Immunohistochemical analysis of MAGE-A4, NY-ESO-1 and HER-2 antigen expression in patients with relapsing invasive ductal breast cancer
Daniela Bandić, Antonio Juretić*, Božena Šarčević, Viktor Šeparović, Mirjana Kujundžić Tiljak1,Tvrtko Hudolin2,Giulio C Spagnoli3, Dinko Čović, Mirko Šamija
UniversityHospital for Tumors, Zagreb, Croatia
1Department for Medical Statistics, Epidemiology and Medical Informatics, Andrija Stampar
School of Public Health, MedicalSchool, University of Zagreb, Zagreb, Croatia
2ClinicalHospitalCenter „Zagreb“, Zagreb, Croatia
3Department of Surgery, Research Laboratory, UniversityHospitalBasel, Switzerland
*Present address: Clinical Hospital Center „Zagreb“, Zagreb, Croatia
Short title: Expression of antigens MAGE-A4, NY-ESO-1 and HER-2 in breast cancer
Correspondence to:
Professor Antonio Juretić, MD, PhD
radiation oncologist
Clinics for Oncology,
Clinical HospitalCenter „Zagreb“
Kispatićeva 12
HR-10000 Zagreb, Croatia.
e-mail:
Summary
Objective: The aim of the study was to evaluate the possible prognostic role of the expression of cancer/testis antigens MAGE-A4 and NY-ESO-1 in patients with invasive ductal breast cancer.
Design:Expression of these antigens together with the HER-2 antigen was evaluated immunohistochemicaly (IHC) on archival paraffin embedded breast cancer tissues from 81 patients. The patients were initially selected on the basis of timing of tumor relapse. Initially all patients (T1 to T3, N0 to N1, M0 tumors) and as adjuvant therapy they all received postoperative irradiation and, if indicated, systemic therapy (chemotherapy and hormonal therapy).The patients who were disease free after a five years of follow up (n=23) were compared with the ones who either had locoregional relapse (n=30) or bone metastases (n=28). After a ten years of follow up patient’s survival was also evaluated.
Results:The three groups of patients were comparable in terms of age, type of operation, tumor size, tumor grade, number of metastatically involved axillary lymph nodes, Nottingham prognostic index (NPI), progesterone receptor (PR)status, and adjuvant hormonal therapy. The “five-year” relapse-free group of patients had a significantly higher number of patients with positive estrogen receptors (ER; p=0.03) and it received significantly less adjuvant chemotherapy (p=0,00002).Expectably, this group also had a a significantly better ten year survival(p0.00000).The three groups of patients did not differ in the NY-ESO-1 or HER-2 expression but in the group of patients with locoregional relapse there was a a statistically significant lower number of patients expressing MAGE-A4 antigen (p=0.006).Furthermore, in all groups MAGE-A4 antigen expression was significantly associated with the NY-ESO-1 antigen expression (p=0,005) but not with tumor size and grade, number of metastatically involved axillary lymph nodes, and the ER and PR status.MAGE-A4 positive patients showed significantly improvedsurvival had the than the MAGE-A4 negative patients (p=0.04). This was not observed regarding the NY-ESO-1 and HER-2 antigen.
Conclusions: The obtained results suggest that the MAGE-A4 antigen might represent a tumor marker of potential prognostic relevance.
Key words:Breast cancer, cancer / testis (C/T) antigens, HER-2 immunoreactivity, immunohistochemistry, MAGE-A4 immunoreactivity, NY-ESO-1 immunoreactivity, tumor antigens.
Introduction
Breast cancer is the most common malignancy in women and its clinical course may vary from indolent and slowly progressive to rapidly metastatic disease. Identification of prognostic and predictive factors that reflect the biology of breast cancer is important for refining our assessment of prognosis and the selection of patients who may benefit from adjuvant and/or systemic therapy. When choosing among prognostic factors suitable for clinicaluse, it is also important to consider aspects such astheir availability, reproducibility, and cost. In routine clinical practice, the standard prognostic factors such as age, menopausal status, tumor size, tumor grade, steroid-hormone receptor status and nodal metastases form the basis for treatment decisions and selection of treatment modalities for individual patients (1-5).
Variability in breast cancer clinical course is undoubtedly related at least in part to tumor cell growth rate and other features such as invasiveness or metastatic potential. Research in molecular biology has identified genes and their products involved in or associated with the malignant celltransformation and behavior. Moreover, expression of some of these molecules, such as p53, Ki-67, nm23, catepsin D, Ep-CAM, HER-2, urokinase-type plasminogen activator and its inhibitor, is usually found to be also associated with the patient’sprognosis. Since it seems that many genes and molecules mightbe involved in, respectively, malignant transformation and malignant cell behavior, additional other molecules can also be tested as potential prognostic factors (1,2,6-11).
