Is Ki-67 Really Useful as a Predictor for Response to Neoadjuvant Chemotherapy in Locally Advanced Breast Cancer?

Abstract

Neoadjuvant chemotherapy (NACT) is routinely offered to operable locally advanced breast cancer (LABC) patients desirous of breast conservation surgery and inoperable LABC patients. Pathological complete response (pCR) following chemotherapy is recognized as a surrogate for survival outcomes in high grade tumour subtypes. Many biological and tumor characters have been shown to predict pCR. The current study was performed with the aim of investigating the ability of Ki-67 in predicting pCR with NACT in breast cancer patients. A total of 105 patients with locally advanced breast cancer who completed NACT followed by surgery were included in this study from January 2020 till December 2022. Patients with advanced metastatic breast carcinoma, who did not give consent for NACT, who did not complete NACT and who did not undergo surgery were excluded. All patients were assessed for Ki-67 score on core-needle biopsy samples and response rate was assessed clinically and by histopathological examination of resected specimen. Quantitative variables were compared using unpaired t-test or Mann–Whitney ‘U’ test and for categorical variables Chi-square or Fisher’s exact test were used. Receiver operating characteristic (ROC) curve analysis was performed to assess the predictive potential of Ki-67 expression levels in predicting pCR. To identify the predictive factors associated with pCR, univariate analysis was performed. The P value < 0.05 was considered as statistically significant. Mean age was 51.57 ± 10.8 years. 51 patients achieved clinical complete response (cCR) and 33 achieved pCR after NACT. Mean Ki-67 index in overall study population, in pCR group and no pCR group was 46.44 ± 22.92%, 51.60 ± 22.3% and 44.06 ± 22.7%, respectively. On univariate analysis, ER negativity, PR negativity and Her 2neu positivity were found predictive of pCR. On subgroup analysis, TNBC and Her 2neu positive sub groups were associated with higher cCR and pCR rate. We found no significant association between Ki-67 and pCR. This result may be confounded by the fact that a significant duration of the study was in the COVID-19 pandemic. Validation of this data is required in a large prospective study.

Introduction

Breast cancer is the most common malignancy in women in the entire world [1]. Locally advanced breast cancer (LABC) constitutes 5 to 15% of newly diagnosed breast cancer cases in the Western countries and 60% of breast cancer cases in developing countries [2]. In India, LABC constitutes half of all breast cancers [3]. Neoadjuvant chemotherapy (NACT) is the initial treatment of both operable LABC and inoperable LABC patients. But response to NACT in terms of survival benefit is difficult to predict [4, 5] although some studies have shown improved short-term outcomes [6]. Clinical and radiological assessment cannot provide an accurate estimate of the amount of residual tumour. Pathological examination of the excised tumour bed is important to identify pathological complete response (pCR). pCR following chemotherapy has been identified as a surrogate for survival outcomes especially in high grade tumour subtypes [7]. Factors which can predict pCR include estrogen receptor (ER) negativity, HER-2 positivity, high grade, high Ki-67 index, small tumour size and node positivity [8–10]. pCR has been shown to predict improved survival in triple negative breast cancer (TNBC) and HER2-enriched subgroups, but there are contradictory data with respect to luminal subtypes [11, 12].

Ki-67 is a nuclear protein expressed during all phases of cell cycle, except G0, and its expression correlates with tumour cell proliferation rate. Many studies have investigated Ki-67 as a prognostic and predictive marker for breast cancer. But none of the studies found consistent results regarding the impact of NACT on status of tumour biomarkers [11, 13]. Some studies have shown a high level of proliferation activity can predict response to NACT [8]. Tumours with high Ki-67 level (> 20%) had better response to NACT than tumours with low Ki-67 levels [9]. Studies have also shown that higher Ki-67 index can better predict achievement of pCR and hence, response to neoadjuvant hormonal therapy and chemotherapy [8–10, 12, 14–18]. The aim of individualized medicine is to identify patients who may benefit from a specific form of treatment. There are contradictory data about Ki- 67 as a predictive factor [19, 20] for response to NACT which raises several questions like the methodology used for Ki-67 detection, or how Ki-67 expression in core biopsies is representative of the whole tumour. Assessment of cellular proliferation by Ki-67 expression is not recommended in routine pathological evaluation by either International Ki-67 in Breast Cancer Working Group of the Breast International Group and North American Breast Cancer Group or the American Society of Clinical Oncology (ASCO) given the fact that it is not totally clear how Ki-67 measurements and thresholds could influence clinical decisions [21].

