JCDR - Register at Journal of Clinical and Diagnostic Research
Journal of Clinical and Diagnostic Research, ISSN - 0973 - 709X
Pathology Section DOI : 10.7860/JCDR/2021/50778.15580
Year : 2021 | Month : Oct | Volume : 15 | Issue : 10 Full Version Page : EC44 - EC47

MCM2 Expression in Different Molecular Subtypes of Epithelial Breast Cancers and its Association with Clinicopathological Parameters and Ki-67 Expression

Meghadipa Mandal1, Anadi Roy Chowdhury2, Susmita Mukhopadhyay3

1 Junior Resident, Department of Pathology, R.G. Kar Medical College and Hospital, Kolkata, West Bengal, India.
2 Professor and Head, Department of Pathology, Murshidabad Medical College and Hospital, Berhampore, West Bengal, India.
3 Assistant Professor, Department of Pathology, R.G. Kar Medical College and Hospital, Kolkata, West Bengal, India.


NAME, ADDRESS, E-MAIL ID OF THE CORRESPONDING AUTHOR: Dr. Meghadipa Mandal, AC 1, Action Area 1, Farsight Housing Cooperative Society, New Town, North 24, Parganas, Kolkata-700156, Bengal, India.
E-mail: meghadipa.mandal41@gmail.com
Abstract

Introduction

Breast cancer is one of the most common malignancies, with few subtypes having a more aggressive outcome and resistance to conventional therapies, Triple Negative Breast Cancers (TNBCs) being one such variant. The Ki-67 lacks reproducibility and a standardised cut-off. MCM2 (Minichromosome Maintenance 2) has role in DNA repair and replication and its role as alternate marker for prognosis has been studied in this case.

Aim

To study MCM2 expression with respect to histologic grade, stage, nodal status and molecular subtypes of breast carcinoma. Also, to look for any correlation between Ki-67 and MCM2 expressions.

Materials and Methods

A cross-divtional, observational study conducted on a group of 20 patients who underwent mastectomy in a Tertiary Care Centre, R.G. Kar Medical College and Hospital, Kolkata, West Bengal, India, for a total duration of six months. Histologic grading, staging, nodal status was evaluated from Haematoxylin and Eosin (H&E) stained divtions. Formalin-Fixed Paraffin-Embedded (FFPE) blocks suitable for Immunohistochemistry (IHC) were selected and MCM2, Estrogen Receptor (ER), Human Epidermal Growth Factor Receptor 2 (HER2), Progesterone Receptor (PR)/neu and Ki-67 were performed. Scores given based on visual examination under light microscope. Analysis was done using IBM Statistical Package for the Social Sciences (SPSS) version 25.0 software.

Results

Most of the subjects belonged to 41-55 years age group. Statistical significance was seen between high MCM2 and Ki-67 expressions (p-value=0.0171) and high histologic grade and TNBCs (p-value=0.009). High MCM2 and Ki-67 expressions also come with increased risk for advanced disease. High Ki-67 is also a risk predictor for lymph node positive cases. Positive correlation was seen between MCM2 and (R)= 0.4318.

Conclusion

The MCM2 is a predictor for adverse outcomes in breast carcinoma cases. It may serve as an alternative to Ki-67 as a proliferation marker, to guide clinicians in treatment strategies. Its role as a therapeutic target in aggressive breast carcinomas may be evaluated with larger study population in the future.

Keywords

Introduction

Breast carcinomas are the most heterogenous group of neoplasms in terms of their clinical, pathological and molecular variability [1] and one of the most common malignancies in current era [2]. While some have indolent course, few like TNBCs are notorious for aggressive clinical outcomes [3]. Different biomarkers and molecular profiling are being attempted to prognosticate outcomes of different breast carcinoma patients and also for implementation of novels molecular targets [4]. The Ki-67 is one such biomarker that utilises proliferation potential of breast carcinoma as one of the main indicators of aggressiveness, which lacks a standardised cut-off for grading the neoplasm [5,6].

Minichromosome Maintenance 2 is a part of complex MCM2-7 that belongs to the family of minichromosome maintenance proteins and is regulated by a group of transcription factors also involved in breast carcinomas [7]. The MCM2 has both cell proliferation associated domain and apoptosis enhancing domain, with important role in DNA damage response, replication and genomic stability [8]. Furthermore, its role in DNA repair and replication initiation has been utilised time and again as novel therapeutic targets in colon cancers and lung cancers [9,10]. Here, it has been evaluated as a proliferation marker with respect to different aspects of breast carcinogenesis.

