Breast Cancer Res Treat (2014) 143:91–98 DOI 10.1007/s10549-013-2786-5
PRECLINICAL STUDY
The potential utility of geminin as a predictive biomarker in breast cancer Sreekumar Sundara Rajan • Andrew M. Hanby Kieran Horgan • Helene H. Thygesen • Valerie Speirs
•
Received: 27 September 2013 / Accepted: 22 November 2013 / Published online: 1 December 2013 Ó Springer Science+Business Media New York 2013
Abstract Compared with other markers of cell proliferation, geminin is unique being expressed selectively during the proliferative phase of the cell cycle, specifically S, G2 and early mitosis, disappearing completely at the metaphase–anaphase transition. We aimed to compare the prognostic significance of geminin to that of Ki67, a proliferation marker which has been investigated in many breast cancer studies. Breast cancer tissue microarrays containing 368 tumours were stained using anti-geminin and Ki67 antibodies. Labelling index (LI) was calculated for geminin, and the percentage of positive cancer nuclei was determined for Ki67. A receiver operation characteristics analysis was used to determine the optimum cut-off value for geminin (LI C 2), and for Ki67, a score of C14 % was considered as positive for survival analysis. Geminin expression correlated positively with Ki67 expression (r = 0.686, p = 0.001). Survival analysis showed only geminin, and not Ki67-positive patients to have poor (breast cancer-specific survival) BCSS [HR 2.85 (1.53–5.32)] and (disease-free survival) DFS [HR 2.63 (1.47–4.71)]. On univariate analysis, along with known clinicopathological variables, both Ki67 and geminin LI were found to be significant predictors of BCSS and DFS. On multivariate analysis, only tumour size, nodal status and adjuvant hormonal therapy were found to be independent predictors for both BCSS and DFS, while geminin positivity (LI C 2 %) was found to be an independent S. Sundara Rajan A. M. Hanby H. H. Thygesen V. Speirs (&) Leeds Institute of Cancer Studies and Pathology, University of Leeds, Leeds LS9 7TF, UK e-mail: [email protected] S. Sundara Rajan A. M. Hanby K. Horgan Leeds Teaching Hospitals NHS Trust, Leeds LS9 7TF, UK
predictor for BCSS [HR 2.27 (1.01–5.06); p = 0.04]. In comparison with Ki67, a more established proliferation marker, geminin expression was a better predictor of adverse outcome in this cohort of breast cancers. Selective expression of geminin during the proliferative phase of the cell cycle and its nuclear specificity increase its potential to be used as an alternative marker of proliferation in breast cancer patients. Keywords Breast cancer Geminin Ki67 Biomarker Prognosis Proliferation
Introduction One of the hallmarks of cancer is its ability to proliferate uncontrollably [1]. This can be evaluated immunohistochemically by measuring the expression of proliferative biomarkers. Of these, Ki67 is the best known, and its expression denotes entry into the cell cycle as it is present in all the phases except G0 [2, 3]. Ki67 is currently considered the assay of choice for measuring and monitoring tumour proliferation in clinical trials. Even in the neoadjuvant endocrine settings, Ki67 is considered as a pharmacodynamic marker of treatment response [3]. Along with ER, PR and HER2, Ki67 is a component part of the IHC-4 assay, a prognostic tool to evaluate clinical outcome in breast cancer patients treated with endocrine therapy [4, 5]. However, due to the well-recognised inability to standardise Ki67 measurement immunohistochemically, and its reported localisation in the membrane and cytoplasm as well as the nucleus [6], there currently exist some limitations to its clinical applicability [3, 7]. Hence, there is a need to critically evaluate other proliferative markers in
123
92
breast cancer to ascertain their prognostic role and potential clinical utility. Geminin is a nuclear protein which is unique to multicellular organisms. Compared with Ki67, geminin is expressed only in the proliferative phase of the cell cycle (S, G2 and M) [8, 9]. This is reflected by its expression only in actively dividing normal cells and over-expression in various malignancies [10–12]. This suggests that geminin may be valuable as a proliferative biomarker in cancer. Recent studies show that geminin is over-expressed in human mammary epithelial cells [13, 14]. When overexpressed, geminin prematurely releases topoisomerase IIa from the DNA replication forks leading to the formation of catenated DNA intertwines [14]. Geminin also competitively inhibits Aurora B, and leads to the inhibition of the cell division phase of mitosis [13]. Hence, it has been postulated that, when over-expressed, geminin leads to the formation of damaged DNA replication forks and prevents separation of daughter cells during mitosis. The end result is the formation of aneuploid and aggressive breast cancer cells which are resistant to chemotherapeutic agents [13]. This would probably explain the poor overall survival observed amongst breast cancer patients with high geminin expression [12]. In this study, we aimed to compare and evaluate the expressions and potential prognostic roles of geminin and Ki67 in a large cohort of breast cancer patients with longterm follow up to determine the potential clinical utility of geminin. REMARK criteria were followed [15].
Methodology
Breast Cancer Res Treat (2014) 143:91–98
fixed paraffin-embedded blocks and were then embedded into the recipient paraffin block in triplicate. The manual tissue microarray instrument (MTA1, Beecher Instrument, Westburg) was used for constructing the TMA. Immunohistochemistry was performed on 4-lm TMA sections. The mouse monoclonal antibody with Leica Envision detection method was used for ERa (clone 1D5, 1:100) as described previously [16]. Conversely, the streptavidin– biotin detection method was used for geminin (mouse monoclonal antibody; clone NCL-L, 1:12.5) and Ki67 (mouse monoclonal antibody; clone MIB1, 1:100). In brief, the sections were deparaffinised in xylene and rehydrated by passing through graded ethanol. The peroxidase block was performed using 10 % hydrogen peroxide, and then antigen retrieval was performed using the pressure cooker. After antigen retrieval, avidin–biotin block was performed before incubating the section with the primary antibody over night at 4 °C in a humid chamber. The next day, sections were counterstained with secondary antibody and avidin-peroxidase before the visualisation of the signal using 3,3-diaminobenzidine chromogen. The stained slides were scanned at 920 magnifications using Aperio ScanScopeTM and visualised for manual scoring using the ImageScopeTM, viewing software. Geminin was scored using labelling index (LI) which accounts for the number of cancer nuclei positively stained with geminin compared to the total number of nuclei in each TMA core [12]. The percentage of Ki67-positive nuclei was determined amongst 500 cancer nuclei in each TMA core as per the recommendation of the International Ki67 in Breast Cancer-Working Group [3]. The Allred method was used for manually scoring ERa, and a score of C2 was considered as ERa positive.
Patient cohort Statistical analysis Following ethical approval (06/Q1206/180), consecutive patients diagnosed with breast cancer from 1992 to 1997 at Leeds Teaching Hospitals NHS Trust, Leeds, UK were included in this study. Electronic pathological database was searched to identify various pathological parameters including, histological type, invasive tumour size, grade, mitotic index, lympho-vascular invasion (LVI) and nodal status. Patient demographics, type of surgery, adjuvant treatment and follow-up information were gathered from electronic case notes. Individual patient case notes were reviewed when there were missing data from the electronic database. Immunohistochemistry Immunohistochemical staining was performed using tissue microarray (TMA). TMAs were constructed from the above cases using 0.6-mm punch biopsies from formalin-
123
The end point for duration of follow-up was pre-defined as death from breast cancer or other causes, final contact of the patient with breast surgeons or oncologists or date of data collection, whichever had happened first. Our primary outcome variables were disease-free survival (DFS) and breast cancer specific survival (BCSS). DFS was calculated from the date of primary surgery to the date of recurrence or metastasis. Similarly, BCSS was defined as the interval from the date of diagnosis to the date of death from breast cancer. The association between continuous variables was examined using Spearman rank correlation (q), while the association between categorical variables and biomarkers was examined using Mann–Whitney U test. We performed cox regression analysis to determine the causative association between various clinico-pathological variables and our primary outcome. The prognostic variables that were significant on univariate analysis were entered into a cox
Breast Cancer Res Treat (2014) 143:91–98
proportional hazard model to identify independent predictors. The effects of various adjuvant treatments on survival of breast cancer patients were accounted for in the multivariate analysis by including them in the cox proportional hazard model. Survival analysis was performed using the Kaplan–Meier method, and survival curves were compared using the Log rank test. A receiver operating characteristic (ROC) curve was plotted, and the area under the curve was considered while determining the optimum cut-off for geminin. The data were analysed using Statistical Package for Social Sciences (SPSS) version 19 software. All statistical tests were two sided with a p value of B0.05 considered statistically significant.
