Cytoplasmic Basigin Localization is Associated with EMT, Tumor Budding, and Adverse Prognosis in Breast Cancer

Abstract

Purpose

To investigate the clinicopathological significance of Basigin subcellular localization in breast invasive ductal carcinoma, with particular emphasis on its association with epithelial–mesenchymal transition (EMT), tumor budding, and patient prognosis.

Methods

A total of 206 cases of primary breast invasive ductal carcinoma were retrospectively collected from Zhengzhou People’s Hospital between January 2014 and January 2024. Immunohistochemistry was performed to evaluate the expression and subcellular localization of Basigin, E-cadherin, and Vimentin. Tumor budding was assessed according to the International Tumor Budding Consensus Conference (ITBCC) criteria. Associations between marker expression, clinicopathological parameters, and disease-free survival (DFS) were analyzed. A prognostic nomogram was subsequently constructed based on independent prognostic factors identified by multivariate Cox regression analysis.

Results

Predominantly cytoplasmic Basigin localization was significantly associated with larger tumor size, higher Nottingham grade, high-grade tumor budding, advanced TNM stage, and HER2-overexpressing/triple-negative subtypes (all P < 0.05). Tumor budding regions exhibited EMT-associated phenotypes characterized by reduced membranous Basigin staining with relative cytoplasmic redistribution, decreased E-cadherin expression, and increased Vimentin expression. Distinct tumor budding patterns were observed among molecular subtypes, with luminal tumors predominantly showing no budding, HER2-overexpressing tumors showing cluster budding, and triple-negative tumors predominantly showing single-cell budding (all P < 0.05). Kaplan–Meier and multivariate Cox regression analyses demonstrated that Basigin localization (HR = 0.380, 95% CI: 0.171–0.844, P = 0.017), Nottingham grade (HR = 4.839, 95% CI: 2.238–10.466, P < 0.001), tumor budding grade (HR = 37.932, 95% CI: 12.850–111.966, P < 0.001), and molecular subtype (HR = 1.523, 95% CI: 1.164–1.991, P = 0.002) were independent prognostic factors for disease-free survival (DFS). The nomogram demonstrated strong predictive performance, with area under the curve (AUC) values of 0.919, 0.921, and 0.980 for predicting 2-, 3-, and 5-year DFS, respectively.

Conclusion

Basigin subcellular redistribution is significantly associated with EMT-associated phenotypes, tumor budding, and adverse prognosis in breast cancer. Assessment of Basigin localization may provide additional prognostic value beyond conventional clinicopathological parameters. A nomogram integrating Basigin localization, tumor budding, histological grade, and molecular subtype may facilitate individualized risk stratification in patients with breast cancer, pending external validation in independent cohorts.

Introduction

Breast cancer remains the most frequently diagnosed malignancy and the leading cause of cancer-related death among women worldwide.¹ Despite substantial advances in molecular subtype-guided therapies, a considerable proportion of patients still experience recurrence and distant metastasis, which remain the principal determinants of adverse clinical outcomes.² Currently available biomarkers—including hormone receptor status, HER2 expression, and the Ki-67 proliferation index—provide essential guidance for treatment selection but do not fully capture the invasive and metastatic potential of individual tumors. Consequently, considerable heterogeneity in clinical outcomes persists among patients within the same traditional risk categories, highlighting an unmet clinical need for additional prognostic markers that reflect tumor biological aggressiveness.

Basigin (also known as EMMPRIN/CD147/HAb18G) is a transmembrane glycoprotein of the immunoglobulin superfamily that has been implicated in tumor invasion, extracellular matrix remodeling, and metastatic progression through the induction of matrix metalloproteinases (MMPs).³ Increasing evidence suggests that Basigin also participates in epithelial–mesenchymal transition (EMT), a critical process during tumor progression in which epithelial cells lose cell polarity and intercellular adhesion while acquiring mesenchymal and migratory phenotypes. Reduced E-cadherin expression and increased Vimentin expression are widely recognized hallmarks of EMT.⁴ Mechanistically, Basigin functions as a chaperone for monocarboxylate transporters (MCTs) at the plasma membrane, facilitating lactate transport and maintaining the acidic tumor microenvironment.^5,6 Loss of membranous Basigin, as observed in the cytoplasmic redistribution pattern, may therefore compromise MCT-mediated metabolic homeostasis and promote metabolic stress, which has been linked to the acquisition of mesenchymal phenotypes. Additionally, membrane-associated Basigin is known to induce MMPs through interactions with integrins and other extracellular partners.³ Cytoplasmic sequestration of Basigin could reduce these membrane-proximal signaling events while potentially exposing or activating alternative intracellular pathways that favor cytoskeletal remodeling and loss of epithelial adhesion. Although the precise molecular link between Basigin redistribution and EMT remains incompletely defined, these observations provide a plausible mechanistic framework linking altered Basigin localization to tumor invasiveness and budding. Tumor budding, defined as isolated single tumor cells or small clusters composed of fewer than five cells at the invasive tumor front, is considered a histomorphological manifestation of EMT-associated tumor invasion.⁷ Although the International Tumor Budding Consensus Conference (ITBCC) established standardized assessment criteria for colorectal carcinoma,⁸ breast cancer-specific budding guidelines are still evolving, and these criteria have not been formally endorsed for routine breast cancer reporting. In the present study, we adopted the ITBCC criteria originally developed for colorectal carcinoma for tumor budding assessment in our breast cancer cohort. Importantly, increasing evidence has demonstrated substantial heterogeneity in Basigin subcellular localization across different tumor types, and its prognostic significance appears to depend not only on expression level but also on subcellular distribution.^9–12 In breast cancer, several studies have investigated the expression of Basigin/CD147. A meta-analysis by Shi et al demonstrated that elevated Basigin expression was significantly associated with poor overall survival and unfavorable clinicopathological features, including lymph node metastasis and advanced TNM stage.¹³ Chen et al further implicated Basigin in triple-negative breast cancer progression through HIF-1α signaling.¹⁴ However, these studies focused almost exclusively on overall expression levels, and none have systematically examined the prognostic implications of Basigin subcellular localization in breast cancer. Whether cytoplasmic redistribution carries similar clinicopathological and prognostic significance in this disease has remained unexplored until the present study.

In the present study, we investigated the clinicopathological and prognostic significance of Basigin subcellular localization in a cohort of 206 breast invasive ductal carcinomas. Specifically, we analyzed the associations between Basigin localization patterns, EMT-related markers (E-cadherin and Vimentin), tumor budding grade, molecular subtype, and disease-free survival. In addition, we evaluated dynamic localization changes in paired liver metastases and constructed a prognostic nomogram integrating Basigin localization with established pathological parameters. Through these analyses, we sought to determine whether Basigin subcellular redistribution may serve as a clinically relevant marker of tumor invasiveness and prognostic stratification in breast cancer.