The objective was to study the expression of MCM2 and Ki-67 in breast carcinomas by Immunohistochemistry (IHC). The grades of MCM2 expression were compared with histologic grade, type, staging, nodal status, molecular subtypes and Ki-67 expression. Any correlation between grades of MCM2 and Ki-67 expression was also evaluated.

Materials and Methods

This cross-sectional, observational study was conducted in a Tertiary Care Centre, R.G. Kar Medical College and Hospital, Kolkata, West Bengal, India after approval from Institutional Ethical Committee (IEC- ECR/322/Inst/WB/2013), for a total duration of six months from October 2019 to April 2020.

Inclusion criteria: The study consisted of patients who had undergone Modified Radical Mastectomy (MRM) in the institute and specimen sent for histopathological evaluation in Department of Pathology during the study period after obtaining the informed consent.

Exclusion criteria: Cases that had received Neo-Adjuvant Chemotherapy (NACT) with no proper tumour foci, no residual tumour or abundant necrosis were excluded from the study population.

A total number of 20 cases were found suitable for the present study within the stipulated time frame. Formalin-Fixed Paraffin-Embedded (FFPE) sections were examined by routine Haematoxylin and Eosin (H&E) stain; histologic type, grade by Elston Ellies Modification of Bloom Richardson Grading (BR Grade), staging and nodal status were evaluated and recorded [11]. Blocks suitable for immunohistochemical analysis of Estrogen Receptor (ER), Progesterone Receptor (PR), HER2/neu, Ki-67 and MCM2 were selected.

The IHC analysis was done using following antibodies: MCM2 (Rabbit Monoclonal IHC Antibody, EP40 (PathnSitu); Lot number R05053NAX), ER (Rabbit Monoclonal Antibody-Cell Marque- Lot number 1504203D), PR (Rabbit Monoclonal Antibody-Cell Marque-Lot number 1514106B), HER2/neu (Mouse Monoclonal Antibody-Cell Marque- Lot number 1515602B) and Ki-67 (Mouse Monoclonal IgG1 Antibody, GM001-PathnSitu). The chromogen used was 3,3′-Diaminobenzidine (DAB). Specificities of anti-MCM2 were confirmed using normal colonic mucosa based on the expression data in the human protein atlas database [Table/Fig-1] [12].

Positive control for MCM2 in colonic mucosa (100x magnification, IHC).

The stained IHC slides were examined under light microscope. The MCM2 expression was recorded by visual method as percentage of positive nuclear stains in hotspot areas (range: 0%-100%) [13]. Similar method was used to record Ki-67 expression status. For reporting ER, PR and HER2/neu status, the World Health Organisation (WHO) recommendations (2019) have been followed [11].

Statistical Analysis

All the data were tabulated in a Master Sheet and analysed using IBM Statistical Package for the Social Sciences Version 25.0 software. Descriptive statistics like mean, median, mode were used wherever applicable. Pearson’s Chi-square test (χ2) and Fisher’s Exact test were used to calculate statistical significance, where p-value is considered less than 0.05 as significant. Pearson’s Correlation Coefficient (R) and Relative Risk (RR) were also used wherever applicable.

Results

The study population belonged to the age group of 41-55 years amongst 10 cases (50%). The histologic type of Invasive Breast Carcinoma No Special Type (IBC-NST) included 18 out of 20 cases. Two other cases were that of metaplastic carcinoma. Majority of 12 cases (60%) were of BR grade 3 and 15 cases (75%) were of Locally Advanced Breast Carcinomas (LABC). The cases with positive nodal status were 12, whereas in four cases the nodal status was indeterminate (Nx). There were 5 ER positive (25%), 3 PR positive (15%) and 2 HER2 positive (10%) cases. Molecular classification showed that a majority of 14 cases (70%) were TNBC, three were Luminal B, two were Luminal A and one was HER2 positive (non luminal). [Table/Fig-2] summarises all the above observations.

Distribution of study population based on demographic variables and various immunohistologic parameters.

