Inhibition of Interleukin-8/C-X-C Chemokine Receptor 2 Signaling Axis Prevents Tumor Growth and Metastasis in Triple-Negative Breast Cancer Cells Open Access

Breast cancer is the most commonly diagnosed cancer, accounting for 32% of all cancers in women [1]. Over the past decade, the incidence rates of breast cancer in young women have increased steeply, and it has also been the leading cause of cancer deaths in women aged 20–49 [1]. In particular, triple-negative breast cancer (TNBC) is characterized by its lack of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression. Accounting for approximately 15–20% of all breast cancer cases, it is recognized as a heterogeneous and malignant disease marked by a challenging clinical prognosis [2]. Renowned for being the most aggressive and invasive breast cancer subtype, the scarcity of effective therapeutic targets for TNBC has spurred a wealth of research endeavors [3]. Efforts to tailor therapies for TNBC through the use of biomarkers have been pursued [4]. Investigating such biomarkers offers a critical theoretical framework for understanding drug efficacy and resistance mechanisms in TNBC patients [5]. Several types of solid tumors, including TNBC, are known to drive cell proliferation and invasion via IL-8 expression, positioning IL-8 as a prognostic and potentially predictive biomarker [6, 7].

Among chemokines, IL-8, also designated as C-X-C motif ligand 8, is identified as a small, soluble protein [8]. IL-8 has been implicated in various cancer types, including those of the breast, lung, melanoma, prostate, and pancreas-colon continuum [9, 10]. It exerts a direct influence on tumors by modulating the inherent characteristics of breast cancer cells [11]. Notably, IL-8 has been highlighted for its significant role in enhancing cancer cell proliferation, angiogenesis, infiltration, and metastatic potential [12]. Indirectly, IL-8 may foster tumor growth and metastasis by reshaping the tumor microenvironment [11]. Moreover, IL-8 is crucial in mediating the differentiation and function of stromal and immune cells within the tumor microenvironment, thereby influencing the effectiveness of immunotherapy [13]. Clinical investigations have established a correlation between serum IL-8 levels and tumor stage, invasiveness, and patient prognosis [14, 15].

IL-8, alongside its receptors C-X-C chemokine receptor 1 (CXCR1) and C-X-C chemokine receptor 2 (CXCR2), has been shown to encourage tumorigenesis in breast cancer cells [16]. Expressed within these cells, CXCR1 and CXCR2 are pivotal in activating the inflammatory response through their chemokines [17] and in modulating tumor cell proliferation, metastasis, and angiogenesis across various cancer types [18, 19]. Despite sharing 76% homology, CXCR1 and CXCR2 display distinct functionalities owing to differential regional regulation [20]. IL-8 has been recognized as an oncogenic factor in several cancers, triggering the activation of Src, signal transducer and activator of transcription (STAT-3), and extracellular signal-regulated kinase (ERK)1/2 pathways. The IL-8/CXCR1 axis, in particular, has recently been highlighted as a promising therapeutic target for combating breast cancer stem cells [21, 22].

This study aimed to explore the pharmacological implications of CXCR1 and CXCR2 inhibition in TNBC cell lines. Prior research has elucidated the role of IL-8 in the progression of breast cancer, with aberrant expression of IL-8 linked to TNBC patient survival rates. Consequently, our focus has been drawn toward CXCR1 and CXCR2 inhibitors as a means to obstruct the IL-8-dependent signaling cascade. Herein, we demonstrate that the specific CXCR2 inhibitor, SB225002, significantly reduces TNBC cell proliferation, invasion, and metastasis. Thus, our findings advocate for the superior efficacy of CXCR2 inhibitors, such as SB225002, over CXCR1 inhibitors in the treatment of TNBC.

Human breast cancer cell lines were obtained from the ATCC and were checked STR profiling data. MCF7 (ER+/PR+/HER2−), Hs578T (ER−/PR−/HER2−), MDA-MB468 (MDA468; ER−/PR−/HER2−), MDA-MB453 (MDA453; ER−/PR−/HER2+), and MDA-MB231 (MDA231; ER−/PR−/HER2−), were cultured in Dulbecco’s Modified Eagle’s Medium (HyClone; Cytiva, Marlborough, MA, USA), whereas T47D (ER+/PR+/HER2−), BT474 (ER+/PR+/HER2+), HCC1419 (ER−/PR−/HER2+), HCC1143 (ER−/PR−/HER2−), and SKBR3 (ER−/PR−/HER2+) were maintained in RPMI1640 medium (HyClone; Cytiva). All culture media were supplemented with 10% fetal bovine serum (FBS; HyClone, Logan, UT, USA), 100 IU/mL penicillin, and 100 μg/mL streptomycin. We housed all cells in a humidified environment containing 95% air and 5% CO₂ at 37°C [23].

