Patients
Participants were selected from patients with oropharyngeal, hypopharyngeal, and uterine cervical cancers at Kawasaki Medical University Hospital and Kawasaki Medical University General Medical Center who consented to have their tumor tissues (stored specimens) used in the study. Furthermore, between 2006 and 2021, 79 patients with oropharyngeal, 44 with hypopharyngeal, and 42 with cervical cancers who underwent surgical treatment or chemoradiation after tissue diagnosis were enrolled in the study. The exclusion criteria included second primary cancer, distant metastasis, and prior irradiation. Patient data were retrospectively collected from electronic medical records. p16 status caused differences in staging; therefore, the histologic type and grade were determined based on the TNM classification (7th edition) defined by the AJCC and UICC [22].
Tissue Preparation and Staining
For pathologic diagnosis, paraffin-embedded sections were prepared and hematoxylin and eosin (H&E) staining and immunohistochemistry were performed.
Immunohistochemistry
After deparaffinization either Target Retrieval Solution (pH 9.0) (Dako, Glostrup, Denmark) or citrated (pH 6.0) buffer at 95°C for 40 min was used for antigen activation in immunohistochemical staining. Antigen activation was performed via heat-induced epitope recovery (HIER). To inactivate endogenous peroxidase activity, the sections were exposed to 3% hydrogen peroxide for 5 min at 15–25°C. Cells were washed with Tris-buffered saline (TBS). Cells were subsequently exposed to the primary anti-bodies, which included: p16 (EPR1473, ab108349, Abcam, Cambridge, UK, dilution 1:500, for 1 h at 15–25°C); ARRDC3 (polyclonal, ab64817, Abcam, dilution 1:100, for 30 min at 15–25°C); and PAR1 (polyclonal, ab32611, Abcam, dilution 1:80, for at 4°C overnight). Sections were rinsed with TBS and incubated with EnVision and the secondary anti-body (Dako, Carpinteria, CA, USA, Cat. No. K4061) for 30 min at 15–25°C. Finally, immunoreactivity was made visible by immersion in 3,3' diaminobenzidine (DAB) for 12 min and counterstained with hematoxylin. As positive controls, p16 represented cervical cancer, and ARRDC3 and PAR1 represented the proximal tubules of normal kidney samples. Negative controls were treated similarly, without the primary antibodies.
Patients with a nuclear expression intensity of ≥ + 2/+3 and a positive distribution of ≥ 75% were p16 positive, based on AJCC. The staining intensities of ARRDC3 and PAR1 were graded into four levels (0, negative; 1, weakly positive [staining intensity lower than the positive control]; 2, moderately positive [staining intensity equal to the positive control]; 3, strongly positive [staining intensity higher than the positive control]). The staining intensity of the entire observed field of view was semi-quantified according to the following expression: H-score (0 to 300 = 0 × of the negative cells + l × of the weak positive cells + 2 × of the intermediate positive cells + 3 × of the strong positive cells). One section per patient was evaluated. The entire tissue section was examined in 200× field of view, and the average H-score of three fields of view was used as the H-score for that case (Fig. 1).
Tumor-Infiltrating Lymphocyte
TIL scoring was determined according to the Immuno-Oncology International TILs Working Group’s definitions [23]. Tumor stromal areas were selected at low magnification. The percentages of stromal areas and mononuclear cells were evaluated at 40× and 100× (200× if necessary) magnifications. Granulocytes and polynuclear cells were excluded and evaluated in accordance with the guidelines of the TILs Working Group using light microscopy (BX53, Olympus, Tokyo, Japan). Tumors were classified into three groups based on TIL infiltration level: Low: 0–10%, Medium: 20–40%, and high: >50%. Regarding prognostic analysis, the three groups were evaluated when divided into low, medium, and high, or when divided into two groups by dividing them into low, medium, and high categories, or alternatively into two groups using the median value as the cutoff (Fig. 2).
Breast Cancer Cell Lines
The MCF-7 and MDA-MB-231 human breast cancer cell lines were provided by Dr. Robert Dickson (Lombardi Cancer Research Center, West Virginia, Washington DC, USA). The breast cancer cell line, KPL-4, was established at the Department of Breast and Thyroid Surgery, Kawasaki Medical University, Kurashiki, Okayama, Japan, [24]. The cells were grown in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) containing 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA) and 0.02% kanamycin (Meiji, Tokyo, Japan) for 3 or 4 d under normal oxygen (5% CO2) conditions and at 37°C.
Western Blotting Assay
For protein extraction, the cells were lysed in Pierce RIPA buffer (Thermo Fisher Scientific). PierceTM BCA Protein Assay Kit (Thermo Fisher Scientific) was used to determine total protein concentration. All proteins were separated by 4–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose membrane (Thermo Fisher Scientific). Membranes were blocked for 1 h at 15–25°C and incubated with primary antibodies 8–16 h at 4°C in blocking buffer consisting of TBS plus 5% BSA and 0.2% Tween 20. Different dilutions of the same antibodies (ARRDC3 1:2000 and PAR1 1:1000) were used for immunohistochemistry. Subsequently, the cells were incubated with a secondary antibody, goat anti-rabbit IgG horseradish peroxidase (Santa Cruz Biotechnology, Dallas, TX, USA, lot no. L2308, dilution 1:10,000), at 15–25°C for 1 h. The membrane was then incubated for 1 h at 15–25°C. An Amersham Imager 680 (Thermo Fisher Scientific) was used to visualize the protein bands transferred to the membrane. As a loading control, β-actin (Sigma-Aldrich, dilution 1:1000) was used.
Statistical Analysis
After dividing the patients into two groups based on p16 expression, the ARRDC3, PAR1, and TIL scores were combined with clinical data and pathological parameters for statistical analysis. ARRDC3 and PAR1 expression levels were evaluated in two groups: high (> median) and low (≤ median), using the median value as the cutoff. TILs were divided into three groups (Low: 0–10%, Medium: 20–40%, High: >50%) and two groups: high (> median) and low (≤ median), using the median value as the cutoff.
The categorical variables are reported as numbers and percentages and were evaluated using the chi-square or the exact Fisher test. The median and 25th and 75th percentiles of continuous variables were calculated using the t-test if they showed a normal pattern of distribution. Otherwise, the Mann–Whitney U test was used. The primary endpoint was disease-specific survival (DSS), and the secondary endpoint was disease-free survival (DFS). DSS and DFS were evaluated using Kaplan–Meier curves and log-rank tests, respectively. Cox regression analysis was used for univariate and multivariate analyses. Oropharyngeal and hypopharyngeal cancers were adjusted for TNM stage (stages I–II vs. III–IV), smoking status, alcohol consumption, venous invasion, lymphatic invasion, duplicate cancer diagnoses, site of origin, and histological differentiation. Cervical cancer was adjusted for TNM stage (stage I–II vs. III–IV), venous invasion, and lymphovascular invasion. The p16-positive and -negative cases were separately analyzed to assess differences in association with the outcome. All tests were two-tailed, and statistical significance was set at p < 0.05.
All the data were analyzed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interfaces for R (R Foundation for Statistical Computing, Vienna, Austria) [25]. To be more precise, it is a customized version of the R program commander that is designed to add statistical functions that are often used in biostatistics.