Results
93 Table 1 Clinical and pathological characteristics of breast cancer patients Clinical and pathological features Median age (IQR)
58 years (22 years)
Median Invasive cancer size (IQR)
20 mm (11 mm)
Median Geminin LI (IQR)
4 % (8.5 %)
Median Ki67 score (IQR)
2.4 % (6.8 %)
Grade G1
64 (22 %)
G2 G3
124 (42.6 %) 103 (35.4 %)
Oestrogen receptor Positive
230 (80.1 %)
Negative
57 (19.9 %)
Not known
4
Nodal status
Our cohort consisted of 368 breast cancer patients. Of these, 77 patients were excluded from the final analysis due to core exhaustion/missing TMA core (n = 57) or missing clinical data (n = 20). The clinical and pathological characteristics in the remaining 291 patients are represented in Table 1. Currently, there are no established criteria for determining geminin positivity within the literature. Hence, a ROC curve was plotted using primary outcome variables (BCSS and DFS) to determine the optimum cut-off for geminin by visual inspection of the sensitivity and specificity values (Fig. 1). A labelling index of C2 % gave a sensitivity of 82 % and a specificity of 61 % using BCSS as the dependent variable, and hence, was taken as the ideal cut-off for determining geminin positivity in our cohort. Based on this, we identified 64 % of our patients to be positive for geminin expression, with the rest being negative. We considered a score of C14 % as Ki67 positive for survival analysis [17], as it was determined against an important distinction in the underlying biology of the breast cancer (Luminal A vs. Luminal B) rather than based on clinical outcomes or median Ki67 values as were used in the past. The expression of geminin was confined to the nuclei of the epithelial cells in invasive breast carcinoma and was uniformly negative in the surrounding stromal elements (Fig. 2a). Conversely, even though Ki67 was expressed mostly in the cancer epithelial cell nuclei, often weak, surrounding cytoplasmic staining was observed (Fig. 2b). However, we did not observe any staining in the stromal elements or inflammatory cells by the Ki67 antibody. Only nuclear staining was considered for Ki67 scoring, and cytoplasmic staining was disregarded as per published recommendations [3]. Geminin expression and its association with various clinicopathological variables were evaluated in our cohort.
Negative
125 (46.8 %)
N1
90 (33.7 %)
N2
24 (9 %)
N3
28 (10.5 %)
Not known
24
Lympho-vascular invasion Yes
130 (47.3 %)
No
145 (52.7 %)
Not known Surgical treatment Wide local excision
16 97 (33.4 %)
Mastectomy
193 (66.6 %)
No surgery
1
Adjuvant radiotherapy Yes
162 (58.7 %)
No
114 (41.3 %)
Not known
15
Adjuvant chemotherapy Yes
91 (33.5 %)
No
181 (66.5 %)
Not known
19
Geminin was positively correlated with tumour size (q = 0.181, p = 0.002) and Ki67 (q = 0.59, p \ 0.01), but negatively associated with age (q = -0.205, p = 0.0004). Both geminin (p \ 0.01) and Ki67 (p \ 0.01) were positively correlated with grade and negatively correlated with ERa status (p \ 0.001). However, there was no difference in the expression of geminin or Ki67 with the nodal status (p = 0.86 and p = 0.51, respectively) or LVI (p = 0.18 and p = 0.91, respectively). One hundred and eighty six patients (63.9 %) were geminin positive, and 105 (36.1 %) were negative (geminin LI C 2 % as cut-off). There were 12 breast cancer-
123
94
Breast Cancer Res Treat (2014) 143:91–98
Co-ordinates of the curve for BCSS y Sensitivity 1Geminin cut-off -1.0 1.000 1.000 2.1 0.824 0.610 63.7 0.000 0.000
Co-ordinates of the curve for DFS Geminin Sensitivity cut-off -1.0 1.000 1.000 2.1 0.808 0.583 63.7 0.000 0.000
Fig. 1 ROC curve to determine the optimum cut-off for geminin. The reference line is depicted in green and geminin curve in blue. Geminin LI of C2 % was taken as the optimum cut-off by visually inspecting the sensitivity and specificity values of the ROC curve coordinates represented in the Table. (Color figure online)
specific deaths amongst geminin negative patients compared to 56 deaths in geminin-positive cases [HR 2.85 (1.53–5.32); mean survival = 188 months]. Similarly, geminin-positive patients (geminin LI C 2 %) were also found to have worse DFS [HR 2.63 (1.47–4.71); mean time to recurrence or metastasis = 183 months]. The BCSS and DFS curves for geminin are represented in Fig. 3a, b. The Ki67 score was unable to be determined in 24 patients due to missing cores in the TMA, a well-recognised issue with TMAs. Ki67 was found to be positive only in 40 patients (15 %) and negative in 227 (85 %) with a cut-off of C14 %. Ki67 positivity did not influence the BCSS (p = 0.088) or DFS (p = 0.149) in our cohort. Cox regression univariate analysis was performed to establish the prognostic role of geminin LI and Ki67 score as percentages (continuous variables) along with other known clinical and pathological prognosticators. Tumour size, lymph nodal status (positive or negative), tumour grade, ER status, Ki67 and geminin LI were identified as statistically significant predictors for BCSS and DFS (Table 2). However, on multivariate analysis after adjusting for adjuvant treatments received, only tumour size, nodal status and adjuvant hormonal treatment were found to be independent predictors for both BCSS and DFS (Table 2). We did not stratify according to the treatment while performing multivariate analysis, as this cohort was
123
Fig. 2 Depiction of nuclear specificity of geminin immunohistochemistry (arrows; a). In contrast, Ki67 immunohistochemistry (b) is often non-specific with granular cytoplasmic staining (arrows) predominating over true nuclear staining (asterisk). Original magnification = 920
predominantly ER-positive (80 %) with patients receiving adjuvant endocrine therapy (mostly tamoxifen). Cox multivariate analysis was repeated using geminin (LI C 2 %) and Ki67 (C14 %) positivity as independent variables (categorical). This confirmed the previously established independent predictive roles of tumour size, nodal status and adjuvant hormonal treatment for BCSS
Breast Cancer Res Treat (2014) 143:91–98
95
Fig. 3 Kaplan–Meier survival curve showing poor BCSS (a) and DFS (b) in gemininpositive breast cancers. Patients were censored, when they ceased to be followed up for any reason but had not died due to breast cancer
and DFS. However, the results also showed geminin positivity [C2 % cut-off; HR 2.27 (1.01, 5.06); p = 0.04) to be a predictor for BCSS. Although there was a trend towards poor DFS with geminin positivity [HR 1.2 (0.95–4.2); p = 0.07], it did not reach statistical significance. Conversely, Ki67 positivity was not found to have any independent predictive role in our cohort.