Materials and Methods

Patient Cohort

A total of 206 cases of primary breast invasive ductal carcinoma of no special type were retrospectively collected from Zhengzhou People’s Hospital between January 2014 and January 2024. The median age of the entire cohort was 56 years (range: 25–87 years), with only 12 cases (5.8%) aged ≥70 years. Corresponding adjacent non-neoplastic breast tissues were available for all cases. All specimens were formalin-fixed, paraffin-embedded (FFPE), and independently reviewed by experienced breast pathologists at the Department of Pathology, Zhengzhou People’s Hospital. The inclusion criteria were as follows: (1) histologically confirmed primary breast invasive ductal carcinoma of no special type; (2) treatment by radical mastectomy or breast-conserving surgery; and (3) availability of complete clinicopathological and follow-up data. Exclusion criteria included: (1) preoperative chemotherapy, radiotherapy, targeted therapy, or neoadjuvant treatment; (2) biopsy-only specimens; (3) recurrent or metastatic breast carcinoma at initial diagnosis; and (4) death from non-breast cancer causes during follow-up. Follow-up data were obtained from institutional medical records and hospital information systems. Disease-free survival (DFS) was defined as the interval between the date of surgery and the first documented local recurrence or distant metastasis. The last follow-up update for data analysis was performed on December 31, 2025. This date refers to the administrative data lock for final statistical analysis rather than to patient enrollment or active follow-up activities, which had been completed prior to ethics submission. Patients without events or those lost to follow-up were censored at the date of last known contact. The median follow-up duration for the entire cohort was 55.0 months (range: 6–120 months). All patients in this cohort received standard postoperative adjuvant therapy according to the St. Gallen consensus guidelines and the recommendations of the Chinese Society of Clinical Oncology (CSCO). All 206 cases had complete clinicopathological and follow-up data, with no missing data for any variables included in this study. This study was approved by the Ethics Committee of Zhengzhou People’s Hospital (Approval No. 2026-KY-007001). Given the retrospective nature of the study and the use of archived anonymized tissue specimens, the requirement for informed consent was waived by the institutional review board. All procedures were conducted in accordance with the Declaration of Helsinki (2013 revision).

Immunohistochemistry and Evaluation Criteria

Immunohistochemical staining was performed using a streptavidin–peroxidase method to evaluate the expression of Basigin, E-cadherin, and Vimentin. Briefly, 4-μm formalin-fixed, paraffin-embedded tissue sections were deparaffinized, rehydrated, and subjected to heat-induced antigen retrieval in citrate buffer. Endogenous peroxidase activity was blocked using 3% hydrogen peroxide. The sections were incubated overnight at 4°C with primary antibodies against Basigin (1:200 dilution; Maixin Biotech, Fuzhou, China), E-cadherin (ready-to-use; Maixin Biotech), and Vimentin (ready-to-use; Maixin Biotech). After incubation with secondary antibodies, immunoreactivity was visualized using diaminobenzidine (DAB), followed by hematoxylin counterstaining.

Immunohistochemical evaluation was independently performed by two experienced pathologists blinded to clinicopathological data. Basigin expression patterns were classified according to predominant subcellular localization as follows: (1) predominantly membranous, defined as ≥50% of tumor cells showing complete membranous or membranous/cytoplasmic staining; and (2) predominantly cytoplasmic, defined as ≥50% of tumor cells showing diffuse cytoplasmic staining accompanied by reduced or discontinuous membranous staining. The 50% cutoff was adopted based on our previously published methodology in non-small cell lung cancer,⁹ which defined “predominant subcellular localization in the majority of tumor cells” as the criterion for classification. Interobserver agreement in the assessment of Basigin localization was evaluated using Cohen’s kappa coefficient. The kappa value was 0.88 (95% CI: 0.82–0.94), indicating excellent interobserver reproducibility (Table S1). Discordant cases were reviewed jointly to achieve consensus. E-cadherin expression was categorized as high or low. High expression was defined as membranous staining in ≥50% of tumor cells with staining intensity comparable to adjacent normal ductal epithelium, whereas reduced membranous staining in <50% of tumor cells was considered low expression. Vimentin expression was evaluated as positive when cytoplasmic staining was observed in ≥5% of tumor cells and negative when staining was present in <5% of tumor cells.

Quantitative image analysis was performed using the Image-Pro Plus software system (Media Cybernetics, Rockville, MD, USA). Five representative high-power fields (400× magnification) were randomly selected for each case, and positive unit (PU) values were measured as relative expression levels.⁹ For analysis of tumor budding regions, budding hotspots were first identified at 200× magnification, followed by manual delineation of budding cells at 400× magnification for quantitative measurement. Non-budding regions were sampled from the central tumor area away from the invasive front.

Assessment of Tumor Budding

Tumor budding was assessed on hematoxylin and eosin-stained sections according to the criteria proposed by the International Tumor Budding Consensus Conference (ITBCC).⁸ Briefly, tumor buds were defined as isolated single tumor cells or clusters composed of fewer than five tumor cells located at the invasive tumor front. Areas with the highest budding density (“hotspots”) were identified at low magnification, followed by evaluation at 200× magnification. Tumor budding was graded as low (0–4 buds), intermediate (5–9 buds), or high (≥10 buds).

In addition to conventional budding grade assessment, tumor budding morphology was further categorized into three patterns: (1) no budding; (2) predominantly cluster budding, defined as ≥50% of tumor buds consisting of clusters containing two or more tumor cells; and (3) predominantly single-cell budding, defined as ≥50% of tumor buds consisting of isolated single tumor cells. Associations between tumor budding morphology and breast cancer molecular subtypes were subsequently analyzed.

Statistical Analysis

Statistical analyses were performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation (SD). Comparisons between two groups were performed using the independent-samples t test, whereas comparisons among multiple groups were analyzed using one-way analysis of variance (ANOVA) followed by either the LSD or Dunnett’s T3 post hoc test, as appropriate. Paired comparisons were analyzed using the paired-samples t test. Correlations between variables were assessed using Spearman rank correlation analysis because the expression data (Basigin, E-cadherin, and Vimentin PU values) did not follow a normal distribution (Shapiro–Wilk test, P < 0.05).

