Ki 67: a Promising Prognostic Marker in Early Breast Cancer—a Review Article

Abstract

Ki67 index is considered to be a reliable indicator of the proliferative activity of breast cancer. Additionally, the Ki67 proliferative marker may play a role in assessing response to systemic therapeutic strategies and can act as a prognostic biomarker. But its limited reproducibility which stems from a lack of standardization of procedures, inter-observer variability, and preanalytical and analytical variabilities all have hampered the use of the Ki67 index in clinical practice. Currently, clinical trials have been evaluating Ki67 as a predictive marker for needing adjuvant chemotherapy in luminal early breast cancer patients receiving neoadjuvant endocrine therapy. But the inconsistencies existing in the estimation of the Ki67 index limit the utility of Ki67 in standard clinical practice. The purpose of this review is to evaluate the benefits and drawbacks of utilizing Ki-67 in early-stage breast cancer to prognosticate the disease and predict the risk of recurrence.

Introduction

Breast cancer (BC) is a heterogeneous disease that is characterized by the presence of discrete molecular subtypes that can be classified by means of genetic array-testing or based on immunohistochemistry (IHC) investigations [1, 2]. Normally some common clinical factors such as grade and size of the tumor, lymph node status, and surgical margins are studied, but they are inadequate to predict prognosis; therefore, it is essential to know the subtype of BC to make a decision on the treatment [3]. Four primary molecular BC subtypes have been identified based on the expression of hormone receptors (HR), human epidermal growth factor receptor 2 (HER2), and the proliferation marker of cancer cells which is usually measured by the Ki67 index. These subgroups are luminal-A, luminal-B, triple-negative, and HER2-enriched subtype BC [1]. Luminal-A BCs are characterized by HR + (estrogen receptor (ER⁺) and/or progesterone receptor-positive (PR⁺)), and HER2 negative with a low Ki67 index. Luminal A cancers are low-grade tumors that have a good prognosis. Luminal-B BC has a higher recurrence rate and is commonly HR + (ER⁺ and/or PR +), along with either HER2 positive or negative with high Ki67 index. The prognosis of luminal B cancers is slightly worse than luminal A cancers. The third subtype of BC with ER, PR, and HER2 negative is termed “triple-negative” BC and is characterized by high-grade tumors. HER2-enriched subtype (HER2⁺/ER⁻/PR⁻) is comparatively less common and are high-grade tumors with poor outcome which tends to grow faster than luminal cancers [4–7].

Proliferative Marker Ki67 as a Prognostic Tool in BC

Apart from the disease stage and patient prognostic factors including lymph node involvement, size of the tumor, and age, the molecular pattern of the tumor is essential to identify the patients, who really benefit from the given treatment. The proliferative marker Ki67 index along with other standard markers such as ER, PR, and HER2 are the most frequently evaluated IHC predictive and prognostic markers and are used to guide treatment decisions. The rate of tumor cell proliferation is known either from the mitotic counting, flow-cytometry for synthesis-phase fraction, or IHC using antibodies sensitive to various proliferating cellular antigens [8]. As one of the most widely utilized IHC proliferation markers, the Ki-67 index is a measure for the quantification of cell proliferation in BC and is considered to be one of the most promising markers [9].

Ki67 is a cancer antigen that is present only in the multiplying cells but not in cells in the resting phase of the cell cycle. Association of Ki67 with proliferation was initially recognized in the early 1980s by Gerdes et al., from a Hodgkin’s lymphoma-descended cell line using a mouse monoclonal antibody directed against a nuclear antigen. Ki67 index has been used as an indicator of cell proliferation for human tumor cells for decades and is useful for selecting the HR-positive breast cancer patients for the addition of chemotherapy to endocrine therapy [8]. IHC is a method of demonstrating the presence and location of proteins in tissue sections. IHC staining is accomplished with antibodies that recognize the target protein and attaches to antigen receptors on the protein. The “antihuman Ki-67 monoclonal antibody” MIB-1 is usually used in IHC to identify the Ki-67 antigen. The higher the concentration of Ki67, the more stained cells will be visible under the microscope. The rate of proliferation is the percentage of positive Ki67 cells among the cell population. Usually, the pathologists find the “hotspot area” which is the area where there is increased tumor proliferation. The Ki 67 index is estimated using various techniques such as “eye-balling estimation,” “vision counting” with a microscope or viewer software, “manual counting” of camera captured or digital image, and automated counting. Higher Ki-67 levels in BC indicate aggressive proliferation and have worse prognosis. Except for the G0 phase, cells in all other phases of the cell cycle such as S, G1, G2, and M show the nuclear antigen Ki-67 [10]. However, Ki67 levels are modest in the G1 and S phases and they surge early in mitosis. A dramatic fall in the Ki67 level occurs later in mitosis. Currently, Ki-67 is primarily used to evaluate prognosis, guide adjuvant therapy decisions, and predict the response to neoadjuvant therapy in ER + /HER2 − BC [11, 12].

