Preoperative prediction of breast cancer Ki-67 status via multimodal ultrasound-clinical nomogram: a single-center study Authors: Xiao-Kai Lu, Nian-Qiu Liu, Yi-Hang Li, Zhi-Yao Li, Dong Chen, Zhi-Rui Chuan, Yin-Xi Qu, Ying-Xian Zhang, Hai-Tao Chen, Xiao-Mao Luo Published: 2025 ABSTRACT AbstractBackgroundBreast cancer is a highly heterogeneous malignancy that poses a major health threat to women. Contemporary clinical management relies heavily on molecular subtyping, in which the Ki-67 proliferation index serves as a critical biomarker for assessing tumor aggressiveness and recurrence risk. However, conventional Ki-67 evaluation depends on invasive biopsy and immunohistochemical analysis, whose accuracy can be compromised by tumor heterogeneity-induced sampling errors and inter-observer variability. Therefore, this study aimed to develop and validate a preoperative nomogram that integrates multimodal ultrasound features with clinical parameters to enable non-invasive preoperative assessment of Ki-67 expression status in breast cancer.MethodsWe retrospectively enrolled 142 consecutive breast cancer patients from The Third Affiliated Hospital of Kunming Medical University’s Breast Center (March 2022 to August 2024). Preoperative multimodal ultrasound parameters (including B-mode, Doppler, and shear-wave elastography) and clinical variables were systematically documented using standardized protocols. Variables demonstrating univariate associations (P<0.10) underwent forward stepwise multivariate regression using likelihood ratio criteria. Model performance was assessed through: (I) calibration curves with Hosmer-Lemeshow test; (II) discrimination via area under the curve (AUC); and (III) clinical utility by decision curve analysis. Internal validation employed bootstrap resampling (1,000 replicates) with optimism correction using Harrell’s method.ResultsUnivariate analysis identified six predictors associated with Ki-67 status (P<0.10): maximum lesion diameter, hyperechoic halo presence, Adler grade, Eratio, posterior echo reduction, and calcifications. Multivariate analysis confirmed four independent predictors of Ki-67 status (P<0.05): hyperechoic halo presence [adjusted odds ratio (aOR) =7.934; 95% confidence interval (CI): 2.604–24.173], posterior echo reduction (aOR =0.245; 95% CI: 0.099–0.601), calcifications (aOR =3.524; 95% CI: 1.466–8.472), and Adler grade (aOR =2.334; 95% CI: 1.222–4.456). The resulting nomogram demonstrated good discrimination (AUC =0.797; 95% CI: 0.722–0.872), with bootstrap-corrected AUC of 0.771 (95% CI: 0.673–0.879).ConclusionsThe validated nomogram provides clinically useful preoperative prediction of Ki-67 status (AUC =0.797; bootstrap-corrected 0.771), with hyperechoic halo presence, posterior echo reduction, calcifications, and high Adler grade as key predictors.Keywords: Breast cancer, multimodal ultrasound, Ki-67, shear-wave elastography (SWE) Highlight box Key findings • A preoperative nomogram integrating multimodal ultrasound and clinical features was developed and validated for non-invasive prediction of Ki-67 status in breast cancer. • Four independent ultrasound predictors were identified: hyperechoic halo, calcifications, Adler grade, and posterior echo reduction. • The nomogram demonstrated good predictive performance with an area under the curve (AUC) of 0.797 (bootstrap-corrected AUC =0.771) and was well-calibrated. What is known and what is new? • The Ki-67 proliferation index is a crucial biomarker for breast cancer molecular subtyping and prognosis. Its current assessment relies on invasive core needle biopsy, which is susceptible to sampling error and inter-observer variability. • This study presents the first nomogram that comprehensively integrates multimodal ultrasound features to preoperatively predict Ki-67 status. It provides a validated, non-invasive tool that may help in clinical decision-making before surgery or biopsy. What is the implication, and what should change now? • This nomogram offers a practical, non-invasive method to assess tumor proliferation potential preoperatively. Clinicians could use it for initial risk stratification, potentially optimizing personalized treatment planning. Future multicenter studies should focus on external validation and integration into clinical workflows. Introduction Breast cancer represents the second most common malignancy globally, accounting for 11.6% of all cancer cases and posing the foremost threat to women’s health (1). As a clinically heterogeneous disease, it exhibits substantial variations in histomorphology, biological