Ribociclib plus fulvestrant for postmenopausal women with hormone receptor-positive, human epidermal growth factor receptor 2-negative advanced breast cancer in the phase III randomized MONALEESA-3 trial: updated overall survival

ORIGINAL ARTICLE Ribociclib plus fulvestrant for postmenopausal women with hormone receptor-positive, human epidermal growth factor receptor 2-negative advanced breast cancer in the phase III randomized MONALEESA-3 trial: updated overall survival D. J. Slamon 1* , P. Neven 2 , S. Chia 3 , G. Jerusalem 4 , M. De Laurentiis 5 ,S.Im 6 , K. Petrakova 7 , G. Valeria Bianchi 8 , M. Martín 9 , A. Nusch 10 , G. S. Sonke 11 , L. De la Cruz-Merino 12 , J. T. Beck 13 ,Y.Ji 14 , C. Wang 15 , U. Deore 14 , A. Chakravartty 14 , J. P. Zarate 14 , T. Taran 15 & P. A. Fasching 16 1 David Geffen School of Medicine at UCLA, Los Angeles, USA; 2 Multidisciplinary Breast Centre, Universitair Ziekenhuis Leuven, Leuven, Belgium; 3 British Columbia Cancer Agency, Vancouver, Canada; 4 CHU Liege and Liège University, Liège, Belgium; 5 Istituto Nazionale Tumori IRCCS ‘Fondazione G. Pascale’, Naples, Italy; 6 Seoul National University Hospital, Cancer Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea; 7 Masaryk Memorial Cancer Institute, Brno, Czech Republic; 8 Fondazione Istituto di Ricovero e Cura a Carattere Scientifico, Istituto Nazionale dei Tumori, Milan, Italy; 9 Instituto de Investigación Sanitaria Gregorio Marañon, Centro de Investigación Biomédica en Red de Cáncer, Grupo Español de Investigación en Cáncer de Mama, Universidad Complutense, Madrid, Spain; 10 Practice for Hematology and Internal Oncology, Velbert, Germany; 11 Netherlands Cancer Institute/Borstkanker Onderzoek Groep Study Center, Amsterdam, The Netherlands; 12 Hospital Universitario Virgen Macarena, Seville, Spain; 13 Highlands Oncology Group, Fayetteville; 14 Novartis Pharmaceuticals Corporation, East Hanover, USA; 15 Novartis Pharma AG, Basel, Switzerland; 16 Department of Gynecology and Obstetrics, University Hospital Erlangen, Comprehensive Cancer Center Erlangen-EMN, Friedrich-Alexander University Erlangen, Erlangen, Germany Available online 5 June 2021 Background: Ribociclib plus fulvestrant demonstrated significant progression-free survival (PFS) and overall survival (OS) benefits in patients with hormone receptor-positive, human epidermal growth factor receptor 2-negative (HRþ/HER2) advanced breast cancer (ABC). Here we present a new landmark in survival follow-up for a phase III cyclin-dependent kinases 4 and 6 inhibitor clinical trial in patients with ABC (median, 56.3 months). Patients and methods: This phase III, randomized, double-blind, placebo-controlled trial was conducted at 174 sites (30 countries). Patients were men and postmenopausal women (age 18 years) with histologically/cytologically confirmed HRþ/HER2ABC. Patients could have received 1 line of endocrine therapy (ET) but no chemotherapy for ABC. Patients, assigned 2:1, were stratified by the presence/absence of liver/lung metastases and previous ET. Patients received intramuscular fulvestrant (500 mg, day 1 of each 28-day cycle plus day 15 of cycle 1) with oral ribociclib (600 mg/day, 3 weeks on, 1 week off) or placebo. Efficacy analyses were by intention to treat. Safety was assessed in patients receiving 1 dose study treatment. OS was a secondary endpoint. MONALEESA-3 is registered with ClinicalTrials.gov (NCT02422615; no longer enrolling). Results: Between 18 June 2015 and 10 June 2016, 726 patients were randomly assigned (484, ribociclib; 242, placebo). At data cut-off (30 October 2020), median OS (mOS) was 53.7 months (ribociclib) versus 41.5 months (placebo) [hazard ratio (HR), 0.73; 95% confidence interval (CI) 0.59-0.90]. Subgroup analyses were consistent with overall population. In the first-line setting, most patients in the ribociclib arm (w60%) lived longer than median follow-up; mOS was 51.8 months in the placebo arm (HR, 0.64; 95% CI 0.46-0.88). In the second-line setting, mOS was 39.7 months (ribociclib) versus 33.7 months (placebo) (HR, 0.78; 95% CI 0.59-1.04). No apparent drugedrug interaction between ribociclib and fulvestrant or new safety signals were observed. Conclusions: This analysis reported extended OS follow-up in MONALEESA-3. mOS was w12 months longer in patients with HRþ/HER2ABC treated with ribociclib plus fulvestrant compared with fulvestrant monotherapy. Key words: ribociclib, CDK4/6 inhibitor, advanced breast cancer, overall survival INTRODUCTION Cyclin-dependent kinases 4 and 6 (CDK4/6) are crucial to cell cycle progression and have become an effective target in the treatment of advanced breast cancer (ABC). 1-3 Ribociclib is a selective small-molecule inhibitor of CDK4/ 6, and is approved in combination with fulvestrant for the *Correspondence to: Dr Dennis J. Slamon, Medicine, Hematology & Oncology at the David Geffen School of Medicine at UCLA, 2825 Santa Monica Blvd., Suite 200, Santa Monica, CA 90404, USA. Tel: þ1310-206-6909 E-mail: [email protected] (D. J. Slamon). 