In a retrospective study, tumor fragments were selected from 63 women aged 30 to 99 years (mean 58 years) attending the Obstetrics and Gynecology service, Hospital de Base, Faculty of Medicine of São José do Rio Preto from 2000 to 2005. The clinicopathological parameters of the patients were obtained from the service medical records.

Patients with no other malignancy who were subjected to mastectomy/quandrantectomy but not to neoadjuvant chemotherapy/radiotherapy were included in the study. All the patients had been diagnosed with invasive ductal carcinoma. A panel of classic prognostic markers were evaluated: estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor-2 (HER-2/neu), p53 tumor-suppressor gene and the cell proliferation marker, Ki-67. Furthermore, the patients were divided in to 4 subgroups according to the immunohistochemistry of the following markers: luminal A (ER⁺PR⁺HER-2/neu⁻), luminal B (ER⁺PR⁺HER-2/neu⁺), HER-2/neu⁺, and triple-negative (ER⁻PR⁻HER-2/neu⁻).

The parameters for histopathological evaluation were those recommended by the International Union Against Cancer (UICC) TNM classification of 1989. Among the entire group, 37 patients (59%) who had received the same treatment including surgery and adjuvant chemotherapy/radiotherapy [5-fluorouracil/epirubicin/cyclophosphamide (FEC)] were designated as group I; 19 patients (30%) who had received only adjuvant chemotherapy were designated as group II and 7 patients (11%) who had received no treatment after surgery were designated as group III. The patients underwent clinical evaluation every 3–4 months. Local tumor recurrence and metastasis in such cases were carefully monitored, as well as the date and cause of death in order to record the survival time of each patient.

Patient follow-up ranged from 120 to 2,704 days, with a median of 1,473 days. For analysis of survival curves, the patients were divided into higher and lower enzyme expression groups using a cut-off value established by the receiver operating characteristic curve (ROC). Only patients who died of tumor disease were included in the final analysis. Patients with an incomplete follow-up were excluded. The study was approved by the Research Ethics Committee of the Faculty of Medicine of São José do Rio Preto (protocol no. 25372009).

For administration of the technique, blocks of paraffin-embedded tumor fragments were cut to provide 3-μm sections. The sections were prepared on silanized slides and subsequently deparaffinized, rehydrated through graded alcohol and incubated with 3% hydrogen peroxidase for 30 min to block endogenous peroxidase activity. Antigen retrieval was carried out in a Pan Steam (ARNO, São Paulo, Brazil) at 95°C with buffer for 35 min for each specific antibody (Table I). After cooling, the slides were covered with bovine serum albumin (BSA) solution for 30 min and incubated at 4°C overnight with the antibodies described in Table I.

Subsequently they were washed with phosphate-buffered saline (PBS) for 15 min and incubated with the Easy Path kit (Erviegas, São Paulo, Brazil), which consisted of the secondary antibody biotinylated anti-mouse, rabbit, and goat immunoglobulins for 1 h and the streptavidin-peroxidase complex for 30 min, followed by washing with PBS for 15 min. We applied 0.5% 3,3′-diaminobenzidine tetrahydrochloride (DAB; Signet Laboratories, Dedham, MA, USA) to the slides for 2–5 min at 20–22°C. The slides were counterstained with Harris’s hematoxylin for 40 sec. Negative controls were obtained by omitting the primary antibody, and liver or prostate tissue served as an internal positive control in every assay.

To assess the immunoexpression of GSH and GPX, multiple fields were examined on each slide, particularly demarcated areas with distinct brown staining. The slides were photographed and the enzymes were quantified by AxioVision software and by observation through a ×40 lens of a Zeiss Axioskop2 microscope.

For each sample, three regions of tumor tissue were selected and 20 spots of the tumor cells were marked within each region, thereby analyzing 60 different spots of every sample to average the relative intensity of immunoreactivity. The values were obtained as arbitrary units (au), and the mean optical density (MOD) indicated the specific staining intensity in the immunoreactive areas.

The staining of HER-2/neu was considered by counting the number of positively stained cells in the membrane and was expressed as a percentage of the total tumor cells. According to ASCO/CAP guidelines (21) the following categories were defined: 0, no immunostaining; 1+, weak or incomplete staining of the membrane in any proportion of tumor cells; 2+, complete membrane staining or weak staining in 10% of the tumor cells stained; 3+, uniform and intense staining of the membrane in more than 30% of tumor cells. Tumors scored with 0 or 1+ were considered negative and those with scores 2+ or 3+ were considered positive. Tumors were categorized into two groups in relation to cellular proliferation according to the staining of Ki-67 based on the count of the number of positively stained cells and expressed as a percentage of the total: 0 or 1, no cellular proliferation; 2 or 3, the presence of cellular proliferation. The same parameter was used in the classification of the ER, PR and p53 tumor-suppressor gene: 0 or 1, no staining; 2, 3 or 4, presence of staining.

