ~ 8 yrs after start of trial The primary endpoint for NSABP B52 is pathologic complete response rate in the breast and lymph nodes Secondary endpoints included pathologic complete response in the breast, clincial complete response, and toxicity. All patients will continue to be followed for the assessment of recurrence-free interval, overall survival, and second primary invasive cancer of any type. These analyses are expected to be performed about 8 years after the initiation of the trial, a period at which all patients should have been followed for at least 5 years. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

Statistical Considerations San Antonio Breast Cancer Symposium, December 6-10, 2016 Statistical Considerations The expected rate of pCR in the group not treated with estrogen deprivation is 45%. Between January 2014 and February 2016, 315 patients were enrolled to provide 80% power to detect a 33% improvement, increasing the path CR rate from 45% to 60%. Our statistical design assumed a pathCR rate of 45% in the control arm, not treated with estrogen deprivation. Between January 2014 and February 2016, 315 patients were enrolled to provide 80% power to detect a 33% improvement, increasing the path CR rate from 45% to 60% with the addition of estrogen deprivation. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

NSABP B-52 Patient Characteristics* San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 Patient Characteristics* Age ≤ 49 46% 50 – 59 32% ≥ 60 22% Race White 79% Black 12% Other/Unk 9% Tumor staging cT0-cT2 74% cT3-cT4c 24% cT4d 2% Clinical Nodal Status Pos. 57% Neg. 43% * Patient characteristics were balanced between treatment regimens Patient characteristics were balanced between treatment groups. 46% of the patients were under age 50%, 79% were Caucasian and 12% were African American. 26% of study participants had T3 or more advanced tumors and 57% had clinically node positive disease. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

NSABP B-52 Toxicity Toxicity TCHP (n=154) TCHP +Est Dep (n=157) Gr 0-1 San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 Toxicity Toxicity TCHP (n=154) TCHP +Est Dep (n=157) Gr 0-1 Gr 2 Gr 3 Gr 4 Diarrhea 42% 34% 23% <1% 43% 35% 22% 0% Nausea 60% 31% 9% 65% 29% 6% Vomiting 82% 10% 8% 13% 5% Dehydration 71% 20% 78% 17% NSABP B52 enrolled the largest cohort treated with TCHP to date. TCHP has considerable toxicity with prominent GI toxicity like diarrhea, nausea, vomiting, and dehydration. There was no difference in toxicity between the two study arms. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

NSABP B-52 Toxicity Toxicity TCHP (n=154) TCHP +Est Dep (n=157) Gr 0-1 San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 Toxicity Toxicity TCHP (n=154) TCHP +Est Dep (n=157) Gr 0-1 Gr 2 Gr 3 Gr 4 Anemia 53% 35% 12% 0% 56% 26% 18% Hypokalemia 83% 5% 10% 2% 80% 8% 1% Febrile Neutropenia - <1% 7% Overall 3% 29% 59% 37% 52% 6% Other toxicities are listed here. Anemia and hypokalemia were common. Febrile neutropenia occurred in 6% in the control arm and 8% in the experimental arm. The overall rate of toxicity was quite high but not different in both study arms. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

NSABP B-52 pCR Breast and Nodes San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 pCR Breast and Nodes P=0.33 Now to efficacy results. NSABP B52 showed a pathologic complete response rate in the overall study population of 41% in the control arm and 46% in the experimental arm. This numeric increase was not statistically significant with a p-valuie of 0.39. Subgroup analysis by menopausal status showed the path CR rate to be 44% vs 46% in premenopausal women and 38% vs. 45% in postmenopausal women, again with non-significant p-values or 0.80 and 0.33 respectively. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

San Antonio Breast Cancer Symposium, December 6-10, 2016 NSABP B-52 pCR Breast P=0.51 I will now present secondary endpoint results Pathologic complete response in the breast was 44% in the control arm and 47% in the experimental arm. This rate was 48% and 49% in premenopausal women, and 40% and 45% in postmenopausal women. All of these rates had no statistically significant differences. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

San Antonio Breast Cancer Symposium, December 6-10, 2016 Conclusion The addition of estrogen deprivation to neoadjuvant chemotherapy was not antagonistic and did not increase toxicity. The combination increased pCR rates numerically, but the improvement was not statistically significant. Correlative science studies, evaluation of residual cancer burden (RCB), and long-term outcomes may help define the role of estrogen deprivation in the treatment of HER2+ early breast cancer. In conclusion, The addition of estrogen deprivation to neoadjuvant chemotherapy was not antagonistic and did not increase toxicity. The combination increased pCR rates numerically, but the improvement was not statistically significant. Correlative science studies, evaluation of residual cancer burden (RCB), and long-term outcomes may help define the role of estrogen deprivation in the treatment of HER2+ early breast cancer. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

San Antonio Breast Cancer Symposium, December 6-10, 2016 Conclusion Given the toxicity of standard chemotherapy observed on this trial, findings from NSABP B52 argue for a tailored de-escalation approach where toxic treatments are omitted or replaced with less toxic ones without compromising outcomes. And finally, Given the toxicity of standard chemotherapy observed on this trial, findings from NSABP B52 argue for a tailored de-escalation approach where toxic treatments are omitted or replaced with less toxic ones without compromising outcomes. This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.

Acknowledgment Patients and their families San Antonio Breast Cancer Symposium, December 6-10, 2016 Acknowledgment Patients and their families Co-authors and all participating NSABP investigators NSABP staff at the Operation and Biostatistical Center Collaborators from Baylor College of Medicine Funding Sources: NCI and Genentech I would like to acknowledge the patients who participated on this trial and their families, my co-authors, all participating NSABP investigators, NSABP staff at the operations and biostatistical center, my colleagues and collaborators at Baylor College of Medicine, the NCI and Genentech for funding the study, and all of you for your attention. Thank you This presentation is the intellectual property of the author. Contact at rimawi@bcm.edu for permission to reprint and/or distribute.