Prognostic Values of Proliferative Markers ki-67 and repp86 in Breast Cancer

Arch Iranian Med 2007; 10 (1): 27 31 Original Article Prognostic Values of Proliferative Markers ki-67 and repp86 in Breast Cancer Jaleh Mohsenifar MD*, Maryam Almassi-Aghdam MD*, Zohreh Mohammad-Taheri MD**, Khandan Zare MD***, Bahar Jafari MD*, Morteza Atri MD, Seyed-Hossein Mortazavi MD, Payam Azadeh MD, Mostafa Bagherzadeh MD, Eznollah Azargashb MSc, Farzaneh Rahimi MD * Background: Breast cancer is the leading cause of carcinoma death in women. Proper treatment depends on the consideration of molecular biology status of tumor cells, which may determine the patient's treatment and prognosis. To determine the prognostic models for this disease, we evaluated the role of cell proliferation-related antigens including ki-67 (a nuclear antigen, expressed in G1, G2, and M phases of cell cycle) and repp86 (an 86-kDa nuclear protein expressed in S, G2, and M phases of cell cycle) for detection of biologic behavior of breast cancer. Methods: We studied 60 women with grade I and II lymph node-negative and 27 with grade III lymph node-positive breast cancers. The mean follow-up periods for these two groups were 60 and 72 months, respectively. Tumor cell proliferation was measured by immunohistochemical methods with monoclonal antibodies directed against the nuclear antigens ki-67 and repp86. Results: The ki-67, repp86 labeling indices (percentage of antibody-stained tumor cell nuclei) were not statistically different between the cases and controls of lymph node-negative patients (ki- 67 with P = 0.33; repp86 with P = 0.40). The odds ratio (the mean chance of ki-67 labeling index > 10%, repp86 labeling index >10%) in patients with recurrence was 4 (CI = 0.2 76.5) for ki-67 and 3.6 (CI = 0.4 32.5) for repp86. Both indices were statistically different in lymph node-positive cases and controls (P < 0.0001). The odds ratio in patients with recurrence was 87 (CI = 4 18.71) for ki-67 and 71.5 (CI = 5.7 899.2) for repp86. Conclusion: The present study confirms the importance of cell proliferation as a determinant of biologic behavior of breast cancer. Measurement of ki-67 and repp86 labeling indices may be very helpful for physicians to detect high-risk patients and to adopt appropriate procedure such as adjuvant therapy. Archives of Iranian Medicine, Volume 10, Number 1, 2007: 27-31. Keywords: Breast cancer immunohistochemistry tumor marker ki-67 repp86 Introduction B reast carcinoma is the most prevalent nondermatologic malignancy in women. A 90-year-old woman may have breast cancer with a chance of 1:8 1 ; on the other hand, the most common cause of carcinoma death among women is breast cancer. The prognosis of this disease can be affected by various clinical and pathologic factors. 1, 2 One factor is the proliferation rate of cells, which is especially appropriate for those patients with lymph node (LN)-negative small-sized tumor for selecting systematicallycorrect treatment methods and to avoid additional treatment. 3 Whether investigated by mitotic count method, or by MIB-1 (ki-67) analogous immunization or by estimating S-shape fraction Authors affiliations: Department of Pathology, *Taleghani Hospital, **Masih Daneshvavri Hospital, Shaheed Beheshti University of Medical Sciences, ***Imam Khomeini Hospital, Tehran University of Medical Sciences, Department of General Surgery, Imam Khomeini Hospital, Tehran University of Medical Sciences, Department of Radiotherapy, Imam Hossein Hospital, Shaheed Beheshti University of Medical Sciences, Imam Khomeini Hospital, Tehran University of Medical Sciences, Department of Pathobiology, Central Pathobiology Laboratory, Department of Statistic, Shaheed Beheshti University of Medical Sciences, Tehran, Iran. Corresponding author and reprints: Farzaneh Rahimi MD, Department of Pathology, Taleghani Hospital, Shaheed Beheshti University of Medical Sciences, Tehran, Iran. E-mail: zhmohssenifar@yahoo.com. Accepted for publication: 14 June 2006 Archives of Iranian Medicine, Volume 10, Number 1, January 2007 27