Categorical variables were compared using the χ²-test. DFS was estimated using the Kaplan–Meier method and compared using the Log rank test. Variables with prognostic significance were further evaluated using multivariate Cox proportional hazards regression analysis to identify independent prognostic factors.

All statistical tests were two-sided, and P values <0.05 were considered statistically significant.

Nomogram Construction and Validation

A prognostic nomogram for predicting 2-, 3-, and 5-year disease-free survival (DFS) was constructed using R software based on independent prognostic variables identified by multivariate Cox regression analysis, including Basigin subcellular localization, tumor budding grade, Nottingham histological grade, and molecular subtype.

The predictive performance of the nomogram was evaluated using receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) was calculated to assess model discrimination. Calibration curves were generated to evaluate the agreement between predicted and observed survival probabilities. Internal validation of the nomogram was performed using 1000 bootstrap resamples with optimism correction to evaluate model stability and reduce overfitting. Decision curve analysis (DCA) was conducted to evaluate the clinical net benefit of the nomogram.

Results

Morphological Observation of Tumor Budding in Breast Carcinoma

A total of 206 patients with invasive breast carcinoma of no special type were stratified into low-grade (n=124) and intermediate-high-grade (n=82) groups based on tumor budding grading criteria. Hematoxylin and eosin (H&E) staining revealed sparse tumor budding foci within the stroma at the invasive front in the low-grade group, whereas abundant dense tumor budding was observed at the invasive front in the intermediate-high-grade group (Figure 1).

Figure 1.

Tumor budding at the invasive front of invasive breast carcinoma of no special type. (A and B) Low-grade tumor budding. (C and D) High-grade tumor budding. Hematoxylin and eosin staining; (A and C) ×200 magnification; (B and D) ×400 magnification. Red circles indicate tumor budding cells at the invasive front.

Expression of Basigin, E-Cadherin, and Vimentin in Breast Invasive Ductal Carcinoma and Adjacent Non-Neoplastic Tissues

Basigin expression was predominantly localized to the cell membrane and cytoplasm in breast invasive ductal carcinoma, whereas little to no expression was observed in adjacent non-neoplastic breast tissues. Quantitative image analysis demonstrated significantly higher Basigin expression levels in tumor tissues compared with adjacent tissues (15.09 ± 5.52 vs 2.62 ± 1.08, respectively; t = 31.60, P < 0.001) (Figure 2A–C and Table 1).

Figure 2.

Expression of Basigin, E-cadherin, and Vimentin in invasive breast ductal carcinoma and corresponding adjacent tissues (immunohistochemistry, ×400 magnification). (A) Basigin is minimally expressed in adjacent tissues. (B) Basigin is highly expressed in invasive breast ductal carcinoma. (C) Quantitative comparison of Basigin expression between breast cancer and adjacent tissues (P < 0.001). (D) E-cadherin is highly expressed in adjacent tissues. (E) E-cadherin expression is decreased in invasive breast ductal carcinoma. (F) Quantitative comparison of E-cadherin expression between breast cancer and adjacent tissues (P < 0.001). (G) Vimentin is minimally expressed in adjacent epithelial cells but is present in stromal cells, including inflammatory cells, vascular endothelial cells, and stromal fibroblasts/myofibroblasts. (H) Vimentin expression is upregulated in breast cancer tissues. (I) Quantitative comparison of Vimentin expression between breast cancer and adjacent tissues (P < 0.001). (J) Basigin expression was negatively correlated with E-cadherin expression (Rs = −0.486, P < 0.001). (K) Basigin expression was positively correlated with Vimentin expression (Rs = 0.669, P < 0.001). Statistical significance is indicated as follows: ***P < 0.001.

Table 1.

Expression of Basigin, E-Cadherin, and Vimentin in Breast Cancer and Adjacent Tissues ()

Groups	N	Basigin	E-Cadherin	Vimentin
Breast cancer	206	15.09±5.52	12.56±5.04	4.06±5.25
Adjacent tissues	206	2.62±1.08	17.47±1.03	0.06±0.07
t		31.60	13.57	10.96
P		0.000	0.000	0.000

E-cadherin expression was predominantly membranous in both tumor and adjacent breast epithelium. However, quantitative expression levels were significantly reduced in breast invasive ductal carcinoma compared with adjacent tissues (12.56 ± 5.04 vs 17.47 ± 1.03, respectively; t = −13.57, P < 0.001) (Figure 2D–F and Table 1).

Vimentin expression was primarily observed in stromal components and was significantly increased in breast invasive ductal carcinoma compared with adjacent tissues (4.06 ± 5.25 vs 0.06 ± 0.07, respectively; t = 10.96, P < 0.001) (Figure 2G–I and Table 1).

Spearman correlation analysis demonstrated that Basigin expression was negatively correlated with E-cadherin expression (Rs = −0.486, P < 0.001) and positively correlated with Vimentin expression (Rs = 0.669, P < 0.001) (Figure 2J and K).

Associations Between Basigin Localization, EMT Marker Expression, and Clinicopathological Features

Among the 206 breast invasive ductal carcinomas, 127 cases (61.65%) demonstrated predominantly membranous Basigin expression, whereas 79 cases (38.35%) demonstrated predominantly cytoplasmic expression. Cytoplasmic Basigin localization was significantly associated with larger tumor size (P = 0.017), higher Nottingham grade (P < 0.001), high-grade tumor budding (P < 0.001), advanced TNM stage (P < 0.001), and molecular subtypes (P < 0.001) (Table 2 and Figure 3A–F).

Table 2.

Relationship Between the Expression of Basigin, E-Cadherin, and Vimentin and Clinicopathological Features in Invasive Breast Carcinoma of No Special Type

Groups	Basigin			E-Cadherin			Vimentin
	Membranous (127)	Cytoplasmic (79)	P	High (122)	Low (84)	P	Negative (137)	Positive (69)	P
Age (year)
<60 (125)	77	48	0.985	74	51	0.993	85	40	0.572
≥60 (81)	50	31		48	33		52	29
Tumor localization
Left (92)	60	32	0.344	58	34	0.316	65	27	0.257
Right (114)	67	47		64	50		72	42
Diameter (cm)
≤2 (94)a	67	27c	0.017	64	30	0.041	74	20b,c	0.003
>2, <5 (77)b	44	33		42	35		44	33a
≥5 (35)c	16	19a		16	19		19	16a
Nottingham grade
I grade (23)a	23	0c	0.000	22	1c	0.000	23	0c	0.000
II grade (114)b	96	18c		92	22c		102	12c
IIIgrade (69)c	8	61a,b		8	61a,b		12	57a,b
Tumor budding grade
Low grade (124)	99	25	0.000	97	27	0.000	102	22	0.000
Intermediate-High grade (82)	28	54		25	57		35	47
Lymph node metastasis
Negative (116)	78	38	0.061	74	42	0.130	83	33	0.081
Positive (90)	49	41		48	42		54	36
TNM stage
I-IIstage (107)	86	21	0.000	81	26	0.000	87	20	0.000
III-IVstage (99)	41	58		41	58		50	49
Molecular Subtype
Luminal A (44)a	32c	12c	0.000	30	14d	0.002	34	10c	0.002
Luminal B (93)b	68	25c,d		68	25c		69	24c
HER2-overexpressing (23)c	6	17a,b		8	15b		10	13a,b
Triple-negative type (46)d	21	25b		16	30a		24	22