0923-7534/© 2021 The Authors. Published by Elsevier Ltd on behalf of European Society for Medical Oncology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Volume 32 -Issue 8 -2021 https://doi.org/10.1016/j.annonc.2021.05.353 1015 treatment of postmenopausal patients with hormone receptor-positive/human epidermal growth factor receptor 2-negative (HRþ/HER2)ABCasfirst- or second-line therapy following progression on endocrine therapy (ET). 4 Ribociclib plus fulvestrant showed a significant benefitin progression-free survival (PFS) versus placebo plus fulvestrant with a median PFS (mPFS) of 20.5 versus 12.8 months [hazard ratio (HR), 0.59; 95% confidence interval (CI) 0.48- 0.73, P<0.001], which was reported as the primary endpoint for the MONALEESA-3 trial. 4 Overall survival (OS) is a secondary endpoint in phase III clinical trials of CDK4/6 inhibitors, 5-10 and improving OS while maintaining quality of life is the ultimate goal of any new therapy in ABC. Three trials have previously reported OS results of CDK4/6 inhibitors in combination with fulvestrant. 6,8,10 The PALOMA-3 trial showed a median OS (mOS) of 34.9 versus 28.0 months for palbociclib plus fulvestrant versus placebo plus fulvestrant (HR, 0.81; 95% CI 0.64-1.03); however, this difference did not reach statistical significance (P¼0.09). 10 A significant OS benefit was demonstrated in the MONARCH-2 trial, with an mOS of 46.7 months for abemaciclib plus fulvestrant versus 37.3 months for placebo plus fulvestrant (HR, 0.76; 95% CI 0.61-0.95; P¼ 0.01). 8 The MONALEESA-3 trial also reported a significant OS benefit for ribociclib plus fulvestrant versus placebo plus fulvestrant, with an mOS of not reached (NR) versus 40.0 months (HR, 0.72; 95% CI 0.57-0.92; P¼0.00455). 6 The MONALEESA-7 trial of ribociclib plus ET in pre-/ perimenopausal women with HRþ/HER2ABC recently reported an exploratory OS analysis with a median followup of 53.5 months. 11 In this exploratory OS update, ribociclib plus ET showed a clinically relevant and significant mOS benefit of 58.7 months compared with 48.0 months in the placebo group (HR, 0.76; 95% CI 0.61-0.96). 11 Similar to MONALEESA-7, we undertook an exploratory OS update for MONALEESA-3 with an extended follow-up (median, 56.3 months) in order to analyze the long-term OS benefitof ribociclib plus fulvestrant versus placebo plus fulvestrant. PATIENTS AND METHODS Study design and participants As previously reported, MONALEESA-3 was a randomized, double-blind, placebo-controlled, phase III trial in which patients were randomized 2:1 to receive fulvestrant with either oral ribociclib or matching placebo. 4 MONALEESA-3 allowed enrollment of men and postmenopausal women of at least 18 years of age, with histologically and/or cytologically confirmed HRþ/HER2ABC that was locoregionally recurrent or metastatic and not amenable to curative therapy. Patients were required to have an Eastern Cooperative Oncology Group performance score of 0 or 1 and measurable disease according to RECIST version 1.1 or at least one predominantly lytic bone lesion. Patients included those who had received no prior ET for advanced disease, up to one line of ET for advanced disease, or had relapsed during or within 12 months of completion of (neo)adjuvant ET. Patients who received prior chemotherapy in the advanced setting, any previous treatment with fulvestrant, or any prior CDK4/6 inhibitor were not enrolled. Patients who received first-line treatment included those who were newly diagnosed (de novo) with ABC, or who had relapsed >12 months from completion of (neo)adjuvant ET with no treatment for advanced or metastatic disease. Patients who received treatment in the second line or early relapse included those who experienced relapse on or within 12 months from completion of (neo)adjuvant ET with no treatment for advanced or metastatic disease (early relapse), relapse >12 months from completion of (neo)adjuvant ET with subsequent progression after one line of ET for advanced or metastatic disease, and finally those with advanced or metastatic breast cancer at diagnosis who progressed after one line of ET for advanced disease with no prior (neo) adjuvant treatment for early disease. Patients were not eligible if they could not receive ET as per investigator judgment or if they had clinically significant cardiac arrhythmias and/or uncontrolled heart disease. Written informed consent was obtained from all patients. The trial was conducted in accordance with the Good Clinical Practice guidelines and Declaration of Helsinki. The study protocol and any modifications were approved by an independent ethics committee or institutional review board at each site. A steering committee, comprising participating investigators and Novartis representatives, supervised the study conduct. An independent data monitoring committee assessed the safety data. Randomization and masking Randomized numbers were generated by the interactive response