Prognostic value of ki-67 and repp86 in breast cancer using flowcytometry, this parameter is one of the most important determinant factors for prognosis. For this reason, it has been included by Elston in Combined Grading, and in practice, some believe that it is the most important component of that system. 1, 4 Ki-67 is an antigen related to nonhistone nuclear proteins that is expressed in G1, G2, M, and S proliferative phases. In general, there is a good relationship between the staining of ki-67 and the mitotic count. 1 Repp86 is an 86-kDa protein that is expressed in S, G2, and M phases. 5 In nucleoplasm of proliferating cells, it is detectable in S boundary. It is compacted at the beginning of mitosis, and continuously moves toward the mitotic spindles and the mid-body during mitosis. Its activity is regulated by phosphorylation of the serine/treonine residues and it is suggested that its overexpression leads to stimulation of cells in G2-M and ultimately to polyploidy. 3 This study was conducted to answer why patients with similar breast cancer in terms of grade and stage, have completely different prognosis, so that some have survival equal to that of healthy people, while others have a 5-year survival of 13%. 3 Patients and Methods In a case-control study, 87 patients with breast cancer, including 60 with LN-negative grade I and II cancers and 27 with LN-positive grade III cancers who referred to various health care centers such as, Department of Radiotherapy of Jorjani and Imam Hossein Hospitals, a private surgical clinic, and a private hospital in Tehran during the past 10 years, were studied. Data sheet was completed for each patient and the patients were then assigned to two groups based on their tumor recurrence; local recurrence or metastasis. They were matched for gender, grade of the disease, stage, and type of treatment they received. Paraffin-embedded blocks and slides of these patients were reinvestigated by two pathologists, independently. After confirmation of the previous diagnosis, new sections were prepared from blocks; at least three sections, one for H&E and two for staining with ki-67 and repp86. Ki-67 and repp86 markers were evaluated by conventional methods. 6 Markers ki-67 and repp86 were evaluated by immunohistochemical (IHC) methods in a conventional manner. Sections of 5-µm thickness were dried at 37ºC during night and were incubated during the next day at 60ºC for 1 hour. Endogenous peroxidase was neutralized by 3% peroxidase-methanol (w+/v) solution. Before using the primary antibodies (ki-s5 and ki-s2), epitopes were revealed by boiling slides in a Tris-HCl solution with ph 9 in autoclave at 120ºC and 1 atmosphere pressure for 10 minutes. After cooling, each side was treated with mab ki-s2 (produced in Pathology Department, Keil University) and mab ki-s5 (Dako) for 50 minutes in a humid media at room temperature. Safety reaction was done by striptavidin complex method using peroxidase and biotinylated rabbit antimouse antibody according to manufacturer s (Dako) instructions (Table 1). In each immunostaining stage, a section of tonsil was considered as positive and negative control. Sections were counterstained by Haris Hematoxylin, were dehydrated, and then a slide was applied on them. Each staining was controlled by three pathologists, independently. IHC staining was studied under light microscope with magnifications of 40x and 400x. First, the most active proliferative regions were scanned with magnification of 40x. Then, in this region, 500 cells were sequentially counted with magnification of 400x. Number of cells stained with proliferative markers with each light intensity were recorded. Labeling index (LI) for each tumor is presented as percentages of immunoreactive cells in <10% and >10%. To investigate