Notes: Superscript letters (a, b, c, d) indicate statistically significant differences after Bonferroni correction for multiple pairwise comparisons among diameter, nottingham grade, and molecular subtypes. Comparisons without a superscript letter are not significantly different (corrected P > 0.05).

Figure 3.

Expression patterns of Basigin in invasive breast carcinoma of no special type with different Nottingham grades and in tumor budding areas (immunohistochemistry, ×400 magnification). (A) Predominantly membranous Basigin expression in Nottingham grade I carcinoma. (B) Predominantly cytoplasmic Basigin expression in Nottingham grade II carcinoma. (C) Predominantly membranous Basigin expression in Nottingham grade III carcinoma. (D) Predominantly cytoplasmic Basigin expression in Nottingham grade III carcinoma. (E) Decreased Basigin expression in low-grade tumor budding areas (red arrows). (F) Reduced membranous localization with partial cytoplasmic accumulation of Basigin in high-grade tumor budding areas (red arrows).

Among HER2-overexpressing tumors, 17 of 23 cases (73.91%) demonstrated cytoplasmic Basigin localization, whereas cytoplasmic expression was observed in 25 of 46 triple-negative tumors (54.35%). In contrast, cytoplasmic Basigin localization was less frequent in Luminal A (12/44, 27.27%) and Luminal B tumors (25/93, 26.88%).

High E-cadherin expression was observed in 122 cases (59.22%), whereas low expression was identified in 84 cases (40.78%). Reduced E-cadherin expression was significantly associated with larger tumor size, higher Nottingham grade, high-grade tumor budding, advanced TNM stage, and molecular subtypes (all P < 0.05) (Table 2 and Figure 4A–F).

Figure 4.

Expression patterns of E-cadherin in invasive breast carcinoma of no special type with different Nottingham grades and in tumor budding areas (immunohistochemistry, ×400 magnification). (A) Strong membranous E-cadherin expression in Nottingham grade I carcinoma. (B) Strong membranous E-cadherin expression in Nottingham grade II carcinoma. (C) Strong membranous E-cadherin expression in Nottingham grade III carcinoma. (D) Reduced E-cadherin expression with weakened membranous localization in Nottingham grade III carcinoma compared with adjacent internal control tumor cells (green arrow). (E) Reduced E-cadherin expression with partial cytoplasmic localization in low-grade tumor budding areas. (F) Reduced E-cadherin expression with partial cytoplasmic localization in high-grade tumor budding areas. Red arrows in (E and F) indicate tumor budding areas with reduced E-cadherin expression.

Vimentin positivity was identified in 69 cases (33.50%), whereas 137 cases (66.50%) were negative. Positive Vimentin expression was significantly associated with larger tumor size, higher Nottingham grade, high-grade tumor budding, advanced TNM stage, and molecular subtypes (all P < 0.05) (Table 2 and Figure 5A–D). Pairwise comparisons with Bonferroni correction are indicated by superscript letters in Table 2.

Figure 5.

Expression patterns of Vimentin in invasive breast carcinoma of no special type with different Nottingham grades and in tumor budding areas (immunohistochemistry, ×400 magnification). (A) Minimal Vimentin expression in Nottingham grade I carcinoma. (B) Minimal Vimentin expression in Nottingham grade II carcinoma. (C) High Vimentin expression in Nottingham grade III carcinoma. (D) Increased Vimentin expression in tumor budding areas (red circles).

Quantitative Expression Analysis in Tumor Budding and Non-Budding Regions

Quantitative image analysis demonstrated significantly higher expression levels in tumors with predominantly cytoplasmic Basigin localization compared with tumors showing predominantly membranous localization (20.16 ± 5.01 vs 11.93 ± 2.80, respectively; P < 0.001). Similarly, E-cadherin expression levels were significantly higher in the high-expression group than in the low-expression group (16.20 ± 2.11 vs 7.25 ± 2.87, respectively; P < 0.001). Vimentin-positive tumors also demonstrated significantly higher quantitative expression levels compared with Vimentin-negative tumors (11.01 ± 2.57 vs 0.56 ± 1.14, respectively; P < 0.001) (Table 3 and Figure 6A–C).

Table 3.

Comparison of Quantitative Expression of Basigin, E-Cadherin, and Vimentin Proteins

Maker	Group	PU value	t	P
Basigin	Membrane (127)	11.93±2.80	−15.11	0.000
	Cytoplasmic (79)	20.16±5.01
E-cadherin	Low (84)	7.25±2.87	25.78	0.000
	High (122)	16.20±2.11
Vimentin	Negative (137)	0.56±1.14	40.37	0.000
	Positive (69)	11.01±2.57

Figure 6.

Quantitative expression analysis of Basigin, E-cadherin, and Vimentin proteins. (A) Quantitative comparison of Basigin expression between membranous and cytoplasmic localization groups (P < 0.001). (B) Quantitative comparison of E-cadherin expression between high- and low-expression groups (P < 0.001). (C) Quantitative comparison of Vimentin expression between negative and positive expression groups (P < 0.001). Statistical significance is indicated as follows: ***P < 0.001.

Comparative analysis between tumor budding and non-budding regions demonstrated significantly lower overall Basigin expression in budding regions than in corresponding non-budding regions (7.15 ± 3.84 vs 15.20 ± 5.23, respectively; P < 0.001). However, tumor budding cells frequently exhibited reduced membranous staining with relative cytoplasmic redistribution of Basigin (Figures 3E, F, 7A and Table 4). In addition, E-cadherin expression was significantly reduced in budding regions, whereas Vimentin expression was significantly increased (both P < 0.001) (Figures 4E, F, 5D, 7B, C and Table 4).

Figure 7.

