REZUMAT
Introducere: Proteina anti-apoptotica bcl-2 este investigata extensiv in relatie cu prognosticul si cu predictia raspunsului la terapie al carcinoamelor mamare. Lucrarea de fata analizeaza, pe un lot de carcinoame mamare diagnosticate la femei in postmenopauza, expresia proteinei bcl-2 in corelatie cu alti factori clinici si imunohistochimici (IHC) de prognostic. Material si metoda: Au fost analizate 40 de carcinoame mamare diagnosticate la femei in postmenopauza. Studiul IHC s-a realizat cu anticorpii monoclonali anti-bcl2, anti-ER, anti-PR, anti-Ki-67, tehnica LSAB2, vizualizare cu DAB. Rezultate: Expresia protooncogenei bcl-2 a fost identificata in 55% din carcinoamele mamare invazive. ER si PR au fost pozitivi in 55% si respectiv 57,5% din cazuri. 30% din carcinoame au prezentat un indice Ki-67=10%, in timp ce restul de 70% au demonstrat un indice Ki-67 > 10%. Bcl-2 a fost exprimat in 100% din tumorile T1 si in 64% din tumorile T1+T2. 63,2% din tumorile GI+GII au prezentat reactie pozitiva pentru bcl-2. 77% si 83% din tumorile ER si respectiv PR–pozitive au fost bcl-2-pozitive. 92% din tumorile cu activitate proliferativa joasa au prezentat reactie pozitiva pentru bcl-2. Tumorile N0 au fost in proportie de 67% bcl-2 pozitive, in timp ce in grupul tumorilor N+, 46% au fost bcl-2 pozitive. Concluzii: La femeile postmenopauzale cu cancer mamar, proteina bcl-2 a fost semnificativ exprimata in tumorile de dimensiuni reduse, bine/mediu diferentiate, cu receptori hormonali prezenti si cu activitate proliferativa redusa. Sunt necesare studii mai largi care sa stabileasca utilitatea includerii markerului in panelul de anticorpi ce stabileste profilul molecular al carcinoamelor mamare.

ABSTRACT
Introduction: The bcl-2 antiapoptosis protein is extensively investigated in correlation with prognosis and prediction of therapy response of breast carcinomas. The current study analyzes the expression of bcl-2 protein linked with other clinical and immunohistochemistry (IHC) prognostic factors in a group of postmenopausal women diagnosed with breast carcinomas. Material and methods: 40 breast carcinomas diagnosed on postmenopausal women were analyzed. IHC study was accomplished by monoclonal antibodies anti-bcl2, anti–ER, anti-PR, anti-Ki-67, LSAB2 technic, DAB visualization. Results: The bcl-2 protooncogene expression was certified in 55% of invasive breast carcinomas. ER and PR were positive in 55% respectively 57.5% of cases. Thirty percent of carcinomas presented a Ki-67 index ≤10%, while the rest had a Ki-67 index >10%. Bcl-2 was expressed in 100% T1 tumors and in 64% T1+T2 tumors. In 63.2% of GI+GII tumors, the reaction for bcl-2 was positive. In 77% and 83% of ER respectively PR positive tumors bcl-2 was positive. Furthermore, 92% of low proliferative activity tumors presented positive reaction for bcl-2. In 67% of N0 tumors and in 46% of N+ tumors, bcl-2 reaction was positive. Conclusions: bcl-2 protein in breast carcinomas of postmenopausal women was significantly expressed by small sized, well/intermediate differentiated, hormone-dependent and low proliferative activity tumors. Larger studies are necessary for demonstration of the utility of this marker in the antibodies panel used for characterization of the breast carcinomas molecular profile.