technology provider using a validated system that automated random assignment of patients into treatment arms, which were in turn linked to medication numbers. A separate medication randomization list was produced using a validated system that automated the random assignment of numbers to medication packs. All patientsdas well as investigators who administered treatment, assessed endpoints, and analyzed datadwere blinded to the trial group assignments until the final OS analysis. Randomization was stratified according to the presence or absence of liver or lung metastases and previous ET in the advanced setting. Unblinding only occurred during the trial for safety reasons, for regulatory reporting purposes, and at the conclusion of the study. Crossover was not permitted until the final OS analysis was completed. Patients and investigators were unblinded after the final OS analysis, following which, patients still receiving study treatment in the placebo arm were given the option to switch to ribociclib. Annals of Oncology D. J. Slamon et al. 1016 https://doi.org/10.1016/j.annonc.2021.05.353 Volume 32 -Issue 8 -2021 Procedures Patients received either oral ribociclib (600 mg/day on a 3- week-on, 1-week-off schedule) or matching placebo. Both groups received intramuscular fulvestrant (500 mg, day 1 of every 28-day cycle, with an additional dose on day 15 of cycle 1). Tumor response was assessed locally as per RECIST 1.1 at screening, every 8 weeks after randomization for 18 months, every 12 weeks until month 36, and then as clinically indicated until disease progression, death, withdrawal of consent, loss to follow-up, or subject/guardian decision; for patients who discontinued for any other reason, assessments continued as per protocol. Survival follow-up continued for patients who discontinued study treatment. Adverse events were monitored and graded according to the Common Terminology Criteria for Adverse Events (version 4.03). 12 Safety follow-up was conducted for at least 30 days after patients’last study treatment dose. Outcomes The primary endpoint of investigator-assessed PFS and the secondary endpoint of OS were previously reported. 4,6 OS was a protocol-specified secondary endpoint and defined as the time from randomization to death from any cause. 6 PFS2, time to chemotherapy, and chemotherapy-free survival were additional exploratory endpoints that have been previously reported. 6 PFS2 was defined as the time from randomization to the first documented disease progression following discontinuation from study treatment while the patient was receiving next-line therapy (as reported by the investigator) or death from any cause, whichever occurred first. Time to chemotherapy was defined as the time from randomization to the beginning of the first subsequent chemotherapy following discontinuation of study treatment, and chemotherapy-free survival was defined as time to first chemotherapy or death. To evaluate ribociclib pharmacokinetics (PK), plasma samples were collected at predose; 2, 4, and 6 h postdose at cycle 1 day 15 and cycle 2 day 15; and predose on cycle 2 day 1. Statistical analysis Statistical methods for the protocol-prespecified final analysis of PFS and OS were previously reported. 4,6 For the original analysis, an estimated total of 660 patients were needed for 95% power to detect an HR of 0.67 for the primary endpoint. The primary analysis of investigatorassessed PFS has been previously reported with the data cut-off date of 3 November 2017. 4 At the time of the final OS analysis (data cut-off date 3 June 2019), 153 patients were still receiving study treatment (121 of 484 in the ribociclib group and 32 of 242 in the placebo group). 6 The mOS was NR for the ribociclib arm at the time of the final OS analysis. In the current follow-up (data cut-off date 30 October 2020), sufficient events were reported in the ribociclib arm to provide a stable estimate of mOS. In this exploratory analysis, mOS and OS rates were estimated using the KaplaneMeier method. The HR for OS was estimated with the use of a stratified Cox proportional hazards model. Patients were censored at the date the patient was last known to be alive. Analyses were carried out on data from the following subgroups: patients receiving first-line therapy, patients receiving second-line therapy, including patients with early relapse [within 12 months after completion of (neo)adjuvant ET]. Additional subgroups of interest included the presence/absence of liver or lung metastases and response to prior ET. Endocrine resistance was defined as a relapse while on the first 2 years of (neo)adjuvant ET or progressive disease within the first 6 months of first-line ET for ABC while receiving ET. Endocrine naïve was defined as patients who did not receive any prior ET in any setting. Patients not described as ET resistant or ET naïve were deemed to be ET sensitive. In addition to the updates of OS results, exploratory analyses for PFS2, time to chemotherapy, chemotherapy-free survival, and PK were also carried out. The rank-preserving structural-failure time (RPSFT) model was used as a sensitivity analysis on OS to assess the effects of crossover and administration of subsequent CDK4/6 inhibitors in the placebo group. 