Discussion The uncontrolled proliferation is a hallmark of malignancy, and the role of established proliferative markers like Ki67 in breast cancer management is limited. This is due to the variation in the analytical methods used for Ki67 immunohistochemistry and the lack of consensus on the ideal cut-off for Ki67 in clinical practice [3]. Compared to Ki67, geminin is expressed only by proliferative cells, either normal (male germ cells, lymphocytes and epithelial cells) or tumour cells (colon, kidney and breast) [10–12, 18]. Moreover, there have been no issues with the analytical methods used for immunohistochemical evaluation of
geminin in various cancer tissues [11, 12, 19], and the staining pattern was shown to be nuclear specific in our cohort (Fig. 2). In this study, the prognostic role of geminin in a large cohort of breast cancer patients with a median follow up of 188 months (range 5–229 months) was determined and compared to that of Ki67 and other clinically relevant prognostic variables. The data differ from other published studies due to the availability of almost complete follow-up data (missing clinical data = 20 patients (6.4 %)) and survival data pertaining to breast cancer-specific events compared to the overall survival reported in the literature so far. Geminin expression was positively correlated with Ki67, tumour size, grade and mitotic index. Conversely, ERa was negatively associated with Ki67 expression and geminin LI. It was also observed that, younger patients have a higher geminin LI compared to that of their older counterparts, and there was a linear association between age and geminin LI. The above associations of geminin with tumour grade, size and the strong positive correlation with K67 strengthen the role of geminin as a potential
123
96
Breast Cancer Res Treat (2014) 143:91–98
Table 2 Cox uni- and multivariate analysis of geminin in disease-free and breast cancer-specific survival Univariate
Multivariate
DFS
BCSS
HR
p
HR
DFS p
HR
BCSS p
HR
p
Grade
5.92 (2.5–13.9)
\0.001
5.4 (2.3–12.8)
\0.0001
2.7 (0.92–7.7)
0.07
2.4 (0.8 –6.9)
0.12
LN
2.48 (1.45–4.22)
0.001
2.8 (1.6–4.9)
\0.001
1.9 (1.06–3.3)
0.03
2.1 (1.2–3.9)
0.016
Size ER
1.02 (1.01–1.02) \0.001 0.44 (0.26–0.72) 0.001
1.02 (1.01–1.02) \0.001 0.4 (0.2–0.8) 0.003
1.02 (1.0–1.03) 1.5 (0.7 –3.2)
0.001 0.28
1.02 (1.0–1.03) 1.7 (0.75–3.7)
0.003 0.21 \0.001
Adjuvant hormone therapy
nd
nd
nd
nd
0.19 (0.08–0.4)
\0.001
0.17 (0.07–0.4)
Adjuvant radiotherapy
nd
nd
nd
nd
0.63 (0.4–1.09)
0.09
1.02 (0.9–1.06)
0.51
Ki67 score (continuous variable)
1.02 (1.01–1.04)
0.001
1.03 (1.01–1.04) \0.001
1.0 (0.98–1.03)
0.7
1.01 (0.9–1.04)
0.45
Geminin LI (continuous variable)
1.04 (1.02–1.06) \0.001
\0.001
1.02 (0.98–1.06)
0.39
1.02 (0.97–1.06)
0.51
1.05 (1.02–1.0)
nd Not done
proliferative marker in breast cancer, and it may also correlate with aneuploidy, given its involvement with defective replication forks [14]. This was reflected even in the survival analysis, with geminin-positive patients (LI C 2 %) found to be three times more likely to succumb to breast cancer-related death during follow-up (HR 2.85; CI 1.53–5.32; p = 0.001). Similarly, geminin-positivity (LI C 2 %) was associated with twice the risk of developing recurrence or metastasis in our cohort (HR 2.63; CI 1.47–4.71; p = 0.001). Furthermore, geminin positivity (LI C 2 %) was shown to be an independent predictor for determining BCSS in our cohort. We found Ki67 was neither an independent predictor nor was related to breast cancer-specific events (BCSS or DFS). The inability to distinguish an optimal cut-off to determine Ki67 positivity is well recognised, with values ranging from 1 to 28.6 % having been used in the literature [20]. Even though various cut-offs have been used, a score in the range of 10–20 % has been used most commonly to dichotomize population into a ‘high’ and ‘low’ Ki67 expression [21]. Hence, we evaluated, whether using the median value of Ki67 and geminin as a cut-off [12, 22] would change the observed results (geminin LI cut-off 4 %; Ki67 cut-off 2.4 %). This analysis showed that both Ki67 and geminin positivity were associated with statistically significant poor BCSS and DFS (results not shown). High Ki67 LI has been reported to be predictive of responsiveness to neo-adjuvant chemotherapy [23, 24] and neo-adjuvant endocrine therapy [25]. Moreover, two recent meta-analyses have also shown that high Ki67 expression to be associated with increased risk of breast cancer relapse and death [21, 26]. However, we did not find Ki67 to be an independent predictor of BCSS or DFS. This may be
123
probably due to the strong positive correlation of geminin with Ki67 observed in our cohort as well as due to the strong influence of geminin on survival. This was confirmed by entering only geminin and Ki67 in the Cox proportion hazard model, which showed geminin to be a stronger predictor compared with Ki67 for both BCSS (Ki67 - B = 0.013, SE = 0.012, p = 0.28; geminin - B = 0.032, SE = 0.016, p = 0.04) and DFS (Ki67 - B = 0.007, SE = 0.012, p = 0.54; geminin - B = 0.036, SE = 0.016, p = 0.02). Identification of biomarkers that can monitor breast cancer progression or serve as a surrogate marker for prognosis will enable us to personalise medicine and to improve care of patients. Individualised recurrence scores like Oncotype DX are heavily weighed for the expression of genes, like that of Ki67, which account for proliferation [27]. However, issues with reproducibility of Ki67 immunohistochemistry limit its potential for efficient translation to the clinical setting for the management of breast cancer patients. Geminin, on the other hand, has been shown to be consistently associated with poor prognosis in various cancers (lung, colon, renal and brain tumours) [11, 19, 28, 29]. Similarly, geminin expression was also shown to be positively correlated with tumour grade in a variety of tumours including breast, kidney, prostate and salivary glands [11, 12, 29–31]. However, there are only two studies which evaluated the prognostic role geminin in breast cancer [12, 32]. Gonzalez et al. [12] showed poor overall survival with high geminin LI, whereas the reliability of the observed results were significantly compromised in the other study due to accounting for cytoplasmic expression of geminin in the immunohistochemical evaluation in spite of geminin being a nuclear antigen [32]. The results
Breast Cancer Res Treat (2014) 143:91–98
presented here support the poor prognosis and adverse breast cancer survival amongst geminin-positive patients (LI C 2 %) agreeing with the findings from other cancers. Moreover, we also showed that geminin was statistically superior to Ki67 in differentiating those patients with poor breast cancer prognosis. Certain limitations of our study should be discussed before summarising the findings. The retrospective nature of the study may have introduced selection bias; however, it provided a large sample of breast cancer patients with longterm follow-up for studying the prognostic significance. The adjuvant treatments received by the patients, especially the adjuvant chemotherapy in the 1990s were indeed different compared to that being currently offered. The effects of various adjuvant treatments on breast cancer-specific survival and prognosis were hence accounted for by adjusting them in the multivariate model during statistical analysis. There was also an element of attrition bias due to the loss of almost 21 % of our initial cohort (77 out of 368). However, only 7 % was lost to follow-up, and the rest due to missing TMA cores (14 %), which is a recognised limitation of studies’ conducted using TMAs. The strong correlation of geminin with Ki67 and other clinicopathological variables (grade, tumour size and ERa) may also have influenced the results of the multivariate analysis. Nevertheless, the results of this study support the prognostic role of geminin in breast cancer. However, in order to improve the validity, independent verification of our results in an external cohort is needed. There is also a need for further research to investigate the correlation between geminin and Ki67 to evaluate whether geminin can reliably substitute Ki67 as a proliferative marker in breast cancer patients and potentially take its place as a predictive as well as prognostic agent or if it is simply a measure of aneuploidy. In conclusion, geminin positivity (LI C 2 %) was shown to be a predictor of poor survival in our breast cancer cohort. The selective expression of geminin during the proliferative phase of the cell cycle and its nuclear specificity increase its potential to be considered as an alternative marker of proliferation in breast cancer patients. Acknowledgments The authors are grateful to the Leeds Teaching Hospitals NHS Trust Special Trustees for funding this work. HHT is funded by the Cancer Research UK. Conflict of interest
97
3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
The authors have no conflicts of interest. 17.