INTRODUCTION

In the past two decades, numerous oncology studies concerned investigations of prognostic valuable factors in malignant tumors with different sites. Classical prognostic factors of breast cancer, represented by tumor size, axillary lymph nodes status, histologic subtype and grade, do not offer enough predictive information for disease evolution and treatment response of all tumors.^1,2 Therefore, researches in the field have focused on a larger range of molecular markers with prognostic and prediction value, that are identified by immunohistochemical (IHC) or biochemical methods, but only few were validated through clinical trials. The breast carcinoma therapeutic planning routinely uses estrogen receptors (ER), progesterone receptors (PR) and HER2/neu oncoprotein determination, but recent data point out the prognostic and predictive importance of some markers involved in apoptosis regulation: p53 protein, bcl-2 family proteins.^3-7
The bcl-2 gene, an anti-apoptosis gene situated on 18th chromosome, encodes a protein involved in programmed cellular death (apoptosis) in various physiologic and pathologic conditions.⁸ The bcl-2 proto-oncogene expression was described in different epithelial tumors (colo-rectal, prostate, pulmonary, etc.), including breast carcinomas.^9-13 The value of bcl-2 identification in breast carcinomas is currently under scrutiny, until now, the majority of researches have signalized a positive relationship between bcl-2 expression and favorable prognostic factors.^14-16
There are certain molecular and therapeutic peculiarities of breast tumors depending on the menopausal status. On one hand, the majority of tumors diagnosed in menopausal women present hormonal receptors, and on the other hand, there are significant differences of adjuvant therapy connected with this menopausal status: tamoxifen vs. aromatase inhibitors.^17,18
The present study analyzes the IHC bcl‑2 protein expression and the relationship with other prognostic factors, on a group of postmenopausal women diagnosed with breast cancer.

MATERIAL AND METHOD

From the Oncology, Surgery and Radiotherapy Services of City Emergency Hospital Timisoara we have selected a group of 40 postmenopausal patients with breast cancer stages I to IV stages. Thirty-six patients underwent histopathological evaluation of axillary lymph node status.
Primary tumor tissue samples after surgical resection or fine needle biopsy were fixed in buffered formalin 24-48 hours, then embedded in paraffin. The standard histopathological examination on HE stain enables the following specifications: the tumor type, the differentiation grade, lymphovascular invasion.
The tumors were diagnosed corresponding to WHO classification of breast tumors, the tumor grade being appreciated base upon the Scarff - Bloom Richardson classification (SBR) modified by Elston and Ellis (grade I – well differentiated tumor; grade II –intermediate differentiated tumor; grade III – poorly differentiated tumor).¹⁹ In situ ductal carcinoma lesions were classified in: “comedo” and “noncomedo”. The tumor pathological stage was specified accordingly to the 6th Edition criteria of AJCC Cancer staging manual.²⁰
The IHC study done on paraffin-embedded tissue, concerned the ER and PR status, the Ki-67 index of cellular proliferation and the bcl-2 anti-apoptosis protein expression. Serial sections, 4 µm thick, were made from paraffin-embedded tumoral representative areas, and then stretched against silanized sides to prevent detachment during antigen unmasking procedures.
From the cases with more paraffin blocks, we selected the tissue section without extended necrotic areas, but with normal mammary gland or benign lesions associated to the carcinoma (for IHC positive control). Tissue sections were deparaffined in xylene, rehydrated in gradual concentrations of ethanol in distillated water, with blocking of endogenous peroxidase with hydrogen peroxide 3%, for 5 min. For unmasking the antigens, we used the heat procedure by boiling 20 min in Target retrieval solution (Dako) pH6 and pH9 in microwaves. Subsequently, we applied immunostain with prediluted DAKO antibodies: anti-Bcl-2 (clone 124), anti-ER (clone 1D5), anti-PR (PgR636), anti-Ki-67 (clone MIB1). The working system used was LSAB2 (labelled streptavidin biotin) – HRP and diaminobenzidine (DAB) was the chromogen used for reactions visualization.
The final reaction product was brown with nuclear site for ER, PR, Ki-67 and cytoplasm and/or membrane for bcl-2. Finally, the sections were counterstained with Mayer hematoxylin, dehydrated, clarified and mounted with Eukitt. Immunostained sections were examined with a Zeiss microscope and images were captured using a Canon digital camera.

IHC reactions results quantification
Only the nuclear stain of the invasive tumor component was considered for ER, PR and Ki-67 regardless the intensity of stain. Randomly chosen 10 high power fields (x400) were evaluated and the final percentage of positive nuclei was the mean value of calculated percentage for each field.
- For ER/PR the threshold was 10% (≥10%) positive nuclei in order to consider ER/PR positive cases.
- For Ki-67 the percentage refers to Ki-67 positive nuclei of the invasive tumor component, using the following semiquantitative score of tumor proliferative activity: 0 – Ki-67 absent; low proliferative activity – positive reaction in ≤ 10% of tumor cells; high proliferative activity – positive reaction in > 10 % Ki-67 positive nuclei.
- bcl-2 expression represented by cytoplasm and/or membrane immunostain was evaluated as follows: negative reaction – absence of tumor cells stain or weak heterogenous positive stain in less than 25% of tumor cells, with an intensity inferior to the lymphocytes’ stain; positive reaction – more than 25% of tumor cells stained, with the same or higher intensity comparative to the lymphocytes’ stain.