Survival analysis was performed in TNBC patients using the Kaplan-Meier plotter database based on IL-8 mRNA expression levels [24]. For this analysis, we utilized the “202859_at” probe set and employed the “Auto select best cutoff” feature to stratify patients based on IL-8 expression. Patient stratification by TNBC subtypes was performed using the “ER status – IHC: ER-negative, ER status – array: ER-negative, PR status – IHC: PR-negative, HER2 status – array: HER2-negative” setting. Calculations and displays included the hazard ratio, 95% confidence intervals, and log-rank p values. We deemed p values <0.05 statistically significant.

In our previous study, we employed TRIzol (Thermo Fisher Scientific) for total RNA extraction from the cells [25]. Following the manufacturer’s guidelines, we synthesized cDNA using a cDNA Synthesis Kit (Thermo Fisher Scientific). The SYBR Green master mix (Bioline Ltd., London, UK) was used for quantitative reverse transcription PCR on an ABI 7900HT real-time PCR detection system (Foster City, CA, USA), normalizing target gene expression against beta-actin (ACTB) using the comparative CT method (2^−ΔΔCT) to determine relative gene expression (fold change). Thermal cycling involved an initial step at 50°C for 2 min, followed by 95°C for 10 min and 40 cycles of 95°C for 15 s, 60°C for 15 s, and 72°C for 15 s. We employed primers for human ACTB (forward, F: TCA CCA ATT GGA TGA GCG GTT; reverse, R: AGT TTC GTG GAT GCC ACA GGA C) and IL-8 (F: AGG GTT GCC AGA TGC AAT AC; R: AAA CCA AGG CAC AGT GGA AC) [25].

To evaluate secreted IL-8 protein levels, we analyzed the conditioned culture media (200 μL) from breast cancer cells using the manufacturer’s protocol of ELISA kit (R&D Systems, MN, USA).

We executed cell migration analyses through the Boyden chamber assay, as previously described [25]. TNBC cells were placed in the upper chamber of Boyden-chamber inserts, having an 8-μm pore size, within 24-well plates (Becton-Dickinson, San Diego, CA, USA). To the lower chamber, fresh culture medium containing 10% FBS was added. The cells (5 × 10⁴ cells/well) in the upper chamber were subjected to 20 ng/mL recombinant human IL-8 treatment for 24 h. Post-treatment, cells on the filter’s upper surface were eradicated, the filter underwent fixation with 100% methanol, followed by staining with 0.1% crystal violet solution. The migrated cells were examined using a CK40 inverted microscope (Olympus, Tokyo, Japan). The quantification of migrated cells was counted using Image J software.

We evaluated sensitivities to CXCR1 and CXCR2 inhibitors, specifically reparixin and SB225002, using the MTT assay. TNBC cells (1.5 × 10³ cells/well) were plated onto 96-well plates. Post 24-h incubation, cells were exposed to DMEM supplemented with 10% FBS, with or without specified concentrations of reparixin or SB225002 for 48 h. To assess cell viability, we mixed equal volumes of serum-free media and MTT solution in each well. Following a 3-h incubation at 37°C, DMSO was added to dissolve the MTT formazan. Absorbance at 590 nm was measured using a tunable microplate reader [26].

For colony formation assays, TNBC cells (2 × 10³ cells/well) were seeded into 12 or 6-well plates and incubated overnight at 37°C. Subsequently, cells received treatment with specified concentrations of reparixin or SB225002, followed by an additional 10-day incubation. Fixed with 100% ethanol and stained with 0.1% crystal violet, cell colonies were visualized using a CK40 inverted microscope (Olympus, Tokyo, Japan) [25, 26].