References 18. 1. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144(5):646–674 2. Isola JJ, Helin HJ, Helle MJ, Kallioniemi OP (1990) Evaluation of cell proliferation in breast carcinoma. Comparison of Ki-67
immunohistochemical study, DNA flow cytometric analysis, and mitotic count. Cancer 65(5):1180–1184 Dowsett M, Nielsen TO, A’Hern R, Bartlett J, Coombes RC, Cuzick J, Ellis M, Henry NL, Hugh JC, Lively T et al (2011) Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer Working Group. J Natl Cancer Inst 103(22):1656–1664 Dowsett M, Allred C, Knox J, Quinn E, Salter J, Wale C, Cuzick J, Houghton J, Williams N, Mallon E et al (2008) Relationship between quantitative estrogen and progesterone receptor expression and human epidermal growth factor receptor 2 (HER-2) status with recurrence in the arimidex, tamoxifen, alone or in combination trial. J Clin Oncol 26(7):1059–1065 Dowsett M, Salter J, Zabaglo L, Mallon E, Howell A, Buzdar AU, Forbes J, Pineda S, Cuzick J (2011) Predictive algorithms for adjuvant therapy: TransATAC. Steroids 76(8):777–780 Faratian D, Munro A, Twelves C, Bartlett JMS (2009) Membranous and cytoplasmic staining of Ki67 is associated with HER2 and ER status in invasive breast carcinoma. Histopathology 54(2):254–257 Yerushalmi R, Woods R, Ravdin PM, Hayes MM, Gelmon KA (2010) Ki67 in breast cancer: prognostic and predictive potential. Lancet Oncol 11(2):174–183 McGarry TJ, Kirschner MW (1998) Geminin, an inhibitor of DNA replication, is degraded during mitosis. Cell 93(6):1043–1053 Wohlschlegel JA, Kutok JL, Weng AP, Dutta A (2002) Expression of geminin as a marker of cell proliferation in normal tissues and malignancies. Am J Pathol 161(1):267–273 Eward KL, Obermann EC, Shreeram S, Loddo M, Fanshawe T, Williams C, Jung HI, Prevost AT, Blow JJ, Stoeber K et al (2004) DNA replication licensing in somatic and germ cells. J Cell Sci 117(Pt 24):5875–5886 Dudderidge TJ, Stoeber K, Loddo M, Atkinson G, Fanshawe T, Griffiths DF, Williams GH (2005) Mcm2, Geminin, and KI67 define proliferative state and are prognostic markers in renal cell carcinoma. Clin Cancer Res 11(7):2510–2517 Gonzalez MA, Tachibana KE, Chin SF, Callagy G, Madine MA, Vowler SL, Pinder SE, Laskey RA, Coleman N (2004) Geminin predicts adverse clinical outcome in breast cancer by reflecting cell-cycle progression. J Pathol 204(2):121–130 Blanchard Z, Malik R, Mullins N, Maric C, Luk H, Horio D, Hernandez B, Killeen J, Elshamy WM (2011) Geminin overexpression induces mammary tumors via suppressing cytokinesis. Oncotarget 2(12):1011–1027 Gardner L, Malik R, Shimizu Y, Mullins N, ElShamy WM (2011) Geminin overexpression prevents the completion of topoisomerase IIalpha chromosome decatenation, leading to aneuploidy in human mammary epithelial cells. Breast Cancer Res 13(3):R53 McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM (2005) Diagnostics ftSSotN-EWGoC: reporting recommendations for tumor marker prognostic studies (REMARK). J Natl Cancer Inst 97(16):1180–1184 Shaaban AM, Ball GR, Brannan RA, Cserni G, Di Benedetto A, Dent J, Fulford L, Honarpisheh H, Jordan L, Jones JL et al (2012) A comparative biomarker study of 514 matched cases of male and female breast cancer reveals gender-specific biological differences. Breast Cancer Res Treat 133(3):949–958 Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, Watson M, Davies S, Bernard PS, Parker JS et al (2009) Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer. J Natl Cancer Inst 101(10):736–750 Hansel DE, Rahman A, Hidalgo M, Thuluvath PJ, Lillemoe KD, Shulick R, Ku JL, Park JG, Miyazaki K, Ashfaq R et al (2003) Identification of novel cellular targets in biliary tract cancers using global gene expression technology. Am J Pathol 163(1):217–229
123
98 19. Nishihara K, Shomori K, Tamura T, Fujioka S, Ogawa T, Ito H (2009) Immunohistochemical expression of geminin in colorectal cancer: implication of prognostic significance. Oncol Rep 21(5):1189–1195 20. Urruticoechea A, Smith IE, Dowsett M (2005) Proliferation marker Ki-67 in early breast cancer. J Clin Oncol 23(28):7212–7220 21. Stuart-Harris R, Caldas C, Pinder SE, Pharoah P (2008) Proliferation markers and survival in early breast cancer: a systematic review and meta-analysis of 85 studies in 32,825 patients. Breast 17(4):323–334 22. Viale G, Giobbie-Hurder A, Regan MM, Coates AS, Mastropasqua MG, Dell’Orto P, Maiorano E, MacGrogan G, Braye SG, Ohlschlegel C et al (2008) Prognostic and predictive value of centrally reviewed Ki-67 labeling index in postmenopausal women with endocrine-responsive breast cancer: results from Breast International Group Trial 1–98 comparing adjuvant tamoxifen with letrozole. J Clin Oncol 26(34):5569–5575 23. Archer CD, Parton M, Smith IE, Ellis PA, Salter J, Ashley S, Gui G, Sacks N, Ebbs SR, Allum W et al (2003) Early changes in apoptosis and proliferation following primary chemotherapy for breast cancer. Br J Cancer 89(6):1035–1041 24. Chang J, Ormerod M, Powles TJ, Allred DC, Ashley SE, Dowsett M (2000) Apoptosis and proliferation as predictors of chemotherapy response in patients with breast carcinoma. Cancer 89(11):2145–2152 25. Dowsett M, Smith IE, Ebbs SR, Dixon JM, Skene A, A’Hern R, Salter J, Detre S, Hills M, Walsh G (2007) Prognostic value of Ki67 expression after short-term presurgical endocrine therapy for primary breast cancer. J Natl Cancer Inst 99(2):167–170
123
Breast Cancer Res Treat (2014) 143:91–98 26. de Azambuja E, Cardoso F, de Castro G Jr, Colozza M, Mano MS, Durbecq V, Sotiriou C, Larsimont D, Piccart-Gebhart MJ, Paesmans M (2007) Ki-67 as prognostic marker in early breast cancer: a meta-analysis of published studies involving 12,155 patients. Br J Cancer 96(10):1504–1513 27. Jonat W, Arnold N (2011) Is the Ki-67 labelling index ready for clinical use? Ann Oncol 22(3):500–502 28. Haruki T, Shomori K, Hamamoto Y, Taniguchi Y, Nakamura H, Ito H (2011) Geminin expression in small lung adenocarcinomas: implication of prognostic significance. Lung Cancer 71(3):356–362 29. Shrestha P, Saito T, Hama S, Arifin MT, Kajiwara Y, Yamasaki F, Hidaka T, Sugiyama K, Kurisu K (2007) Geminin: a good prognostic factor in high-grade astrocytic brain tumors. Cancer 109(5):949–956 30. Dudderidge TJ, McCracken SR, Loddo M, Fanshawe TR, Kelly JD, Neal DE, Leung HY, Williams GH, Stoeber K (2007) Mitogenic growth signalling, DNA replication licensing, and survival are linked in prostate cancer. Br J Cancer 96(9):1384–1393 31. Vargas PA, Cheng Y, Barrett AW, Craig GT, Speight PM (2008) Expression of Mcm-2, Ki-67 and geminin in benign and malignant salivary gland tumours. J Oral Pathol Med 37(5):309–318 32. Di Bonito M, Cantile M, Collina F, Scognamiglio G, Cerrone M, La Mantia E, Barbato A, Liguori G, Botti G (2012) Overexpression of cell cycle progression inhibitor geminin is associated with tumor stem-like phenotype of triple-negative breast cancer. J Breast Cancer 15(2):162–171
PRECLINICAL STUDY
The potential utility of geminin as a predictive biomarker in breast cancer Sreekumar Sundara Rajan • Andrew M. Hanby Kieran Horgan • Helene H. Thygesen • Valerie Speirs
•
Received: 27 September 2013 / Accepted: 22 November 2013 / Published online: 1 December 2013 Ó Springer Science+Business Media New York 2013
Abstract Compared with other markers of cell proliferation, geminin is unique being expressed selectively during the proliferative phase of the cell cycle, specifically S, G2 and early mitosis, disappearing completely at the metaphase–anaphase transition. We aimed to compare the prognostic significance of geminin to that of Ki67, a proliferation marker which has been investigated in many breast cancer studies. Breast cancer tissue microarrays containing 368 tumours were stained using anti-geminin and Ki67 antibodies. Labelling index (LI) was calculated for geminin, and the percentage of positive cancer nuclei was determined for Ki67. A receiver operation characteristics analysis was used to determine the optimum cut-off value for geminin (LI C 2), and for Ki67, a score of C14 % was considered as positive for survival analysis. Geminin expression correlated positively with Ki67 expression (r = 0.686, p = 0.001). Survival analysis showed only geminin, and not Ki67-positive patients to have poor (breast cancer-specific survival) BCSS [HR 2.85 (1.53–5.32)] and (disease-free survival) DFS [HR 2.63 (1.47–4.71)]. On univariate analysis, along with known clinicopathological variables, both Ki67 and geminin LI were found to be significant predictors of BCSS and DFS. On multivariate analysis, only tumour size, nodal status and adjuvant hormonal therapy were found to be independent predictors for both BCSS and DFS, while geminin positivity (LI C 2 %) was found to be an independent S. Sundara Rajan A. M. Hanby H. H. Thygesen V. Speirs (&) Leeds Institute of Cancer Studies and Pathology, University of Leeds, Leeds LS9 7TF, UK e-mail: [email protected] S. Sundara Rajan A. M. Hanby K. Horgan Leeds Teaching Hospitals NHS Trust, Leeds LS9 7TF, UK