RESULTS

Forty patients with breast carcinoma aged between 51 and 80 years old (mean age: 62 years) were included in the study. Only patients with installed menopause at the time of breast carcinoma diagnosis were selected.
The majority of breast carcinomas of postmenopausal women were diagnosed between 55‑64 years old (37.5%) and 65-74 years old (35%) as resulting from Table 1.
From the 40 primary breast tumors, 34 (85%) were ductal invasive carcinomas and 6 (15%) nonductal or mixt carcinomas (infiltrative lobular, medullary, papillary, invasive ductal and lobular).
Due to the primary tumor size, the 40 tumors were designated: pT1- 6 tumors (15%); pT2 – 21 tumors (52.5%); pT3 – 0, pT4 -12 tumors (30%); in one case (2.5%) data on primary tumor dimensions was lacking (pTx). Pathological evaluation of axillary lymph nodes status (in 36 cases out of 40), indicated the presence of lymph node metastases in 24 (60%) of 40 cases. Accordingly to the histologic grade, tumors were classified as follows: GI - 0, GII - 31 (77.5%) and GIII – 9 (22.5%).
Table 2 reveals the main clinical and pathological features of IHC studied cases.

Table 1. Age-related incidence of breast carcinoma (n = 40).

Table 2. Clinical and pathological characteristics of the studied cases.

IHC studied markers expression and their interrelations
Bcl-2
The bcl-2 protein was expressed in 22 of 40 tumors (55%). The bcl-2 expression from carcinomas invasive component had heterogenous distribution and variable stain intensity within the same tumor section, even in the same tumor area. (Fig. 1) The staining pattern was cytoplasmic and membranous, sometimes with perinuclear enhancement. (Fig. 2)

Figure 1. bcl-2 positive reaction in a breast invasive ductal carcinoma. Anti-bcl-2, LSAB2 system, DAB visualization, hematoxylin counte [...]

Figure 2. Cytoplasm and membrane stain for bcl-2. Anti-bcl2, LSAB2 system, DAB visualization, hematoxylin counterstain.

The bcl-2 expression was present in the majority of ductal and myoepithelial cells of nonmalignant peritumoral component, except few cystic dilated ductal structures and apocrine metaplastic ducts, which were non-reactive. In ductal carcinoma in situ “non-comedo” type, foci associated to invasive tumor presented bcl-2 overexpression of variable intensity, while ductal carcinoma in situ “comedo” type was without reaction. The tumor stroma lymphocytes with positive reaction for bcl-2 represented the intern positive control of IHC reaction in cases without bcl-2 reactivity in tumor cells.

ER, PR
ER and PR were identified in the tumor cells nuclei of invasive and intraductal components, together with ductal epithelial cells of nonmalignant component associated to some of analyzed tumors. The majority of cases presented a heterogenous distribution of ER and PR within the same section, and variable stain intensity: from weak to moderate or strong/intense. Some cases had a positive cytoplasmic reaction, in the absence of nuclear stain, which was not counted. ERs were positive in 22 cases (55%) and PRs in 23 cases (57.5%).

Ki-67
The positive reaction for Ki-67 antigen, exclusively intranuclear, presented an important stain intensity variation, heterogenous expression as topographic distribution and diverse stain patterns: fine granular, nucleolus isolated stain, nucleolus and nuclear membrane, intense and complete stain and perichromosome stain.
Twelve (30%) of the 40 analyzed tumors demonstrated low proliferative activity (≤10% nuclei stained for Ki-67), while 28 tumors (70%) were characterized by high proliferative activity (>10% Ki-67 positive nuclei).
Table 3 is presenting the immunohistochemically determined molecular characteristics of the analyzed tumor group.
The results of the analysis concerning the relationship between bcl-2 expression and other prognostic factors reveal that bcl-2 was significantly expressed by small sized tumors (100% T1 and 60% T1+T2) comparative with larger tumors (T3+T4) up to 42% positive reaction for bcl-2. (Table 4)

Table 3. IHC molecular markers of the analyzed tumors (n = 40).

Table 4. bcl-2 expression associated with other prognostic factors (n= 40 cases).

