Ultraview detection kit

Immunohistochemical staining was performed using the BenchMark XT autostainer (Ventana Medical Systems, Tucson, AZ, USA) with an UltraView detection kit (Ventana Medical Systems, Basel, Switzerland). Commercially-available antibodies were used for immunohistochemical staining; CD8 (Clone C8/144B, ready to use; Dako, Carpinteria, CA, USA), hMLH1 (Clone M1, ready to use; Ventana, Basel, Switzerland), hMLH2 (Clone G219-1,129, 1:100 dilution; Abnova, Taipei City, Taiwan), hMLH6 (Clone 44, 1:100 dilution; Cell Marque, Rocklin, CA, USA), and PMS2 (Clone A16-4, ready to use; Ventana Medical Systems, Basel, Switzerland). The normal lymph node was used as positive control of CD8 stain. For MLH1, MLH2, MLH6, and PMS2, inflammatory cells and stromal tissue were used as internal positive control.

All of the 21 patients had standard biomarker information including ER, PR, and HER2 status, and Ki-67 for the largest-index tumour (tumour #1); however, IHC had not been performed for tumour #2. Thus, we used the following antibodies that had been used for evaluation of tumour #1 and performed IHC for tumour #2: ER (1:100; clone SP1; Labvision, Fremont, CA), PR (1:70; PgR 636; Dako, Carpinteria, CA), HER2 (ready to use; 4B5; Ventana Medical Systems, Tucson, AZ), and Ki-67 (1:250; MIB-1; Dako). Immunohistochemical staining on representative tissue sections was carried out in a BenchMark XT autostainer (Ventana Medical Systems) using an UltraView detection kit (Ventana Medical Systems).
Immunohistochemical expression of the standard biomarkers was used to categorise the tumour samples into breast cancer subtypes according to the 2011 St. Gallen Expert Consensus as follows^{16 (link)}: luminal A (ER+ and/or PR+, HER2−, Ki-67 <14%), luminal B/HER2-negative (ER+ and/or PR+, HER2−, Ki-67 ≥14%), luminal B/HER2-positive (ER+ and/or PR+, HER2+), HER2-positive (ER−, PR−, HER2+), and triple-negative subtype (ER−, PR−, HER2−). For ER and PR, 1% or greater nuclear staining in tumour cells was considered positive. For HER2, 3+ on IHC or the presence of gene amplification on in situ hybridisation was considered positive.

The data for the CD4+, CD8+, and FOXP3+ TILs and PD-L1+ immune cells were adopted from our previous studies^{11 (link),36 (link)} for 288 cases of DCIS and 339 cases of IBC. Immunohistochemical staining had been performed with a BenchMark XT autostainer (Ventana Medical Systems) using an UltraView detection kit (Ventana Medical Systems). The following antibodies were used: CD4 (clone SP35; ready to use; Dako), CD8 (clone C8/144B; ready to use; Dako), FOXP3 (clone 236A/E7; 1:100; Abcam) and PD-L1 (clone E1L3N; 1:100; Cell Signaling, Danvers, MA, USA).
CD4+, CD8+, and FOXP3+ T cells had been counted in intratumoral and stromal areas as absolute numbers per high-power field. Detailed information on the counting method of TILs is described in the previous studies^{11 (link),36 (link)}. PD-L1+ immune cells were considered to be present when at least 1% of the tumor stromal area was occupied by PD-L1+ immune cells.

Immunohistochemical staining for IDO (Millipore, Billerica, MA, USA), CD68 (Dako, Carpinteria, California, USA), CD163 (Novocastra, Newcastle, UK), CD4 (Novocastra), CD8 (Dako), and FOXP3 (Abcam, Cambridge, UK) was performed on the TMA blocks following a standard protocol using a Ventana Automated Immunostainer (Ventana, Benchmark, Tuscan, AZ USA). After deparaffinization, heat-induced antigen retrieval was performed using citrate buffer, pH 6.0 (CC1 protocol, Ventana). Reactivity was detected using the Ultra-View detection kit (Ventana).

Immunohistochemistry (IHC) was performed using whole‐tissue slides and antibodies against PD‐L1 (SP263; Ventana Medical System, Tucson, AZ, USA), CD8 (C8/144B, 1:200; Agilent Technologies, Santa Clara, CA, USA), and Ki‐67 (clone MIB‐1, 1:100; Dako, Carpinteria, CA, USA). IHC results were obtained using the OptiView Detection Kit (Ventana Medical System) and the UltraView Detection Kit (Ventana Medical System).

Immunohistochemistry for EZH2, SUZ12, BMI1, and H3K27me3 was performed on the tissue microarray blocks following a standard protocol, using a Ventana automatic immunostainer (Ventana, Benchmark, Tuscan, AZ). The primary antibodies used in this study and the specific immunohistochemistry conditions are listed in Supplementary Table 7. After deparaffinization, heat-induced antigen retrieval was performed using citrate buffer (CC1 protocol; Ventana) of pH 6.0. Reactivity was visualized using the Ultra-View detection kit (Ventana). The positive rate for each marker was scored independently by two pathologists (S.H.K and S.O.Y). EZH2, SUZ12, BMI1, and H3K27me3 all showed nuclear expression. The percentage and intensity of positive-stained tumor cells were recorded by manually counting at least 500 tumor cells from representative fields in each case. The cut-off value for high expression of EZH2, SUZ12, H3K27me3, and BMI1 was 75% of tumor cells showing moderate to strong intensity as described in our previous study [25 (link)].