predictor for BCSS [HR 2.27 (1.01–5.06); p = 0.04]. In comparison with Ki67, a more established proliferation marker, geminin expression was a better predictor of adverse outcome in this cohort of breast cancers. Selective expression of geminin during the proliferative phase of the cell cycle and its nuclear specificity increase its potential to be used as an alternative marker of proliferation in breast cancer patients. Keywords Breast cancer Geminin Ki67 Biomarker Prognosis Proliferation
Introduction One of the hallmarks of cancer is its ability to proliferate uncontrollably [1]. This can be evaluated immunohistochemically by measuring the expression of proliferative biomarkers. Of these, Ki67 is the best known, and its expression denotes entry into the cell cycle as it is present in all the phases except G0 [2, 3]. Ki67 is currently considered the assay of choice for measuring and monitoring tumour proliferation in clinical trials. Even in the neoadjuvant endocrine settings, Ki67 is considered as a pharmacodynamic marker of treatment response [3]. Along with ER, PR and HER2, Ki67 is a component part of the IHC-4 assay, a prognostic tool to evaluate clinical outcome in breast cancer patients treated with endocrine therapy [4, 5]. However, due to the well-recognised inability to standardise Ki67 measurement immunohistochemically, and its reported localisation in the membrane and cytoplasm as well as the nucleus [6], there currently exist some limitations to its clinical applicability [3, 7]. Hence, there is a need to critically evaluate other proliferative markers in
123
92
breast cancer to ascertain their prognostic role and potential clinical utility. Geminin is a nuclear protein which is unique to multicellular organisms. Compared with Ki67, geminin is expressed only in the proliferative phase of the cell cycle (S, G2 and M) [8, 9]. This is reflected by its expression only in actively dividing normal cells and over-expression in various malignancies [10–12]. This suggests that geminin may be valuable as a proliferative biomarker in cancer. Recent studies show that geminin is over-expressed in human mammary epithelial cells [13, 14]. When overexpressed, geminin prematurely releases topoisomerase IIa from the DNA replication forks leading to the formation of catenated DNA intertwines [14]. Geminin also competitively inhibits Aurora B, and leads to the inhibition of the cell division phase of mitosis [13]. Hence, it has been postulated that, when over-expressed, geminin leads to the formation of damaged DNA replication forks and prevents separation of daughter cells during mitosis. The end result is the formation of aneuploid and aggressive breast cancer cells which are resistant to chemotherapeutic agents [13]. This would probably explain the poor overall survival observed amongst breast cancer patients with high geminin expression [12]. In this study, we aimed to compare and evaluate the expressions and potential prognostic roles of geminin and Ki67 in a large cohort of breast cancer patients with longterm follow up to determine the potential clinical utility of geminin. REMARK criteria were followed [15].
Methodology
Breast Cancer Res Treat (2014) 143:91–98
fixed paraffin-embedded blocks and were then embedded into the recipient paraffin block in triplicate. The manual tissue microarray instrument (MTA1, Beecher Instrument, Westburg) was used for constructing the TMA. Immunohistochemistry was performed on 4-lm TMA sections. The mouse monoclonal antibody with Leica Envision detection method was used for ERa (clone 1D5, 1:100) as described previously [16]. Conversely, the streptavidin– biotin detection method was used for geminin (mouse monoclonal antibody; clone NCL-L, 1:12.5) and Ki67 (mouse monoclonal antibody; clone MIB1, 1:100). In brief, the sections were deparaffinised in xylene and rehydrated by passing through graded ethanol. The peroxidase block was performed using 10 % hydrogen peroxide, and then antigen retrieval was performed using the pressure cooker. After antigen retrieval, avidin–biotin block was performed before incubating the section with the primary antibody over night at 4 °C in a humid chamber. The next day, sections were counterstained with secondary antibody and avidin-peroxidase before the visualisation of the signal using 3,3-diaminobenzidine chromogen. The stained slides were scanned at 920 magnifications using Aperio ScanScopeTM and visualised for manual scoring using the ImageScopeTM, viewing software. Geminin was scored using labelling index (LI) which accounts for the number of cancer nuclei positively stained with geminin compared to the total number of nuclei in each TMA core [12]. The percentage of Ki67-positive nuclei was determined amongst 500 cancer nuclei in each TMA core as per the recommendation of the International Ki67 in Breast Cancer-Working Group [3]. The Allred method was used for manually scoring ERa, and a score of C2 was considered as ERa positive.
Patient cohort Statistical analysis Following ethical approval (06/Q1206/180), consecutive patients diagnosed with breast cancer from 1992 to 1997 at Leeds Teaching Hospitals NHS Trust, Leeds, UK were included in this study. Electronic pathological database was searched to identify various pathological parameters including, histological type, invasive tumour size, grade, mitotic index, lympho-vascular invasion (LVI) and nodal status. Patient demographics, type of surgery, adjuvant treatment and follow-up information were gathered from electronic case notes. Individual patient case notes were reviewed when there were missing data from the electronic database. Immunohistochemistry Immunohistochemical staining was performed using tissue microarray (TMA). TMAs were constructed from the above cases using 0.6-mm punch biopsies from formalin-
123
The end point for duration of follow-up was pre-defined as death from breast cancer or other causes, final contact of the patient with breast surgeons or oncologists or date of data collection, whichever had happened first. Our primary outcome variables were disease-free survival (DFS) and breast cancer specific survival (BCSS). DFS was calculated from the date of primary surgery to the date of recurrence or metastasis. Similarly, BCSS was defined as the interval from the date of diagnosis to the date of death from breast cancer. The association between continuous variables was examined using Spearman rank correlation (q), while the association between categorical variables and biomarkers was examined using Mann–Whitney U test. We performed cox regression analysis to determine the causative association between various clinico-pathological variables and our primary outcome. The prognostic variables that were significant on univariate analysis were entered into a cox
Breast Cancer Res Treat (2014) 143:91–98
proportional hazard model to identify independent predictors. The effects of various adjuvant treatments on survival of breast cancer patients were accounted for in the multivariate analysis by including them in the cox proportional hazard model. Survival analysis was performed using the Kaplan–Meier method, and survival curves were compared using the Log rank test. A receiver operating characteristic (ROC) curve was plotted, and the area under the curve was considered while determining the optimum cut-off for geminin. The data were analysed using Statistical Package for Social Sciences (SPSS) version 19 software. All statistical tests were two sided with a p value of B0.05 considered statistically significant.