behavior, therapeutic response, and molecular profiles (2). Contemporary management employs multimodal strategies integrating surgical resection with tailored adjuvant therapies (chemotherapy, radiotherapy, endocrine therapy) based on tumor biology and patient status to optimize oncological outcomes while preserving quality of life. Nevertheless, delayed diagnosis remains a critical contributor to breast cancer mortality (3). Consequently, early detection is paramount for improving patient prognosis. Conventional prognostic markers—including tumor size, lymph node involvement, and histological grade—predict recurrence risk and overall survival (4). However, molecular characterization has revealed that biomarker expression patterns critically guide personalized treatment selection and prognostic stratification (2,5). Based on expression profiles of estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and Ki-67, breast cancers are classified into four molecular subtypes: Luminal A, Luminal B, HER2-enriched, and basal-like. The 2011 St. Gallen Consensus established Ki-67 index as the definitive criterion for distinguishing Luminal B from Luminal A subtypes in ER-positive/HER2-negative disease (6). Ki-67, a non-histone nuclear protein encoded by MKI67, regulates nucleolar organization during interphase (G1, S, G2). Its expression progressively increases throughout the cell cycle, peaking during mitosis (M phase) while remaining undetectable in quiescent cells (G0 phase) (7). Expression levels peak during mitosis (M phase), reflecting direct correlation with proliferative activity (8). These characteristics establish Ki-67 as the gold-standard indicator of cellular proliferation. As a proliferation marker, Ki-67 predicts tumor aggressiveness and therapeutic response across multiple malignancies including breast, prostate, and gastrointestinal cancers (9). Capitalizing on its biological role, Ki-67-targeted antisense oligonucleotides (ASOs) have been developed to suppress proliferation and induce apoptosis across multiple malignancies. This phenomenon has similar results in various cell lines, including IM-9, RT-4, RM-11, MB-49, 4T1, U2OS, and MHT (10,11). Clinically, Ki-67 serves as a critical prognostic biomarker owing to its specificity for proliferating cells and whole-cell-cycle detectability. Large-scale clinical evidence demonstrates progressive Ki-67 upregulation across the pathological continuum from benign lesions to ductal carcinoma in situ (DCIS) and invasive breast carcinoma. Multicenter studies confirm that elevated Ki-67 indices independently predict reduced disease-free survival and increased recurrence risk (12). Collectively, these findings provide the mechanistic rationale for Ki-67-guided personalized therapeutic strategies. Conventional Ki-67 assessment requires invasive biopsy for immunohistochemical analysis, which suffers from inherent sampling limitations and poor reproducibility due to tumor heterogeneity (13). These issues are compounded by the limited analytical validity of Ki-67, which arises from inter-observer variability, non-standardized counting methods, and ambiguous cut-off values. This underscores the need for non-invasive quantification methods. Ultrasonography provides a cost-effective, non-invasive modality for preoperative breast lesion characterization with excellent reproducibility. Shear-wave elastography (SWE) quantifies tissue stiffness—a biomechanical property reflecting extracellular matrix remodeling associated with tumorigenesis and progression (14). Emerging evidence correlates increased SWE values with adverse prognostic factors including advanced tumor stage and lymphovascular invasion (15). This study therefore investigates multimodal ultrasound features combined with clinical parameters to develop a preoperative nomogram for non-invasive Ki-67 prediction. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/rc). Methods Data collection and patient selection Between March 2022 and August 2024, 173 consecutive female breast cancer patients treated at our Breast Cancer Center were initially screened. Thirty-one patients were excluded based on predefined criteria. The final cohort (n=142) underwent preoperative multiparametric ultrasound examination, with comprehensive clinical and imaging datasets enabling complete lesion characterization. Inclusion criteria: (I) no prior neoadjuvant therapy or diagnostic biopsy. (II) Pathologically confirmed breast cancer. (III) Axillary lymph node status histologically verified. (IV) High-quality ultrasound images permitting clear feature extraction. (V) Complete clinical and imaging records. Exclusion criteria: (I) indeterminate pathological diagnosis. (II) Receipt of neoadjuvant therapy. (III) Incomplete clinical or imaging data. (IV) Male gender. Therefore, a complete-case analysis was employed, and no data imputation was necessary. This approach ensured a robust dataset for model development without the potential biases introduced by imputation methods. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of Yunnan Cancer Hospital (No. KYLX2023-021). The ethics committee granted a waiver of informed consent for this retrospective analysis, with all data anonymized prior to analysis. Ultrasound protocol Ultrasound examinations were performed using the Aixplorer V system (Supersonic Imagine, Aix-en-Provence, France) equipped with an SL15-4 linear transducer (4–15 MHz). Patients underwent standardized scanning in supine and contralateral decubitus positions. Comprehensive assessment included: lesion size, internal echogenicity, posterior acoustic features, margins, shape, hyperechoic halo, calcifications, Adler blood flow grading, SWE parameters [Emax (maximum stiffness) and Eratio (lesion-to-fat stiffness ratio)], and ipsilateral axillary lymph nodes. Two board-certified ultrasound physicians (>5 years breast imaging experience) independently performed and interpreted all examinations. Clinical data collection The patient’s medical records and pathological reports were retrospectively investigated, and clinical information such as age, menopausal status, body mass index (BMI), and number of pregnancies were collected. In this study, Ki-67 expression was evaluated based on core needle biopsy specimens obtained at the time of initial diagnosis. All Ki-67 staining was performed uniformly using the same automated platform and reagents. Specifically, the antibody used was the CONFIRMTM anti-Ki-67 [30-9] Rabbit Monoclonal Primary Antibody (Ventana Medical Systems, Inc., Tucson, Arizona, USA). The staining was performed on a Ventana BenchMark series automated stainer strictly following the manufacturer’s recommended protocol, ensuring consistent staining conditions (including antibody clone, dilution, incubation time, antigen retrieval, and detection system) across all cases. The Ki-67 proliferation index was assessed using the global counting method. Pathologists systematically evaluated the entire tumor section at low power to select at least three regions representing the spectrum of staining heterogeneity. A consecutive count of no less than 1,000 tumor cells was performed at high power, and the final index was calculated as the average percentage of positive cells across these regions (16). In accordance with international consensus (including the St. Gallen International Expert Consensus and the monarchE trial), and as defined by the 2023 Chinese Society of Clinical Oncology (CSCO) Breast Cancer Guidelines, high Ki-67 expression was classified as ≥20% positively stained nuclei (17-19). Statistical analysis Statistical analyses were conducted using SPSS 26.0 (IBM Corp.) for regression modeling and GraphPad Prism 10 (GraphPad Software) for graphical representation. Univariate logistic regression first evaluated associations between multimodal ultrasound features/clinical variables and Ki-67 status (high vs. low expression). Variables showing univariate associations (P<0.10) underwent multivariate forward stepwise logistic regression with likelihood ratio criterion for variable selection. Statistical significance was defined as P<0.05. Significant independent predictors (P<0.05) were incorporated into a nomogram constructed with the ‘rms’ package in R 4.3.1 (R Foundation). Model performance was assessed through: ❖ Calibration curves with Hosmer-Lemeshow goodness-of-fit test; ❖ Calibration curves with Hosmer-Lemeshow goodness-of-fit test; ❖ Discriminatory capacity via area under the curve (AUC); ❖ Discriminatory capacity via area under the curve (AUC); ❖ Clinical net benefit by decision curve analysis (DCA). ❖ Clinical net benefit by decision curve analysis (DCA). Internal validation employed 1,000 bootstrap replicates with optimism correction using Harrell’s method. Results Univariate analysis demonstrated no significant associations (P≥0.1) between Ki-67 status and demographic variables (age, menopausal status, BMI, parity), lymph node involvement, lesion morphology, or Emax values (Table 1). Significant univariate predictors (P<0.10) included maximum lesion diameter, hyperechoic halo presence, Adler grade, Eratio, posterior echo reduction, and microcalcifications (Table 2). Table 1A comparison of the characteristics of ultrasound imaging features and clinical features between the Ki-67 high-expression group and the low-expression groupCharacteristicsHigh expression group (n=100)Low expression group (n=42)OR95% CIP valueAge (years)50.6±10.350.5±10.31.1150.3679–3.0620.84BMI (kg/m2)23.48±3.4623.53±3.481.0280.9271–1.1480.61Menopausal status1.0000.5788–2.4650.63 Premenopausal52 (52.0)20 (47.6) Postmenopausal48 (48.0)22 (52.4)Number of pregnancy0.83110.3955–1.7190.62 0–145 (45.0)17 (40.5) ≥255 (55.0)25 (59.5)Lymph node status1.4480.6889–3.1410.34 Positive58 (58.0)28 (66.7) Negative42 (42.0)14 (33.3)Maximum diameter (mm)33.61±22.6433.4±22.631.0321.007–1.0660.03*Shape0.74510.2888–1.7700.52 Regular25 (25.0)8 (19.0) Irregular75 (75.0)34 (81.0)Hyperechoic halo3.6211.539–9.5950.005* Positive42 (42.0)7 (16.7) Negative58 (58.0)35 (83.3)Posterior acoustic0.39290.1810–0.82310.02* Non-attenuation56 (56.0)14 (33.3) Attenuation44 (44.0)28 (66.7)Calcification3.1111.486–6.6450.003* Yes70 (70.0)18 (42.9) No30 (30.0)24 (57.1)Adler grade2.0991.223–3.7140.008* I15 (15.0)7 (16.7) II41 (41.0)30 (71.4) III44 (44.0)5 (11.9)Emax198.89±63.82198.61±68.321.0030.9972–1.0080.35Eratio12.37±8.1812.30±8.151.0470.9978–1.1070.08Data are presented as mean ± standard deviation or n (%). *, statistically significant. BMI, body mass index; CI, confidence interval; OR, odds ratio. A comparison of the characteristics of ultrasound imaging features and clinical features between the Ki-67 high-expression group and the low-expression group Data are presented as mean ± standard deviation or n (%). *, statistically significant. BMI, body mass index; CI, confidence interval; OR, odds ratio. Table 2Multivariable logistic regression analysis of factors significantly associated with Ki-67 expression status selected from comparative characteristicsCharacteristicsHigh expression group (n=100)Low expression group (n=42)OR95% CIPHyperechoic halo7.9342.604–24.173<0.001* Positive42 (42.0)7 (16.7) Negative58 (58.0)35 (83.3)Adler grade2.3341.222–4.4560.01* I15 (15.0)7 (16.7) II41 (41.0)30 (71.4) III44 (44.0)5 (11.9)Posterior acoustic0.2450.099–0.6010.002* Non-attenuation56 (56.0)14 (33.3) Attenuation44 (44.0)28 (66.7)Calcification3.5241.466–8.4720.005* Yes70 (70.0)18 (42.9) No30 (30.0)24 (57.1)Data are presented as n (%) unless otherwise stated. *, statistically significant. CI, confidence interval; OR, odds ratio. Multivariable logistic regression analysis of factors significantly associated with Ki-67 expression status selected from comparative characteristics Data are presented as n (%) unless otherwise stated. *, statistically significant. CI, confidence interval; OR, odds ratio. Multivariate analysis established four independent predictors of high Ki-67 expression: hyperechoic halo presence [adjusted odds ratio (aOR) =7.934; 95% confidence interval (CI): 2.604–24.173], posterior echo reduction (aOR =0.245; 95% CI: 0.099–0.601), calcifications (aOR =3.524; 95% CI: 1.466–8.472), and Adler grade (aOR =2.334; 95% CI: 1.222–4.456) (all P<0.05; Table 2). These predictors were integrated into a clinically applicable nomogram quantifying individualized risk of high Ki-67 expression (Figure 1). The nomogram was used to predict the probability of multimodal ultrasound predicting Ki-67 status in breast cancer. Calculate the probability, find out the prediction point on the highest point scale corresponding to each patient variable, and sum it. The total number of points projected onto the bottom scale represents the probability of multimodal ultrasound predicting the Ki-67 status of breast cancer. The nomogram demonstrated moderate discrimination (AUC =0.796, 95% CI: 0.722–0.872; Figure 2) and excellent calibration (Hosmer-Lemeshow, P=0.79). Bootstrap validation (1,000 replicates) yielded apparent C-index of 0.797 (95% CI: 0.734–0.875) and optimism-corrected C-index of 0.771 (95% CI: 0.673–0.879), with calibration slope 0.92 (95% CI: 0.85–0.99) indicating minimal overfitting (Figure 3). This indicates moderate discriminative capacity with well-maintained calibration accuracy. The prediction of the expression level of Ki-67 using the logistic regression analysis of certain ultrasonographic parameters. AUC, area under the curve; CI, confidence interval. The calibration curve illustrates the observation and predictive diagnostic rate of multimodal ultrasound for Ki-67 status of breast cancer. DCA demonstrated significant net clinical benefit across threshold probabilities (10–70%), supporting clinical utility for Ki-67 stratification (Figure 4). The