13 Role of the funding source The funder of the study, in conjunction with the authors and the study steering committee members, designed this study. Representatives of the trial sponsor carried out data collection and analysis. All authors reviewed and approved the data, contributed to the development and approval of the manuscript, and accepted the decision to submit the manuscript for publication. RESULTS Overall survival From 18 June 2015 to 10 June 2016, a total of 726 patients were randomly assigned: 484 to the ribociclib group and 242 to the placebo group (Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2021.05.353). Previously published efficacy analyses included the baseline characteristics. 4 At the cut-off date for this exploratory OS analysis (30 October 2020), 68 (14.0%) of 484 patients in the ribociclib group and 21 (8.7%) of 242 patients in the placebo group were still receiving study treatment. Following the final OS analysis, two patients crossed over from placebo to ribociclib. At a median 56.3-month followup (minimum, 52.7 months), 222 (45.9%) of 484 patients receiving ribociclib and 142 (58.7%) of 242 patients receiving placebo had died. A significant OS benefit, with an mOS of 53.7 months (95% CI 46.9-NR months) in the ribociclib arm versus 41.5 months (95% CI 37.4-49.0 months) in the placebo arm (HR, 0.73; 95% CI 0.59-0.90), was observed (Figure 1A). KaplaneMeier-estimated 4-year survival rates were 54% (95% CI 49% to 58%) and 45% (95% CI 38% to 51%) for ribociclib and placebo, respectively, while the 5-year survival rates were estimated to be 46% D. J. Slamon et al. Annals of Oncology Volume 32 -Issue 8 -2021 https://doi.org/10.1016/j.annonc.2021.05.353 1017 0 2 4 6 8 10 12 14 16 18 20 22 26 2824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Overall survival (%) Months 100 484 242 470 233 454 227 444 223 436 218 428 213 414 208 402 200 397 195 389 188 374 185 365 175 334 160 326 156 348 170 310 149 301 143 290 137 285 133 273 127 256 120 245 112 241 107 233 102 224 99 218 96 191 82 90 29 145 56 50 15 20 5 0 0 4 1 Ribociclib + fulvestrant Placebo + fulvestrant 222 142 53.7 41.5 (n = 484) (n = 242) Hazard ratio = 0.726 (95% CI 0.588-0.897) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Overall survival (%) Months 100 237 128 229 126 222 125 217 122 214 121 210 119 207 117 206 114 205 111 202 107 194 105 190 100 174 94 173 92 182 98 167 87 164 86 159 84 157 81 151 76 146 73 141 69 139 67 136 65 133 63 129 62 116 55 57 17 89 38 32 8 10 4 0 0 1 0 Ribociclib + fulvestrant Placebo + fulvestrant 84 67 NR 51.8 (n = 237) (n = 128) Hazard ratio = 0.640 (95% CI 0.464-0.883) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant NR, not reached. 0 2 4 6 8 10 12 14 16 18 20 22 26 2824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Overall survival (%) Months 100 237 110 231 104 222 99 218 98 213 94 210 91 199 89 188 84 184 82 179 79 172 78 167 73 152 64 145 62 158 70 135 60 129 55 123 51 120 50 114 49 103 45 97 41 96 38 92 35 87 34 85 32 71 26 32 11 53 17 17 7 10 1 3 1 0 0 Ribociclib + fulvestrant Placebo + fulvestrant 134 74 39.7 33.7 (n = 237) (n = 110) Hazard ratio = 0.780 (95% CI 0.587-1.037) A B C No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant Median overall survival months Events, n Median overall survival months Events, n Median overall survival months Events, n Figure 1. Overall survival. (A) Overall survival in all patients. (B) Overall survival in patients receiving ribociclib plus fulvestrant as first-line therapy. (C) Overall survival in patients receiving ribociclib plus fulvestrant as second-line therapy. CI, confidence interval; NR, not reached. Annals of Oncology D. J. Slamon et al. 1018 https://doi.org/10.1016/j.annonc.2021.05.353 Volume 32 -Issue 8 -2021 (95% CI 40% to 52%) and 31% (95% CI 23% to 40%), respectively. Analysis of subgroups according to prior lines of ET was also carried out. Of the 365 patients who received study treatment as a first-line therapy, 84 (35.4%) of 237 patients in the ribociclib group and 67 (52.3%) of 128 patients in the placebo group died (Figure 1B). Patients in the first-line subgroup had an mOS of NR (95% CI 59.9-NR months) in the ribociclib group and 51.8 months (95% CI 40.4-57.6 months) in the placebo group (HR, 0.64; 95% CI 0.46-0.88). At 56 months (approximately the median follow-up time), the OS rate in the ribociclib group was 60.5%. In the firstline subgroup, the 4-year OS rates were estimated to be 66% (95% CI 59% to 72%) and 53% (95% CI 44% to 62%) in the ribociclib and placebo groups, respectively. The 5-year OS rates were estimated to be 54% (95% CI 41% to 65%) and 36% (95% CI 23% to 49%) in the ribociclib and placebo groups, respectively. Of the 347 