the frequency distribution of markers in each group, Fisher exact test, χ 2 test, and kappa coefficient of congruency were used. Results The age of patients with LN-negative grade I and II cancers ranged between 27 and 78 years (mean: 50). During 1990 2004, they had followups ranging from 3 to 12 years (mean: 5). Of 60 LN-negative patients investigated, seven (12%) had recurrence (case group) all of whom had ki-67 Table 1. Primary antibodies, antibody dilutions, commercial sources, and positive control. Antibody Antigen Clone Dilution Source Positive Retrieval mab ki-s2 repp86 Monoclonal 0 Keil University Tonsil Tris HCl ph: 9 autoclave mab ki-s5 ki-67 Monoclonal 1.25 Dako Tonsil Tris HCl ph: 9 autoclave 28 Archives of Iranian Medicine, Volume 10, Number 1, January 2007

J. Mohsenifar, M. Almassi-Aghdam, Z. Mohammad-Taheri, et al Table 2. Frequency distribution of ki-67 and repp86 in cases and controls of low- and high-grade, lymph node-negative Low-grade n (%) High-grade n (%) ki-67 repp86 ki-67 repp86 Case Control Total Case Control Total Case Control Total Case Control Total <10% 0 11 11 1 20 21 0 10 10 1 11 12 (0) (20.8) (18.3) (14.3) (37.7) (35) (0) (76.9) (37) (7.2) (84.6) (44.4) >10% 7 42 49 6 33 39 14 3 17 13 2 15 (100) (79.2) (81.7) (85.7) (62.3) (65) (100) (23.1) (63) (92.8) (15.4) (55.6) Total 7 53 60 7 53 60 14 13 27 14 13 27 (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) LI >10%; six (86%) of them had repp86 LI >10%. Of 53 patients in the control group, 42 (79%) had a ki-67 LI >10%. Furthermore, of these 53 patients, 33 (62%) had a repp86 LI >10% (Table 2). The age of patients with LN-positive grade III cancer ranged from 30 to 62 years (mean: 43). During 1990 2004, they had follow-ups ranging from two to 10 years (mean: 5.6) including those who died of the disease. Of 27 LN-positive patients, 14 (48%) had recurrence (case group). Of the 14 patients in the case group, all had ki-67 LI >10%, and 13 (93%) had repp86 LI >10%. Of these 13 patients, 3 (12%) had ki-67 LI>10%, and two (15%) had repp86 LI >10% (Table 2). The age of patients with LN-positive grade III cancer ranged from 30 to 62 years (mean: 43). During 1990 2004, they had follow-ups ranging from two to 10 years (mean: 5.6) including those who died of the disease. Of 27 LN-positive patients, 14 (48%) had recurrence (case group). Of the 14 patients in the case group, all had ki-67 LI >10%, and 13 (93%) had repp86 LI >10%. Of these 13 patients, 3 (12%) had ki-67 LI>10%, and two (15%) had repp86 LI >10% (Table 2). Simultaneous investigation of the frequency distribution of markers ki-67 and repp86 in both LN-positive and LN-negative case patients showed an almost complete correlation between the two markers (i.e., all patients but one in each case group had ki-67 and repp86 LI >10% for LNpositive and LN-negative patients). Control group patients, on the other hand, did not show a statistically significant correlation between the two markers (Tables 3, 4, and 5). Investigation of the frequency distribution of each marker based on conditions of lymph nodes indicated that there is no statistically significant difference between LN-positive and LN-negative patients for ki-67 and repp86 incidence (Table 6). Discussion In this study, the main objective was to assess the predictive value of cell proliferation factors ki- 67 and repp86 in prognosis of patients with breast cancer. To the best of our knowledge, this study is the first one investigating this issue among Iranian Our results indicated that both the case and the control groups in LN-negative patients had no statistically significant difference in ki-67 (P = 0.33) and repp86 (P = 0.40). However, on average, the chance of having ki-67 LI >10% and repp86 LI >10% (odds ratio) in the case group was 4-fold (CI 95% : 0.21 76.5) and 3.6-fold (CI 95% : 0.4 Table 3. Simultaneous investigation of frequency distribution of markers ki-67 and repp86 in lymph nodenegative and -positive Lymph node-negative patients Lymph node-positive patients <10% 9 12 21 7 5 12 >10% 2 37 39 2 13 15 Total 11 49 60 9 18 27 Table 4. Frequency distribution of markers ki-67 and repp86 in lymph node-negative case and control Case Control <10% 0 1 1 9 11 20 >10% 0 6 6 1 32 23 Total 0 7 7 10 43 53 Archives of Iranian Medicine, Volume 10, Number 1, January 2007 29