Quantitative expression of Basigin, E-cadherin, and Vimentin in tumor budding and non-budding regions of invasive breast carcinoma. (A) Basigin expression was significantly higher in non-budding regions than in corresponding tumor budding regions (P < 0.001). (B) E-cadherin expression was significantly higher in non-budding regions than in corresponding tumor budding regions (P < 0.001). (C) Vimentin expression was significantly lower in non-budding regions than in corresponding tumor budding regions (P < 0.0001). Statistical significance is indicated as follows: ***P < 0.001.

Table 4.

Quantitative Expression of Basigin, E-Cadherin, and Vimentin in Tumor Budding and Non-Budding Areas of Invasive Breast Carcinoma of No Special Type

Groups	N	Basigin	E-Cadherin	Vimentin
Tumor budding	206	7.15±3.84	6.10±3.05	6.09±6.30
Non-budding areas	206	15.20±5.23	12.84±4.81	4.03±5.17
t		20.39	19.60	−12.65
P		0.000	0.000	0.000

Comparison Between Primary Tumors and Paired Liver Metastases

To further investigate localization changes during metastatic progression, we analyzed Basigin localization and EMT marker expression in 56 paired primary breast invasive ductal carcinomas and corresponding liver metastases. Representative immunohistochemical findings are shown in Figure 8. Primary tumors predominantly demonstrated membranous Basigin staining, whereas paired liver metastases more frequently exhibited reduced membranous staining with predominant cytoplasmic localization. In parallel, E-cadherin expression was reduced and Vimentin expression was increased in metastatic lesions compared with matched primary tumors.

Figure 8.

Expression of Basigin, E-cadherin, and Vimentin in primary invasive breast ductal carcinoma and paired liver metastases (immunohistochemistry, ×400 magnification). (A) Predominantly membranous Basigin expression in primary breast carcinoma. (B) Predominantly cytoplasmic Basigin expression in paired liver metastasis. (C) Strong membranous E-cadherin expression in primary breast carcinoma. (D) Reduced E-cadherin expression in paired liver metastasis. (E) Negative Vimentin expression in primary breast carcinoma. (F) Positive Vimentin expression in paired liver metastasis.

Quantitative analysis demonstrated that the proportion of tumors showing predominantly cytoplasmic Basigin localization increased from 41.07% (23/56) in primary tumors to 75.00% (42/56) in liver metastases, whereas the proportion showing predominantly membranous localization decreased from 58.93% (33/56) to 25.00% (14/56) (P < 0.001) (Table 5 and Figure 9A). Similarly, the frequency of low E-cadherin expression increased from 44.64% (25/56) in primary tumors to 71.43% (40/56) in liver metastases (P < 0.001) (Table 5 and Figure 9B). Vimentin positivity also increased from 53.57% (30/56) in primary tumors to 69.64% (39/56) in liver metastases (P = 0.012) (Table 5 and Figure 9C).

Table 5.

Comparison of Basigin, E-Cadherin, and Vimentin in Primary Breast Invasive Carcinoma of No Special Type and Paired Liver Metastases (56 Cases)

Group (N=56)	Basigin		E-Cadherin		Vimentin
	Cytoplasmic	Membranous	Low	High	Negative	Positive
Primary lesion	23 (41.07%)	33 (58.93%)	25 (44.64%)	31 (55.36%)	26 (46.42%)	30 (53.57%)
Liver metastases	42 (75.00%)	14 (25.00%)	40 (71.43%)	16 (28.57%)	17 (30.35%)	39 (69.64%)
P	0.000		0.000		0.012

Figure 9.

Distribution of Basigin localization patterns, E-cadherin expression levels, and Vimentin status in primary invasive breast ductal carcinoma and paired liver metastases. (A) Basigin localization (cytoplasmic vs membranous). (B) E-cadherin expression (low vs high). (C) Vimentin status (negative vs positive).

Association Between Tumor Budding Morphology and Molecular Subtypes

Among the 206 breast invasive ductal carcinomas, tumor budding morphology was categorized as no budding in 129 cases (62.62%), predominantly cluster budding in 30 cases (14.56%), and predominantly single-cell budding in 47 cases (22.82%). The distribution of tumor budding morphologies differed significantly among molecular subtypes (P < 0.001) (Table 6 and Figure 10). Luminal A and Luminal B tumors were predominantly associated with the absence of tumor budding, accounting for 41 of 44 cases (93.18%) and 82 of 93 cases (88.17%), respectively. In contrast, HER2-overexpressing tumors more frequently demonstrated predominantly cluster budding (18/23 cases, 78.26%), whereas triple-negative tumors predominantly exhibited single-cell budding morphology (42/46 cases, 91.30%). Pairwise comparisons with Bonferroni correction are indicated by superscript letters in Table 6.

Table 6.

Relationship Between Tumor Budding Status and Molecular Subtypes in 206 Cases of Invasive Breast Carcinoma of No Special Type

Molecular Subtype	Tumor Budding Status
	No Tumor Budding (129)	Clustered Tumor Budding (30)	Single-Cell Tumor Budding (47)
Luminal A(44)a	41 (31.78%)c,d	2 (6.67%)d	1 (2.13%)c
Luminal B(93)b	82 (63.57%)c,d	8 (26.67%)d	3 (6.38%)c
HER2-overexpressing(23)c	4 (3.10%)a,b	18 (60.00%)d	1 (2.13%)a,b,d
Triple-negative type(46)d	2 (1.55%)a,b	2 (6.67%)a,b,c	42 (89.36%)c
P	0.000

Notes: Superscript letters (a, b, c, d) indicate statistically significant differences after Bonferroni correction for multiple pairwise comparisons among molecular subtypes. Comparisons without a superscript letter are not significantly different (corrected P > 0.05).

Figure 10.

Association between tumor budding patterns and molecular subtypes in invasive breast carcinoma.

Survival Analysis and Prognostic Factors

Kaplan–Meier survival analysis demonstrated significant associations between clinicopathological variables and disease-free survival (DFS). Patients with TNM stage I–II tumors demonstrated more favorable DFS compared with those with stage III–IV disease (68.22% vs 52.53%, respectively; log-rank χ² = 4.36, P = 0.037). DFS differed significantly according to Nottingham grade, with lower survival observed in higher-grade tumors (grade I: 95.65%, grade II: 73.68%, grade III: 27.54%; log-rank χ² = 83.31, P < 0.001).