clinical decision curve showed that the nomogram model had a higher net benefit in predicting the risk of high Ki-67 expression. Mode 1: represents the nomogram of this study. Discussion Ultrasonography offers distinct advantages for breast cancer screening, including absence of ionizing radiation, cost-effectiveness, non-invasiveness, and excellent procedural reproducibility. Critical knowledge gaps persist regarding the association between sonographic biomarkers and key biological prognosticators in breast cancer. This study therefore investigated specific sonographic signatures—hyperechoic halo, microcalcifications, posterior acoustic alteration, maximum diameter, Adler grade, and Eratio to establish their correlation with Ki-67 status, aiming to develop a non-invasive preoperative predictor for clinical decision support. This approach provides a clinically actionable framework for personalizing therapeutic strategies, monitoring treatment response, and ultimately improving breast cancer outcomes. This study established hyperechoic halo presence as an independent predictor of elevated Ki-67 expression, demonstrating a significant positive correlation. Conventional two-dimensional (2D) ultrasonography defines hyperechoic halo as a circumferential band of increased echogenicity enveloping the lesion periphery. Histopathologically, this feature correlates with spiculated tumor borders and peritumoral infiltration patterns. As a validated malignancy indicator (20). Pathological studies have confirmed that this feature represents the active proliferation of cancer cells, lymphocyte infiltration, and reactive hyperplasia of fibrous connective tissue around malignant tumors (21). Furthermore, subsequent studies demonstrate that an increased hyperechoic halo width significantly correlates with poor patient prognosis. These evidences establish hyperechoic halo as a quantifiable imaging biomarker of tumor invasiveness. Further analysis reveals that elevated Ki-67 expression—a marker protein of cellular proliferation—is strongly associated with aggressive tumor behavior. The imaging features of hyperechoic halo essentially reflect the proliferative active area at the edge of the tumor and the surrounding matrix reaction, which is intrinsically consistent with Ki-67-driven cell cycle progression. Thus, hyperechoic halo provides dual validation: imaging confirmation of Ki-67-associated proliferation dynamics and cross-modal evidence of tumor biology through radiology-pathology correlation. In this study, calcification is also an independent factor affecting the expression of Ki-67, and there is a positive correlation between them. Two distinct pathomechanisms underlie breast cancer microcalcifications: dystrophic calcification: resulting from ischemic necrosis in rapidly proliferating tumors with compromised vascular supply (22). The other is cell-mediated breast cancer microcalcification formation. This pathological calcification may be related to the active secretion process of tumor cells. Some breast cells undergo epithelial-mesenchymal transition (EMT) under specific stimulation to obtain osteogenic function. Pathological calcification is formed by simulating ossification process with the participation of osteogenic related proteins such as bone morphogenetic protein (BMP) and alkaline phosphatase (ALP). Comprehensive analysis found that the mechanisms of these two calcifications are potentially associated with excessive cell proliferation. The first type of necrotic calcification directly reflects the microenvironment disorder caused by excessive tumor proliferation. The second type of cell-mediated calcification forms a cross-regulatory network with proliferation signaling pathways (such as Wnt/β-catenin) through the EMT process to promote cell invasion and metastasis (23). It has been reported that microcalcification in breast cancer is closely related to the recurrence of breast ductal carcinoma and the invasion and metastasis of breast invasive ductal carcinoma (24). In summary, we believe that microcalcification is not only an important marker of breast cancer imaging diagnosis, but also the proliferation-related pathways involved in its formation mechanism may provide a new perspective for the study of the biological function of Ki-67. A significant inverse correlation was observed between posterior acoustic attenuation and Ki-67 expression. As one of the typical ultrasonic features of breast cancer, posterior echo attenuation is due to the abnormal deposition of collagen and collagen fibers in tumor tissue, the absorption