Prognostic value of ki-67 and repp86 in breast cancer Table 5. Frequency distribution of markers ki-67 and repp86 in lymph node-positive case and control Case Control <10% 0 1 1 8 3 11 >10% 0 13 13 2 0 2 Total 0 14 14 10 3 13 32.5), respectively. All LN-negative patients in the case group represented ki-67 LI >10% and all but one represented repp86 LI >10%. It is necessary to note that the above single case had involved surgical margin at the time of initial diagnosis, with no special procedure done for the situation, as mentioned in her file. The patient was followed for 10 years. Three years after the initial diagnosis, she had local recurrence and seven years later, she developed osseous metastasis; nevertheless, she is alive now. The surgical margin involvement in this case can explain her recurrence in spite of having repp86 LI <10%, especially by considering her long-term follow-up. However, almost 79% pf the cases and 62% of the controls showed ki-67 LI >10% and repp86 LI >10%. The issues proposed in this field are as follows: Follow-up period was approximately three years in about half of the During this period, the patients cannot be judged for recurrence; these patients may be found positive to have tumor recurrences by long-term followup. In some studies, the cut-off value used for these indices is reported to be 20%, 7 while in other studies and ours, the value of 10% has been used as the cut-off value. Considering the individual-geographic predictive factors and conditions, increasing the cut-off value may provide more acceptable results. All of these patients were LN-negative and had low stages of the disease. It is necessary to note that the facilities available for detection of LN involvement include tissue limited dissections and H&E staining. In cases with deeper dissections and by applying more specialized staining, some of these LN-negative patients may be found positive and they may correlate with high percentage of tumor proliferation. 8 On the other hand, case and control groups of LN-positive patients showed statistically significant differences for ki-67 and repp86 (p < 0.0001 for both markers). On average, the chance of having ki-67 LI >10% and repp86 LI >10% in the case group were 87-flod (CI 95% : 4 1871) and 71.5-fold (CI 95% : 5.7 899), respectively. Statistical results of this study are very similar to other studies conducted in this field. In a study by Wiekel, a clear relationship was found between ki-67 score and the latency of the disease 9 Crosier indicated that increase in proliferation level in tumor cells can lead tumors to be symptomatic in cancer follow-up period. 6 Finally, ki-67 as an alternative (for mitotic count) can have a predictive value in proliferation of breast tumor injuries, 5 so that its increased expression percentage can lead to progression of proliferation in tumor cells and a worse prognosis. Very limited studies are available on the importance of repp86 LI as the most important factor predicting the survival of patients with breast cancer. 10 If this maker is less than 10% in tumor cells of the affected patient, there is no difference in survival rate with healthy population with similar age. On the other hand, if it is >10%, the mortality rate will be increased by 20-fold. By examining telomerase relative activity levels in melanocytic lesions, Rudolph et al concluded that measurement of cell proliferation (ki-67, repp86) together with telomerase activity can make it possible to detect malignant transformation of the lesions in the initial stages of the disease. 