Cells were disrupted using PRO-PREP™ Protein Extraction Solution (Intron Biotechnology, Inc., Gyeonggi-do, Korea) on ice, centrifuged (16,100 g) for 15 min, and the resulting supernatants were harvested. Prepared samples, consisting of equal amounts of proteins boiled for 5 min in Laemmli sample buffer, were deployed into SDS-polyacrylamide gels under denaturing conditions and subsequently transferred to polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). Blocked with 5% skim milk for 1 h at room temperature (RT), membranes were incubated overnight at 4°C with primary antibodies against β-actin (Abfrontier, Seoul, South Korea, LF-PA0207, 1:2,000), PARP-1 (CST, Danvers, MA, USA, 9542S, 1:1,000), pro-caspase3 (Abcam, Waltham, MA, USA, ab32351, 1:5,000), cleaved-caspase3 (CST, 9661S, 1:1,000), p-STAT3 (Abcam, ab76315, 1:10,000), t-STAT3 (Santa Cruz Biotechnology, CA, USA, sc-8019, 1:1,000), p-AKT (Abcam, ab81283, 1:5,000), t-AKT (CST, 9272S, 1:1,000), p-ERK (Santa Cruz, sc-7383, 1:1,000), and t-ERK (CST, 9102S, 1:1,000). Following washing, membranes were treated with horseradish peroxidase-conjugated secondary antibodies (Cell Signaling Technology, Danvers, MA, USA) at room temperature for 1 h. Visualization of transferred proteins was achieved using the ECL^™ prime reagent (GE Healthcare, Bucks, UK), and the protein bands on the X-ray film were sensitized using a JPI automatic X-ray film processor (JPI Healthcare Co., Ltd, Seoul, South Korea, JP-33).

We conducted flow cytometry with propidium iodide-stained cells to analyze the cell cycle. Post-harvesting, cells were fixed with 70% ethanol at room temperature for 20 min. After fixation, cells were treated with 100 µg/mL DNase-free RNase A (Thermo Fisher Scientific, Waltham, MA, USA) at 37°C for 30 min. Following a PBS wash, cells were resuspended in fluorescence-activated cell sorting (FACS) buffer containing 50 µg/mL propidium iodide (Sigma-Aldrich, St. Louis, MO, USA) and analyzed using a BD FACSCalibur flow cytometer (BD, Becton-Dickinson, San Diego, CA, USA) [25].

For apoptosis assays, we utilized the Annexin V PE Apoptosis Detection Kit (BioGems, Via Colinas, Westlake Village, USA), following the manufacturer’s protocol. In short, cells (1 × 10⁶ cells/mL) were collected, washed twice with PBS, and then resuspended in 500 μL of staining solution containing 5 μL Annexin V-PE and PE-Cy7. Following a 15-min incubation at RT in a dark environment, cells were immediately subject to analysis on a flow cytometer. Apoptotic cells, identified by dual staining with Annexin V-PE and PE-Cy7, were assessed using the FACS Vantage system (Becton-Dickinson, San Diego, CA, USA), with the percentage of cells undergoing apoptosis quantified thereafter.

Utilizing the Proteome Profiler Human Apoptosis array kit (R&D Systems, Minneapolis, MN, USA), we analyzed cell lysates from 10 μm SB225002-treated and -untreated vehicles at 48 h post-treatment for apoptosis-related proteins, adhering strictly to the manufacturer’s guidelines. Cell lysates were prepared using lysis buffer from the kit, and protein concentration was ascertained through the Bradford method. Following a blockage phase in buffer, equilibrated proteins were incubated overnight at 4°C on a rocking platform shaker with precoated array membranes ready for analysis. Post-incubation, the membranes underwent three washes in 1× washing buffer to discard unattached proteins, followed by detection antibody cocktail treatment for 1 h at room temperature while shaking. After additional washes, the membranes were incubated with diluted streptavidin-HRP for 30 min at room temperature under shaking. Detection was performed by X-ray film exposure, leveraging enhanced chemiluminescence. Quantification of each protein level was relative to positive control spots on the same membrane.

Female BALB/c nude mice (Orient Bio, Seoul, Korea) at 6–8 weeks were used to establish a nude mouse xenograft model. All animal care and experimental procedures complied with guidelines approved by the Institutional Animal Care and Use Committee (IACUC) of Samsung Medical Center (Protocol No. 20240403001) [26].

To investigate the effect of SB225002 on lung metastasis, 4T1 cells (1.8 × 10⁷ cells/120 μL) were resuspended in Matrigel (BD Biosciences, Bedford, MA, USA) and injected directly into the right secondary mammary fat pads. The mice were randomly divided into two groups (n = 4/group), which were treated with vehicle (25% PEG 400 and 5% Tween 80) or 10 mg/kg SB225002 starting on the day of inoculation and continuing until the end of the experiment. Once the tumors reached a volume of approximately 50 mm³, the tumor size was measured using digital calipers at set time points, and the volume was determined using the formula V = 1/2 × length × (width)². Growth curves were calculated using the average relative tumor volume per group at the set time points [26].