Therefore, the current study was performed with the aim of assessing the ability of Ki-67 to predict pCR to NACT. Also, the response rate in different subtypes of LABC and different clinicopathological parameters was evaluated.

Material and Methods

This was a prospective observational study carried out at Apollo CBCC Cancer Care, Bhat GIDC Estate, Gandhinagar as well as Apollo CBCC Cancer Care City Hospital ‘Akshara’, Ellisbridge, Ahmedabad. The study was carried out for a period of three years from January 2020 to December, 2022. During this period, all biopsy proven breast cancer patients who completed NACT and underwent surgery in the oncology department of the above-mentioned hospitals were included in the study. Patients with advanced metastatic breast carcinoma, who did not give consent for NACT, patients who did not complete NACT and patients who did not undergo surgery were excluded.

Data Collection

Bio data, chief complaints, history, local, general and systemic examination were noted. Routine blood investigations like complete blood count, serum creatinine, SGPT (Serum alanine transaminase) and bilirubin were checked. The diagnosis of invasive breast cancer was proven pathologically via core needle biopsy. The clinical and pathologic characteristics of all patients were recorded including age, menopausal status, tumour size and nodal status. Distant metastases were excluded with appropriate imaging modality like PET-CT, CT Chest, CT Abdomen and Bone scintigraphy. Pathological characteristics like histology, grade of tumor, ER/progesterone receptor (PR)/HER-2 neu status and Ki-67 index on the preoperative biopsy specimen were documented.

Pathological Procedures

Formalin-fixed paraffin embedded tissues were cut at 4-μm thickness and the sections were mounted on poly-L-lysine coated slides. Staining of specimen was done using the Ventana HX automatic system Benchmark (Ventana Medical Systems, SA, Illkirch Cedex, France), with the following antibodies anti-ER (clone SP1; Ventana Medical Systems, SA) rabbit monoclonal primary antibody; anti-PR (cloneIE2; Ventana Medical Systems, SA) rabbit monoclonal primary antibody; anti HER-2/neu (clone 4B5; Ventana Medical Systems, SA) rabbit monoclonal primary antibody; anti-Ki-67 (MIB-1; 1:50; DAKO) mouse monoclonal antibody. Antigen retrieval and blocking of endogenous peroxidase activity were performed automatically by the Benchmark system. The tissue sections were incubated with primary antibody for 32 min at 42 °C. Immuno-peroxidase staining was performed using the LSAB system Neu Vision, according to the manufacturer’s instructions (Ventana) and sections were counterstained with hematoxylin eosin.

Immunohistochemical Determination

The ER and PR positive cell were identified and both the percentage of positive tumor cells and the intensity of staining were determined. The tumors were considered to be positive for the estrogen and progesterone receptors if ≥ 1% of the cells showed positive staining. For the evaluation of HER-2 expression, semi-quantitative analysis was performed according to scoring guidelines from the Herceptest instruction guide. Only membranous staining intensity and patterns was evaluated as defined in the guidelines. A score of 0 was given to sections showing no staining or membrane staining in < 10% of tumor cells; 1 + was given to sections showing faint or barely perceptible membrane staining in more than 10% of tumor cells; 2 + was given to sections showing weak to moderate complete membrane staining in more than 10% of tumor cells; and 3 + was given to specimens showing strong complete membrane staining in more than 10% of tumor cells. Scores of 0 and 1 + were considered negative for HER-2 over expression, whereas a score of 3 + was considered as HER-2 over expression. If 2 + results was obtained using the Herceptest scoring system, HER-2 testing was performed by validated dual-probe ISH assay. HER-2 by ISH was considered positive only if either the HER-2/CEP17 ratio is ≥ 2 or average HER-2 copy number were ≥ 6 signals/cell (33). The cells with Ki-67 were counted and expressed as the percentage of cells with positive nuclear staining among the total number of tumor cells (at least 1000 were counted). If the staining was homogenous throughout the section, three high-power fields were examined and score was calculated. If Hot spots (areas with increased numbers of Ki67 positive cells) were present, whole section was assessed and the overall average score was recorded.