Demographic variablesFrequency (n)Percentage (%)
Age (years n=20)
≤40315
41-551050
56-70735
Histologic type (n=20)
IBC-NST1890
Others (Metaplastic carcinoma)210
BR Grade (n=20)
Grade 2840
Grade 31260
Stage (n=20)
Early525
LABC1575
Nodal status (n=16)
Node positive1275
Node negative425
Molecular types (n=20)
TNBC1470
Luminal A210
Luminal B315
HER2 positive (non luminal)15

IBC NST: Invasive breast carcinoma no special type; LABC: Locally advanced breast carcinomas; TNBC: Triple negative breast cancers; BR Grade: Bloom richardson grading


The mean MCM2 expression amongst the study population was 66.5% and median was 70%. For the purpose of grading, median value has been taken as cut-off [14]. So, an expression of <70% has been graded as low MCM2 expression and similarly ≥70% taken as high MCM2 expression [Table/Fig-3,4]. Following this criterion, 60% of the study population, i.e., 12 cases showed a high MCM2 expression. As per study conducted by Lombardi A et al., 20% is taken as cut-off for Ki-67 expression [5], which is also the median value in the present case. Cut-off >20% is graded as high Ki-67 expression and similarly, ≤20% is considered as low Ki-67 expression [Table/Fig-5]. Accordingly, 55% (11 out of 20 cases) showed low Ki-67 proliferation index in this study population, mean value of expression being 31.5%.

High MCM2 expression in BR Grade 3 breast cancer. (400x magnification);

Low MCM2 expression in BR Grade 2 breast cancer. (400x magnification).

High Ki-67 expression in TNBC. (400x magnification).

On performing Chi-square test, it was seen that high MCM2 expression was associated with high histologic grade of the tumour and the difference was statistically significant (p<0.05). Similar association was seen between high MCM2 expression and hormone receptors (ER/PR) negative tumours, with p-values less than 0.05 in both the cases. The TNBC cases also showed a very high MCM2 expression and the relationship was statistically significant. Such significant association (p<0.05) was also seen between high MCM2 and concurrently high Ki-67 expressions, with a weak but positive correlation (Pearson’s correlation coefficient (R)= 0.4318). The Ki-67 was also seen to be significantly high amongst high histologic grade and triple negative cases [Table/Fig-6].

Statistical associations between Ki-67 and MCM2 expressions and histologic type, grade, stage, nodal status and molecular variants of breast carcinoma.

MCM2High MCM2 (≥70%)Low MCM2 (<70%)p-value (significant at <0.05)Comment
Histologic type
IBC-NST1080.4947 (Fisher-Exact Test)Not significant
Others (Metaplastic carcinoma)20
BR Grade
Grade 2260.0090Significant
Grade 3102
Stage
Early230.29181. Not significant.2. r (Pearson’s correlation coefficient)=0.2645 (weak positive correlation).3. RR=1.33
LABC105
Lymph Node status
Positive750.5509Not significant
Negative31
ER status
Positive050.0036 (Fisher’s-Exact Test)Significant
Negative123
PR status
Positive030.0491 (Fisher’s-Exact Test)Significant
Negative125
Molecular types
TNBC1130.009Significant
Others15
Ki-67 index
High (>20%)810.01711. Significant2. r=0.4318 (weak but positive correlation)
Low (≤20%)47
Ki-67High Ki-67 (>20%)Low Ki-67 (≤20%)p-value (significant at <0.05)Comment
BR Grade
Grade 2170.0171Significant
Grade 384
Stage
Early230.79521. Not significant2. R=0.5986 (moderate positive correlation)3. RR=1.07
LABC78
Lymph node status
Positive660.38271. Not significant2. RR=1.29
Negative13
Molecular types
TNBC950.0141 (Fisher-Exact Test)Significant
Others (Luminal A, Luminal B, HER2 positive)06

Statistical tests used are- Chi-square test; Fisher’s-exact test; RR: Relative risk; r: Pearson’s correlation coefficient


No significant association was seen between MCM2 expression and histologic type, stage and nodal status of the cases. The Ki-67 expression and stage and nodal status of the patients were also not significantly associated. However, weak but positive correlation (r=0.2645) was seen between high MCM2 expression and advanced stage of the disease with a RR=1.33 >1. This indicates that a higher MCM2 expression is predisposed to have an advanced stage of cancer. Similar moderate positive correlation (r=0.5986) and RR=1.07 >1 was also seen between high Ki-67 expression and late stage of the disease. Risk ratio of 1.29 indicates that high Ki-67 index cases have higher risk of node positive diseases. All these statistical conclusions have been summarised in [Table/Fig-6].

