Trefoil factors peptide-3 is associated with residual invasive breast carcinoma following neoadjuvant chemotherapy

Archival paraffin blocks of invasive breast carcinoma cases before and after neoadjuvant chemotherapy were obtained from the surgical pathology files in the anatomic pathology division at Tawam hospital in Al Ain city, which is the main oncology center in the UAE. In total 133 paraffin blocks of core needle biopsies of invasive breast carcinoma before neoadjuvant chemotherapy and 133 paraffin blocks from mastectomy and lumpectomy specimens of similar cases after neoadjuvant chemotherapy were included in this study. In addition, 16 paraffin blocks of non-neoplastic breast tissue were also be included in this study. The protocol of the present study conformed to the ethical guidelines of the World Medical Association, Declaration of Helsinki, and was approved by Al Ain Medical District Human Research Ethics Committee, (Protocol No. 14/05). Patients or their caregivers signed informed written consent allowing using their anonymous material for research purposes.

Five-micrometer sections were prepared and mounted on aminopropyltriethoxysilane (APES) coated slides. After dewaxing with xylene and rehydrating with graded alcohol, slides were placed in a 0.01 M citrate buffer solution (pH = 6.0) and pre-treatment procedures to unmask the antigens was performed in a microwave oven for 10 min. Then, sections were treated with peroxidase block for 20 min followed by protein block for 10 min. Then the sections were stained using streptavidin-biotin immunohistochemical method for TFF3 (mouse monoclonal, clone EP107, Cellmarque, USA), AKT1 (rabbit polyclonal, Santa Cruz biotechnology, USA), NFқB (rabbit polyclonal, Thermo Fisher Scientific, USA), Bcl2 (mouse monoclonal, clone 124, Cellmarque, USA), Bax (rabbit polyclonal, Dako, Agilent, USA), ki-67 (mouse monoclonal, Dako, Agilent, USA), and cleaved casapase-3 (rabbit polyclonal, Asp175, Cell signaling biotechnology, USA). All primary antibodies were applied on sections with appropriate concentration for one hour at room temperature, followed by washing the sections with phosphate buffer saline (PBS) for 15 min in three changes then After conjugation with primary antibodies, sections were incubated with secondary antibody (EnVisionTM Detection System, DAKO, Agilent, USA) for 20 min at room temperature followed by addition of DAB chromogen (EnVisionTM Detection System, DAKO, Agilent, USA) and counter staining done with haematoxylin. Appropriate positive controls were used. For negative control, the primary antibody was not added to sections and the rest of the protocol was carried out as other sections. Positive and negative controls were used in every batch of slides that were stained (not shown in figures).

Double-Immunofluorescence analysis were done to show co-localization of TFF3 with AKT1, Bcl2 and NFқB in various parts of the tissue sections in different groups. Five-micrometer sections were deparaffinised with xylene and rehydrated with descending concentrations of ethanol. The sections were later incubated with TFF3 (mouse monoclonal, 1:50, Abcam, USA), overnight at room temperature. After 3 times of washing with PBS, sections were incubated with donkey anti-mouse Alexa Fluor 488 (1: 100, Abcam, USA) for 1 h at room temperature. After washing several times, the same sections were incubated overnight at room temperature with a second primary antibodies AKT1 (rabbit polyclonal,1:50, Santa Cruz biotechnology, USA), NFқB (rabbit polyclonal, 1:50, Thermo Fisher Scientific, USA), Bcl2 (rabbit polyclonal, 1:50, Abcam, USA). Sections then were washed in PBS and incubated for 1 h at room temperature with donkey anti rabbit rhodamine (1, 100, Abcam, USA). The sections were then mounted in water-soluble mounting media and viewed with Olympus Fluorescence microscope. Positive and negative controls were used in every batch of slides that were stained (not shown in figures).

Sections were evaluated by using light microscope with X40 objective. The positive tumor cells were expressed in percentage in each case using ImageJ software (http://​rsbweb.​nih.​gov/​ij/​). Cases were considered to be positive when more than 1% of malignant cells are positive for particular antibody. The percentage of positive cells were calculated by dividing the number of positive cancer cells per 100 cancer cells. The calculation was repeated in at least 10 random high power fields and the total were divided by 10 to get the mean in each case. The frequency of expression is graded low when expression of particular protein is seen in 2–25% of tumor cells. The frequency of expression is graded intermediate when expression of particular protein is seen in 26–75% of tumor cells. The frequency of expression is graded high when expression of particular protein is seen in > 75% of tumor cells.

Since our aim in this study is to assess expressions of TFF3 in invasive breast carcinoma before neoadjuvant and after neoadjuvant therapy and whether TFF3 is expressed in residual malignant cells, we have graded the pathological response to neoadjuvant chemotherapy into two groups through reviewing H&E sections on the basis of the parameters used by Chevallier in his study [11].

The Apoptotic index was calculated by dividing the number of positive cancer cells for anti-casapase-3 per 100 cancer cells. The calculation was repeated in at least 10 random high power fields and the total were divided by 10 to get the apoptotic index.

The proliferative index was calculated by dividing the number of positive cancer cells for ki-67 per 100 cancer cells. The calculation was repeated in at least 10 random high power fields and the total were divided by 10 to get the proliferative index.

In total 133 cases of surgically resectable breast carcinomas treated with neoadjuvant chemotherapy were included in this study. The mean age was 46.8 ± 11.8 years; the median was 45 years and the range 26–85 years. The peak incidence was in the fifth decade (Fig. 1).

