Study outline. This prospective observational study was approved by the Ethics Committee of the University Hospital Center Sestre milosrdnice (EP-1402/18-3). All participants provided a signed informed consent. In the preliminary part, plasma sTGFβrIII levels were determined in a smaller group of healthy women at different phases of the ovarian cycle, and in a smaller group of breast cancer patients at different times of the day (Figure 1). The main study assessed plasma sTGFβrIII concentrations in: (i) healthy women in whom breast lesions were excluded during preventive check-ups; (ii) women with benign breast lesions, as identified by biopsy; (iii) women with breast cancer classified as AJCC stage 0 to IIB. Breast cancer patients were sampled at three time points: 1-10 days before surgery, at 10-30 and 160-180 days after surgery (Figure 1). Part of the removed tissues underwent immunohistochemical (IHC) staining for TGFβrIII (Figure 1).
Preliminary evaluations. In order to evaluate the dependence of plasma sTGFßrIII levels on the menstrual cycle, 13 healthy women provided blood samples (between 09:00 and 11:00 a.m.) between days 2 and 8 of the ovarian cycle (follicular phase), between days 12 and 16 (mid-phase, peri-ovulatory) and between days 21 and 26-28 (luteal phase). To exclude the possibility of diurnal variation, 15 hospitalized women with diagnosed breast cancer (before treatment) provided blood samples at 10:00, 13:00 and 18:00 hours on the same day.
Patient management. A common inclusion criterion was signed informed consent. Women diagnosed with breast cancer (“cases“) AJCC stage 0-IIB were surgically treated. Primary surgical procedures included: mastectomy (with or without primary reconstruction), breast-conserving surgery, sentinel lymph node biopsy, or axillary lymph node dissection. Indication for the type of primary surgery was determined by a multidisciplinary team depending on the size of the tumor, localization, its biology and the patient's preference. Therapeutic goals included complete resection of the primary tumor, with negative margins to reduce the risk of local recurrences, and pathologic staging of the tumor and axillary lymph nodes to provide necessary prognostic information. Based on patient’s age and cardiovascular status, and in line with the tumor characteristics, adjuvant chemotherapeutic protocols were individually selected and commenced between the 4th and 6th postoperative week. Premenopausal women with hormone receptor-positive tumors were treated with tamoxifen, whereas postmenopausal women were treated with aromatase inhibitors. IHC was used to test for human epidermal growth factor receptor 2 (HER2)-positive tumors and borderline-significant specimens were further analyzed by the fluorescence in situ hybridization (FISH). Patients with HER2-positive tumors were treated with trastuzumab for one year. Locoregional irradiation was administered in women who met the radiation criteria. Patient follow-up included in this study occurred at two time points: 10 – 30 days and 160 – 180 days after surgery, i.e. tumor mass removal.
Detection of sTGFßrIII. Blood samples for sTGFßrIII plasma measurements in the main part of the study were taken during a.m. All procedures and evaluation of the results were conducted by researchers blinded to clinical and pathological patient data. sTGFßrIII in plasma was detected using an indirect ELISA kit (Human TGF-beta RIII DuoSet DY242, R&D, Minneapolis, MN), according to manufacturer’s instructions. Results were obtained with a plate reader (Molecular Devices – SpectraMax i3x) at 540 nm. All samples and standards were analyzed in duplicates and samples with an individual coefficient of variation (CV) greater than 25% were retested in duplicates. Tissue sTGFβrIII expression levels in benign and malignant breast lesions (N=10 randomly chosen samples per group) were determined on formalin-fixed paraffin-embedded tissue sections by IHC. Selected tissues were cut in 3-4 μm sections, mounted on slides, and dried at 60 °C for 60 minutes. Dewaxing and target heat retrieval were performed simultaneously in automated PTLink (Dako) for 20 minutes at 97 °C in Target retrieval solution (3 in 1), pH 9.0 (S2367; Dako, Glostrup, Denmark). After blocking peroxidase with 5% hydrogen peroxide for 5 minutes, sections were incubated with a primary murine monoclonal antibody against TGFβrIII, (clone A4, sc-74511, Santa Cruz Biotechnology, USA) at a dilution of 1:50, at room temperature for 60 minutes. Thereafter, a secondary antibody conjugated to horseradish peroxidase (EnVision Flex / HRP high pH; Dako, Denmark) was applied for 50 minutes. Finally, sections were incubated with 3,3'-diaminobenzidine chromogen, contrasted with hematoxylin, and cover-slipped. Benign lesions included fibroadenomas, tubular adenomas, adenosis, usual ductal hyperplasia, and mastitis. Malignant lesions included in situ and invasive cancers. Most invasive cancers were ductal invasive carcinoma (no special type) and one case was a lobular invasive cancer. Adenosis as a benign breast condition served as a positive control, whereas tissue sections stained with murine immunoglobulin G instead of a primary antibody served as a secondary antibody only control. Stained slides were analyzed using an optical microscope (Zeiss Axiostar plus, magnification range 20X and 40X). IHC evaluation was performed by two investigators blinded to clinical and pathologic data (GL, SG) and reconfirmed by a second evaluation by a board-certified pathologist (MPB) blinded to the interpretations of the first set of evaluations. Immunostaining results were compared, and discrepancies were reviewed. There was significant agreement between the three observers (98% correspondence); thus, the pathologist’s scores are presented.
Data analysis. In preliminary evaluations, ln-transformed plasma sTGFßrIII concentrations were analyzed by fitting generalized linear mixed models with time (ovarian cycle phase or time of the day) and age as fixed effects. Plasma concentrations in healthy women, women with benign lesions and women with breast cancer before surgery were compared by fitting a generalized linear model with health status, age and health status*age interaction, and effects were expressed as adjusted mean differences. Plasma sTGFßrIII concentrations in women with breast cancer determined at different time points after surgery were compared to values before surgery by fitting generalized linear mixed models with time point, age, time point*age interaction and AJCC stage as fixed effects. All confidence intervals and P-values were adjusted for multiplicity by the simulation method. We used SAS for Windows 9.4 (SAS Inc., Cary, NC), procedure GLIMMIX.