For immunohistochemistry assays, paraffin-embedded lung tissue sections (4 μm) were deparaffinized in xylene, dehydrated in graded alcohol, and hydrated in water. Tissue sections were evaluated by H&E staining for detection of lung metastasis. The slides were analyzed using a ScanScope XT apparatus (Aperio Technologies, CA, USA) [26].

Graph construction was carried out using Microsoft Excel 2016 (Microsoft, Redmond, WA, USA) and GraphPad Prism 8 software (GraphPad Software, La Jolla, CA, USA). We computed p values using the Student’s t test (unpaired, two-tailed) and one-way ANOVA. All experimental operations were independently replicated a minimum of three times. Data presentation adheres to the mean ± standard error of the mean (SEM) format, with significance ascribed to differences when p < 0.05.

To assess the role of IL-8 expression on survival rates across different breast cancer subtypes, we utilized the Kaplan-Meier plotter dataset. Our analysis revealed a significant correlation between IL-8 expression and survival outcomes in TNBC patients (Fig. 1a–c). Patients with elevated IL-8 levels exhibited substantially worse RFS (Fig. 1a, p = 0.00058), OS (Fig. 1b, p = 0.00049), and DMFS (Fig. 1c, p = 0.011) compared to those with lower IL-8 expression. Hence, these findings underscore the direct impact of IL-8 expression on the prognosis of TNBC.

Evaluating the expression levels of IL-8 in breast cancer cells, we discovered a pronounced upregulation of IL-8 mRNA and protein expression in TNBC cells, in contrast to other breast cancer subtypes (Fig. 2a, b). To further comprehend the consequences of inherent IL-8 expression in TNBC cells, we analyzed the metastatic potential through a Boyden chamber migration assay in HCC1143, Hs578T, and MDA231 cells. Recombinant human IL-8 significantly heightened migration across cell culture inserts (Fig. 2c), indicating that aberrant IL-8 expression is critical for increasing TNBC cell migration ability. As shown in Figure 2c, cell invasiveness by IL-8 treatment was increased 2.8-fold (in HCC1143), 3.1-fold (in Hs578T), and 4.6-fold (MDA231) compared to the control groups, respectively.

Figure 3a illustrates the chemical structures and functions of reparixin and SB225002. In our effort to curb IL-8-mediated signaling in TNBC cells, we administered a specific CXCR1 inhibitor (reparixin) or a CXCR2 inhibitor (SB225002) for 48 h at various concentrations. Interestingly, reparixin showed no effect on cell viability (Fig. 3b), whereas SB225002 distinctly reduced cell viability in a dose-dependent fashion (Fig. 3c). Additionally, we evaluated the pharmacological effectiveness of both inhibitors on cell growth using a colony-forming assay. As anticipated, our experiments confirmed that SB225002 notably hindered cell proliferation (Fig. 3d), spotlighting CXCR2 inhibitors like SB225002 as more efficacious than CXCR1 inhibitors in the treatment of TNBC.

Following promising results, we proceeded to analyze apoptosis-related protein expressions using the Proteome Profiler Human Apoptosis Array, post-SB225002 treatment. The treatment resulted in a visible decline in green square indicators (Fig. 4a). Key proteins associated with apoptosis, including Bcl-2, cIAP-1, cIAP-2, Survivin, XIAP, HIF-1α, and HO-1 experienced significant downregulation upon SB225002 treatment. Furthermore, Annexin V-PE and PE-Cy7 staining showcased a notable increase in apoptosis rates in SB225002-treated groups compared to the control, whereas reparixin showed no observable effect in TNBC cells (Fig. 4b, c). Additionally, we examined cell cycle alterations induced by SB225002 or reparixin. Our results showed that the subG1 population significantly increased in both HCC1143 and MDA231 cells following treatment with SB225002 (online supplement 1A, B; for all online suppl. material, see https://doi.org/10.1159/000545659), but not reparixin (online supplement 1C, D). As shown in Figure 4d, pro-PARP-1 expression was decreased by SB225002 while cleaved-PARP-1 expression was increased. Also, pro-caspase 3 and cleaved caspase 3 expression showed a similar pattern, as did PARP-1 expression in Hs578T cells. The phosphorylation of intracellular signaling molecules such as STAT-3, AKT, and ERK was also decreased by SB225002 but not by reparixin (Fig. 4e). Collectively, these findings illustrate that SB225002, unlike reparixin, can trigger apoptotic cell death in TNBC cells.