patients who received study treatment as a second-line therapy, 134 (56.5%) of 237 patients in the ribociclib group and 74 (67.3%) of 110 patients in the placebo group died (Figure 1C). Patients in the second-line subgroup had an mOS of 39.7 months (95% CI 37.4-46.9 months) in the ribociclib group and 33.7 months (95% CI 27.8-41.3 months) in the placebo group (HR, 0.78; 95% CI 0.59-1.04). In the second-line subgroup, the 4-year OS rates were estimated to be 42% (95% CI 35% to 48%) and 34% (95% CI 25% to 43%) in the ribociclib and placebo groups, respectively, while the 5-year OS rates were estimated to be 37% (95% CI 31% to 44%) and 24% (95% CI 15% to 36%) in the ribociclib and placebo groups, respectively. Of 364 patients with lung or liver metastases, 118 (48.8%) of 242 patients in the ribociclib group and 77 (63.1%) of 122 patients in the placebo group died (Figure 2). Patients with lung or liver metastases had an mOS of 46.9 months (95% CI 38.1-NR months) with ribociclib versus 39.4 0.125 0.25 0.5 1 2 4 8 16 32 n of deaths/total N (%) Ribociclib + fulvestrant Placebo + fulvestrant Hazard ratio (95% Cl) Placebo betterRibociclib better First-line Early relapse or second line Yes No Yes No <3 ≥3 (Neo) adjuvant Metastatic <65 ≥65 1 0 Asian Caucasian Other Asia Europe and Australia Latin America North America Other − + ++ Other Endocrine naive Endocrine resistance Endocrine sensitive All patients Treatment line of ET for advanced disease Liver or lung involvement Bone lesion only Number of metastasis sites Most recent therapy Age, years ECOG score Race Region PGR ER + PGR Prior endocrine therapy Subgroup 142/242 (58.7) 67/128 (52.3) 74/110 (67.3) 77/122 (63.1) 65/119 (54.6) 29/51 (56.9) 113/190 (59.5) 84/147 (57.1) 58/94 (61.7) 94/152 (61.8) 25/41 (61.0) 75/129 (58.1) 67/113 (59.3) 89/158 (56.3) 53/83 (63.9) 7/18 (38.9) 130/214 (60.7) 2/5 (40.0) 7/16 (43.8) 107/173 (61.8) 1/3 (33.3) 24/43 (55.8) 3/7 (42.9) 97/167 (58.1) 41/69 (59.4) 97/167 (58.1) 45/74 (60.8) 39/74 (52.7) 18/25 (72.0) 85/140 (60.7) 0.73 (0.59-0.90) 0.64 (0.46-0.88) 0.78 (0.59-1.04) 0.73 (0.55-0.98) 0.74 (0.54-1.01) 0.67 (0.42-1.08) 0.74 (0.58-0.93) 0.73 (0.55-0.96) 0.74 (0.53-1.04) 0.78 (0.60-1.01) 0.67 (0.42-1.09) 0.73 (0.54-0.97) 0.72 (0.53-0.99) 0.67 (0.51-0.88) 0.78 (0.55-1.10) 0.99 (0.40-2.43) 0.69 (0.55-0.86) 1.26 (0.23-6.83) 0.82 (0.32-2.05) 0.69 (0.54-0.88) 2.33 (0.24-22.78) 0.76 (0.43-1.33) 0.84 (0.16-4.40) 0.71 (0.55-0.92) 0.79 (0.53-1.19) 0.70 (0.54-0.91) 0.81 (0.56-1.19) 0.62 (0.41-0.95) 0.82 (0.45-1.47) 0.73 (0.56-0.96) 222/484 (45.9) 84/237 (35.4) 134/237 (56.5) 118/242 (48.8) 104/242 (43.0) 45/102 (44.1) 177/382 (46.3) 132/308 (42.9) 90/176 (51.1) 130/264 (49.2) 57/112 (50.9) 116/258 (45.0) 106/226 (46.9) 131/311 (42.1) 90/172 (52.3) 21/45 (46.7) 185/407 (45.5) 10/17 (58.8) 17/40 (42.5) 164/347 (47.3) 5/6 (83.3) 30/69 (43.5) 6/22 (27.3) 155/353 (43.9) 56/113 (49.6) 152/350 (43.4) 70/134 (52.2) 48/139 (34.5) 34/53 (64.2) 139/289 (48.1) Figure 2. Exploratory analyses of overall survival in subgroups. ER and PGR status þþ means that patients were positive for both ER and PGR. CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; ER, estrogen receptor; ET, endocrine therapy; PGR, progesterone receptor. D. J. Slamon et al. Annals of Oncology Volume 32 -Issue 8 -2021 https://doi.org/10.1016/j.annonc.2021.05.353 1019 months (95% CI 29.9-44.9 months) with placebo (HR, 0.73; 95% CI 0.55-0.98). Overall, 507 study patients had received prior ET (342 in the ribociclib group and 165 in the placebo group); 53 (15.5%) and 25 (15.2%) patients were endocrine resistant in the ribociclib and placebo groups, respectively, while 289 (84.5%) and 140 (84.8%) were endocrine sensitive. In the endocrine-resistant subgroup, the mOS was 35.6 months (95% CI 28.6-42.2 months) with ribociclib compared with 31.7 months (95% CI 13.0-41.5 months) with placebo. In the endocrine-sensitive subgroup, the mOS was 49.0 months (95% CI 40.7-NR months) with ribociclib and 41.8 months (95% CI 33.0-48.6 months) with placebo. Of the 213 patients who were ET naïve, 48 (34.5%) of 139 in the ribociclib group and 39 (52.7%) of 74 in the placebo group died (Supplementary Figure S1, available at https://doi. org/10.1016/j.annonc.2021.05.353). Patients in the ET naïve subgroup had an mOS of 59.9 months (95% CI 59.9- NR months) in the ribociclib group and 50.9 months (95% CI 37.4-NR months) in the placebo group (HR, 0.62; 95% CI 0.41-0.95). Results for OS analyses in other subgroups were generally consistent with the overall population. Subsequent therapy The percentage of patients reporting subsequent therapies after discontinuation of study treatment was comparable between study arms: 340 patients (81.9%) in the ribociclib group and 190 (86.4%) in the placebo group (Table 1). As reported in the final OS analysis, the most common first subsequent therapies were hormonal therapy alone (28% and 21%), chemotherapy alone (23% and 20%), and hormonal therapy plus targeted therapy/other (21% and 