Treatment Given

Patients received 6–8 cycles of chemotherapy depending on standard guidelines for neoadjuvant settings. In HER-2 positive patients, option of targeted therapy with trastuzumab and/or Pertuzumab was also given. All patients underwent surgery after 3 to 4 weeks of the last dose of NACT.

Evaluation of Response to NACT

Response was assessed clinically after each chemotherapy cycle before next cycle. Patients were evaluated by imaging for response (in selected cases) before surgery after completion of decided neoadjuvant chemotherapy. The Response Evaluation Criteria in Solid Tumours (RECIST) guidelines version 1.1 was used for the quantitative analysis of clinical response. The residual tumour size and lymph node status were noted. pCR was defined as no pathologic evidence of a residual invasive carcinoma in the breast or axillary lymph nodes (ypT0N0 or ypTisN0). Residual ductal carcinoma in situ was included under pCR.

Statistical Analysis

Descriptive analysis was done using mean or median with standard deviation (SD) or inter-quartile range for quantitative variables and frequencies along with respective percentages for categorical variables. Graphical presentation of data was done using appropriate pie or bar diagrams. The statistical comparisons for quantitative variables were done using unpaired t-test or Mann–Whitney ‘U’ test while Chi-square or Fisher’s exact test were used for categorical variables. Receiver operating characteristic (ROC) curve analysis was performed to assess the predictive potential of Ki-67 expression levels in predicting pathological complete response and to assess the optimal cut-off value for Ki-67. Univariate analysis was performed to identify the predictive factors associated with pCR. The data was entered in MS EXCEL spreadsheet and analysis was done using Statistical Package for Social Sciences (SPSS) (Version 21.0, SPSS Inc, Chicago, IL, USA) and Medcalc version 12.5 for windows (MedCalc Software, Ostend, Belgium). The P value < 0.05 was considered as statistically significant.

Ethical Issues

The study was approved by the Institutional Ethics Committee Biomedical Research, Apollo Hospitals International Ltd., Ahmedabad, Gujarat, India on February 28, 2020. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was taken from all patients. No harm was intended for the subjects. The participants were not subjected to any extra cost because of the study.

Results

A total of 105 patients were included in this study. 53 (50.5%) patients were postmenopausal while 52 (49.5%) patients were premenopausal. Mean age was 51.57 ± 10.8 years. 2 patients were tumour stage T1 (< 2 cm), 31 patients were T2 (2–5 cm), 17 patients were T3 (> 5 cm), 3 were T4a (chest wall fixation) and 52 were T4b(involvement of skin). 7 patients had N0 disease, 55 patients had N1 disease, 27 had N2 disease and 16 had N3 disease.

All patients had Infiltrating duct carcinoma with 12 Grade II and 93 Grade III tumour. 56 patients were ER positive and 49 ER negative, 58 patients were PR positive and 47 PR negative, 37 patients were Her 2 positive and 68 Her 2 negative (Table 1).

Table 1.

ER, PR, HER-2 status

	Status	Frequency	Percentage
ER	Negative	49	46.7
	Positive	56	53.3
PR	Negative	47	44.8
	Positive	58	55.2
HER2	Negative	68	64.8
	Positive	37	35.2

With regard to treatments, all patient received atleast 6–8 cycles of NACT. Her 2 neu positive patients also received Trastuzumab &/or Pertuzumab based chemotherapy regimen. Most commonly used regimen was 4 cycles of Doxorubicin followed by 4 cycles of paclitaxel with or without Trastuzumab (4 AC – 4 T ± H). Other used regimens were 6 cycles of TAC – Docetaxel + Doxorubicin + Cyclophosphamide or 6 cycles of TCH ± P – Docetaxel + Carboplatin + Trastuzumab with or without Pertuzumab.