Discussion

Breast cancer is one of the leading causes of cancer-related mortalities and deaths round the world, with very high incidence rate [15]. The majority of study population belonged to the age group of 41-55 years which corroborated with other study populations [16,17]. A major population in the present study was of locally advanced disease (Stage III), with no stage IV cases as the metastatic status of our cases were unknown. This was discordant with the findings of Kim EJ et al., where majority were stage I cases of breast cancer [18]. This can be attributed to the fact that there might be lack of awareness amongst women regarding breast cancer and screening by self-breast examination that leads to late presentation of the disease. Hospital bias may also be one of the contributing factors, where only the critically ill patients with advanced disease were referred from periphery to tertiary centre for further management. A substantial proportion of 70% was TNBC in this case, which was in concordance with Thakur KK et al., who evaluated the alarming burden of TNBCs among Indian population, influenced by various socio-demographic and genetic factors [19].

Statistical significance was established between high MCM2 expression and higher histologic grades of breast cancers. Similar association was also established by Bukholm IR et al., and many other studies [20-23]. Such statistical association was also seen between high MCM2 expression and the aggressive triple negative phenotype of breast carcinoma, which was also seen by Yousef EM et al., and others [1,7,21]. Tumours having high Ki-67 expression, which are aggressive in nature were also seen to have high MCM2 expressions as was also demonstrated by Wojnar A et al., and others [24,25]. High MCM2 was also associated with increased risk (RR>1) for advanced stage of disease, although statistical significance could not be established possibly due to small sample size of the study. However, previous works with a higher study population could conclusively prove that higher MCM2 expression was associated with node positive and advanced pTNM staging [26]. The Ki-67 was seen to be expressed more in higher histologic grades, as was established in various previous works [27-29]. Just like MCM2, TNBCs were also seen to have a high Ki-67 expression, which is concordant with Keam B et al., [30]. High Ki-67 cases were seen to have a higher risk of advanced stage and node positive disease as compared to low Ki-67 proliferative index, which is concordant with the findings of de Azambuja E et al., [31]. However, statistical significance could not be established between grades of Ki-67 expression and their adverse prognostic factors, small study population being the main contributory factor in this case.

Limitation(s)

Thus, it can be seen that small study population is the major limiting factor in this case. In many cases, statistical association could not be established due to small sample size. Also, the study needs to be conducted on a larger population for generalising the outcome and implementing it in our routine practice.

Conclusion(s)

Thus, it can be concluded by saying that MCM2 is a predictor for adverse prognostic factors like high histologic grade, high Ki-67 index and triple negative phenotypes. Its high expression also comes with increased risk of advanced stage of disease. The Ki-67 is also a predictor of poor outcome with increased risk of LABCs and node positive cases and its association with higher histologic grade and triple negative cases. However, in few scenarios statistical significance could not be established, small sample size being the main limiting factor.

The MCM2 can serve as an attractive alternative as a predictor of adverse outcomes, guiding a tailored strategy among patients and also avoid harmful overtreatments. Just like MCM2 has proved its worth as a novel therapeutic target in colon and breast cancers, its conclusive role in breast carcinogenesis may pave the way for newer therapeutic agents in treating more aggressive cancers like the triple negative subtypes, which are resistant to the conventional therapies.

IBC NST: Invasive breast carcinoma no special type; LABC: Locally advanced breast carcinomas; TNBC: Triple negative breast cancers; BR Grade: Bloom richardson gradingStatistical tests used are- Chi-square test; Fisher’s-exact test; RR: Relative risk; r: Pearson’s correlation coefficient

References

[1]Yousef EM, Furrer D, Laperriere DL, Tahir MR, Mader S, Diorio C, MCM2: An alternative to Ki-67 for measuring breast cancer cell proliferation Mod Pathol 2017 30(5):682-97.10.1038/modpathol.2016.23128084344  [Google Scholar]  [CrossRef]  [PubMed]

[2]Ibrahim T, Mercatali L, Amadori D, A new emergency in oncology: Bone metastases in breast cancer patients (review) Oncol Lett 2013 6:306-10.10.3892/ol.2013.137224137321  [Google Scholar]  [CrossRef]  [PubMed]