Continuing our exploration, we assessed the pharmacological impact of SB225002 on 4T1 mouse TNBC cells. Treatment with SB225002 notably amplified the total apoptotic cell population, a contrast not seen with reparixin treatment (Fig. 5a). Specifically, the apoptotic cell population increased by 12.75 ± 0.59% (early) and 7.63 ± 0.36% (late) relative to control levels (Fig. 5b). Conforming to expectations, cell proliferation markedly decreased upon SB225002 treatment (Fig. 5c). Observations also included cell morphological changes toward roundness, swelling, and the formation of a balloon-like structure (Fig. 5c). Consistent with the results in Figure 4d, the expression of apoptosis-related proteins such as PARP-1 and caspase 3 was also regulated by SB225002 in 4T1 murine TNBC cells (Fig. 5d). To investigate the pharmacological efficacy of SB225002 in vivo, we performed xenograft model studies, which are schematically illustrated in Figure 5e. Moreover, examining the suppressive effects of SB225002 on tumor growth and metastatic potential via orthotopic xenograft models, we noted a significant reduction in primary tumor size at the second fat pad due to SB225002 treatment (Fig. 5f). The size of tumors in the SB225002-treated group was about half of that in the vehicle-treated group (Fig. 5f), and lung metastasis was dramatically reduced (Fig. 5g). This compilation of results highlights the efficacy of SB225002 as a potent antagonist in curbing growth and metastasis of TNBC cells.

TNBC is characterized by a highly aggressive phenotype and an increased propensity for metastasizing to the central nervous system and visceral organs, such as the liver, lungs, and bones due to the absence of effective therapeutic targets [27]. Furthermore, TNBC patients often experience a significantly shortened survival period following metastasis, with an early recurrence rate of 25% after surgical intervention; notably, the predominant cause of mortality in breast cancer emanates from metastatic progression rather than primary tumor growth [27, 28]. Despite these challenges, the search for an effective targeted therapy for TNBC remains ongoing. In our investigation, we identified a marked reduction in survival rates among TNBC patients with aberrant induction of IL-8. Moreover, the expression levels of IL-8 mRNA and protein were found to be substantially elevated in TNBC cells compared to other subtypes, leading us to explore the potential of targeting IL-8’s binding receptors, specifically CXCR1 and CXCR2, as therapeutic strategies for TNBC.

Prior research by Kim et al. [29] elucidated that IL-8 is secreted by various cell types and exhibits highest expression within the TNBC stromal environment relative to other breast cancer subtypes [30]. The overexpression of IL-8 correlates with several oncogenic characteristics, including angiogenesis, invasion, and enhanced metastatic capacity across multiple carcinoma types, serving as a prognostic marker for adverse clinical outcomes [31, 32]. Zhang et al. [33] reported that the expression level of serum IL-8 was 79.68 ± 9.53 ng/L in benign group and 220.54 ± 12.49 ng/L in malignant group, and both were higher than the 54.31 ± 10.26 ng/L in healthy control group in ovarian cancer. The basal expression of IL-8 is modulated via a mitogen-activated protein kinase)/ ERK1/2-dependent pathway in TNBC cells [29]. Contrarily, employing a specific MEK inhibitor, UO126, led to reduced anchorage-independent growth, cell invasion, and migration in TNBC cultures [29]. Our results were also revealed that the levels of ERK, AKT, and STAT-3 phosphorylation were decreased by SB225002 but not by reparixin. Additionally, we observed an augmentation in cell invasiveness following treatment with recombinant human IL-8 in a spectrum of TNBC cell lines. Thus, it is posited that disrupting IL-8-mediated signaling pathways could offer a novel approach for curbing TNBC proliferation and metastasis.