31%) for ribociclib and placebo, respectively. The use of CDK4/6 inhibitors any time after discontinuation of study treatment was lower in the ribociclib (14%) than in the placebo (30%) group. Subsequent chemotherapy, alone or in combination, at any time after study treatment was received by 215 (44.4%) patients in the ribociclib group and 131 (54.1%) patients in the placebo group. The median time to chemotherapy (time from randomization to the beginning of the first subsequent chemotherapy following discontinuation of study treatment) was 48.1 months (95% CI 38.2-NR months) versus 28.8 months (95% CI 24.3-37.5 months) with ribociclib versus placebo (HR, 0.70; 95% CI 0.57-0.88), respectively (Figure 3A). The median chemotherapy-free survival (time to first chemotherapy or death) was 32.3 months (95% CI 28.1-38.5 months) in patients receiving ribociclib versus 22.4 months (95% CI 19.4-26.1 months) in patients receiving placebo (HR, 0.69; 95% CI 0.57-0.83) (Figure 3B). Progression-free survival 2 Overall, 265 (54.8%) and 163 (67.4%) patients treated with ribociclib and placebo, respectively, had disease progression while receiving subsequent therapy or died of any cause (Figure 4A). The mPFS2 was 37.4 months (95% CI 31.1-42.6 months) in the ribociclib group and 28.1 months (95% CI 24.0-31.6 months) in the placebo group (HR, 0.7069; 95% CI 0.57-0.84). In the subgroup of patients who received study treatment as first-line therapy, 107 (45.1%) patients in the ribociclib group and 82 (64.1%) patients in the placebo group had disease progression while receiving subsequent therapy or died of any cause (Figure 4B). The mPFS2 was 53.7 versus 35.5 months with ribociclib and placebo, respectively (HR, 0.63; 95% CI 0.47-0.84). For patients who received study treatment in the second line, 153 (64.6%) patients treated with ribociclib and 80 (72.7%) patients treated with placebo had progression while receiving subsequent therapy or died of any cause. The mPFS2 in the ribociclib group was 26.0 months compared with 20.5 months in the placebo group (HR, 0.73; 95% CI 0.56-0.96) (Figure 4C). Sixty-eight patients in the placebo group went on to receive ribociclib or another CDK4/6 inhibitor as a subsequent therapy after discontinuation, including 66 patients after discontinuing study treatment and 2 patients who crossed over from placebo to ribociclib (Supplementary Table S1, available at https://doi.org/10.1016/j.annonc. 2021.05.353). A RPSFT model-based sensitivity analysis was used to account for this. After adjusting for the subsequent CDK4/6 inhibitor treatment, the mOS in the placebo group was estimated to be 40.4 months (95% CI 35.4-47.2 months) (HR, 0.70; 95% CI 0.55-0.88) compared with 41.5 months (95% CI 37.4-49.0 months) in the main analysis (HR, 0.73; 95% CI 0.59-0.90). Safety Adverse events were consistent with those previously reported in the primary PFS and final OS analyses 4,6 and were generally more common in patients treated with ribociclib (Supplementary Table S2, available at https://doi.org/10. Table 1. Subsequent antineoplastic therapies among patients who discontinued the trial regimen Variable Ribociclib D fulvestrant n[484 Placebo D fulvestrant n[242 Patients who discontinued the trial regimen, n 415 220 Patients who received any subsequent therapy, n(%) 340 (81.9) 190 (86.4) First subsequent antineoplastic therapy Chemotherapy alone 96 (23.1) 44 (20.0) Chemotherapy plus hormone therapy or other therapy a 36 (8.7) 29 (13.2) Hormone therapy alone 115 (27.7) 47 (21.4) Hormone therapy plus other therapy b 88 (21.2) 55 (18.0) Targeted therapy alone 5 (1.2) 1 (0.5) Patients who received any subsequent CDK4/6 inhibitor, n(%) 58 (14.0) 66 (30.0) Palbociclib 36 (8.7) 52 (23.6) Ribociclib 14 (3.4) 11 (5.0) Abemaciclib 10 (2.4) 5 (2.3) CDK4/6, cyclin-dependent kinases 4 and 6. a This category includes patients who received chemotherapy in combination with any non-chemotherapy. b This category includes patients who received hormone therapy plus another medication without chemotherapy. Annals of Oncology D. J. Slamon et al. 1020 https://doi.org/10.1016/j.annonc.2021.05.353 Volume 32 -Issue 8 -2021 1016/j.annonc.2021.05.353). Neutropenia (58.2%, ribociclib; 0.8%, placebo) was the most frequent grade 3 or 4 adverse event. Grade 3 or 4 adverse events of special interest included hepatobiliary toxicity (13.9%, ribociclib; 6.2%, placebo) and prolonged QT interval (3.1%, ribociclib; 1.2%, placebo). Pharmacokinetics In this study, the geometric mean trough plasma concentrations of ribociclib when combined with fulvestrant were largely consistent with the exposure of ribociclib as a single agent, 14 suggesting that fulvestrant had no effect on the PK of ribociclib (Supplementary Table S3, available at https:// doi.org/10.1016/j.annonc.2021.05.353). There is no known drugedrug interaction (DDI) with fulvestrant or in vivo interaction with cytochrome P-450 3A4 (CYP3A4) substrates or modulators. 