Although Ki-67 is a commonly used measure of cellular proliferation in breast cancer tissue, its utility as a biomarker for helping to guide therapy decisions has been clouded by technical and clinical questions.
Although Ki-67 is a commonly used measure of cellular proliferation in breast cancer tissue, its utility as a biomarker for helping to guide therapy decisions has been clouded by technical and clinical questions. However, recent efforts to develop a framework for analyzing Ki-67 and data from several clinical trials suggest the biomarker may play a more extensive role in breast cancer treatment in the future.
Despite a lack of support for the practice from national oncology guidelines, clinicians have made wide use of Ki-67 staining with immunohistochemistry (IHC) as a prognostic biomarker in breast cancer.1 In December 2020, the International Ki-67 in Breast Cancer Working Group (IKWG) set forth recommendations for establishing analytical validity for IHC evaluation for measuring Ki-67. But the panel of breast cancer experts said that evidence for Ki-67’s clinical utility is limited to its being a prognostic biomarker, with specified cutoffs, to determine whether patients with earlystage estrogen receptor (ER)-positive disease should receive adjuvant chemotherapy.2
Although Ki-67’s recommended clinical utility does not extend beyond this specific setting, recent data from several clinical trials indicate significant promise regarding its potential for guiding therapeutic decisions.
The ADAPT trial (NCT01779206), featured in several presentations at the 2020 San Antonio Breast Cancer Symposium (SABCS 2020), bolstered evidence that, after a short course of neoadjuvant endocrine therapy (NET), an on-treatment Ki-67 index can identify patients who can be spared intensive chemotherapy in the adjuvant setting.3,4
Meanwhile, the monarchE trial (NCT03155997) was the first to prospectively investigate Ki-67 as a biomarker in a phase 3 trial of cyclin-dependent kinase inhibitors (CDKIs) in the adjuvant setting. Data reported at SABCS 2020 suggested that high Ki-67 levels in conjunction with high-risk features could be used to select patients who would benefit from the addition of abemaciclib (Verzenio) to endocrine therapy (ET) in the adjuvant treatment of ER-positive, HER2- negative stage I or II breast cancer.5
Unchecked proliferation is a hallmark of cancer cells,6 but not all tumors grow at the same rate; the spectrum ranges from indolent to aggressive. The ability to determine the proliferation rate could provide important information about a patient’s prognosis and even guide therapeutic decisions.
Among several methods for assessing cancer cell proliferation, Ki-67 assessment is probably the best known.2 A monoclonal antibody is used for IHC staining of the proliferation-associated nuclear protein Ki-67 in tumor cells to determine the percentage of Ki-67–positive cells among the total population of tumor cells in formalin-fixed paraffin-embedded sections obtained from a core-cut biopsy sample; this is the Ki-67 index.2,7,8
Tumors are classified as having a high or low Ki-67 index based on a prespecified cutoff. Tumors with a high Ki-67 index have a larger number of proliferating cells and are therefore likely to grow more quickly.
The Ki-67 index has a well-established role in the grading of neuroendocrine tumors of the pancreas and gastrointestinal tract. The World Health Organization has incorporated the Ki-67 index into the classification of these tumor types.9 The index also has potential utility in other types of neuroendocrine tumors9 and a variety of other cancer types, including bladder, prostate, and liver cancer; soft tissue sarcomas; and meningiomas10-15; however, these uses are much less well defined.
Most notably, the Ki-67 index has growing importance in breast cancer. This highly heterogeneous disease is characterized by the presence of discrete molecular subtypes with distinct clinical behaviors that can be classified according to gene expression signatures.16
Findings from numerous studies have demonstrated the prognostic value of measuring Ki-67 in breast cancers.17 In one of the largest retrospective studies, in which investigators analyzed data from a regional population-based cancer registry (N = 3568), the Ki-67 index was a prognostic factor for both disease-free survival (DFS) and overall survival (OS).