Ki 67 in Neoadjuvant Therapy

Clinically, neoadjuvant therapy in early-stage BC is beneficial in reducing tumor burden, increasing the breast-conserving surgery rate, and also helps to decide on patients who will require switch-over therapy. For triple-negative and HER2 + BC, neoadjuvant chemotherapy (NACT) is the first choice of therapy indicated but for HR + HER2 − BC, the choice of therapy is still debated [13]. The evaluation of the Ki67 index on treatment alongside other features that predict risk, to triage patients and avoid unnecessary chemotherapy is being studied in clinical trials.

The “Perioperative Endocrine Therapy-Individualized Care” (POETIC) trial in a single central laboratory measured the short time variation in Ki-67 among patients on perioperative aromatase inhibitors (AIs) and assessed the Ki67 index prior to therapy and after 14 days of therapy [14]. BC patients of HER2-ve type who had Ki67 higher than 10% before therapy and who exhibited a reduction in Ki67 index less than 10% had an 8.4% risk of recurrence in a 5-year period. Those BC patients who did not show a reduction in the Ki67 index below 10% after 14 days of AIs had an absolute risk of 21.5% after 2 weeks of AI therapy. Ki67 estimation after 14 days of AI therapy would help to predict prognosis, especially in patients who had a high Ki67 index before therapy [14, 15]. In another trial (Z1031 trial) patients with ER + , primary breast tumors were administered with preoperative AIs for a period of 2–4 weeks [16, 17]. The patients with a Ki67 index of more than 10% after the preoperative treatment period were advised adjuvant chemotherapy assuming that the therapy had not shown the desired benefit. Despite the reasonably large residual Ki67 index, only 5.7% of patients had a pathological full response after completing chemotherapy, arguing against the prognostic effect of the Ki67 index for chemotherapy response [15]. The “preoperative prognosis index” (PEPI) in the Z1031 study is a weighted multi-factorial method that comprises residual tumor size, lymph node involvement, ER expression level, and Ki67 index. In this trial, 25% of patients had zero PEPI score (greatly favorable), indicating that their prognosis was excellent [17]. The Ki67 triage and PEPI evaluating procedures have been studied further in the ALTERNATE trial. The ALTERNATE trial assessed the endocrine-sensitive disease rate of patients receiving fulvestrant or fulvestrant + anastrozole with anastrozole alone. The 5-year recurrence-free survival rate for patients with modified-preoperative prognosis index (mPEPI) zero on anastrozole alone without chemotherapy was 95%. For patients who had a low Ki67 (< 10) after their 4 weeks of preoperative endocrine therapy, their endocrine-sensitive disease rate and the breast-conserving surgery rate were 27.7% and 70.3% with anastrozole; 29.6% and 68.1% with fulvestrant; and 26.8% and 69.9% with anastrozole + fulvestrant, respectively [18]. In the IMPACT trial with anastrozole or tamoxifen alone or anastrozole along with tamoxifen treatment in postmenopausal patients, anastrozole-treated group had resulted in a better reduction in Ki67 index at 2 weeks and at 12 weeks of therapy which was similar to the results of ATAC trial which evaluated disease-free survival after adjuvant endocrine therapy in long-term follow-up. The patients who participated in the IMPACT trial were assessed as requiring mastectomy at baseline, among them 44% of patients received breast-conserving surgery after anastrozole compared with 31% of patients after tamoxifen [19–21]. But the hypothesis that clinical outcomes might predict long-term outcomes in adjuvant therapy was not fulfilled.