The cancer/ testis (C/T) genes encode tumor associated antigens (TAA) found in various tumors of different histological origins, but not in normal tissues other than testis. Their physiological function is unknown. Peptides derived from these antigens could be used as targets for active immunotherapy. Expression analysis of these genes or of their products in malignancies could also be of potential diagnostic and/or prognostic relevance (12,13).We present here therefore data on the immunohistochemical expression of antigens MAGE-A4 and of NY-ESO 1. Since the HER-2 antigen has a prognostic and predictive role it was also included into our analysis.
Patients and methods
This is a retrospective study which included 81 patientswho were diagnosed with the invasive breast cancerwithout distant metastasis (pT1-3pN0-1M0). Patients were identified retrospectively on the year 2000 from the medical records at the Department of Radiotherapy, UniversityHospital for Tumors, Zagreb, Croatia. Before the inclusion into the study they all already had primary surgical treatment and adjuvant treatment: radiotherapy in all patients and if indicated systemic treatment (chemotherapy and hormonal therapy).
Patients were selected based on a five year disease (relapse) free period or relaps for which irradiation was used (medical records at the Department of Radiotherapy). Therefore, the year 1995 was chosen as the year for the retrospective identification of patients without breast cancer recurrence (five-years of follow up). For the disease relapse patients having either bone metastases or locoregional relapses were selected. To have approximately a similar number of patients in these two groups (between 20 to 30), additional patients diagnosed in other calendar years were then also included. Locoregional recurrence was defined as the first recognized recurrence in the chest wall or breast or axilla and supraclavicular region. Before reirradiation some of these patients were also treated surgically (reoperation). Patients with bone metastases might also have metastases at other sites. Moreover, they might also have received systemic anticancer treatment. Patients’ survival analysis was performed in the year 2005. Their survival data were also checked in the Croatian national cancer registry and rechecked by making, whenever possible,personal telephone callsto the presumed alive patients. The study protocol was approved by the Ethics Committee of the Hospital.
The following clinical, pathological, and laboratory data were presented and analyzed (Table 1): age, year of diagnosis, type of surgical operation, median time of disease relapse, tumor size, histological grade, axillary node status, Nottingham prognostic index (NPI), estrogen and progesteron receptor positivity, administration of adjuvant chemotherapy or of adjuvant tamoxifen and the patients’ survival.
Pathological examination of primary tumors andaxillary lymph nodes was performed in the same hospital at the Department of Pathology. Forroutine histological analysis resected material was fixed in 10% buffered formalin, embedded in paraffin and stained with haematoxylin and eosin. The histological grade of tumors was determined according to the method by Elston (14). Tumors were divided into three groups on two criteria, regarding their tumor size (0,1-2 cm (pT1), 2,1-5 cm (pT2), more than 5 cm (pT3)) and regarding the ipsilaterally axillary lymph node status (patients without positive lymph node (N0), and positive lymph nodes (N1)) (15). NPI scores were calculated according to Rampault et al. (16): NPI =0.2 x tumor size (cm) + lymph-node stage (1, 2 or 3) + histological grade (1, 2 or 3), wheresize is measured in centimeters; lymph node stage 1 is lymphnode–negative, stage 2 is one to three positive lymphnodes, stage 3 is ≥ four positive lymph nodes; and the scoringof histologic grade is 1 to 3 (see below). For prognostic considerations,NPI was categorized into three groups: low (good prognosis), NPI ≤ 3.4; intermediate (moderate prognosis),NPI 3.41 to ≤ 5.4; and high (poor prognosis), NPI > 5.4. Concentrations of estrogen and progesteron receptors in tumor cytosol were evaluated by the dextran-coated charcoal assay as described (17). For estrogen receptors (ER) and progesterone receptors (PR), levels of 5 fmol/mg of protein or more and of 10 fmol/mg of protein or more, respectively, were considered to be positive (17).
Adjuvant therapies were based and prescribed according to the University Hospital Treatment Protocol for Breast Cancer (unpublished document for internal usage). Adjuvant radiotherapy consistedof external megavoltage irradiation delivered from the linear accelerator(18).Adjuvant chemotherapy included either the “CMF protocol” (cyclophosphamide, 600 mg/m2 i.v. on day 1; methotrexate, 40 mg/m2 i.v. on day 1; 5-fluorouracil, 600 mg/m2 i.v. on day 1) or the “FAC protocol” (5-fluorouracil, 500 mg/m2 i.v. on day 1; doxorubicine, 50 mg/m2 i.v. on day 1; cyclophosphamide, 500 mg/m2 i.v. on day 1). Cycles were repeated every 3 weeks, 6 cycles in total. Tamoxifen was administered as systemic hormonal therapy (2x10 mg over 5 years) (19).
