Abstract PD2-01: High Ki-67 as a biomarker for identifying patients with high risk early breast cancer treated in monarchE

Abstract

Background monarchE, a phase 3, open-label, randomized study evaluating endocrine therapy with or without abemaciclib in patients with node positive, HR+, HER2-, high risk early breast cancer, resulted in a statistically significant improvement in invasive disease-free survival (IDFS) at a pre-planned interim analysis. Ki-67, a marker of cellular proliferation in breast cancer, was used in addition to other clinical and/or pathological features to identify patients whose cancer may be at higher risk of recurrence. A key secondary endpoint was to evaluate IDFS in patients with high (≥20%) Ki-67 tumors. Methods Patients with ≥4 positive nodes, or 1-3 nodes and either grade 3 disease, tumor size ≥5 cm, or central Ki-67 ≥20% were eligible for monarchE. Ki-67 was centrally assessed for all eligible patients with suitable untreated breast tissue using a standardized kit produced by Agilent/Dako (MIB-1). A sequential gatekeeping strategy was utilized to test the statistical significance in IDFS for patients with high Ki-67 tumors. Data on this subgroup are presented. Results Primary outcome results in the ITT population are presented separately and resulted in a significant 28.7% reduction in the risk of developing invasive disease with abemaciclib plus ET versus ET alone (HR = 0.713; 95% CI = 0.583, 0.871). Of the 5637 patients enrolled in monarchE, 4425 (78.5%) had Ki-67 samples eligible for testing. Of those tested, 2498 patients (56.5%) had Ki-67 ≥20% (Ki-67H). Results from the Ki-67H population are presented separately and showed that abemaciclib plus ET demonstrated superior IDFS versus ET alone with a 30.9% reduction in the risk of invasive disease (p=.0111; HR = 0.691; 95% CI = 0.519, 0.920). There was an absolute improvement of 4.5% in IDFS rate at 2 years (91.6% in the abemaciclib arm and 87.1% in the control arm). An additional planned analysis was performed evaluating efficacy in 2003 patients with high Ki-67 tumors enrolled in cohort 1 (patients with ≥4 positive nodes or 1-3 nodes and either tumor size ≥5 cm and/or grade 3 disease). The IDFS treatment benefit in this group was statistically significant with a HR of 0.643 (95% CI = 0.475, 0.872) corresponding to a 35.7% reduction in the risk of developing invasive disease. Two-year IDFS rates in this group were 91.3% in the abemaciclib group and 86.1% in the control arm, representing a 5.2% absolute improvement at 2 years. A clinically meaningful improvement was also observed in distant relapse-free survival (DRFS) in both populations. Baseline characteristics were balanced across arms in both Ki-67H populations. An exploratory analysis was conducted evaluating patients in cohort 1 with low Ki-67 (<20%) and will be presented. Conclusion This represents the first time a prespecified threshold of ≥20% for Ki-67 has been used to prospectively evaluate the utility of Ki-67 in a phase III registration trial with a standardized assay. There was a statistically significant improvement in IDFS for patients with high Ki-67 tumors across the ITT population (HR = 0.691) and in cohort 1 (HR = 0.643). These results suggest that Ki-67 ≥20% is an additional clinicopathological feature that can be used in conjunction with high risk features of nodal involvement, tumor size, and grade, to select patients with a higher risk of recurrence who may benefit from treatment with abemaciclib in the adjuvant setting. ClinicalTrials.gov: NCT03155997 Ki-67H Ki-67H Cohort 1EndpointAbemaciclib + ETN=1262ET aloneN=1236Abemaciclib + ETN=1017ET aloneN=986IDFS# events, n (%)82 (6.5)115 (9.3)71 (7.0)106 (10.8)log rank Pvalue, HR (95% CI)p=.0111 0.691 (0.519, 0.920)p=.00420.643 (0.475, 0.872)Rate (%) at 2-years (95% CI)91.6(89.4, 93.4)87.1(84.3, 89.5)91.3(88.9, 93.2)86.1(83.1, 88.7)DRFS# events, n (%)65 (5.2)102 (8.3)56 (5.5)96 (9.7)log rank Pvalue,HR (95% CI)p=.0018 0.609 (0.445, 0.833)p=.00040.554 (0.397, 0.773)Rate (%) at 2 years (95% CI)93.6(91.6, 95.1)88.5(85.7, 90.7)93.3(91.2, 95.0)87.384.4, 89.8) Citation Format: Nadia Harbeck, Stephen Johnston, Peter Fasching, Miguel Martin, Masakazu Toi, Priya Rastogi, Chuangui Song, David Molthrop, Jacqueline Vuky, Toshinari Yamashita, Georgina Garnica Jaliffe, Mahmut Gumus, Desiree Headley, Ran Wei, Susana Barriga, Maria Munoz, Michael Method, Valerie Andre, Hans Kreipe, Joyce O'Shaughnessy. High Ki-67 as a biomarker for identifying patients with high risk early breast cancer treated in monarchE [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PD2-01.

Reply to K. Hashimoto and A. Shimomura

“…As presented at SABCS 2020, the addition of abemaciclib to endocrine therapy (ET) resulted in improved IDFS for patients with high clinicopathologic risk factors and with either high or low Ki-67, indicating that all patients benefit from the addition of abemaciclib regardless of Ki-67 index. 4 Patients with high clinicopathological risk factors and high Ki-67 tumors had a higher risk of recurrence (2-year IDFS rate ET alone arm of 86.1%) than patients with high-risk factors and low Ki-67 tumors (92.0%), confirming the prognostic value of Ki-67. 4 We do agree that pathologist interpretation can often be problematic when reading Ki-67 slides.…”

Using Breast Cancer Gene Expression Signatures in Clinical Practice: Unsolved Issues, Ongoing Trials and Future Perspectives

“…In monarchE trial, high-risk patients were selected according to cTNM, tumor grade and centrally assessed Ki-67 [ 102 ]. In subgroup analyses, patients with Ki-67 ≥ 20% had a numerically better hazard ratio for efficacy outcome [ 104 ]. Specific predictive biomarkers are needed to optimize the selection of patients who might benefit most from CDK4/6 inhibitors in adjuvant or neoadjuvant setting.…”

Update Breast Cancer 2021 Part 1 – Prevention and Early Stages

“…In addition to this principal analysis, data was also presented relating to the subgroup of patients with Ki-67 ≥ 20 % and other clinically unfavourable prognostic factors [83]. In this group with a very high risk of relapse, the hazard ratio was 0.643 (95 % CI: 0.475-0.872), with a similar effect in the group with a high clinical risk of relapse and Ki-67 below 20 % (HR = 0.685; 95 % CI: 0.462-1.017) [83].…”