[3]da Silva JL, Cardoso Nunes NC, Izetti P, de Mesquita GG, de Melo AC, Triple negative breast cancer: A thorough review of biomarkers Crit Rev Oncol Hematol 2020 145:10285510.1016/j.critrevonc.2019.10285531927455  [Google Scholar]  [CrossRef]  [PubMed]

[4]Sørlie T, Perou CM, Tibshirani R, Gene expression patterns of breast carcinomas distinguish tumour subclasses with clinical implications Proc Natl Acad Sci USA 2001 98(19):10869-74.10.1073/pnas.19136709811553815  [Google Scholar]  [CrossRef]  [PubMed]

[5]Lombardi A, Lazzeroni R, Bersigotti L, Vitale V, Amanti C, The proper Ki-67 cut-off in hormone responsive breast cancer: a monoinstitutional analysis with long-term follow-up Breast Cancer (Dove Med Press) 2021 13:213-17.10.2147/BCTT.S30544033854368  [Google Scholar]  [CrossRef]  [PubMed]

[6]Polley MY, Leung SC, McShane LM, Gao D, Hugh JC, Mastropasqua MG, International Ki67 in Breast Cancer Working Group of the Breast International Group and North American Breast Cancer GroupAn international Ki-67 reproducibility study J Natl Cancer Inst 2013 105(24):1897-906.10.1093/jnci/djt30624203987  [Google Scholar]  [CrossRef]  [PubMed]

[7]Kwok FH, Zhang SD, McCrudden CM, Yuen HF, Ting KP, Wen Q, Prognostic significance of minichromosome maintenance proteins in breast cancer Am J Cancer Res 2015 5(1):52-71.10.1073/pnas.040414310115263078  [Google Scholar]  [CrossRef]  [PubMed]

[8]Shechter D, Gautier J, MCM protein and checkpoint kinases get together at the fork PNAS 2004 101(30):10845-46.  [Google Scholar]

[9]Liu Y, He G, Wang Y, Guan X, Pang X, Zhang B, MCM-2 is a therapeutic target of Trichostatin A in colon cancer cells Toxicol Lett 2013 221(1):23-30.10.1016/j.toxlet.2013.05.64323770000  [Google Scholar]  [CrossRef]  [PubMed]

[10]Zhang X, Teng Y, Yang F, Wang M, Hong X, Ye L, MCM2 is a therapeutic target of lovastatin in human non small cell lung carcinomas Oncol Rep 2015 33:2599-605.10.3892/or.2015.382225738322  [Google Scholar]  [CrossRef]  [PubMed]

[11]Allison KH, Brogi F, Ellis IO, Fox SB, WHO classification of Breast tumours 2018 edition 5:01-147.  [Google Scholar]

[12]https://www.proteinatlas.org/accessed on 4/6/2021 at 11:45 am  [Google Scholar]

[13]Issac MSM, Yousef E, Tahir MR, MCM2, MCM4, and MCM6 in breast cancer: Clinical utility in diagnosis and prognosis Neoplasia 2019 21(10):1015-35.10.1016/j.neo.2019.07.01131476594  [Google Scholar]  [CrossRef]  [PubMed]

[14]Woo YS, Kim S, Determination of cut-off values for biomarkers in clinical studies Precision and future Medicine 2020 4(1):02-08.10.23838/pfm.2019.00135  [Google Scholar]  [CrossRef]

[15]Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries CA Cancer J Clin 2021 71:209-49.10.3322/caac.2166033538338  [Google Scholar]  [CrossRef]  [PubMed]

[16]Makki J, Diversity of breast carcinoma: Histological subtypes and clinical relevance. clinical medicine insights Pathology 2015 8:23-31.10.4137/CPath.S3156326740749  [Google Scholar]  [CrossRef]  [PubMed]

[17]Erbas B, Amos A, Fletcher A, Kavanagh AM, Gertig DM, Incidence of invasive breast cancer and ductal carcinoma in situ in a screening program by age: Should older women continue screening? Cancer Epidemiol Biomarkers Prev 2004 13(10):1569-73.  [Google Scholar]