Immunohistochemistry (IHC)
Expression of MAGE-A4 or of NY-ESO-1 tumor associated cancer/testis antigens in primary breast cancer tissue was studied by using “57B” monoclonal antibody (mab)(20) or “B9.8.1.1” mab (21), respectively. Briefly, tissue sections from paraffin-embedded breast tumor samples (0,5 mm thick) were placed on Silane (3-aminopropyltriethoxysilane, A 3648, Sigma, St Louis, MO, USA) treated microscope glass slides. After deparaffinization the sections were heated in a 800 W household microwave oven at maximum power for 8,5 and 5 minutes in 10 mmol/L citric buffer (pH 6,0) and washed with phosphate buffered saline (PBS; pH 7,2). The sections were H2O2 treated in order to suppress endogenous peroxidase activity. Following an additional PBS wash the sections were also incubated for 20 minutes with 1:10 diluted normal rabbit sera (DAKO X0902, DAKO A/S, Glostrup, Denmark) at room temperature in a humidified chamber to prevent nonspecific immunoglobulin binding. Mab “57B” or mab “B9.8.1.1.”, in the form of undiluted hybridoma supernatant, were applied to themfor 90 minutes at room temperature. Specific binding was revealed by using a streptoavidin-biotinylated horseradish peroxidase based detection system (DAKO K 0355, DAKO A/S, Glostrup, Denmark) (22,23).
Immunoreactivity was scored in the following way: 0, no positive tumor cells; +, up to20% positive cells („mild reaction“); ++, 21 to 50% positive cells („middle strong reaction“); +++, over 50% positive tumor cells (strong reaction). In all specimens non neoplastic cells, such as normal ductal epithelial cell, fibroblasts, etc., were indeed present but were not stained, and thus served as internal negative controls (22,23). In tables, for the sake of simplicity, we presented the immunoreactivity scores only as either “negative” or “positive” which encompasses mild, middle and strong reactions.
DAKO Hercept TestTM kit (FDA-approved reagent) was used for the HER-2 immunohistochemical staining in accordance with producer instructions. Samples with staining intensity score 3+ (standard control slides were included in the Hercept TestTM kit) were considered to be HER-2 positive. When using the HERCEP test scoring system a strong positive reaction implied a complete(diffuse) membrane staining in more than 10% of tumor cells (24).
Statistical Analysis
The complete statistical analysis was performed by the use of the Statistics 6.1 software package (StatSoft, Inc.; Tulsa, USA). The following statistical tests were used: frequency tables, crosstabulation tables, ANOVA (analysis of variance) and survival analysis. The p value of less than 0,05 was considered as statistically significant.
Results
Patients characteristics
The characteristics of the 81 patients studied are listed in Table 1. The patients’ age ranged from 36 to 82 years. The majority of patients were diagnosed with breast cancer in the year 1995 and modified radical mastectomy was the predominant operational procedure. In the group of patients with locoregional relapse the median time of relapse was 18 months (3-49 months), while in the group with bone metastases it was 13 months (2-44 months) (p=0.766). When these three groups of patients were compared regarding the standard clinical parameters no statistically significant difference was found in any parameter exceptin ER status (p=0.03) and in adjuvant chemotherapy (p=0.00002). In the five relapse-free group of patients there was a significantly higher number of patients with the positive ER. Moreover, this group also received significantly less adjuvant chemotherapy.This group of patients (as expected) had also significantly better survival outcome (p0.00000).After the ten years of follow up (the year 2005.) 14 out of 23 patiens were alive (61%; 2 patients with uknown survival status). In the other two groups of patiens only one patien per group was alive.