15 Ribociclib is unlikely to impact fulvestrant PK based on its metabolism and DDI data (a substrate and moderate-to-strong inhibitor of CYP3A4). 16,17 DISCUSSION This extended follow-up (median, 56.3 months) analysis of MONALEESA-3 is the longest reported follow-up for any CDK4/6 inhibitor clinical trial in a purely postmenopausal patient population and demonstrates a lasting benefitofOS with ribociclib plus fulvestrant versus fulvestrant alone in patients with HRþ/HER2ABC. This benefit was maintained when ribociclib was given in both the first and second lines and was maintained across most subgroups. In patients who received ribociclib either as a first-line therapy 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Patients not yet receiving chemotherapy (%) Months 100 484 242 461 227 417 198 393 191 381 181 370 174 349 166 331 150 315 144 295 135 285 126 277 116 248 100 240 94 261 109 223 86 215 80 208 78 202 75 196 72 186 67 179 66 174 64 167 63 162 61 158 57 142 51 69 18 111 38 36 9 14 5 0 0 3 1 Ribociclib + fulvestrant Placebo + fulvestrant 215 131 48.1 28.8 (n = 484) (n = 242) Hazard ratio = 0.704 (95% CI 0.566-0.876 months) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Chemotherapy-free survival (%) Months 100 484 242 461 227 417 198 393 191 381 181 370 174 349 166 331 150 315 144 295 135 285 126 277 116 248 100 240 94 261 109 223 86 215 80 208 78 202 75 196 72 186 67 179 66 174 64 167 63 162 61 158 57 142 51 69 18 111 38 36 9 14 5 3 1 0 0 Ribociclib + fulvestrant Placebo + fulvestrant 287 178 32.3 22.4 (n = 484) (n = 242) Hazard ratio = 0.688 (95% CI 0.570-0.830 months) A B No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant Months, n Events, n Months, n Events, n Figure 3. Time to first subsequent chemotherapy and chemotherapy-free survival. (A) Time to first chemotherapy. (B) Chemotherapy-free survival. CI, confidence interval. D. J. Slamon et al. Annals of Oncology Volume 32 -Issue 8 -2021 https://doi.org/10.1016/j.annonc.2021.05.353 1021 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Progression-free survival 2 (%) Months 100 237 128 229 125 221 123 215 117 211 115 206 112 199 107 190 98 184 95 177 90 170 87 162 82 151 75 148 72 155 81 139 69 134 65 129 60 126 57 124 55 120 52 116 48 111 47 107 47 105 45 103 43 94 38 43 13 71 28 25 5 7 2 1 0 0 0 Ribociclib + fulvestrant Placebo + fulvestrant 107 82 53.7 35.5 (n = 237) (n = 128) Hazard ratio = 0.626 (95% CI 0.469-0.835 months) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Progression-free survival 2 (%) Months 100 237 110 231 104 222 98 210 96 201 89 190 82 173 74 155 69 145 63 140 54 132 50 124 42 102 34 95 33 112 38 85 29 81 24 79 22 74 20 72 19 67 19 66 19 64 19 61 19 58 19 58 17 50 14 23 5 41 8 11 4 7 1 2 1 0 0 Ribociclib + fulvestrant Placebo + fulvestrant 153 80 26.0 20.5 (n = 237) (n = 110) Hazard ratio = 0.730 (95% CI 0.557-0.957 months) No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant 0 2 4 6 8 10121416182022 262824 30 32 34 36 38 40 42 44 46 48 50 52 5654 58 60 6462 0 20 40 60 80 Progression-free survival 2 (%) Months 100 484 242 470 232 453 224 434 216 421 207 403 207 379 183 352 169 336 160 324 146 308 139 292 126 259 111 249 107 273 121 230 100 221 91 214 84 206 79 202 76 193 73 187 69 180 68 172 68 166 66 164 62 147 51 66 19 113 37 36 9 14 3 0 0 3 1 Ribociclib + fulvestrant Placebo + fulvestrant 265 163 37.4 28.1 (n = 484) (n = 242) Hazard ratio = 0.693 (95% CI 0.570-0.844 months) B C A No. at risk Ribociclib + fulvestrant Placebo + fulvestrant Ribociclib + fulvestrant Placebo + fulvestrant Months, n Events, n Months, n Events, n Months, n Events, n Figure 4. Progression-free survival 2. (A) Progression-free survival 2 in all patients. (B) Progression-free survival 2 in patients receiving ribociclib plus fulvestrant as first-line therapy. (C) Progression-free survival 2 in patients receiving ribociclib plus fulvestrant as second-line therapy. CI, confidence interval. Annals of Oncology D. J. Slamon et al. 1022 https://doi.org/10.1016/j.annonc.2021.05.353 Volume 32 -Issue 8 -2021 or as a second-line therapy, the OS benefit remained consistently strong from the final OS analysis throughout the duration of this follow-up. 6 Moreover, treatment with ribociclib had a positive effect on subsequent chemotherapy use, and the benefit of ribociclib during study treatment was maintained over subsequent lines of therapy. Furthermore, ribociclib treatment delayed the time until chemotherapy was required regardless of censoring for death, prolonging the time these patients were chemotherapy-free. Safety signals were consistent with those previously reported for MONALEESA-3, 4,6 and PK analyses indicated no apparent DDI between ribociclib and fulvestrant. These results strengthen the previously published final OS analysis of MONALEESA-3 and support a lasting effect of ribociclib treatment, including beyond study treatment. Currently, almost all patients with ABC will receive a CDK4/6 inhibitor during the course of their disease. In MONALEESA-3, investigators and