The patients were separated into groups according to the clinicopathological variables. The averages of the densitometric analysis referring to the densitometry results and the quantification of the different breast tumor groups were compared by Student’s t-test or ANOVA, followed by the Bonferroni test. The values are expressed as means ± SEM.

The cut-off for the risk of death was determined by ROC. Survival curves were plotted by the Kaplan-Meier method, and the differences between the curves were evaluated by a log-rank test and hazard function. Multivariate logistic regression was performed to evaluate the factors that influenced death. For all tests, P<0.05 was considered to indicate a statistically significant result. All analyses were performed using GraphPad Prism 4 and StatsDirect software.

Twelve core biopsies or tumor fragments >2 cm were collected by surgery from the breast cancer patients who were diagnosed with ductal carcinoma, who had not undergone chemotherapy and/or radiation therapy prior to the point of collection. The tumor fragments were washed in PBS containing 1% penicillin/streptomycin, cut into microfragments with a scalpel and incubated at 37°C in 5% CO₂ with RPMI-1640 supplemented with 20% BSA, 1% penicillin/streptomycin and 1% L-glutamine. The cells were cultivated until they reached 80% confluence and were subjected to immunocytochemistry using the primary antibodies anti-cytokeratin, anti-vimentin and anti-calponin to confirm epithelial origin.

Cells were divided into two groups: control (no treatment), and cells treated with 0.2 mg routine chemotherapy (doxorubicin) for 24 h. At the end of the treatment, cell viability was verified by cell counting in a Neubauer chamber (LaborOptik, Bad Homburg, Hessen, Germany) with trypan blue dye (0.4%).

Total RNA was extracted from the cell culture with TRIzol (Invitrogen Life Technologies, São Paulo, Brazil) as recommended by the laboratory and the total RNA from each sample was reverse transcribed to complementary DNA (cDNA) using a High Capacity cDNA kit (Applied Biosystems, Foster City, CA, USA).

Quantitative real-time polymerase chain reactions were performed in triplicate using StepOnePlus™ System (Applied Biosystems). Briefly, the reactions were performed in a 20 μl volume with 10 μl of Power SYBR^®-Green PCR Master Mix (Applied Biosystems), 250 nM of each primer, and 10 ng of cDNA. PCR conditions consisted of 50°C for 2 min, 95°C for 10 min, followed by 35 cycles of 95°C for 15 sec and 60°C for 1 min. Following PCR, dissociation curve analysis was performed to confirm the desired single gene product: one cycle of 95°C for 15 sec, 60°C for 1 min, and 95°C for 15 sec was carried out.

Each transcript level was normalized by division with the expression values of HPRT1 used as an internal control. Gene expression stability was analyzed by geNorm software. The transcript level was calculated using the 2^−Ct method (22). The Ct was the difference between the threshold cycles of the target and the internal control and Ct was the difference between the average Ct of the sample and the average Ct of the calibrator sample. The fold difference (relative abundance) was calculated using the 2^−ΔΔCT formula and plotted as mean ± SD of the triplicate reactions. At least 3-fold differences were considered significant.

A negative control was included in each reaction, and one sample was chosen for reaction calibration. Experiments were repeated when the coefficient of variation was higher than 5%. After each reaction, the products were analyzed on 2% agarose gel stained with ethidium bromide.

The gene was searched, selected in PubMed (http://www.ncbi.nlm.nih.gov/entrez), and synthesized from human messenger RNA already sequenced and confirmed. Its design was made using the program PRIMER 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.Cgi). The HPRT1 gene was used as an endogenous control. Primers used for amplification include: GSS foward, 5′-TGCTAAAGCCCCAGAGAG AG-3′ and reverse, 5′-AGCAGGCAATTCTCAAAAGG-3′; GCLC foward, 5′-GCACAACGTTCTCAAGTG-3′ and reverse, 5′-TGGTTTGGGTTTGTCCTTTC-3′; GPX foward, 5′-ATGGCGCAATTGTCCAAG-3′ and reverse, 5′-CTGGCC TCCCCTTACAGTG-3′; HPRT1 foward, 5′-TTATAGTCAAG GGCATATCC-3′ and reverse, 5′-AGCTTGCTGGTGAAA AGGAC-3′.

The genes were classified as underexpressed (samples with less than -1 log3 measurement) and overexpressed (samples with quantification greater than 1 log3).