Just like other tumor cell proliferative markers, the Ki67 score at baseline is greatly associated with the probability of pathological complete response (pCR) to chemotherapy. Ki67 estimation at baseline alone has not been considered a golden standard for choosing patients for chemotherapy as it is unclear whether the lower pCR rate for tumors with a lower Ki67 index is adequate to withhold it [15].

Ki67 Index to Predict the Prognosis in Early-Stage BC

Fear of recurrence and absence of fool-proof prognostic predictors result in the overtreatment of cancer, subjecting many patients to the adverse effects of chemotherapy. Adjuvant chemotherapy profits only a few patients with early cancer [22]. Thus, every effort should be made to avoid cancer over treatment by recognizing those patients in whom the administration of these risky treatment modalities can be safely avoided. However, careful case by case analysis must be done so that no patient who may benefit from adjuvant therapy is denied the same [21, 22]. If a patient’s residual risk of recurrence is thought to be high enough, the clinician can choose the best suitable adjuvant systemic therapy.

The choice of ET and anti-HER2 therapy depends on HR and HER2 expression levels and clinicians usually do not depend on the Ki67 index or other proliferative markers [23]. Tumors showing a low Ki67 index indicating low proliferation usually do not respond to chemotherapy because most chemotherapeutic agents can work only in cells that are actively dividing [11]. ER-ve BC which are aggressive in nature respond to chemotherapy comparatively better in metastatic and neoadjuvant settings than those who are ER + ve. On the other hand, the choice of therapy for ER + HER2 − patients is still challenging [24, 25].

Numerous genomic assays such as “OncotypeDX,” “Prosigna,” “EndoPredict,” “Mammaprint,” and “CanAssist Breast” have been established to predict the risk of distant recurrence in ER + early-stage BC patients who are treated with endocrine therapy (ET) alone [26]. The Oncotype DX™ Breast Recurrence Score Test is a genomic assay that analyzes the activity of a group of 21 genes which can forecast the probability of recurrence and effectiveness of chemotherapy in women with early-stage node-negative, ER⁺ BC [3], in which Ki 67 is one of the markers analyzed.

However, the demand for such advanced and expensive genomic assays in lower-middle-income countries is minimal (mainly due to cost constraints); also Ki67 estimation plays a similar role, in which, a high Ki67 index itself indicates a poor prognosis of ER + BC and would make them candidates for adjuvant chemotherapy. Researchers have suggested the prognosis of BC can be predicted accurately using 4 core proteins estimated through IHC analysis, specifically the Ki67 index, ER, PR, and HER2. Thus, IHC4 is found to be more feasible and offer similar results as that of multi-genomic assays at a lower cost if these assays are carefully optimized and performed accurately [27].

The investigators of the “International Ki67 in Breast cancer working group” (IKWG) agree on the decision of withholding chemotherapy in patients with a Ki67 index of less than 5% and giving chemotherapy in patients with a Ki67 index value greater than 30%. Investigators also suggest that the necessity of going for any additional expensive genomic assays for making this decision can be avoided if the results of Ki-67 IHC are below or above these thresholds, respectively [15, 28].

Issues for Assessment of Ki67 Index in Neoadjuvant Therapy

Like any other analytical test, the analytical procedure of Ki67 also requires validation. In the absence of validity, the Ki67 index will be influenced by a number of variables such as sample collection method, processing of samples, specimen staining procedure, analysis, and reporting. Ongoing quality assessment throughout the analysis of samples can ensure a flawless Ki67 score. IKWG has acknowledged the possibility of deviations in scoring at the preanalytical, analytical, interpretation, and data analytical phases of the Ki67 assessment.

Before Ki67 estimation from IHC is introduced into regular clinical practice, there are numerous essential concerns to consider. In order to acquire an optimal Ki67 score in all of the investigational preoperative studies, IHC analysis has to be carried out by highly trained analytical experts and practiced on standard unambiguous protocols on preanalytical and analytical procedures. In addition, a series of Ki67 analyses are performed to compare the biopsies taken at the surgery with those taken prior to or during early treatment [15].