Clinical complete response (cCR) was defined as per RECIST 1.1 and pCR was defined as no invasive tumour in breast or axilla. 51 patients achieved cCR and pCR was observed in 33 patients (Table 2).

Table 2.

Rates of pCR and cCR

	Response	Frequency	Percentage
pCR	No	72	68.6
	Yes	33	31.4
cCR	No	54	51.4
	Yes	51	48.6

Patients were classified into 4 major molecular subtypes based on ER, PR, Her 2 status. 25 (23.8%) patients were in TNBC group, 43 (41%) were in Hormone positive Her 2 neu negative group, 21 (20%) patients were in Hormone positive Her 2 positive group, 16 (15.2%) were in Hormone negative Her 2 neu positive group.

cCR and pCR rates were assessed in these groups. 10/43 (23.2%) patients with HR ( +)/HER-2 ( −); 14/21 (66.6%) patients with HR ( +)/HER-2 ( +); 13/16 (81.2%) patients with HR ( −)/HER-2 ( +) and 13/25 (52%) patients with TNBC tumours achieved cCR. 4/43 (9.3%) patients with HR ( +)/HER-2 ( −); 8/21 (38.1%) patients with HR ( +)/HER-2 ( +); 11/16 (68.7%) patients with HR ( −)/HER-2 ( +) and 9/25 (36%) patients with TNBC tumours achieved pCR (Fig. 1).

Fig. 1.

Correlation of subgroups and pCR

With regard to Ki-67 labeling index, mean Ki-67 index was 46.44 ± 22.92% in all patients. However, mean Ki-67 index was 50.90 ± 21.83% in those who achieved cCR while it was 42.24 ± 23.1% in those who did not achieve cCR. Mean Ki-67 index was 51.60 ± 22.3% in those who achieved pathological complete response while it was 44.06 ± 22.7% in those who did not achieve pathological complete response.

ROC curve analysis was used to determine the optimal cut-off value for categorization of Ki-67 (into high or low) to predict response to NACT. The area under the ROC curve for Ki-67 expression and pCR was 0.59 (P = 0.14; 0.47–0.71) (Fig. 2) which indicated that baseline Ki-67 expression level is not a useful parameter for prediction of achievement of pathological complete response. Similarly, the area under the ROC curve for Ki-67 expression and cCR was 0.60 (P = 0.07; 0.50–0.71) which indicated that baseline Ki-67 expression level is not a useful parameter for prediction of achievement of pathological complete response.

Fig. 2.

ROC Ki-67 by pCR

Univariate analysis was performed to find the association of baseline clinico-pathologic factors with cCR and pCR. ER negativity, PR negativity and HER-2 positivity were the factors associated with better pCR and cCR rates (Table 3).

Table 3.

