This cross-sectional study was conducted from December 2022 to April 2023 at Wahidin Sudirohusodo Hospital and Hasanuddin University Hospital, 2 major referral centers in Makassar, Indonesia. All laboratory analyses were performed at the Department of Anatomic Pathology, Faculty of Medicine, Hasanuddin University.
A consecutive sampling method was used to recruit patients, resulting in a final sample of 74 women. The inclusion criteria were a diagnosis of luminal breast cancer (ER+ and/or progesterone receptor [PR+], HER2−); age between 17 and 65 years; a body mass index (BMI) within the normal range for Asian populations (18.5–22.9 kg/m²); and provision of informed consent. Patients were excluded if they had other concurrent malignancies, if their clinicopathological data were incomplete, or if their formalin-fixed, paraffin-embedded (FFPE) tissue blocks were unsuitable for immunohistochemical analysis due to poor fixation or extensive necrosis.
Clinicopathological data, including patient age, histopathological grade (using the Nottingham Grading System), molecular subtypes, and metastatic status (confirmed by imaging or biopsy), were retrieved from patient medical records. All breast cancer patients included in the study provided informed consent, which included an explanation of the benefits and procedures of the research. Medical history taking was conducted to record patient information and examination results according to the research form prepared. Data in this study were divided based on histopathological grading (I, II, or III) and metastasis (presence or absence). Breast tissue samples were collected from patients under sterile conditions and placed in bottles containing 10% buffered formalin solution. All samples were examined via IHC using the CDK4 (DCS–31) monoclonal antibody to assess the nuclear expression of the CDK4 protein within the tumor cells.
Molecular subtypes were determined based on immunohistochemical results for ER, PR, HER2, and the Ki-67 proliferation index. Each staining batch included external positive and negative tissue controls to ensure staining validity. The subtypes were defined as luminal A–like and luminal B–like according to a previous study.
ER and PR expression was considered positive if distinct nuclear staining was detected in 1% or more of tumor cells, using antibodies from GenomeME Lab Inc (Cat No. IHC423-100 for ER and Cat No. IHC751-100 for PR), respectively. The assessment of HER2 status was conducted in accordance with the latest American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) guidelines using the HER2/neu antibody (GenomeME Lab Inc, Cat No. IHC042-100). Cases were considered positive for HER2 if they exhibited an IHC score of 3+ (strong, complete, circumferential membrane staining in >10% of tumor cells) and negative for scores of 0 or 1+. All cases showing an equivocal (IHC 2+) score were reflexively evaluated with in situ hybridization (ISH) to determine the definitive HER2 gene amplification status.
To determine the Ki-67 proliferation index, tissue slides were treated for 25 minutes with the Ki-67 antibody (clone MIB-1) diluted 1:500. The slides were visualized with diaminobenzidine (DAB) chromogen and counterstained with hematoxylin. The index was calculated as the percentage of positively stained cells among all invasive cells counted. Based on a 20% cutoff point, the results were then categorized as either low Ki-67 or high Ki-67.
Four-micrometer (4–μm) sections were cut from FFPE tissue blocks. The sections were deparaffinized and rehydrated. Antigen retrieval was performed using a heat-induced epitope retrieval method with citrate buffer (pH 6.0). Slides were then incubated with a concentrated primary monoclonal antibody against CDK4 (Clone: DCS–31; Cat No. CMC47829000; Cell Marque, Rocklin, USA). The antibody was optimized in-house through serial dilutions using control tissue to achieve the best signal-to-noise ratio, and a final dilution of 1:100. A polymer-based detection system was applied, and the reaction was visualized using DAB chromogen, followed by counterstaining with Mayer hematoxylin. Human tonsil tissue was used as a positive control for CDK4 expression,
These controls were included in each staining run to ensure specificity, consistency, and reliability of the immunostaining process.
Two independent pathologists, blinded to the clinical data, evaluated CDK4 nuclear staining. The initial interobserver agreement was substantial, with a Cohen κ coefficient of 0.88. Any cases with discordant initial scores were jointly reevaluated by both pathologists using a multihead microscope to reach a final consensus score. A semiquantitative, binary scoring method was used. Positive expression (high expression or overexpression) was defined as tumor cells exhibiting moderate to strong (2+ to 3+) nuclear staining in 25% or more of the tumor area. All other staining patterns were categorized as negative expression (low expression). This threshold was adapted from previously published scoring systems for cell cycle and nuclear proteins in breast cancer that have shown prognostic relevance.
CDK4 expression was assessed semi-quantitatively by evaluating the staining intensity and percentage of stained carcinoma nuclear cell and comparing the results to those of all carcinoma cells. Immunostaining intensity was scored as follows (Figure 1): 0 (negative), 1 (weak), 2 (moderate), and 3 (strong) according to the percentage of area. CDK4 expression was considered overexpressed if staining expression (2+ to 3+) was found in 25% or more of tumor cells in cancer specimens. CDK4 expression was considered not overexpressed if staining expression (0 to 1+) was found in less than 25% of cancer cells in tumor specimens.
Representative Images of CDK4 Immunohistochemical Staining. A, Low expression, showing negative (0) nuclear staining. B, Low expression, with weak (1+) nuclear staining. C, High expression, demonstrating moderate (2+) nuclear staining. D, High expression, with strong (3+) and diffuse nuclear staining in the majority of tumor cells. All images are shown at ×400 magnification.
Data were analyzed using SPSS version 24.0 (IBM Corp). For the purpose of analysis, histopathological grade was dichotomized into low grade (combining grades I and II) and high grade (grade III). This grouping was implemented for 2 key reasons. First, from a statistical standpoint, it ensures more robust analysis by preventing small expected cell counts in the χ2 test, which could otherwise compromise statistical power given the study's sample size. Second, this approach is clinically and biologically meaningful, as Grade III tumors represent a distinct, highly aggressive phenotype, making the comparison against lower-grade tumors a relevant measure of aggressive biology. The association between CDK4 expression (low vs high) and categorical variables (histopathological grade and metastatic status [no vs yes]) was assessed using the χ2 test or Fisher exact test where appropriate. The association between CDK4 expression and the Ki-67 proliferation index (categorized as low [<20%] vs high[≥20%]) was also assessed using the χ2 test. The odds ratio (OR) with a 95% CI was calculated using binary logistic regression to estimate the risk of metastasis. A P value of < 0.05 was considered statistically significant.
To assess whether CDK4 expression was an independent predictor of metastasis, a multivariable logistic regression analysis was performed. Variables with a P value of 0.20 or less in the initial univariable analysis were selected for inclusion in the multivariable model to control for potential confounding.
The final model included CDK4 expression status, histopathological grade (high vs low/moderate), molecular subtype (luminal B vs luminal A), and patient age (≥50 years vs <50 years). The adjusted odds ratio (aOR) with a 95% CI was calculated.