In patients with a Ki-67 index of 15% or less, 5-year DFS and OS rates were 86.7% and 89.3%, respectively, compared with 75.8% and 82.8% in patients with a high Ki-67 index (> 45%).1
The vast majority of breast cancers are ER positive and can be classified into 2 of the aforementioned molecular subtypes: luminal A and luminal B. Compared with their luminal B counterparts, luminal A tumors generally are lower grade, have higher expression of estrogen-related genes, have a better prognosis, and are more sensitive to ET.18
Distinguishing between these 2 subtypes can guide decisions about the need for added chemotherapy. Because the cost of gene expression analysis has limited its adoption in clinical practice, protein expression levels of ER, progesterone receptor (PR), and HER2 (determined by IHC analysis) can be used as surrogate markers for subtype classification.18 For ER-positive, HER2-negative breast cancers, Ki-67 has emerged as another important surrogate marker because luminal A tumors typically have a lower rate of proliferation than luminal B tumors.18,19
The Ki-67 index was shown to discriminate between the luminal A and luminal B subtypes most effectively when a cutoff of 14% was used.19,20 This finding resulted in the 14% cutoff being adopted by the experts at the St Gallen International Breast Cancer Conference in 2011.21 In 2013, the majority of the St Gallen panel voted to change this threshold to 20% or greater.22
This highlights a major issue that has hampered the clinical adoption of the Ki-67 index: a multiplicity of assays and lack of standardization in analysis and established cutoffs. To date, neither the American Society of Clinical Oncology nor the National Comprehensive Cancer Network clinical guidelines on breast cancer recommend using the Ki-67 index because of these issues.2,23,24
The IKWG was established in 2011 to develop internationally acceptable standards relating to Ki-67 measurement. Recently updated recommendations from the group detailed a standardized visual scoring method2,25 that should be adopted, stressing the importance of quality assurance and quality control programs to maintain the analytical validity of the assay.2
Among the controversies surrounding the Ki-67 assay are the multiple counting methods.9 The IKWG has endorsed global (average) counting as opposed to “hot spot” counting, in which only the areas of highest proliferative activity are counted, as the latter has been associated with higher variability.2
The IKWG concluded that the clinical utility of Ki-67 as a prognostic marker was “evident only for prognosis estimation in anatomically favorable ER-positive and HER2-negative patients, to identify those who do not need adjuvant chemotherapy.”
The group’s consensus was that Ki-67 levels of 5% or below or 30% or greater could be used to estimate prognosis and determine the advisability of adjuvant chemotherapy. For tumors with a Ki-67 index between these values, the group recommended using a commercially available multigene expression panel.2
Neoadjuvant chemotherapy (NAC) has become standard of care in the treatment of early-stage breast cancer because it promotes achievement of pathological complete response (pCR), which correlates with favorable long-term clinical outcomes.26 Hormone receptor–positive, HER2-negative breast cancers were far less likely to exhibit a pCR to preoperative chemotherapy26; however, some patients may be offered NET alone.24
Several studies have looked at baseline levels of Ki-67 as a biomarker for NET and found conflicting results. Results from IMPACT, a relatively small study (N= 330) conducted at oncology centers in the United Kingdom and Germany, showed that the baseline Ki-67 level did not significantly predict outcome following NET.27,28 More recently, however, investigators from the POETIC trial (NCT02338310), which was conducted in a much larger patient population (N = 4480) and used the IKWG standardized assay, reported that patients with a low baseline Ki-67 index had a lower risk of recurrence and could likely skip NET.29
Although the role of baseline Ki-67 is unclear, these studies have highlighted the powerful potential of the on-treatment Ki-67 index to predict responsiveness to ET and guide decisions about subsequent adjuvant ET. The Ki-67 index at 2 to 4 weeks after starting NET indicates the level of persistent cell proliferation and resistance or response to ET and is significantly associated with risk of recurrence.27,29