Fifty-six percent of ductal invasive tumors presented bcl-2 overexpressed,
while 50% of nonductal and mixt carcinomas expressed bcl-2. The majority of GIII tumors (78%) had no reactivity for bcl-2, while 77% and 83% of ER respectively PR-positive tumors demonstrated bcl-2 protein overexpression. (Figs. 3 - 5)

Figure 3. bcl-2 expression in an intermediate differentiated invasive ductal carcinoma. Anti-bcl-2, LSAB2 system, DAB visualization,

Figure 4. Intense expression of ER in an invasive ductal carcinoma (same case as previous Fig. 3). Anti-ER, LSAB2 system, DAB visualizat [...]

Figure 5. Intense expression of PR in an invasive ductal carcinoma (same case as Fig. 3). Anti-PR, LSAB2 system, DAB visualization, hema [...]

Figure 6. Proliferation index Ki-67 <10%. MIB1, LSAB2 system, DAB visualization, hematoxylin cuonterstain.

An important percentage of low proliferative activity tumors (92%) showed positive reaction for bcl-2 (Fig. 6), compared to high proliferative activity tumors, which were positive for bcl-2 up to 39%. Analyzing the relationship between bcl-2 expression and tumor differentiation grade, we observed that 64.5% of well and moderate differentiated tumors were positive for bcl-2, in contrast with poorly differentiated tumors, with only 22% reactive for bcl-2. The tumors without lymph nodes metastases (N0) were positive for bcl-2 in 67% cases, while tumors with lymph nodes metastases (N+) were bcl-2 positive in 46% cases.

DISCUSSIONS

The proteins involved in the cellular cycle regulation can provide useful information regarding the initiation, progression and response to therapy of neoplastic disease. Therefore, many recent studies investigated genes/proteins involved in apoptosis regulation. The role of anti-apoptosis bcl-2 genes/proteins family is intensely examined in the breast carcinomas, both in oncogenesis and progression of the disease and in the response to various therapies: hormones, cytotoxic agents, radiotherapy, targeted therapies.^21-24 bcl-2 gene, initially identified in the B-cell follicular lymphomas as a tumor cell apoptosis inhibitor, therefore correlated with unfavorable outcome, seems to be, unexpectedly and apparently unexplainably, associated with a favorable prognostic of the breast cancer.^5,8,16,25
The percentage of bcl-2 positive invasive tumors in our group (55%) is lower compared with other studies results, but closer to the values reported by other authors.^5,6,22-24,26 The review of published data shows that this percentage varies significantly, from 45% to 93%.^5,25 These differences can be explained by the heterogeneity of examined breast tumors and by extremely different methods to quantify the results of IHC reactions: threshold value 10% vs. 25% stained cells, stain score resulting from positive cells percentage multiplied with intensity stain vs. moderate and intense stain22 vs. any intensity stain.^{5,7,16,22-24,26-28} The lower percentage of bcl-2 positive tumors in our study could be connected with the bcl-2 IHC reaction results quantification criteria - positive reaction with the same or higher intensity than lymphocytes stain, over 25% of tumor cells. The IHC bcl-2 protein expression analysis of investigated group revealed a cytoplasmic stain pattern, which is compatible with bcl-2 protein location within external mitochondrial membrane. For the nonmalignant epithelial structures associated to tumor, our results do not confirm the 100% intense stain of ductal epithelial and myoepithelial cells.²² In our study group, a part of ductal structures cystically dilated or with apocrine metaplasia did not present bcl-2 reactivity. The absence of bcl-2 reaction in apocrine metaplasia foci seems to correlate with the reduced percentage of the reported apocrine breast carcinomas with bcl-2 overexpression (Leal et al., quoted by ref. 22). Less investigated was the relationship between bcl-2 expression and the histologic type of the tumor. We identified bcl-2 protein expression both in ductal invasive carcinomas (56%) and in nonductal tumors (50%), aspects also found by other authors (Luna More and al).^22,29 Papadimitriou et al. suggest that bcl-2 could represent a multicentric tumor marker, based on increased bcl-2 expression in multicentric carcinomas, regardless their histologic subtype.³⁰
For the survival estimation and selection of optimal adjuvant therapy, it is extremely important to use prognostic algorithms. The most reliable prognostic models are the TNM system and Nottingham prognostic index (NPI). NPI combines LN status with tumor grade and size, being the best independent predictor of evolution for each patient (NPI good, moderate and poor).³¹ An attempt has been made to improve the prediction accuracy by evaluation of lymphovascular invasion and hormone receptors, but the results were mediocre for patients with good and poor NPI. Recent researches show that bcl-2 represents an independent prognostic factor that improves the predictive value of NPI,16 but these aspects must be confirmed by larger studies that compare bcl-2 value with other markers. Other studies point out that bcl-2 together with ER, PR, Ki-67, c-erb-b2, myc, p53, belongs to a large genes and markers panel implied in the prediction of the tamoxifen treated breast tumors evolution.^16,23 As bcl-2 is an tumor cells apoptosis inhibitor, its overexpression should be associated with an unfavorable prognosis. However, in breast cancer, unexpectedly, it is associated with favorable prognostic factors (small tumor size, low nuclear grade), aspects also confirmed by our results which indicate bcl-2 overexpression in an important proportion of small size tumors (100% of T1 and 60% of T1 + T2), of more differentiated cancers (54.5% of GI + GII tumors shown bcl-2 expression in contrast with 20% of GIII tumors) and in those negative for lymph nodes metastases (67%). Moreover, our data support the idea that increased bcl-2 expression positively correlates with the presence of hormone receptors and with a low proliferative activity.^{6,15,22,24,25} Nowadays there is not a clear explanation for the role of breast cancer bcl-2 overexpression in an apparently favorable prognostic, but is proven that some bcl-2 proteins with anti-apoptosis activity have an inhibitor effect over cellular proliferation.³² Furthermore, it is to be determined if the bcl-2 prognostic value is a consequence of apoptosis involvement or is due to its nonapoptotic abilities.¹⁶
Today, it is largely admitted that there is a close relationship between the bcl-2 immonopositivity and a better therapeutic response: hormones, cytostatics, radiation, with a longer disease free interval and a better survival, but these aspects must be confirm by randomized clinical trials.^7,8,21,22,33