Preanalytical Variabilities

Many factors with respect to biopsy specimens such as the process adopted for fixation, type of fixative, the time lag between collection of biopsy specimen and placing the specimen in fixative to the fixation of the tissue can impact “cold ischemia time” and duration of fixation of biopsy, can all change the Ki67 index [27].

The type of fixatives used varies with the type of specimen and this can lead to variation in the Ki67 index. In addition, the surgical method adopted serves as an important practical consideration. A mastectomy yields significantly more tissue than that of wide local excision, which may inhibit central tumor tissue fixation if it is not handled accurately. The prolonged period for complete fixation of excisional biopsies compared with core needle biopsies may result in a decrease in Ki67 values. For accurate results, minimum time should be used for fixation in the case of excision biopsy [15, 27].

Improper fixation or absence of fixation would prolong cold ischemia time and so the cells will no longer be in the cell cycle and would ultimately result in variation of the Ki67 index. However, this could be prevented by a standard histopathology tissue handling practice. Fixed specimens remain constant in the paraffin block for quite a long time than a cut section; conducting Ki67 analysis shortly after tissue fixation can yield more accurate results than performing with a sample stored in the paraffin block for a long time [29].

Ki67 scores also decrease with the use of fixatives, other than “neutral buffered formalin,” with delays of 16 h or more before fixation, or with overly short (3 h) or long (14 days) time of fixation with neutral buffered formalin [9]. If the tissue has not been properly fixed or embedded in paraffin, it results in the shedding of tissue and antigenicity will be lost.

IKWG has put enormous efforts to standardize the staining procedure to get near accurate values. Techniques adopted for staining of tissues vary between laboratories in terms of method of staining, type of platform used, the process of retrieval of antigen, primary antibody, detection system, and counter stain and all these will all influence the Ki67 results. Hence, periodical evaluation of staining protocols is mandatory to get consistent results in Ki67 analysis (Table 1).

Table 1.

Applications and limitations of Ki67 assay

Applications of Ki-67 assay

Limitations of Ki-67 assay

To predict the prognosis of early-stage BC To decide on the need for further adjuvant chemotherapy To predict whether chemotherapy may or may not be effective To determine whether the regimen chosen is working or an alternative choice should be considered

Poor analytical reproducibility Time of fixation changes with specimen type Time to fixation alters the cold ischemia time Fixatives other than neutral buffered formalin decrease Ki67 value Overly short (3 h) or long (14 days) fixation times decrease ki67 value Inter-laboratory variation exists due to differences in staining methodologies, including staining platform, antigen retrieval, primary antibody, detection system, and counter stain Substantial variation in Ki67 measurement across a single hot spot Standard cutoff for distinguishing low and high Ki-67 index does not exist Inter-observer variability

Analytical Variabilities

Ki67 index plays a role in deciding the need for adjuvant therapy in early-stage BC. Inconsistencies across laboratories have limited the use of the Ki67 value. Its value is inconsistent due to inter-observer variability, especially in midrange values (6–29). Moreover, the cut-off values may vary considerably between different institutions. Ki-67 scores and cutoff for clinical decision-making cannot be transferred among the laboratories without standardizing the scoring procedure, since its analytical validity is limited. However due to inconsistency in the interlaboratory methodology, it becomes unreliable and despite the apparent prognostic utility of the Ki-67 index, routine use of this tumor biomarker has not been widely recommended [30, 31].

Similar to other biomarkers, there occurs a large difference across a single hot spot used for evaluation in Ki67 analysis. In order to reach good reproducibility of Ki67 scoring among pathologists, IKWG has recorded “weighted global score” outputs without making use of “specific cell counting” or using “hot spot scoring” methods. This makes use of an online scoring app which can be accessed from the IKWG website “https://www.ki67inbreastcancerwg.org/.” Ki67 stained BC slide is reviewed and a score of 100 negative or positive nuclei will be recorded as directed by the software which accounts for the Ki67 index for that respective slide. A high level of attention is required to perform analysis which can be attained through training and practice so that they complete the evaluation within 9 min time. Though it is difficult to attain this refinement in the given laboratory setting, sustained attempts should be made toward perfection [15].