In the POETIC trial, for patients with high Ki-67 levels at baseline and after 2 weeks of NET, the 5-year absolute risk of recurrence was 21.5%. The risk was 8.4% for those with high baseline but low 2-week Ki-67 levels, and 4.3% for those with low levels at both time points.29
This idea of using on-treatment Ki-67, alongside other features that predict risk, to triage patients and avoid unnecessary chemotherapy is being prospectively evaluated in several trials. In the phase 2/3 umbrella trial ADAPT, patients underwent 3 weeks of NET and, based on their Oncotype DX Recurrence Score (RS) and Ki-67 index, were assigned to subsequent adjuvant ET or dose-dense NAC followed by adjuvant ET.3,4 The Oncotype DX Breast RS test is a multigene assay that assesses the expression of 16 genes associated with breast cancer, including those encoding ER, PR, HER2, and Ki-67, in addition to 5 reference genes, to generate an RS ranging from 0 to 100 that indicates the risk of cancer recurrence.30,31
Patients with high-risk disease (n = 864; defined as cN2-3 disease; grade 3 disease with post-ET Ki-67 > 40%; cN0-1 disease with RS > 25; or RS 12-25 with post-ET Ki-67 > 10%) were treated with NAC. The probability of achieving pCR was low for patients with RS of 25 or less and no response to NET (7.2% pCR), and for those with an RS greater than 25 and a response to NET (5.6% pCR).3
Patients with cN1 disease and an RS of 0 to 11 and those with RS 12 to 25 and a response to NET skipped NAC and were treated with adjuvant ET alone. Rates of 5-year invasive DFS (IDFS), OS, and distant DFS were 92.6%, 97.3%, and 95.6%, respectively, in the first group and 93.9%, 98%, and 96.3% in the second group. These data demonstrate that the on-treatment Ki-67 index complements the Oncotype DX RS in selecting patients for whom chemotherapy can be omitted.4
In the phase 3 ALTERNATE trial (NCT01953588), patients (N = 1362) were treated with 4 to 12 weeks of NET. Investigators recommended that those with a Ki-67 index greater than 10% at week 4 or 12 (n= 286) switch to NAC. Meanwhile, endocrine responsiveness was evaluated using the modified Preoperative Endocrine Prognostic Index (mPEPI), which incorporates tumor size, number of involved lymph nodes, and on-treatment Ki-67 index. Patients with an mPEPI score of 0 (pT1-2, pN0, Ki-67< 2.7%) received adjuvant ET, whereas those with an mPEPI score above 0 received chemotherapy and ET.32
An mPEPI score of 0 has been shown to be associated with low risk of recurrence without adjuvant chemotherapy.33 The ALTERNATE trial is looking to prospectively validate this biomarker, and results are eagerly awaited.
An analysis of patients with a high on-treatment Ki-67 index who receive NAC in this trial was recently presented at SABCS 2020. The results indicated that salvage treatment led to very low pCR rates, suggesting that alternative treatment options will be needed for these patients.34
To help overcome resistance to ET in breast cancer, which occurs almost universally, a number of combination strategies have been explored. Several CDKIs are FDA approved for use in combination with ET for the treatment of advanced/metastatic ER-positive, HER2-negative disease, and these successes have prompted the evaluation of this therapeutic approach in early-stage breast cancer. Investigators are also exploring whether the Ki-67 index can select patients who might benefit from the addition of a CDKI in this setting.
In the ongoing monarchE trial, 5637 patients with high-risk, node-positive, ER-positive, HER2-negative breast cancer are randomized to receive 5 to 10 years of adjuvant ET, with or without the addition of 2 years of the CDKI abemaciclib. Patients were enrolled in 2 cohorts according to clinicopathologic risk factors (cohort 1) or Ki-67 index (cohort 2).
In an exploratory subanalysis of cohort 1, the addition of abemaciclib to ET reduced the risk of IDFS by 35.7% compared with ET alone in patients with a high baseline Ki-67 index (≥ 20%). Although patients with a low Ki-67 index (<20%) also benefited from the addition of abemaciclib, the magnitude of treatment benefit was numerically lower.5
Investigators are also examining the combination of a CDKI and NET, and in several recent clinical trials, a change in the Ki-67 index from baseline after therapy served as a primary end point. The addition of a CDKI resulted in greater suppression of Ki-67 levels in the MONALEESA-1 (NCT01919229), neoMONARCH (NCT02441946), and PALLET (NCT02296801) trials.35-38 Whether this corresponds to improved response rates, and thus whether Ki-67 levels could serve as a surrogate marker of response in this setting, remain to be determined.