Association between study variables and pCR

		pCR		Total	Chi square test
		No	Yes		Chi square value	p-value
Age	30 – 39	13	5	18	5.22	0.27(NS)
		72.2%	27.8%	100.0%
	40 – 49	24	9	33
		72.7%	27.3%	100.0%
	50 – 59	14	13	27
		51.9%	48.1%	100.0%
	60 – 69	15	5	20
		75.0%	25.0%	100.0%
	70 – 79	6	1	7
		85.7%	14.3%	100.0%
Menopausal status	Postmenopausal	35	18	53	0.32	0.57(NS)
		66.0%	34.0%	100.0%
	Premenopausal	37	15	52
		71.2%	28.8%	100.0%
Clinical stage	T1	2	0	2	-	0.35(NS)#
		100.0%	0.0%	100.0%
	T2	17	13	30
		56.7%	43.3%	100.0%
	T2(m)	0	1	1
		0.0%	100.0%	100.0%
	T3	12	5	17
		70.6%	29.4%	100.0%
	T4	1	0	1
		100.0%	0.0%	100.0%
	T4a	3	0	3
		100.0%	0.0%	100.0%
	T4b	37	14	51
		72.5%	27.5%	100.0%
Nodal stage	N0	1	6	7	-	0.03*#
		14.3%	85.7%	100.0%
	N1	39	16	55
		70.9%	29.1%	100.0%
	N2	14	2	16
		87.5%	12.5%	100.0%
	N2a	7	3	10
		70.0%	30.0%	100.0%
	N2b	1	0	1
		100.0%	0.0%	100.0%
	N3a	1	2	3
		33.3%	66.7%	100.0%
	N3b	2	1	3
		66.7%	33.3%	100.0%
	N3c	7	3	10
		70.0%	30.0%	100.0%
Histology	IDC	72	33	105	-	-
		68.6%	31.4%	100.0%
Grade	2	10	2	12	1.37	0.24(NS)
		83.3%	16.7%	100.0%
	3	62	31	93
		66.7%	33.3%	100.0%
ER	Negative	25	24	49	13.13	< 0.001*
		51.0%	49.0%	100.0%
	Positive	47	9	56
		83.9%	16.1%	100.0%
PR	Negative	26	21	47	6.93	0.008*
		55.3%	44.7%	100.0%
	Positive	46	12	58
		79.3%	20.7%	100.0%
HER2	Negative	54	14	68	10.52	0.001*
		79.4%	20.6%	100.0%
	Positive	18	19	37
		48.6%	51.4%	100.0%

^#Fisher's exact test;  *p < 0.05 Statistically significant; NS Non Significant

Discussion

This study evaluated Ki-67 as a predictive factor for response to NACT. Clinical response was evaluated by palpation of the breast and axilla in all cases and in selected cases one imaging modality either MRI or PET-CT was done at the discretion of the treating physician. cCR was observed in 51 patients (48.6%) in this study which was comparatively higher to cCR rates of 32.1% observed by Macgrogan et al. [8] but lower to cCR rates of 63.6% observed by Bear et al. [22] in their studies. This difference might be due to higher proportion of T4 tumours in our study.

33 out of 51 (64.7%) patients who had cCR also had pCR. 18 patients who achieved cCR on clinical examination and imaging were later on found to have residual foci of tumor on final histopathological examination of the excised specimen. This consolidates the fact that achieving cCR may not always translate into pCR and highlights the importance of histopathological examination as the gold standard.

Various definitions of pCR have been used in different studies. Some have included absence of residual invasive and non-invasive tumor in breast as well as axilla (ypT0N0 status) [8, 23], while some included non-invasive residual tumor (ypT0-is N0 status) in definition of pCR [24–26] while some didn’t include nodal status (ypT0-is N0-3 status) in definition of pCR [22–24]. This study defined pCR as absence of residual invasive tumor in breast or axilla (ypT0-is N0 status). Wide range of pCR rates of 13–65% have been reported in different studies [8, 14, 17, 22, 23, 27–29]. Such huge variations in pCR may be due to varied definitions of pCR, selection of the breast cancer subtypes with high proliferation and chemo-sensitivity. pCR in this study was 31.4% which was similar to that observed in two of the largest trials of NACT [22, 23].

ER and PR negative tumours are highly proliferating and chemo sensitive as shown in previously published studies [7, 30]. ER and PR negative tumours are also associated with higher pCR and cCR rates as observed in this study and previously published studies [8, 9, 14, 17, 29, 31, 32].

In this study there was positive correlation of HER-2neu with pCR rates. pCR rates were 51.35% and 20.58% in patients who were HER-2 positive and HER-2 negative, respectively. Several studies present conflicting data on HER-2neu as a predictive factor in breast cancer. Initial studies had found that HER-2 is not a predictive factor of response to NACT [33], as they did not include trastuzumab in the neoadjuvant regimen. This study showed results similar to more recent studies where the efficacy of trastuzumab and or pertuzumb in the neoadjuvant setting has been evaluated. In the first randomized trial to evaluate the addition of trastuzumab to chemotherapy in the neoadjuvant setting, Buzdar et al. [34] reported a 66.7% pCR rate among patients receiving chemotherapy plus trastuzumab versus 25% among patients receiving chemotherapy alone (P = 0.02). After this, the GeparQuattro [35] and NOAH [36] (NeO Adjuvant Herceptin) trials have shown improved pCR rates among patients receiving trastuzumab, with an event-free survival benefit seen at 3 years among patients treated with trastuzumab in the NOAH trial [36].