On other hand, the “Ki67 IHC MIB-1 PharmDx” is a laboratory test, designed to identify the expression of proliferative marker Ki67 protein in tumor cells. This is the first IHC assay measuring Ki67 expression to receive FDA approval for treatment with abemaciclib in combination with endocrine therapy in patients with early-stage breast cancer. Ki67 IHC MIB-1 PharmDx is applicable for automated staining using Dako Omnis instrument and has been quality controlled by IHC using required reagents and staining procedures. This diagnostic test aids in evaluating recurrence risk in early-stage breast cancer helping the physician to identify patients who may benefit from the drug abemaciclib combined with standard additional endocrine therapy. The Ki67 IHC MIB-1 PharmDx protocol is used with the primary antibody Ki67 IHC MIB-1 PharmDx, and the Ki67 IHC NCR PharmDx protocol is used with the isotype-matched negative control reagent. Ki67 protein expression is determined by assessing the percentage of viable tumor cells showing convincing nuclear staining at intensities 1 + and higher. But the use of Ki67 IHC NCR PharmDx on fine-needle aspirates has not been validated. The user should always ensure adherence to their protocol while using Dako Omnis instrument to obtain optimum results [32].

Interpretation of Data

In IHC analysis from a clinically accredited laboratory, antigen retrieval, antibody selection, colorimetric identification, along with adequate counter-staining of the nuclei that are not positive, all necessitate standardization to confirm the consistency of the Ki-67 index. Interpersonal variability may lead to variations in recording the Ki67 index by misinterpretation of scoring. Inter-observer concordance of visual assessment of Ki67 index for clinical decision-making is low. Due to the uncertainty in the selection of relevant cutoff points in the Ki67 index, controversy in data analysis exists. Apart from the inconsistencies in Ki67 scoring arising from different means of staining and evaluating procedures, variabilities in inter-laboratory procedures can also limit the reproducibility of this biomarker. While using MIB1 staining, the accuracy in assessment probably gets lost due to the heterogenicity of its expression pattern. The differences in staining across “tumor hot spots” and tumor peripheries can cause problems in judging the most representative part of the tumor overall. Presently, MIB1 is considered the gold standard for Ki-67 evaluation, since it is the most frequently used clone and has built up a long and validated track record [30]. The steps to be followed to limit the discrepancies in the Ki67 estimation are presented in Table 2.

Table 2.

Steps to be carried out to minimize the discrepancies in the estimation of Ki67 index

• Minimization of prefixation delay

• Division of surgical specimen to 5 to10-mm slices

• Fixation of specimen with neutral buffered formalin for 6 to 72 h

• In accordance with the type of specimen, fixation time should be altered

• Minimize the cold ischemia time

• Adopting “weighted global score” method recommended by IKWG instead of making use of “specific cell counting” or using “hot spot scoring”

• Standardize the cutoff for distinguishing low and high Ki-67 index

Conclusion

Ki-67 estimation is considered to be a valuable proliferative marker in predicting response, prognosis, and long-term outcome for the regimen preferred. Although it is extensively used in the histopathological evaluation, inconsistencies in the methodology adopted in Ki67 evaluation, lack of standardized guidelines, and inter-laboratory variations, all negatively impact the reliability and standardization of this biomarker in clinical practice. Standardizing the preanalytical and analytical variabilities in IHC Ki67 evaluation would minimize the inconsistencies in Ki67 estimation. Once standardized, Ki67 estimation in IHC could determine whether additional adjuvant chemotherapy will benefit early-stage BC, predict whether adjuvant chemotherapy will be effective or not, and monitor response during or after neoadjuvant endocrine therapy (NET) or neoadjuvant chemotherapy (NACT) to see if the preferred choice of therapy is effective or if an alternative should be considered [15]. This can potentially help thousands of patients belonging to low- and middle-income countries, who cannot afford genomic assays to determine the need for adjuvant therapy.

Declarations

Conflict of Interest

The authors declare no competing interests.