[18]Kim EJ, Park HS, Kim JY, Kim SI, Cho YU, Park BW, Assessment of the prognostic staging system of American Joint Committee on Cancer 8th Edition for Breast Cancer: Comparison with the Conventional Anatomic Staging System J Breast Cancer 2020 23(1):59-68.10.4048/jbc.2020.23.e932140270  [Google Scholar]  [CrossRef]  [PubMed]

[19]Thakur KK, Bordoloi D, Kunnumakkara AB, Alarming burden of triple-negative breast cancer in India Clin Breast Cancer 2018 18(3):e393-99.10.1016/j.clbc.2017.07.01328801156  [Google Scholar]  [CrossRef]  [PubMed]

[20]Bukholm IR, Bukholm G, Holm R, Nesland JM, Association between histology grade, expression of HsMCM2, and cyclin A in human invasive breast carcinomas J Clin Pathol 2003 56(5):368-73.10.1136/jcp.56.5.36812719458  [Google Scholar]  [CrossRef]  [PubMed]

[21]Joshi S, Watkins J, Digital imaging in the immunohistochemical evaluation of the proliferative markers Ki-67, MCM2 and Geminin, in early breast cancer, and their putative prognostic value BMC Cancer 2015 15:54610.1186/s12885-015-1531-326205655  [Google Scholar]  [CrossRef]  [PubMed]

[22]Ali HR, Dawson SJ, Blows FM, Aurora kinase A outperforms Ki-67 as a prognostic marker in ER-positive breast cancer Br J Cancer 2012 106(11):1798-806.10.1038/bjc.2012.16722538974  [Google Scholar]  [CrossRef]  [PubMed]

[23]Wojnar A, Kobierzycki C, Correlation of Ki-67 and MCM-2 proliferative marker expression with grade of histological malignancy (G) in ductal breast cancers Folia Histochem Cytobiol 2010 48(3):442-46.10.2478/v10042-010-0069-021097442  [Google Scholar]  [CrossRef]  [PubMed]

[24]Wojnar A, Pula B, Piotrowska A, Jethon A, Kujawa K, Kobierzycki C, Correlation of intensity of MT-I/II expression with Ki-67 and MCM-2 proteins in invasive ductal breast carcinoma Anticancer Research 2011 31:3027-34.  [Google Scholar]

[25]Gonzalez MA, Pinder SE, Callagy G, Minichromosome maintenance protein 2 is a strong independent prognostic marker in breast cancer J Clin Oncol 2001 21:4306-13.10.1200/JCO.2003.04.12114645419  [Google Scholar]  [CrossRef]  [PubMed]

[26]Gou K, Liu J, Feng X, Li H, Yuan Y, Xing C, Expression of Minichromosome Maintenance Proteins (MCM) and cancer prognosis: A meta-analysis Journal of Cancer 2018 9(8):1518-26.10.7150/jca.2269129721062  [Google Scholar]  [CrossRef]  [PubMed]

[27]Soliman NA, Yussif SM, Ki-67 as a prognostic marker according to breast cancer molecular subtype Cancer Biol Med 2016 13(4):496-504.10.20892/j.issn.2095-3941.2016.006628154782  [Google Scholar]  [CrossRef]  [PubMed]

[28]Inwald EC, Klinkhammer-Schalke M, Ki-67 is a prognostic parameter in breast cancer patients: Results of a large population-based cohort of a cancer registry Breast Cancer Res Treat 2013 139:539-52.10.1007/s10549-013-2560-823674192  [Google Scholar]  [CrossRef]  [PubMed]

[29]Trihia H, Murray S, Price K, Ki-67 expression in breast carcinoma: Its association with grading systems, clinical parameters, and other prognostic factors- a surrogate marker? Cancer 2003 97:1321-31.10.1002/cncr.1118812599241  [Google Scholar]  [CrossRef]  [PubMed]

[30]Keam B, Im SA, Lee KH, Ki-67 can be used for further classification of triple negative breast cancer into two subtypes with different response and prognosis Breast Cancer Res 2011 13(2):R2210.1186/bcr283421366896  [Google Scholar]  [CrossRef]  [PubMed]

[31]de Azambuja E, Cardoso F, de Castro G, Ki-67 as prognostic marker in early breast cancer: A meta-analysis of published studies involving 12,155 patients Br J Cancer 2007 96:1504-13.10.1038/sj.bjc.660375617453008  [Google Scholar]  [CrossRef]  [PubMed]