Results
93 Table 1 Clinical and pathological characteristics of breast cancer patients Clinical and pathological features Median age (IQR)
58 years (22 years)
Median Invasive cancer size (IQR)
20 mm (11 mm)
Median Geminin LI (IQR)
4 % (8.5 %)
Median Ki67 score (IQR)
2.4 % (6.8 %)
Grade G1
64 (22 %)
G2 G3
124 (42.6 %) 103 (35.4 %)
Oestrogen receptor Positive
230 (80.1 %)
Negative
57 (19.9 %)
Not known
4
Nodal status
Our cohort consisted of 368 breast cancer patients. Of these, 77 patients were excluded from the final analysis due to core exhaustion/missing TMA core (n = 57) or missing clinical data (n = 20). The clinical and pathological characteristics in the remaining 291 patients are represented in Table 1. Currently, there are no established criteria for determining geminin positivity within the literature. Hence, a ROC curve was plotted using primary outcome variables (BCSS and DFS) to determine the optimum cut-off for geminin by visual inspection of the sensitivity and specificity values (Fig. 1). A labelling index of C2 % gave a sensitivity of 82 % and a specificity of 61 % using BCSS as the dependent variable, and hence, was taken as the ideal cut-off for determining geminin positivity in our cohort. Based on this, we identified 64 % of our patients to be positive for geminin expression, with the rest being negative. We considered a score of C14 % as Ki67 positive for survival analysis [17], as it was determined against an important distinction in the underlying biology of the breast cancer (Luminal A vs. Luminal B) rather than based on clinical outcomes or median Ki67 values as were used in the past. The expression of geminin was confined to the nuclei of the epithelial cells in invasive breast carcinoma and was uniformly negative in the surrounding stromal elements (Fig. 2a). Conversely, even though Ki67 was expressed mostly in the cancer epithelial cell nuclei, often weak, surrounding cytoplasmic staining was observed (Fig. 2b). However, we did not observe any staining in the stromal elements or inflammatory cells by the Ki67 antibody. Only nuclear staining was considered for Ki67 scoring, and cytoplasmic staining was disregarded as per published recommendations [3]. Geminin expression and its association with various clinicopathological variables were evaluated in our cohort.
Negative
125 (46.8 %)
N1
90 (33.7 %)
N2
24 (9 %)
N3
28 (10.5 %)
Not known
24
Lympho-vascular invasion Yes
130 (47.3 %)
No
145 (52.7 %)
Not known Surgical treatment Wide local excision
16 97 (33.4 %)
Mastectomy
193 (66.6 %)
No surgery
1
Adjuvant radiotherapy Yes
162 (58.7 %)
No
114 (41.3 %)
Not known
15
Adjuvant chemotherapy Yes
91 (33.5 %)
No
181 (66.5 %)
Not known
19
Geminin was positively correlated with tumour size (q = 0.181, p = 0.002) and Ki67 (q = 0.59, p \ 0.01), but negatively associated with age (q = -0.205, p = 0.0004). Both geminin (p \ 0.01) and Ki67 (p \ 0.01) were positively correlated with grade and negatively correlated with ERa status (p \ 0.001). However, there was no difference in the expression of geminin or Ki67 with the nodal status (p = 0.86 and p = 0.51, respectively) or LVI (p = 0.18 and p = 0.91, respectively). One hundred and eighty six patients (63.9 %) were geminin positive, and 105 (36.1 %) were negative (geminin LI C 2 % as cut-off). There were 12 breast cancer-
123
94
Breast Cancer Res Treat (2014) 143:91–98
Co-ordinates of the curve for BCSS y Sensitivity 1Geminin cut-off -1.0 1.000 1.000 2.1 0.824 0.610 63.7 0.000 0.000
Co-ordinates of the curve for DFS Geminin Sensitivity cut-off -1.0 1.000 1.000 2.1 0.808 0.583 63.7 0.000 0.000
Fig. 1 ROC curve to determine the optimum cut-off for geminin. The reference line is depicted in green and geminin curve in blue. Geminin LI of C2 % was taken as the optimum cut-off by visually inspecting the sensitivity and specificity values of the ROC curve coordinates represented in the Table. (Color figure online)
specific deaths amongst geminin negative patients compared to 56 deaths in geminin-positive cases [HR 2.85 (1.53–5.32); mean survival = 188 months]. Similarly, geminin-positive patients (geminin LI C 2 %) were also found to have worse DFS [HR 2.63 (1.47–4.71); mean time to recurrence or metastasis = 183 months]. The BCSS and DFS curves for geminin are represented in Fig. 3a, b. The Ki67 score was unable to be determined in 24 patients due to missing cores in the TMA, a well-recognised issue with TMAs. Ki67 was found to be positive only in 40 patients (15 %) and negative in 227 (85 %) with a cut-off of C14 %. Ki67 positivity did not influence the BCSS (p = 0.088) or DFS (p = 0.149) in our cohort. Cox regression univariate analysis was performed to establish the prognostic role of geminin LI and Ki67 score as percentages (continuous variables) along with other known clinical and pathological prognosticators. Tumour size, lymph nodal status (positive or negative), tumour grade, ER status, Ki67 and geminin LI were identified as statistically significant predictors for BCSS and DFS (Table 2). However, on multivariate analysis after adjusting for adjuvant treatments received, only tumour size, nodal status and adjuvant hormonal treatment were found to be independent predictors for both BCSS and DFS (Table 2). We did not stratify according to the treatment while performing multivariate analysis, as this cohort was
123
Fig. 2 Depiction of nuclear specificity of geminin immunohistochemistry (arrows; a). In contrast, Ki67 immunohistochemistry (b) is often non-specific with granular cytoplasmic staining (arrows) predominating over true nuclear staining (asterisk). Original magnification = 920
predominantly ER-positive (80 %) with patients receiving adjuvant endocrine therapy (mostly tamoxifen). Cox multivariate analysis was repeated using geminin (LI C 2 %) and Ki67 (C14 %) positivity as independent variables (categorical). This confirmed the previously established independent predictive roles of tumour size, nodal status and adjuvant hormonal treatment for BCSS
Breast Cancer Res Treat (2014) 143:91–98
95
Fig. 3 Kaplan–Meier survival curve showing poor BCSS (a) and DFS (b) in gemininpositive breast cancers. Patients were censored, when they ceased to be followed up for any reason but had not died due to breast cancer
and DFS. However, the results also showed geminin positivity [C2 % cut-off; HR 2.27 (1.01, 5.06); p = 0.04) to be a predictor for BCSS. Although there was a trend towards poor DFS with geminin positivity [HR 1.2 (0.95–4.2); p = 0.07], it did not reach statistical significance. Conversely, Ki67 positivity was not found to have any independent predictive role in our cohort.