Patients with low-grade tumor budding demonstrated significantly more favorable DFS than those with intermediate/high-grade budding (96.77% vs 6.10%, respectively; log-rank χ² = 203.40, P < 0.001). Similarly, DFS differed according to tumor budding morphology, with the most favorable outcomes observed in tumors without budding, intermediate outcomes in tumors with predominantly cluster budding, and the least favorable outcomes in tumors with predominantly single-cell budding (78.29%, 46.67%, and 21.28%, respectively; log-rank χ² = 61.42, P < 0.001).

Predominantly membranous Basigin localization was associated with more favorable DFS compared with predominantly cytoplasmic localization (76.38% vs 35.44%, respectively; log-rank χ² = 44.80, P < 0.001). Reduced E-cadherin expression and positive Vimentin expression were also associated with less favorable DFS (both P < 0.001). In addition, significant survival differences were observed among molecular subtypes, with Luminal A and Luminal B tumors demonstrating more favorable outcomes compared with HER2-overexpressing and triple-negative tumors (log-rank χ² = 88.75, P < 0.001) (Figure 11A–H).

Figure 11.

Kaplan–Meier analysis of disease-free survival in patients with invasive breast carcinoma. (A) Disease-free survival according to TNM stage. (B) Disease-free survival according to Nottingham grade. (C) Disease-free survival according to tumor budding grade. (D) Disease-free survival according to tumor budding pattern. (E) Disease-free survival according to molecular subtype. (F) Disease-free survival according to Basigin subcellular localization. (G) Disease-free survival according to E-cadherin expression status. (H) Disease-free survival according to Vimentin expression status.

Multivariate Cox proportional hazards regression analysis identified Nottingham grade (HR = 4.839, 95% CI: 2.238–10.466, P < 0.001), tumor budding grade (HR = 37.932, 95% CI: 12.850–111.966, P < 0.001), molecular subtype (HR = 1.523, 95% CI: 1.164–1.991, P = 0.002), and Basigin localization (HR = 0.380, 95% CI: 0.171–0.844, P = 0.017) as independent prognostic factors for DFS. Age, tumor size, TNM stage, lymph node metastasis, tumor budding morphology, E-cadherin expression, and Vimentin expression were not independently associated with DFS in the multivariate model (Table 7).

Table 7.

Multivariate Cox Regression Analysis of Independent Prognostic Factors in Patients with Invasive Breast Carcinoma of No Special Type

Parameters	B	SE	Wald	P	Exp(B)	95.0% CI for Exp(B)
						Lower	Upper
TNM stage	−0.328	0.331	0.984	0.321	0.720	0.376	1.378
Nottingham grade	1.577	0.394	16.053	0.000	4.839	2.238	10.466
Tumor budding grade	3.636	0.552	43.342	0.000	37.932	12.850	111.966
Molecular Subtype	0.421	0.137	9.442	0.002	1.523	1.164	1.991
Basigin localization	−0.968	0.407	5.651	0.017	0.380	0.171	0.844
E-cadherin	0.001	0.388	0.000	0.999	1.001	0.467	2.143
Vimentin	0.641	0.514	1.556	0.212	1.898	0.693	5.197

Abbreviations: B, regression coefficient; SE, standard error; Wald, Wald chi-square statistic; Exp(B), Exponential of the regression; HR, hazard ratio; CI, confidence interval.

Nomogram Construction and Validation

A prognostic nomogram incorporating Basigin subcellular localization, tumor budding grade, Nottingham histological grade, and molecular subtype was constructed to estimate 2-, 3-, and 5-year disease-free survival (DFS) in breast cancer patients (Figure 12A). Calibration analysis showed acceptable agreement between predicted and observed survival probabilities in the study cohort (Figure 12B).

Figure 12.

Development and validation of a prognostic nomogram based on Basigin localization, Nottingham grade, molecular subtype, and tumor budding. (A) Nomogram for predicting 2-, 3-, and 5-year disease-free survival in patients with invasive breast carcinoma. (B) Calibration curves showing agreement between predicted and observed disease-free survival probabilities at 2, 3, and 5 years. (C–E) Receiver operating characteristic (ROC) curve analysis evaluating the discriminative performance of the nomogram and individual predictors for predicting 2-year (C), 3-year (D), and 5-year (E) disease-free survival. (F–H) Decision curve analysis (DCA) demonstrating the net clinical benefit of the nomogram and individual predictors for predicting 2-year (F), 3-year (G), and 5-year (D) disease-free survival.

Receiver operating characteristic (ROC) curve analysis demonstrated bootstrap-corrected AUC values of 0.919, 0.921, and 0.980 for predicting 2-, 3-, and 5-year DFS, respectively, derived from 1000 internal validation resamples to correct for optimism bias (Figure 12C–E).

To further evaluate the clinical utility and incremental value of integrating Basigin localization, Decision Curve Analysis (DCA) was performed across all three time points (24, 36, and 60 months) (Figure 12F–H). At 24 months, the Nomogram yielded a net benefit (NB) of approximately 0.064 within the threshold range of 0.01–0.05, exceeding that of the Basigin-alone model (NB≈0.033). At 36 months, this advantage expanded, with the Nomogram maintaining superior (NB≈0.114) across thresholds of 0.01–0.06, compared with Basigin localization alone (NB≈0.058). At 60 months, the clinical utility gap widened substantially: within the clinically meaningful threshold range of 0.05–0.35, the Nomogram’s NB consistently ranged from 0.35 to 0.42, while the Basigin-alone model peaked at only 0.22 and rapidly declined to negative values at higher thresholds. Collectively, these DCA findings confirm that integrating Basigin localization confers time-dependent incremental clinical benefit, with the most pronounced advantage observed at the 60-month endpoint—the primary timepoint of clinical interest for long-term DFS prediction.

Discussion

In this retrospective cohort of 206 breast invasive ductal carcinomas, cytoplasmic redistribution of Basigin was significantly associated with EMT-related marker alterations, increased tumor budding, and adverse clinical outcome. In contrast to previous studies focusing primarily on overall Basigin expression levels, the present study evaluated Basigin according to its subcellular localization pattern and demonstrated that predominantly cytoplasmic localization was independently associated with reduced disease-free survival. These findings suggest that Basigin redistribution may represent a previously underrecognized pathological feature associated with tumor invasiveness in breast cancer.