CONCLUSIONS

On a group of postmenopausal women with breast carcinomas, bcl-2 protein was expressed more by the small sized, well and moderately differentiated, hormone-dependent and with low proliferative activity tumors. The data must be confirmed on larger studies, which based upon correlation of the marker expression with evolution and therapy response, should determine the usefulness of including bcl-2 within the antibodies panel design for evaluation of the breast carcinomas molecular profile.

ACKNOWLEDGEMENT

This study was financed by the Research Grant CNCSIS 275/2005.

REFERENCES

1. Goldhirsch A, Glick JH, Gelber RD, et al. Meeting highlights: International Consensus Panel on the treatment of Primary Breast Cancer. J Natl Cancer Inst 1998;90:1601-8.
2. Caldas C, Aparicio SA. The molecular outlook. Nature 2002;415:484‑5.
3. Bast RCjr, Ravdin P, Hayes DF, et al. 2000 update of recommendations for use of tumor markers in breast and colorectal cancer: clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 2001;19:1865-78.
4. Haldar S, Negrini M, Monne M, et al. Downregulation of Bcl-2 by p53 in breast cancer cells. Cancer Res 1994;54:2095-7.
5. Zhang G-J, Kimijima I, Abe R, et al. Correlation between the expression of apoptosis-related bcl-2 and p53 oncoproteins and the carcinogenesis and progression of breast carcinomas. Clin Cancer Res, 1997;3:2329-35.
6. Bottini A, Berruti A, Bersiga A, et al. P53 but not bcl-2 immunostaining is predictive of poor clinical complete response to primary chemotherapy in breast cancer patients. Clin Cancer Res. 2000;6:2751-8.
7. Kroger N, Milde-Langosch K, Riethof S, et al. Prognostic and predictive effects of immunohistochemical factors in high-risc primary breast cancer patients. Clin Cancer Res 2006;12(1):159-68.
8. Daidone MG, Luisi A, Veneroni S, et al. Clinical studies of bcl-2 and treatment benefit in breast cancer patients. Endocr Relat Cancer 1999;6:61-8.
9. Shabaik ASW, Krajewski S, Burgan A, et al. Bcl-2 protooncogene expression in normal, hyperplastic, and neoplastic prostate tissue. J Urol Pathol 1994;3:17-27.
10. Sinicrope FA, Ruan SB, Cleary KR, et al. Bcl-2 and p53 oncoprotein expression during colorectal tumorigenesis. Cancer Res 1995;55:237‑41.
11. Ghosh M, Crocker J, Morris AG. CD40 and Bcl-2 expression in squamous cell carcinoma of the lung: correlation with apoptosis, survival, and other clinicopathological factors. J Pathol 1999;189(3):363-7.
12. Nathan B, Gusterson B, Jadayel D, et al. Expression of Bcl-2 in primary breast cancer and its correlation with tumour phenotype. Ann Oncol 1994;5:409-14.
13. Viacava P, Naccarato AG, Bevilacqua G. Different proliferative patterns characterize different preinvasive breast lesions. J Pathol 1999;188:245-51.
14. Alsabeh R, Wilson CS, Ahn CW, et al. Expression of bcl-2 by breast cancer: a possible diagnostic application. Mod Pathol 1996;9(4):439‑44.
15. Yang Q, Sakurai T, Jing X, et al. Expression of Bcl-2, but not Bax, correlates with estrogen receptor status and tumor proliferation in invasive breast carcinoma. Pathol Int 1999;49:775-80.
16. Callagy GM, Pharoah SE, Pinder SE, et al. Bcl-2 is a prognostic marker in breast cancer independently of the Nottingham Prognostic Index. Clin Cancer Res 2006;12(8):2468-75.