11 Bonatz et al also evaluated the above variables in the initial stages of ovarian carcinoma and reached the same conclusion. 12 Role of repp86 in patients with neuroblastoma was examined by Krams et al. 5 According to the results obtained in this study, measurement of repp86 LI makes it possible to determine patients with favorable, and Table 6. Frequency distribution of markers ki-67 and repp86 based on condition of lymph node. ki-67/lymph node status Repp86/lymph node status + - Total + - Total <10% 10 11 21 12 21 33 >10% 17 49 66 15 39 54 Total 27 60 86 27 60 87 30 Archives of Iranian Medicine, Volume 10, Number 1, January 2007

J. Mohsenifar, M. Almassi-Aghdam, Z. Mohammad-Taheri, et al adverse prognosis in groups of specified stage, grade, age, and MYCN gene conditions. 5 Cell proliferation plays an important role in the clinical behavior and aggressiveness of breast cancer. Increased proliferation is strongly associated with poor prognosis of the disease, although proliferative markers as prognostic factors predicting response to the treatment have been weakly considered. Evaluation of ki-67, repp86 LI in relation with a tumor sample can improve clinician s ability to predict the prognosis of the patients and to differentiate low-risk patients from high recurrent ones (for receiving adjunct therapy). Surely, more studies should be conducted to shed light over this area of cancer research. Acknowledgment We would like to express our regards to the eternal soul of the high-ranked Professor Parwaresh of Molecular Pathology Department of Kiel University, for his sincere assistance in this study. References 1 Rosai J, Ackerman LV. Rosai and Ackerman s Surgical Pathology. 9th ed. Edinburgh: New York: Mosby; 2004. 2 Kumar V, Abbas AK, Fausto N, Robbins SL. Robbins and Cotran, Pathologic Basis of Disease. 7th ed: Philadelphia: Elsevier Saunders; c2005. 3 Heidebrecht HJ, Adam-Klages S, Szczepanowski M, Pollmann M, Buck F, Endl E, et al. Repp86: a human protein associated in the progression of mitosis. Mol Cancer Res. 2003; 1: 271 279. 4 van Diest PJ, van der Wall E, Baak JP. Prognostic value of proliferation in invasive breast cancer: a review. J Clin Pathol. 2004; 57: 675 681. 5 Krams M, Heidebrecht HJ, Hero B, Berthold F, Harms D, Parwaresch R, et al. Repp86 expression and outcome in patients with neuroblastoma. J Clin Oncol. 2003; 21: 1810 1818. 6 Crosier M, Scott D, Wilson RG, Griffiths CD, May FE, Westley BR. Differences in ki-67 and c-erbb2 expression between screen-detected and true interval breast cancers. Clin Cancer Res. 1999; 5: 2682 2688. 7 Breast Cancer at Baylor College of Medicine. Prognostic biomarkers ki-67 proliferation index [cited 2005]. Available from: URL: http://www.breastcenter.tmc.edu/ research/cores/path/services/ki67.htm. 8 Kuhling H, Alm P, Olsson H, Ferno M, Baldetorp B, Parwaresch R, et al. Expression of cyclins E, A, and B, and prognosis in lymph node-negative breast cancer. J Pathol. 2003; 199: 424 431. 9 Weikel W, Beck T, Rosenthal H, Herzog R. The immunohistochemical growth fraction (ki-67) of breast cancers: relations to tumor spread, tumor morphology and receptor testing. Geburtshilfe Frauenheilkd. 1989; 49: 277 282. 10 Rudolph P, Alm P, Heidebrecht HJ, Bolte H, Ratjen V, Baldetorp B, et al. Immunologic proliferation marker Ki- S2 as prognostic indicator for lymph node-negative breast cancer. J Natl Cancer Inst. 1999; 91: 271 278. 11 Rudolph P, Schubert C, Tamm S, Heidorn K, Hauschild A, Michalska I, et al. Telomerase activity in melanocytic lesions: a potential marker of tumor biology. Am J Pathol. 2000; 156: 1425 1432. 12 Bonatz G, Frahm SO, Klapper W, Helfenstein A, Heidorn K, Jonat W, et al. High telomerase activity is associated with cell cycle deregulation and rapid progression in endometrioid adenocarcinoma of the uterus. Hum Pathol. 2001; 32: 605 614. Archives of Iranian Medicine, Volume 10, Number 1, January 2007 31