patients were unblinded following the final OS analysis, after which two patients in the placebo group crossed over to receive ribociclib by the data cut-off date (30 October 2020). Additionally, a greater percentage of patients in the placebo group (30.0%) compared with the ribociclib group (14.0%) went on to receive a CDK4/6 inhibitor as a subsequent therapy. Despite more patients in the placebo arm going on to receive a subsequent CDK4/6 inhibitor, the mOS in the ribociclib arm remained significantly longer than in the placebo arm. When adjusted for subsequent CDK4/6 inhibitor use, the OS benefit was maintained. This OS analysis is an extended follow-up analysis (median, 56.3-month follow-up) for the MONALEESA-3 trial of ribociclib in combination with fulvestrant for treatment of HRþ/HER2ABC. The PALOMA-3 and MONARCH-2 trials for CDK4/6 inhibitors in combination with fulvestrant also reported OS benefits. PALOMA-3 described an mOS of 34.9 months in the palbociclib group versus 28.0 months in the placebo group at the 44.8-month follow-up point. 10 At a follow-up of 47.7 months, MONARCH-2 reported an mOS of 46.7 versus 37.3 months with abemaciclib and placebo, respectively. 8 With a median follow-up of 56.3 months, the current analysis of MONALEESA-3 reports the longest mOS in a clinical trial of a CDK4/6 inhibitor in combination with fulvestrant. However, cross-trial comparisons should be interpreted with caution due to differences in patient population characteristics, including prior treatment. 6,8,10 These results complement the extended follow-up results of the MONALEESA-7 trial, which reported a 58.7-month mOS for pre-/perimenopausal women treated with ribociclib in combination with ET (nonsteroidal aromatase inhibitor or tamoxifen) in the first-line setting. 11 Similarly, MONALEESA-3 has continued to demonstrate a larger OS benefit with ribociclib versus placebo over this period of extended follow-up, in the first-line and also in the secondline subgroups. These results confirm the prolonged and consistent benefit of ribociclib for the treatment of HRþ/HER2ABC, regardless of the ET partner used in combination with ribociclib or the patient’s menopausal status, and further strengthen the use of ribociclib in firstline as well as in second-line therapy. ACKNOWLEDGEMENTS This study was sponsored by Novartis. We thank the patients who participated in this trial, their families, and their caregivers; members of the data monitoring committee; members of the study steering committee; staff members who helped with the trial at each site; and Daniele Cary, PhD, of MediTech Media for medical editorial assistance with this manuscript. Ribociclib was discovered by Novartis Institutes for BioMedical Research in collaboration with Astex Pharmaceuticals. FUNDING The study was funded by Novartis Pharmaceuticals Corporation. DISCLOSURE DJS reports board of directors (stock) and travel expenses from BioMarin; stock ownership, research funding, and travel expenses from Pfizer; advisory board, consulting, research funding, and travel expenses from Novartis; consulting from Eli Lilly; and stock ownership from Amgen and Seattle Genetics. SC reports that his institution received grants and personal fees for advisory boards from Novartis, Pfizer, Hoffmann-La Roche, and Eli Lilly, outside the submitted work. GJ reports personal fees from Novartis, during the conduct of the study; grants, personal fees, and nonfinancial support from Novartis, Roche, and Pfizer; personal fees and non-financial support from Lilly, Amgen, BMS, and AstraZeneca; personal fees from AbbVie and Daiichi-Sankyo; and non-financial support from Med-immune and Merck KGaA, outside the submitted work. MDL reports personal fees for speaker honoraria and advisory board honoraria from Novartis, AstraZeneca, Eli Lilly, and Pierre Fabre, outside the submitted work. SI reports grants from AstraZeneca, Pfizer, Eisai, and Daewoong; advisory and personal fees from AstraZeneca, Novartis, Hanmi, Pfizer, Eisai, Amgen, MediPacto, Roche, Eli Lilly, MSD, and GlaxoSmithKline; and non-financial support from Novartis. KP reports personal fees for advisory board from Novartis, AstraZeneca, Roche, Pfizer, and BMS, outside the submitted work. GVB reports personal fees for advisory board from Novartis and Eli Lilly, outside the submitted work. MM reports personal fees for speaker honoraria and honoraria for participation in advisory boards from Lilly and Pfizer; honoraria for participation in advisory boards from AstraZeneca, GlaxoSmithKline, Pharmamar, and Taiho Oncology; and research grants and honoraria for participation in advisory boards from Novartis and Roche- Genentech, outside the submitted work. AN reports consulting/advisory role, travel/accommodation/expenses, and research funding from Novartis and consulting/advisory role from Amgen during the conduct of the study. GSS D. J. Slamon et al. Annals of Oncology Volume 32 -Issue 8 -2021 https://doi.org/10.1016/j.annonc.2021.05.353 1023