Previous studies have documented heterogeneous patterns of Basigin subcellular localization across multiple tumor types, although the biological and prognostic implications of these localization patterns remain incompletely understood.^9–12 In lung adenocarcinoma, the cellular localization pattern of EMMPRIN/CD147 has been reported to correlate with patient prognosis independently of overall expression levels.^9,10 In esophageal squamous cell carcinoma, Zhu et al¹¹ reported predominantly membranous Basigin expression with significant associations with tumor invasion depth and lymph node metastasis. In hepatocellular carcinoma, Zhao et al¹² observed both membranous and cytoplasmic patterns, though the significance of cytoplasmic redistribution was not further explored. Collectively, these findings suggest that the pathological significance of Basigin localization patterns may be context-dependent, varying across tumor types and microenvironments. Notably, while membranous localization was associated with adverse prognosis in lung adenocarcinoma,^9,10 Our study in breast cancer identified cytoplasmic localization as an adverse prognostic pattern. This discrepancy underscores the importance of tumor-specific context in interpreting Basigin subcellular localization and highlights the need for further mechanistic studies to elucidate the underlying biological determinants.

Previous studies have provided limited histopathological evidence linking Basigin localization patterns to EMT-associated tumor budding. In the present study, tumor budding regions and paired liver metastases consistently exhibited a phenotype characterized by cytoplasmic Basigin accumulation accompanied by reduced E-cadherin expression and increased Vimentin expression. Quantitative analysis further demonstrated that although the overall Basigin expression level was lower in budding regions than in non-budding regions (7.15 vs 15.20, P < 0.001), cytoplasmic redistribution appeared more prominent within budding cells. These findings suggest that the biological significance of Basigin may depend not only on total protein expression level but also on its subcellular localization pattern. Given Basigin’s established role as a chaperone for monocarboxylate transporters (MCTs) at the plasma membrane,^5,6 the cytoplasmic redistribution observed in budding and metastatic cells may compromise MCT-mediated metabolic homeostasis, potentially contributing to metabolic stress that favors the acquisition of a mesenchymal phenotype. In malignant melanoma, Basigin has additionally been reported to undergo dynamic intracellular redistribution from the plasma membrane to mitochondria, with mitochondrial localization associated with more aggressive biological behavior.¹⁵ Although the mechanistic basis of Basigin redistribution was not investigated in the present study, our findings support a potential association between cytoplasmic Basigin redistribution, EMT-related phenotypic alterations, and invasive tumor budding in breast cancer.

Our results demonstrated significant differences in tumor budding morphology among molecular subtypes of breast invasive ductal carcinoma (P < 0.001). Luminal A and Luminal B tumors predominantly exhibited no evident tumor budding, whereas the HER2-overexpressing subtype was characterized mainly by cluster budding and triple-negative breast cancer (TNBC) predominantly exhibited single-cell budding morphology. Tumor budding is widely regarded as a histomorphological manifestation of EMT,¹⁶ although increasing evidence suggests that EMT within budding cells often represents a partial EMT state characterized by co-expression of epithelial and mesenchymal features.¹⁶ In the present study, the predominance of cluster budding in HER2-overexpressing tumors may reflect a collective migratory pattern with partial preservation of intercellular adhesion, whereas the predominance of single-cell budding in TNBC may indicate a more pronounced loss of epithelial cohesion. These observations suggest that distinct molecular subtypes may exhibit different EMT-associated invasive patterns. Previous studies investigating the association between tumor budding and breast cancer molecular subtypes have reported inconsistent findings.^17–20 Our study further extends these observations by demonstrating subtype-specific budding morphologies, which may provide additional morphological insight into the heterogeneous invasive behavior of different breast cancer subtypes. However, the biological basis and clinical significance of these distinct budding patterns require further validation in larger cohorts and mechanistic studies.

In 56 paired primary breast tumors and corresponding liver metastases, we observed a significant shift in Basigin subcellular localization during metastatic progression, with the proportion of predominantly cytoplasmic Basigin increasing from 41.07% in primary tumors to 75.00% in liver metastases (P < 0.001). This redistribution was accompanied by an increase in low E-cadherin expression (44.64% vs 71.43%) and Vimentin positivity (53.57% vs 69.64%) in metastatic lesions (all P < 0.05). Notably, these phenotypic alterations were highly consistent with those observed in tumor budding regions, suggesting a potential association between cytoplasmic Basigin redistribution, EMT-related marker alterations, and metastatic progression in breast cancer.

Lu et al¹⁵ reported in malignant melanoma that Basigin may undergo dynamic intracellular redistribution from the plasma membrane to mitochondria, with mitochondrial localization associated with enhanced oxidative phosphorylation and more aggressive tumor behavior. In the present study, although Basigin predominantly exhibited cytoplasmic accumulation in breast cancer liver metastases, whether additional mitochondrial redistribution occurs remains unclear. Gallagher et al⁵ further demonstrated that co-expression of membrane-associated Basigin and MCT4 is important for maintaining tumor cell survival within an acidic microenvironment, suggesting that loss of membranous Basigin may be associated with altered metabolic adaptation during tumor progression. Although the mechanistic basis of these localization changes was not investigated in the current study, our findings provide histopathological evidence supporting a potential association between dynamic Basigin redistribution and metastatic progression in breast cancer. In addition, the observed differences between primary and metastatic lesions suggest that reassessment of Basigin localization in metastatic tumors may provide additional prognostic information beyond evaluation of the primary lesion alone.

Kaplan–Meier survival analysis demonstrated that patients with predominantly cytoplasmic Basigin localization had significantly shorter disease-free survival than those with predominantly membranous localization. This finding differs from previous reports in lung cancer, in which membranous Basigin localization was associated with adverse clinical outcome,^9,10 suggesting that the pathological significance of Basigin localization patterns may vary among tumor types. Multivariate Cox regression analysis further identified Basigin localization, Nottingham histological grade, tumor budding grade, and molecular subtype as independent prognostic factors for disease-free survival. Whereas previous studies in breast cancer have primarily focused on the relationship between overall Basigin expression level and prognosis,^13,21 our findings suggest that subcellular localization assessment may provide additional prognostic information beyond total protein expression alone.

In the present study, we integrated Basigin subcellular localization with tumor budding grade, Nottingham histological grade, and molecular subtype to construct a prognostic nomogram for breast cancer. The nomogram demonstrated favorable discriminatory performance within the study cohort, with AUC values exceeding 0.9 for prediction of 2-, 3-, and 5-year disease-free survival. Compared with individual clinicopathological parameters alone, incorporation of Basigin localization appeared to improve risk stratification for recurrence. Given the established prognostic significance of molecular subtype classification systems proposed by Perou et al²² and Sørlie et al,²³ as well as the St Gallen consensus recommendations emphasizing integrated pathological and molecular assessment,²⁴ evaluation of Basigin localization may represent a simple and potentially cost-effective adjunctive pathological parameter, particularly in settings where multigene testing is not routinely available.