17. Malara NM, Leotta A, Sidoti A, et al. Ageing, hormonal behaviour and cyclin D1 in ductal breast carcinomas. Breast 2006;15(1):81-9.
18. Lin NU, Winer EP. Advances in adjuvant endocrine therapy for postmenopausal women. J Clin Oncol 2008;26(5):798-805.
19. Elston CW, Ellis IO. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long- term follow-up. Histopathology 1991;19:403‑10.
20. Greene FL, Page DL, Fleming ID, et al. AJCC cancer staging manual. 6 th edition. New York: Springer, 2002.
21. Krajewski S, Krajewska M, Turner BC, et al. Prognostic significance of apoptosis regulators in breast cancer. Endocr Relat Cancer 1999;6:29-40.
22. Park S-H, Kim H, Song B-J. Down regulation of bcl-2 expression in invasive ductal carcinomas is both estrogen-and progesterone-receptor dependent and associated with poor prognostic factors. Pathol Oncol Res 2002;8:26-30.
23. Linke SP, Bremer TM, Herold CD, et al. A multimarker model to predict outcome in tamoxifen-treated breast cancer patients. Clin Cancer Res 2006;12(4):1175-83.
24. Lee K-H, Im S-A, Oh D-Y, et al. Prognostic significance of bcl-2 expression in stage III breast cancer patients who had received doxorubicin and cyclophosphamide followed by paclitaxel as adjuvant chemotherapy. BMC Cancer 2007;7:63.
25. Tang S-C, Jessalyn B, Murphy S, et al. BAG-1 expression correlates with Bcl-2, p53, differentiation, estrogen and progesterone receptors in invasive breast carcinoma. Breast Cancer Res Treat 2004;84:203-13.
26. Malamou-Mitsi V, Gogas H, Dafni U, et al. Evaluation of the prognostic and predictive value of p53 and Bcl-2 in breast cancer patients participating in a randomized study with dose-dense sequential adjuvant chemotherapy. Ann Oncol 2006;17(10):1504‑11.
27. Cobleigh MA, Tabesh B, Bitterman P, et al. Tumor gene expression and prognosis in breast cancer patients with 10 or more positive lymph node. Clin Cancer Res 2005;11(24):8623-31.
28. Barbareschi M, Caffo O, Veronese S, et al. Bcl-2 and p53 expression in node-negative breast carcinoma: A study with long-term follow-up. Hum Pathol 1996;27:1149-55.
29. Frolik D, Caduff R, Varga Z. Pleomorphic lobular carcinoma of the breast: its cell kinetics, expression of oncogenes and tumor suppressor genes compared with invasive ductal carcinomas and classical infiltrating lobular carcinoma. Histopathology 2001;39(5):503-13.
30. Papadimitriou CS, Costopoulos JS, Christoforidou BP, et al. Expression of Bcl-2 protein in human breast carcinomas ant its correlation with multifocality, histopathological types and prognosis. Eur J Cancer 1997;33(8):1275-80.
31. Haybittle JL, Blamey RW, Elston CW, et al. A prognostic index in primary breast cancer. Br J Cancer 1982;45:361-6.
32. Reed JC, Miyashita T, Takayama S, et al. Bcl-2 family proteins : regulators of cell death involved in the pathogenesis of cancer and resistance to therapy. J Cell Bioch 1996;60:23-32.
33. Hori M, Nogami T, Itabashi M, et al. Expression of Bcl-2 in human breast cancer: correlation between hormone receptor status, p53 protein accumulation and DNA strand breaks associated with apoptosis. Pathol Int 1997;47:757-62.