Discussion The uncontrolled proliferation is a hallmark of malignancy, and the role of established proliferative markers like Ki67 in breast cancer management is limited. This is due to the variation in the analytical methods used for Ki67 immunohistochemistry and the lack of consensus on the ideal cut-off for Ki67 in clinical practice [3]. Compared to Ki67, geminin is expressed only by proliferative cells, either normal (male germ cells, lymphocytes and epithelial cells) or tumour cells (colon, kidney and breast) [10–12, 18]. Moreover, there have been no issues with the analytical methods used for immunohistochemical evaluation of
geminin in various cancer tissues [11, 12, 19], and the staining pattern was shown to be nuclear specific in our cohort (Fig. 2). In this study, the prognostic role of geminin in a large cohort of breast cancer patients with a median follow up of 188 months (range 5–229 months) was determined and compared to that of Ki67 and other clinically relevant prognostic variables. The data differ from other published studies due to the availability of almost complete follow-up data (missing clinical data = 20 patients (6.4 %)) and survival data pertaining to breast cancer-specific events compared to the overall survival reported in the literature so far. Geminin expression was positively correlated with Ki67, tumour size, grade and mitotic index. Conversely, ERa was negatively associated with Ki67 expression and geminin LI. It was also observed that, younger patients have a higher geminin LI compared to that of their older counterparts, and there was a linear association between age and geminin LI. The above associations of geminin with tumour grade, size and the strong positive correlation with K67 strengthen the role of geminin as a potential
123
96
Breast Cancer Res Treat (2014) 143:91–98
Table 2 Cox uni- and multivariate analysis of geminin in disease-free and breast cancer-specific survival Univariate
Multivariate
DFS
BCSS
HR
p
HR
DFS p
HR
BCSS p
HR
p
Grade
5.92 (2.5–13.9)
\0.001
5.4 (2.3–12.8)
\0.0001
2.7 (0.92–7.7)
0.07
2.4 (0.8 –6.9)
0.12
LN
2.48 (1.45–4.22)
0.001
2.8 (1.6–4.9)
\0.001
1.9 (1.06–3.3)
0.03
2.1 (1.2–3.9)
0.016
Size ER
1.02 (1.01–1.02) \0.001 0.44 (0.26–0.72) 0.001
1.02 (1.01–1.02) \0.001 0.4 (0.2–0.8) 0.003
1.02 (1.0–1.03) 1.5 (0.7 –3.2)
0.001 0.28
1.02 (1.0–1.03) 1.7 (0.75–3.7)
0.003 0.21 \0.001
Adjuvant hormone therapy
nd
nd
nd
nd
0.19 (0.08–0.4)
\0.001
0.17 (0.07–0.4)
Adjuvant radiotherapy
nd
nd
nd
nd
0.63 (0.4–1.09)
0.09
1.02 (0.9–1.06)
0.51
Ki67 score (continuous variable)
1.02 (1.01–1.04)
0.001
1.03 (1.01–1.04) \0.001
1.0 (0.98–1.03)
0.7
1.01 (0.9–1.04)
0.45
Geminin LI (continuous variable)
1.04 (1.02–1.06) \0.001
\0.001
1.02 (0.98–1.06)
0.39
1.02 (0.97–1.06)
0.51
1.05 (1.02–1.0)
nd Not done
proliferative marker in breast cancer, and it may also correlate with aneuploidy, given its involvement with defective replication forks [14]. This was reflected even in the survival analysis, with geminin-positive patients (LI C 2 %) found to be three times more likely to succumb to breast cancer-related death during follow-up (HR 2.85; CI 1.53–5.32; p = 0.001). Similarly, geminin-positivity (LI C 2 %) was associated with twice the risk of developing recurrence or metastasis in our cohort (HR 2.63; CI 1.47–4.71; p = 0.001). Furthermore, geminin positivity (LI C 2 %) was shown to be an independent predictor for determining BCSS in our cohort. We found Ki67 was neither an independent predictor nor was related to breast cancer-specific events (BCSS or DFS). The inability to distinguish an optimal cut-off to determine Ki67 positivity is well recognised, with values ranging from 1 to 28.6 % having been used in the literature [20]. Even though various cut-offs have been used, a score in the range of 10–20 % has been used most commonly to dichotomize population into a ‘high’ and ‘low’ Ki67 expression [21]. Hence, we evaluated, whether using the median value of Ki67 and geminin as a cut-off [12, 22] would change the observed results (geminin LI cut-off 4 %; Ki67 cut-off 2.4 %). This analysis showed that both Ki67 and geminin positivity were associated with statistically significant poor BCSS and DFS (results not shown). High Ki67 LI has been reported to be predictive of responsiveness to neo-adjuvant chemotherapy [23, 24] and neo-adjuvant endocrine therapy [25]. Moreover, two recent meta-analyses have also shown that high Ki67 expression to be associated with increased risk of breast cancer relapse and death [21, 26]. However, we did not find Ki67 to be an independent predictor of BCSS or DFS. This may be
123
probably due to the strong positive correlation of geminin with Ki67 observed in our cohort as well as due to the strong influence of geminin on survival. This was confirmed by entering only geminin and Ki67 in the Cox proportion hazard model, which showed geminin to be a stronger predictor compared with Ki67 for both BCSS (Ki67 - B = 0.013, SE = 0.012, p = 0.28; geminin - B = 0.032, SE = 0.016, p = 0.04) and DFS (Ki67 - B = 0.007, SE = 0.012, p = 0.54; geminin - B = 0.036, SE = 0.016, p = 0.02). Identification of biomarkers that can monitor breast cancer progression or serve as a surrogate marker for prognosis will enable us to personalise medicine and to improve care of patients. Individualised recurrence scores like Oncotype DX are heavily weighed for the expression of genes, like that of Ki67, which account for proliferation [27]. However, issues with reproducibility of Ki67 immunohistochemistry limit its potential for efficient translation to the clinical setting for the management of breast cancer patients. Geminin, on the other hand, has been shown to be consistently associated with poor prognosis in various cancers (lung, colon, renal and brain tumours) [11, 19, 28, 29]. Similarly, geminin expression was also shown to be positively correlated with tumour grade in a variety of tumours including breast, kidney, prostate and salivary glands [11, 12, 29–31]. However, there are only two studies which evaluated the prognostic role geminin in breast cancer [12, 32]. Gonzalez et al. [12] showed poor overall survival with high geminin LI, whereas the reliability of the observed results were significantly compromised in the other study due to accounting for cytoplasmic expression of geminin in the immunohistochemical evaluation in spite of geminin being a nuclear antigen [32]. The results
Breast Cancer Res Treat (2014) 143:91–98
presented here support the poor prognosis and adverse breast cancer survival amongst geminin-positive patients (LI C 2 %) agreeing with the findings from other cancers. Moreover, we also showed that geminin was statistically superior to Ki67 in differentiating those patients with poor breast cancer prognosis. Certain limitations of our study should be discussed before summarising the findings. The retrospective nature of the study may have introduced selection bias; however, it provided a large sample of breast cancer patients with longterm follow-up for studying the prognostic significance. The adjuvant treatments received by the patients, especially the adjuvant chemotherapy in the 1990s were indeed different compared to that being currently offered. The effects of various adjuvant treatments on breast cancer-specific survival and prognosis were hence accounted for by adjusting them in the multivariate model during statistical analysis. There was also an element of attrition bias due to the loss of almost 21 % of our initial cohort (77 out of 368). However, only 7 % was lost to follow-up, and the rest due to missing TMA cores (14 %), which is a recognised limitation of studies’ conducted using TMAs. The strong correlation of geminin with Ki67 and other clinicopathological variables (grade, tumour size and ERa) may also have influenced the results of the multivariate analysis. Nevertheless, the results of this study support the prognostic role of geminin in breast cancer. However, in order to improve the validity, independent verification of our results in an external cohort is needed. There is also a need for further research to investigate the correlation between geminin and Ki67 to evaluate whether geminin can reliably substitute Ki67 as a proliferative marker in breast cancer patients and potentially take its place as a predictive as well as prognostic agent or if it is simply a measure of aneuploidy. In conclusion, geminin positivity (LI C 2 %) was shown to be a predictor of poor survival in our breast cancer cohort. The selective expression of geminin during the proliferative phase of the cell cycle and its nuclear specificity increase its potential to be considered as an alternative marker of proliferation in breast cancer patients. Acknowledgments The authors are grateful to the Leeds Teaching Hospitals NHS Trust Special Trustees for funding this work. HHT is funded by the Cancer Research UK. Conflict of interest
97
3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
The authors have no conflicts of interest. 17.