The interpretation criteria used in the present study, in which the cytoplasmic pattern was defined as ≥50% of tumor cells showing predominantly cytoplasmic staining, were readily applicable in routine pathological assessment and demonstrated high interobserver reproducibility between two pathologists (kappa = 0.88). Given the successful integration of tumor budding assessment into pathological reporting systems following the ITBCC recommendations in colorectal carcinoma,⁸ incorporation of Basigin localization pattern together with tumor budding grade may provide additional information regarding tumor invasiveness in breast cancer. Notably, HER2-overexpressing and triple-negative tumors demonstrated relatively high frequencies of cytoplasmic Basigin localization, suggesting that localization assessment may be particularly relevant in these aggressive molecular subtypes. However, the potential clinical utility of routine Basigin localization evaluation requires further validation in independent cohorts and prospective studies.

Several limitations of this study should be acknowledged. First, this was a retrospective single-center study, and the findings require validation in larger multicenter cohorts. Second, we excluded patients who received neoadjuvant or targeted therapies to minimize treatment-related confounding and to allow assessment of the natural biological association between Basigin localization and tumor aggressiveness. While this approach was methodologically justified, it limits the generalizability of our findings to patient populations treated with contemporary standard-of-care regimens. Future studies incorporating treated patients are needed. Third, the mechanistic investigation was limited to histopathological observations, and the molecular mechanisms underlying Basigin subcellular redistribution were not experimentally explored. Fourth, although dynamic alterations in Basigin localization were observed in paired liver metastases, the metastatic cohort size was relatively limited (56 cases), and other metastatic sites, such as lung and bone metastases, were not evaluated. In addition, the subgroup analyses—particularly the HER2-overexpressing group (n=23) and the paired liver metastasis cohort (n=56)—had limited statistical power. These exploratory findings should therefore be interpreted with caution and require validation in larger independent cohorts. Fifth, our IHC evaluation focused on the percentage of cells with specific subcellular localization rather than staining intensity. While this approach demonstrated excellent interobserver reproducibility (κ=0.88) and was consistent with our previous methodology,⁷ we acknowledge that combined assessment using H-score or other quantitative methods may provide additional information. Future studies incorporating both localization pattern and staining intensity are warranted. Sixth, while all patients received standard adjuvant therapy according to guidelines, detailed information on specific regimens was not available due to the retrospective design, representing an important unmeasured confounder. Seventh, the nomogram was constructed and validated in a single-center cohort without external validation; therefore, its predictive performance should be considered preliminary and requires independent validation before clinical application.

In addition, previous studies in malignant melanoma have reported dynamic “membrane-to-mitochondria” translocation of Basigin.¹⁵ Whether similar organelle-specific redistribution occurs in breast cancer and contributes to metabolic reprogramming during metastasis remains speculative and requires further investigation using advanced techniques such as immuno-electron microscopy or organelle fractionation. Future studies should explore whether Basigin localization could serve as a predictive biomarker for therapies targeting MCT-mediated metabolic pathways, as membrane-associated Basigin is required for MCT function. Additionally, the findings of the present study were obtained from invasive breast carcinoma of no special type; whether similar associations exist in other breast cancer subtypes, such as lobular carcinoma or rare histologic types, remains to be investigated. Future studies integrating spatial pathology, metabolic profiling, and functional experiments are therefore warranted to further elucidate the biological significance of Basigin translocation during breast cancer progression and metastasis.

Conclusion

Translocation of Basigin from the cell membrane to the cytoplasm was significantly associated with reduced E-cadherin expression, increased Vimentin expression, and high-grade tumor budding, indicating that Basigin redistribution correlates with EMT-associated tumor invasion and metastatic progression. Cytoplasmic Basigin localization was identified as an independent adverse prognostic factor in breast cancer, with predictive significance comparable to established pathological parameters, including molecular subtype, histological grade, and tumor budding grade. Notably, although HER2-overexpressing and triple-negative breast cancers are driven by distinct molecular mechanisms, both demonstrated enrichment of a cytoplasmic Basigin-associated EMT/tumor budding phenotype, suggesting that Basigin localization may serve as a cross-subtype marker of tumor invasiveness. Assessment of Basigin localization may provide additional prognostic information that warrants further investigation, beyond conventional clinicopathological parameters. A prognostic nomogram integrating Basigin localization, tumor budding, histological grade, and molecular subtype may serve as a potential adjunctive pathological parameter, pending prospective validation and standardized interpretation protocols. This nomogram showed favorable predictive performance within the present cohort; however, external validation using independent patient cohorts is required prior to any clinical implementation.

Funding Statement

This study was supported by the Zhengzhou City Technological Innovation Project (No. 2025YLZDJH311) and the Zhengzhou People’s Hospital Internal Research Project (No. zy20250003).

Abbreviations

AUC, Area under the curve; CI, Confidence interval; CSCO, Chinese Society of Clinical Oncology; DAB, Diaminobenzidine; DCA, Decision curve analysis; DFS, Disease-free survival; EMT, Epithelial–mesenchymal transition; EMMPRIN, Extracellular matrix metalloproteinase inducer; FFPE, Formalin-fixed, paraffin-embedded; H&E, Hematoxylin and eosin; HER2, Human epidermal growth factor receptor 2; HIF-1α, Hypoxia-inducible factor 1-alpha; HR, Hazard ratio; IHC, Immunohistochemistry; ITBCC, International Tumor Budding Consensus Conference; MCT, Monocarboxylate transporter; MMP, Matrix metalloproteinase; NB, Net benefit; PU, Positive unit; ROC, Receiver operating characteristic; SD, Standard deviation; TNBC, Triple-negative breast cancer.

Data Sharing Statement

All data are included in the manuscript.

Ethics Approval and Consent to Participate

All procedures involving human participants were conducted in accordance with the ethical standards of the Research Committee of Zhengzhou People’s Hospital and the Declaration of Helsinki and its subsequent amendments or comparable ethical standards. The study was approved by the Bioethics Committee of Zhengzhou People’s Hospital (Approval No. 2026-KY-007001). All methods were performed in accordance with relevant guidelines and regulations.

Author Contributions

Xiaoyan Xu: Conceptualization, Methodology, Formal analysis, Project administration, Writing – original draft. Yaozhong Zhang: Methodology, Investigation, Data curation, Formal analysis, Writing – original draft. Huijuan Fang and Xianwei Xu: Conceptualization, Supervision, Project administration, Writing – review & editing. Huang Jiang and Shuguang Liu: Conceptualization, Methodology, Formal analysis, Project administration, Writing – review & editing. All authors made a significant contribution to the work reported, whether in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no competing interests in this work.