Ki-67 is a nuclear antigen which is expressed in all phases of cell cycle except G0 phase [10, 13] and is considered a marker of proliferation [10]. It has been seen that high proliferative activity has predictive value for response to NACT [7]. Overall, evidence indicates that levels of Ki-67 are low during G1 and early S-phase and progressively increase to reach a maximum during mitosis [37]. In this study, mean Ki-67 index in overall study population, in pCR group and no pCR group was 46.44 ± 22.92%, 51.60 ± 22.3% and 44.06 ± 22.7%, respectively, which was statistically not significant. Possible reasons being, small sample size, lack of standardization for ki 67 estimation method, different chemotherapy regimens used, different dose intensities used (due to toxicities or logistic reasons), variable interval between the cycles. This study was conducted during era of COVID-19 pandemic. So many patients defaulted treatment, many deferred surgery after taking NACT and many patients faced delay in surgery post chemotherapy in view of pandemic. All these confounding factors also affected the response in terms of pathological outcome post NACT. Our study shows the nuances of treating patients with breast cancer in the pandemic period. As our numbers are small, larger studies are required to confirm the findings of the present study.

Patients were classified into 4 major molecular subtypes based on ER, PR, Her 2 status. These subtypes are, TNBC, Hormone positive Her2neu negative, Hormone positive Her 2 positive, Hormone negative Her 2 positive. 25 (23.8%) patients were in TNBC group, 43 (41%) were in Hormone positive Her 2 negative group, 21 (20%) patients were in Hormone positive Her 2 positive group, 16 (15.2%) were in Hormone negative Her 2 positive group. 4/43 (9.3%) patients with HR ( +)/HER-2 ( −); 8/21 (38.1%) patients with HR ( +)/HER-2 ( +); 11/16 (68.7%) patients with HR ( −)/HER-2 ( +) and 9/25 (36%) patients with TNBC tumours achieved pCR.

Similar observation of higher pCR rates in TNBC and HER-2 positive breast cancer had been shown in previously published studies [8, 9, 28]. St Gallen Consensus 2013 [38] recommended the use of Ki-67 as an additional factor to distinguish luminal A and luminal B type of breast cancer. It was seen that Ki-67 levels > 20% had better concordance with molecular studies in distinguishing luminal A from luminal B type. This distinction between HR positive breast cancer is important from treatment perspective, as patients with luminal B type breast cancer are more chemo sensitive than luminal A type.

There are several problems in detection of Ki-67 expression levels like inter and intra-observer variation in reporting because different monoclonal antibodies (MIB-1, MM-1, Ki-S5,SP-6) used have produced varied results and heterogeneous expression of Ki-67 in tumors. This variability in Ki-67 reporting has been confirmed by Polley et al. [39] in their study. Moreover, there is no validated cut-off for Ki-67 in predicting response to NACT and pathological complete response. Nishimura et al. [15] used the cut off of 12%, Li XR et al. [40] used the cut-off of 20%, and Kim et al. [10] used the cut-off of 25%, which suggested the variable cut-off used for Ki-67. But all these studies suggested positive correlation of high Ki-67 and pCR. We used ROC method to calculate the best cut-off level of Ki-67 in predicting pCR but our cut-off for Ki-67 was statistically not significant.

We also performed univariate analysis and found that ER negativity, PR negativity, HER-2 positivity are predictive of pCR.

Conclusion

In this study, no significant association was found in between Ki 67 score and pCR rate. Thus Ki 67 index was apparently not a good marker to predict response to neo adjuvant chemotherapy. This result may be confounded by the fact that a significant duration of the study was in the COVID-19 pandemic. Our study shows the nuances of treating patients with breast cancer in the pandemic period. Validation of this data is required in a large prospective study.

Funding

None.

Data Availability

The materials are available upon request from the corresponding author.

Declarations

Conflict of Interest

None declared.