Ki-67 Identified as Biomarker for iDFS in HR+/HER2– High-Risk Early Breast Cancer

Nadia Harbeck, MD, PhD

In patients with hormone receptor–positive, HER2-negative, node-positive, early breast cancer whose tumors have high clinicopathological risk factors, treatment with abemaciclib (Verzenio) in combination with endocrine therapy (ET) reduced the risk of invasive disease recurrence or death, independent of Ki-67 level, according to results from a subanalysis of the phase 3 monarchE clinical trial.¹

A poster shared as part of the 2020 San Antonio Breast Cancer Symposium explored Ki-67 as a biomarker for the determination of the risk of invasive disease recurrence in patients. For the analysis, investigators centrally assessed Ki-67 in all eligible patients who had untreated breast tissue available for testing. This analysis marks the first time Ki-67 was prospectively investigated as a biomarker with a prespecified threshold of ≥20% with the use of a standardized assay in a phase 3 registration trial.

In total, 5637 patients were randomized 1:1 in the study and divided into 2 cohorts based on either clinicopathological risk factors (cohort 1) or Ki-67 level (cohort 2). Of the patients enrolled in cohort 1, 39.1% had a high Ki-67 (defined as ≥ 20%). Patients with a low Ki-67 (<20%) represented 37.4% of the study, and Ki-67 data were missing or unavailable for the remaining 23.5%. In cohort 2, 95.7% of patients had high Ki-67 and 2.5% had low Ki-67. For 1.7% of cohort 2, information was missing or unavailable.

The analysis was powered to assess invasive disease-free survival (iDFS) in the overall population with a 2-sided P-value boundary of 0.0424 and for iDFS in high Ki-67 patients from cohort 1 with a 2-sided P-value boundary of 0.0426. In exploratory analyses, the study explored iDFS in patients with low Ki-67 from cohort 1, as well as iDFS in patients in cohort 2.

The primary end point explored in the study was iDFS, and the secondary end points included iDFS in patients with high Ki-67, distant relapse-free survival (DRFS), overall survival (OS), safety, patient-reported outcomes, and pharmacokinetics.

Subjects in cohort 1 received abemaciclib 150 mg twice daily for up to 2 years in combination with ET for 5 to 10 years. In cohort 2, patients received ET alone for 5 to 10 years. To assess the end points, patients in both cohorts were stratified by whether they had received prior chemotherapy, menopause status, and by region.

A gated analysis of cohort 1 showed that abemaciclib plus ET reduced the risk of patients developing an iDFS event by 35.7% compared ET alone. The 2-year iDFS rate was 91.3% with the addition of abemaciclib to ET versus 86.1% with ET alone for a difference of 5.2 percentage points (HR, 0.643; 95% CI, 0.475-0.872; 2-sided P = .0042). The improvement in iDFS with the combination of abemaciclib and ET was considered statistically significant.

An exploratory analysis of cohort 2 demonstrated a 31.5% reduction in the risk of an iDFS event with abemaciclib plus ET versus ET alone. The iDFS rates at 2 years were 94.7% and 92.0%, respectively, for a difference of 2.7 percentage points (HR, 0.685; 95% CI, 0.462-1.017; P = .0591). It was noted that even though patients with low Ki-67 had a lower magnitude of treatment, all of cohort 1 derived benefit from abemaciclib in combination with ET.

“Among patients with high clinicopathological risk factors, patients with high Ki-67 tumors had an even greater risk of recurrence than those with low Ki-67 tumors, confirming the prognostic value of Ki-67,” wrote the study authors, led by Nadia Harbeck, MD, PhD, in their poster.

Findings from this subanalysis are consistent with the iDFS benefit for the addition of abemaciclib observed in the preplanned primary outcome analysis of the monarchE trial. It was shown that abemaciclib reduced the risk of invasive disease recurrence or death by 28.7% in patients with hormone receptor–positive, HER2-negative, node-positive, high-risk early breast cancer.2

Patients in the subanalysis with available Ki-67 data included 2498 patients in the intent-to-treat (ITT) population, 2003 in the high Ki-67 group from cohort 1, and 1914 in the low Ki-67 group from cohort 1. At baseline, 1 to 3 positive nodes were found in 53.6%, 42.9%, and 26.2% of patients in the 3 groups, respectively. At least 4 positive nodes were found in 46.2% of patients in the ITT population, 57.0% of the high Ki-67 cohort 1 group, and 73.5% in the low Ki-67 cohort 1 group. Information was missing for the remaining patients in the 3 groups.

Most patients enrolled had either grade 2 disease including 43.1% of the ITT group, 34.7% of the high Ki-67 cohort 1, and 57.9% in the low Ki-67 cohort 1 group, or grade 3 disease in 47.8%, 58.0%, and 25.7%, respectively. Few patients had grade 1 histology, and histological information was missing for a small percentage of the study population.

The majority of the study subjects had a pathological tumor size of 2 to 5 cm including 53.1% of the ITT population, 52.9% in the high Ki-67 cohort 1 population, and 46.8% in the low Ki-67 cohort 1 group. A large proportion of patients also displayed a <2 cm pathologic tumor size at baseline, including 30.2%, 27.45%, and 24.6% respectively. Other patients had ≥5 cm tumors which included 15.4% in the ITT group, 18.6% in the high Ki-67 cohort 1 group, and 27.3% in the low Ki-67 cohort 1 population. Tumor size information was missing for the remaining 76 patients.

Both women and men with hormone receptor–positive, HER2-negative, node-positive, high-risk early breast cancer were eligible to enroll in the study. Women were required to be either pre- or post-menopausal. All patients could enroll with or without prior neoadjuvant or adjuvant chemotherapy. Patients could not have distant metastases. For cohort 1, patients were required to have ≥4 axillary lymph nodes (ALNs), and 1 to 3 of the ALNs must have shown a histologic grade of 3 and a tumor size ≥5 cm. Cohort 2 was required to have 1-3 ALNs and centrally tested Ki-67 ≥20%. Patients in cohort 2 could not have grade 3 disease or have a tumor size ≥5 cm.

“Our results suggest that Ki-67 greater than or equal to 20% can be used together with clinicopathological features of nodal involvement, tumor size, and grade, to identify patients with HR+, HER2-, early breast cancer at high risk of recurrence, said Harbeck , head of the Breast Center of the University of Munich, Germany, during the ASH poster presentation.

_References:

_{1. Harbeck N, Johnston S, Fasching P, et al. High Ki-67 as a biomarker for identifying patients with high risk early breast cancer treated in monarchE. Presented at: 2020 San Antonio Breast Cancer Symposium; December 8-12, 2020; Virtual. Abstract PD2-01.}

_{2. O’Shaughnessy J, Johnston SRD, Harbeck N, et al. Primary outcome analysis of invasive disease-free survival for monarchE: abemaciclib combined with adjuvant endocrine therapy for high-risk early breast cancer. Presented at: 2020 San Antonio Breast Cancer Symposium; December 8-11, 2020; Virtual. Abstract GS1-01.}