References 18. 1. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144(5):646–674 2. Isola JJ, Helin HJ, Helle MJ, Kallioniemi OP (1990) Evaluation of cell proliferation in breast carcinoma. Comparison of Ki-67
immunohistochemical study, DNA flow cytometric analysis, and mitotic count. Cancer 65(5):1180–1184 Dowsett M, Nielsen TO, A’Hern R, Bartlett J, Coombes RC, Cuzick J, Ellis M, Henry NL, Hugh JC, Lively T et al (2011) Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer Working Group. J Natl Cancer Inst 103(22):1656–1664 Dowsett M, Allred C, Knox J, Quinn E, Salter J, Wale C, Cuzick J, Houghton J, Williams N, Mallon E et al (2008) Relationship between quantitative estrogen and progesterone receptor expression and human epidermal growth factor receptor 2 (HER-2) status with recurrence in the arimidex, tamoxifen, alone or in combination trial. J Clin Oncol 26(7):1059–1065 Dowsett M, Salter J, Zabaglo L, Mallon E, Howell A, Buzdar AU, Forbes J, Pineda S, Cuzick J (2011) Predictive algorithms for adjuvant therapy: TransATAC. Steroids 76(8):777–780 Faratian D, Munro A, Twelves C, Bartlett JMS (2009) Membranous and cytoplasmic staining of Ki67 is associated with HER2 and ER status in invasive breast carcinoma. Histopathology 54(2):254–257 Yerushalmi R, Woods R, Ravdin PM, Hayes MM, Gelmon KA (2010) Ki67 in breast cancer: prognostic and predictive potential. Lancet Oncol 11(2):174–183 McGarry TJ, Kirschner MW (1998) Geminin, an inhibitor of DNA replication, is degraded during mitosis. Cell 93(6):1043–1053 Wohlschlegel JA, Kutok JL, Weng AP, Dutta A (2002) Expression of geminin as a marker of cell proliferation in normal tissues and malignancies. Am J Pathol 161(1):267–273 Eward KL, Obermann EC, Shreeram S, Loddo M, Fanshawe T, Williams C, Jung HI, Prevost AT, Blow JJ, Stoeber K et al (2004) DNA replication licensing in somatic and germ cells. J Cell Sci 117(Pt 24):5875–5886 Dudderidge TJ, Stoeber K, Loddo M, Atkinson G, Fanshawe T, Griffiths DF, Williams GH (2005) Mcm2, Geminin, and KI67 define proliferative state and are prognostic markers in renal cell carcinoma. Clin Cancer Res 11(7):2510–2517 Gonzalez MA, Tachibana KE, Chin SF, Callagy G, Madine MA, Vowler SL, Pinder SE, Laskey RA, Coleman N (2004) Geminin predicts adverse clinical outcome in breast cancer by reflecting cell-cycle progression. J Pathol 204(2):121–130 Blanchard Z, Malik R, Mullins N, Maric C, Luk H, Horio D, Hernandez B, Killeen J, Elshamy WM (2011) Geminin overexpression induces mammary tumors via suppressing cytokinesis. Oncotarget 2(12):1011–1027 Gardner L, Malik R, Shimizu Y, Mullins N, ElShamy WM (2011) Geminin overexpression prevents the completion of topoisomerase IIalpha chromosome decatenation, leading to aneuploidy in human mammary epithelial cells. Breast Cancer Res 13(3):R53 McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM (2005) Diagnostics ftSSotN-EWGoC: reporting recommendations for tumor marker prognostic studies (REMARK). J Natl Cancer Inst 97(16):1180–1184 Shaaban AM, Ball GR, Brannan RA, Cserni G, Di Benedetto A, Dent J, Fulford L, Honarpisheh H, Jordan L, Jones JL et al (2012) A comparative biomarker study of 514 matched cases of male and female breast cancer reveals gender-specific biological differences. Breast Cancer Res Treat 133(3):949–958 Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, Watson M, Davies S, Bernard PS, Parker JS et al (2009) Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer. J Natl Cancer Inst 101(10):736–750 Hansel DE, Rahman A, Hidalgo M, Thuluvath PJ, Lillemoe KD, Shulick R, Ku JL, Park JG, Miyazaki K, Ashfaq R et al (2003) Identification of novel cellular targets in biliary tract cancers using global gene expression technology. Am J Pathol 163(1):217–229
123
98 19. Nishihara K, Shomori K, Tamura T, Fujioka S, Ogawa T, Ito H (2009) Immunohistochemical expression of geminin in colorectal cancer: implication of prognostic significance. Oncol Rep 21(5):1189–1195 20. Urruticoechea A, Smith IE, Dowsett M (2005) Proliferation marker Ki-67 in early breast cancer. J Clin Oncol 23(28):7212–7220 21. Stuart-Harris R, Caldas C, Pinder SE, Pharoah P (2008) Proliferation markers and survival in early breast cancer: a systematic review and meta-analysis of 85 studies in 32,825 patients. Breast 17(4):323–334 22. Viale G, Giobbie-Hurder A, Regan MM, Coates AS, Mastropasqua MG, Dell’Orto P, Maiorano E, MacGrogan G, Braye SG, Ohlschlegel C et al (2008) Prognostic and predictive value of centrally reviewed Ki-67 labeling index in postmenopausal women with endocrine-responsive breast cancer: results from Breast International Group Trial 1–98 comparing adjuvant tamoxifen with letrozole. J Clin Oncol 26(34):5569–5575 23. Archer CD, Parton M, Smith IE, Ellis PA, Salter J, Ashley S, Gui G, Sacks N, Ebbs SR, Allum W et al (2003) Early changes in apoptosis and proliferation following primary chemotherapy for breast cancer. Br J Cancer 89(6):1035–1041 24. Chang J, Ormerod M, Powles TJ, Allred DC, Ashley SE, Dowsett M (2000) Apoptosis and proliferation as predictors of chemotherapy response in patients with breast carcinoma. Cancer 89(11):2145–2152 25. Dowsett M, Smith IE, Ebbs SR, Dixon JM, Skene A, A’Hern R, Salter J, Detre S, Hills M, Walsh G (2007) Prognostic value of Ki67 expression after short-term presurgical endocrine therapy for primary breast cancer. J Natl Cancer Inst 99(2):167–170
123
Breast Cancer Res Treat (2014) 143:91–98 26. de Azambuja E, Cardoso F, de Castro G Jr, Colozza M, Mano MS, Durbecq V, Sotiriou C, Larsimont D, Piccart-Gebhart MJ, Paesmans M (2007) Ki-67 as prognostic marker in early breast cancer: a meta-analysis of published studies involving 12,155 patients. Br J Cancer 96(10):1504–1513 27. Jonat W, Arnold N (2011) Is the Ki-67 labelling index ready for clinical use? Ann Oncol 22(3):500–502 28. Haruki T, Shomori K, Hamamoto Y, Taniguchi Y, Nakamura H, Ito H (2011) Geminin expression in small lung adenocarcinomas: implication of prognostic significance. Lung Cancer 71(3):356–362 29. Shrestha P, Saito T, Hama S, Arifin MT, Kajiwara Y, Yamasaki F, Hidaka T, Sugiyama K, Kurisu K (2007) Geminin: a good prognostic factor in high-grade astrocytic brain tumors. Cancer 109(5):949–956 30. Dudderidge TJ, McCracken SR, Loddo M, Fanshawe TR, Kelly JD, Neal DE, Leung HY, Williams GH, Stoeber K (2007) Mitogenic growth signalling, DNA replication licensing, and survival are linked in prostate cancer. Br J Cancer 96(9):1384–1393 31. Vargas PA, Cheng Y, Barrett AW, Craig GT, Speight PM (2008) Expression of Mcm-2, Ki-67 and geminin in benign and malignant salivary gland tumours. J Oral Pathol Med 37(5):309–318 32. Di Bonito M, Cantile M, Collina F, Scognamiglio G, Cerrone M, La Mantia E, Barbato A, Liguori G, Botti G (2012) Overexpression of cell cycle progression inhibitor geminin is associated with tumor stem-like phenotype of triple-negative breast cancer. J Breast Cancer 15(2):162–171