IHC results
Within our panel of monoclonal antibodies, the mab “57B” (20) recognizes a number of related MAGE-A gene products, including MAGE-A2, MAGE-A3, MAGE-A4, MAGE-A6, and MAGE-A12 (25). In paraffin-embedded specimens however it has been shown to predominantly recognize MAGE-A4 TAA (26). The mab “B9.8.1.1” is specific to NY-ESO-1 TAA (21) and the monoclonal antibody in the DAKO Hercept TestTM kit is FDA approved as thespecific reagent for the HER-2 detection. Representative examples of immunohistochemical staining obtained with these mab are presented in Figure1. Immunohistochemical staining of cancer-testis protein, MAGE-A4 and NY-ESO-1 is predominatly visible as cytoplasmatic staining limited to tumor cells. HER-2 staining is visible asHER-2 staining is visible as membrane staining in normal and tumor cells.HER-2 positivity is based in relation to the percentage of tumor cells and intensity ofmembrane staining.The scoring system was the following: negative reaction (-) - up to 10% of tumor cells with negative or weak membrane staining, weak reation (+) - more than 10% tumor positive cells but partial membrane staining, moderate reaction (++) - more than 10% tumor positive cells with moderate membrane staining and strong reaction (+++) - strong positive completly membrane staining in more than 10% of tumor cells. Theresults from the immunohistochemical expression of MAGE-A4, NY-ESO-1 C/T and HER-2 protein in the three analyzed groups of patients are presented in table 2. The overall positivity of mab “57” and of mab “B9.8.1.1” was 60/81 (74,0%) and 32/81 (40%), respectively. A statistically significantly lower expression of MAGE-A4 antigen was detected in the group of patients with locoregional relapse (p=0.013). No significant difference was observed when these three groups of patients were compared for NY-ESO-1 antigen expression. Positive HER-2 reaction was found in 18/81 patients (22.2% ) with equal distribution in all three groups.
Survival of patients was also analyzed in the relation to the expression of MAGE-A4, NY-ESO-1 and of HER-2 antigen. It was found that MAGE-A4 positive patients had significantly better survival than MAGE-A4 negative patients (p =0.04569; figure 2).In contrast NY-ESO-1 and HER-2 antigen expression was not correlated with survival (data / figures not shown).
Relation of MAGE-A4 antigen to the standard prognostic and predicitive factors
Since the three groups of breast cancer patients analyzed differed significantly in MAGE-A4 antigen expression,the relationship between MAGE-A4 antigen and standard prognostic and predictive factors was analyzed (Table 3). MAGE-A4 antigen expression was found to be associated to a significant degree only with the NY-ESO-1 antigen expression (p=0,006), but not with tumor size and grade, number of metastatically involved axillary lymph nodes, or with ER and PR status. Such results suggest that in our analyzed sample the MAGE-A4 antigen might behave as a prognostic factor unrelatedto the above standard prognostic and predictive factors.
Discussion
Breast cancer biology is complex, with multiple factors contributing to breast cancer development, tumor growth, and metastatic progression. The clinical data of the follow-up and studies of the biology of breast carcinoma can be used, for example, in order to identify parameters which could serve as prognostic or predictive factors. Decision making is usually based on a combination of clinical and tumor characteristics, such as age, tumor size and histology (type, grade), lymph-node and ER and PR status (1-5). However, since the prognostic value of these criteria is imperfect, it is apparent that additional and still unidentified molecular factors influence and determine the clinical course of the breast cancer disease. By identifying these additional factors therapeuticapproaches to patients with breast cancer could be further individualized thus increasing both their survival rate and quality of life. Novel high-performance screening methods, such as the DNA microarray, analyzing simultaneously in a single experiment the expressions of thousands of genes in a tissue may allow the identification of disease subsets that correlate with clinical outcomes. Clearly such gene-expression profiling (holistic approach) will provide highly useful prognostic information but at the moment there is still no routine clinical use of this new technology as of yet (6-11).
Since detection of MAGE-A4 and NY-ESO-1 antigen expression is based on IHC expression, as a control IHC detection of HER-2 antigen expression was employed. HER-2 molecule belongs to a family of four homologousreceptors involved in the tyrosine kinase-mediated regulation of normal breast tissue growth and development. Overexpression of HER-2 molecule in breast cancer cells was associated with poor prognosis (7,8,24). C/T TAA antigens were discovered in the 1990s, initially as targets in CD8 T cell recognition of autologous human melanoma cells (27). To date, 44 C/T gene have been identified and their expression has been studied in numerous cancer types. Briefly, they code for products with the following characteristics: (i) mRNA expression in normal tissues appears to be restricted to testis, fetal ovary, and placenta. (ii) mRNA expression in cancers of diverse origin is common - up to 30 - 40% of a number of different cancer types, e.g., melanoma, bladder cancer, sarcoma express one or more C/T antigens. (iii) The X chromosome codes for the majority of C/T antigens, but a number of more recently defined C/T coding genes have non-X chromosomal loci. (iv) The function of most C/T antigens is unknown, although some role in regulating gene expression appears likely. One possibility to account for the aberrant C/T expression in cancer relates to the global demethylation associated with certain cancers. Another important issue is whether expression of these genes in the cancer cell contributes to its malignant behavior. (v) There is increasing evidence that C/T expression is correlated with tumor progression and takes place in tumors of higher malignant potential (28-30)