There was a variable expression of TFF3 by non-neoplastic breast epithelial cells ranging from absence of any expression (Fig. 2a), to low expression (Fig. 2b and c), to intermediate expression (Fig. 2d) and high expression (Fig. 2e). There was also increased expression of TFF3 by epithelial cells forming the lining of cysts in fibrocystic disease of the breast. The fluid of the cyst also shows high TFF3 content (Fig. 2f).

There was a variable expression of TFF3 by malignant epithelial cells in intraductal carcinoma in situ ranging from absence of any expression (Fig. 3a), to low expression (Fig. 3b), to intermediate expression (Fig. 3c) and high expression (Fig. 3d).

There was a variable expression of TFF3 by invasive breast carcinomas. There was no expression of TFF3 in 57 (43%) cases (Fig. 4a). There was variable expression of TFF3 in 76 (57%) of the cases. Low expression of TFF3 is seen in 13 cases (10%) (Fig. 4b), while intermediate (Fig. 4c) and high expression (Fig. 4d) are seen in 32 (24%) and 31 (23%) respectively.

In total 133 cases of breast carcinoma were treated with neoadjuvant chemotherapy.

There was complete response to neoadjuvant chemotherapy with no residual tumor in 33(25%) cases (Fig. 5a and b, while 100 (75%) show incomplete response to neoadjuvant therapy and show residual tumor (Fig. 5c-h).

In total 27 (85%) of breast carcinoma cases that have complete response to neoadjuvant therapy have shown either no expression (Fig. 6a) or low expression of TFF3 in neoplastic cells at pre-neoadjuvant biopsies (Fig. 6c). Only 6 (15%) cases that have complete response to neoadjuvant therapy have shown moderate TFF3 expression. In total, 22 biopsies show no expression of TFF3 in neoplastic cells at pre-neoadjuvant biopsies while 11 biopsies show low or moderate expression of TFF3 in neoplastic cells at pre-neoadjuvant biopsies.

While 58 (43%) cases of breast carcinoma biopsies with incomplete response or no response to neo-adjuvant therapy have shown moderate (Fig. 6e) to high (Fig. 6g) expression of TFF3 in pre-neoadjuvant biopsies and post neoadjuvant biopsies (Fig. 6f and h).

There was a significant correlation between the expression of TFF3 in breast carcinoma cells and response to chemotherapy (p = 0.0165) as 58% (58) of residual carcinomas express TFF3 and 66% (22) of cases with complete response did not express TFF3 in pre-neoadjuvant biopsies (Fig. 7a) (Additional file 1: Table S1). In addition, the cases with residual carcinoma that express TFF3 in resection specimens, they also show expression of TFF3 in pre-neoadjuvant biopsies, since the expression of TFF3 is similar in pre-neoadjuvant core needle biopsies and resection specimens with residual carcinoma. This indicates that TFF3 expression is associated with resistance to chemotherapy as it is highly expressed in residual carcinoma.

Co-localization of TFF3 with Bcl2, p-AKT-1 and NF kappa-B was done by double labeling immunofluorescent technique and it show high expression of Bcl2, p-AKT-1 and NF Kappa-B in neoplastic cells of residual tumors following neoadjuvant therapy which also show high expression of TFF3 (Fig. 8).

There was significant co-expression of TFF3 and BCl2 in breast carcinoma cases treated with neoadjuvant therapy (p = 0.0152) suggesting anti-apoptotic role of TFF3 (Fig. 7b and 8c), (Additional file 1: Table S2).

There was significant co-expression of TFF3 and AKT1 in breast carcinoma cases treated with neoadjuvant therapy (p = 0.0365) which supports the anti-apoptotic role of TFF3 (Fig. 7c and 8f), (Additional file 1: Table S3).

There was significant co-expression of TFF3 and NF Kappa-B in breast carcinoma cases treated with neoadjuvant therapy (p = 0.0243). NF Kappa-B is a growth signaling pathway and when it is stimulated it can also inhibit apoptosis which also supports the anti-apoptotic role of TFF3 (Fig. 7d and 8i), (Additional file 1: Table S4).

There was a significant decrease in proliferative index in post-neoadjuvant breast carcinoma cells when compared with pre-neoadjuvant proliferative activity (P < 0.0001) (Fig. 9a). In many cases of residual carcinoma there was no expression of ki-67 by malignant cells indicating loss proliferative activity by these cells.

There was a significant decrease in apoptotic index in post-neoadjuvant breast carcinoma cells when compared with pre-neoadjuvant proliferative activity (P < 0.0001) (Fig. 9b). In many cases of residual carcinoma there was no expression of cleaved caspase 3 by malignant cells indicating increased antiapoptotic activity in these cells.

There was very low or no expression of ki67 protein (Fig. 10f) in residual tumor as compared to high expression of TFF3 in residual tumor cells (Fig. 10a) suggesting that residual tumor cells have a low proliferating activity, this is supported by the high expression of Bcl2 (Fig. 10b) and p-AKT-1 (Fig. 10c) by residual tumor cells.

There is also a very low apoptotic activity in residual tumor cells supported by very low expression of Bax (Fig. 10d) and cleaved caspase 3 (Fig. 10e) by residual tumor cells and high expression Bcl2 and AKT-1 by residual tumor cells (Fig. 10b and c).