IL-8 exerts its biological actions through binding to G-protein-coupled receptors, specifically the cysteine-X-cysteine receptors CXCR1 and CXCR2 [34]. The IL-8/CXCR complex initiates a cascade of intracellular signaling pathways, including phosphoinositide-3-kinase/protein kinase B Akt (PI3K/Akt), mitogen-activated protein kinase, and STAT-3, implicated in tumorigenesis [35‒37]. Although our results did not explicitly demonstrate this, we noted that the mRNA expression levels of CXCR1 and CXCR2 did not exhibit significant variance across different breast cancer cell types. Nonetheless, the abnormal induction of IL-8 and its interaction with respective receptors in TNBC prompted an investigation into the efficacy of specific inhibitors. Intriguingly, our data unveiled distinct responses to CXCR1 and CXCR2 inhibitors in TNBC cells, with prior studies highlighting CXCR2’s involvement in tumor growth, metastasis, and chemoresistance in various cancers, including breast carcinoma [34‒38]. Consequently, we provide evidence that the IL-8/CXCR2 signaling axis, unlike CXCR1, may represent a viable therapeutic target in TNBC by promoting cancer cell apoptosis to inhibit metastasis.

Many studies have revealed that CXCR2 expression is associated with their proliferation, invasion, and migration in a variety of cancer cells including lung and pancreatic cancers [38, 39]. In addition, the inhibition of CXCR1/2 by a IL-8 analog or the CXCR2 antagonists decreased lung tumor cell proliferation, angiogenesis, and metastasis in vivo [18, 40]. Prajapati et al. [41] reported that SCH-479833, a CXCR1/2 antagonist, significantly decreased cell proliferation of tumors derived from antagonist-treated mice were and increased apoptosis through immunostaining of Ki-67 and CC3. The anticancer agents induce the intrinsic pathway for apoptosis, whereas direct activation of extrinsic apoptotic pathway may not be needed to kill the cells downstream of antineoplastic agents [42]. Based on these results, our results revealed that mice treated with the SB225002 significantly reduced tumor volume and incidence of lung metastasis when compared to the control group in vivo study using xenografts model. These findings reflect that the apoptotic potential of CXCR2 antagonists such as SB225002 is effective in preventing tumor growth or lung metastasis in TNBC.

Elevated secretion levels of IL-8 were predominantly observed in TNBC cells compared to other subtypes. The aberrant induction of IL-8 was strongly linked with poor survival outcomes in TNBC patients. Leveraging these insights, our study evaluated the therapeutic potentials of CXCR1- and CXCR2-specific inhibitors using TNBC models. Notably, SB225002, a selective CXCR2 inhibitor, exhibited pronounced effectiveness in inducing apoptotic cell death in TNBC, whereas reparixin, a CXCR1 inhibitor, demonstrated negligible impact. Subsequent investigations confirmed that SB225002 substantially inhibited tumor growth and lung metastasis in 4T1 cell-based lung metastasis models. Here, we demonstrated that CXCR2 antagonist is effective for the treatment of TNBCs that are not subject to targeted therapy against ER, PR, and HER2. Based on these preclinical findings, the inhibitory effects of the CXCR2 antagonist resulted from decreased tumor growth, progression, and metastasis of TNBC cells. A significant point of further exploration includes the downstream signaling pathways of IL-8/CXCR2 complex or the mechanism of intrinsic or extrinsic apoptosis pathway in TNBC cells. Collectively, these findings underscore the critical role of the IL-8/CXCR2 signaling pathway in TNBC tumor proliferation and metastatic dissemination. We cautiously propose that therapies targeting CXCR2 may offer a more promising treatment modality compared to CXCR1-based strategies for TNBC management.

All animal care and experimental procedures were performed in accordance with the principles and guidelines of the Korea Council for Animal Care. This study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of Samsung Medical Center (Approval No. 20240403001). Experiments were performed in accordance with the relevant guidelines and regulations. Studies with humans were collected through the Kaplan-Meier plotter database (https://kmplot.com/analysis/index.php?p=service&cancer=breast). This database is open source for easy access for all researchers, so no separate ethical approval is required for data collection.

The authors declare no conflicts of interest.

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2024-00452849).

Y.J., S.Y.Y., and S.P.J. contributed to the investigation, data curation, formal analysis, and writing – original draft. S.K. and S.J.N contributed to the manuscript review and editing. S.K and J.E.L. contributed to the study conceptualization, data curation, formal analysis, and validation. All authors have read and agreed to the published version of the manuscript.

Yisun Jeong, Sun Young Yoon, and Seung Pil Jung contributed equally to this work.

The data that support the findings of the current study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (Sangmin Kim: [email protected]) on reasonable request.