and scattering of acoustic energy by necrotic hemorrhage and calcification (25). Well-differentiated tumors demonstrate higher prevalence of posterior attenuation compared to poorly-differentiated/invasive carcinomas. Low-grade tumors usually show slow growth rate, low metastatic tendency and usually inert clinical course. The biological behavior of this phenomenon often represents a limited demand for cell proliferation, which is consistent with the low expression of Ki-67. This further confirms that the posterior echo attenuation can be used as an imaging marker to assist in predicting the low expression of Ki-67 in tumors. Adler blood flow grade independently predicted elevated Ki-67 expression, demonstrating significant positive correlation. Relevant studies suggest that Ki-67 plays an important role in cell proliferation and differentiation, and is positively correlated with vascular endothelial growth factor (VEGF). VEGF enhances tumor angiogenesis through dual mechanisms: neovascularization induction and vascular permeability elevation, ultimately increasing microvessel density (MVD). These processes manifest sonographically as hypervascularity (high Adler grade) and accelerated tumor growth. However, some researchers believe that HER2, as a proto-oncogene, plays an important role in inhibiting tumor apoptosis, promoting cell proliferation, and promoting the formation of malignant tumor microvessels and lymphatic vessels. While our model established Adler grade as an independent predictor, the complex interrelationships among tumor size, vascularity, and proliferation markers warrant further mechanistic investigation. This study analyzed several SWE parameters, including Emax and Eratio. Eratio quantifies relative stiffness as lesion stiffness divided by adjacent fat stiffness. During the development of breast cancer, cancer-related growth factors in the tumor microenvironment directly or indirectly stimulate the expression of VEGF, connective tissue growth factor (CTGF) and fibroblast growth factor (FGF), and promote fibrosis, collagen fiber deposition and tumor angiogenesis (26). Lesion stiffness directly correlates with collagen content. Studies have suggested that breast cancer with poor prognosis may have higher Emax and Eratio values according to histological prognostic characteristics, immunohistochemical characteristics and subtypes (27). However, in this study, there was no significant difference between Emax and Eratio values and Ki-67 expression status. This discrepancy warrants validation in larger cohorts to clarify SWE’s role in proliferation assessment. There are several limitations in this study. First of all, the interpretation of breast ultrasound features is largely influenced by the personal experience of ultrasound doctors. The ultrasound section images stored in the database may cause the features of other ultrasound sections to be ignored. Secondly, there are many pathological types of breast cancer, but this study only focuses on invasive breast cancer, so its representativeness is limited; finally, this study is a retrospective, single-center study with a small total sample size. Future multicenter, prospective studies with larger sample sizes are warranted to externally validate our model and improve its general applicability. Conclusions In summary, the validated nomogram provides clinically useful preoperative prediction of Ki-67 status, with the presence of a hyperechoic halo, posterior echo reduction, calcifications, and a high Adler grade on multimodal ultrasound identified as key predictors. The model developed in this study demonstrates good predictive performance and generalizability for estimating Ki-67 expression levels. In future clinical practice, utilizing these ultrasound features could enable a preliminary assessment of breast cancer prognosis, thereby enhancing the clinical utility of ultrasonography in the diagnosis and management of breast cancer. Supplementary The article’s supplementary files as Acknowledgments None. Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of Yunnan Cancer Hospital (No. KYLX2023-021). The ethics committee granted a waiver of informed consent for this retrospective analysis, with all data anonymized prior to analysis. Footnotes Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/rc Funding: This study was supported by scientific research foundation of Yunnan Education Department (No. 2023Y0654), Yunnan Provincial Department of Education Youth Talent Basic Research Project (No. 2024J0248), Yunnan Provincial Department of Science and Technology-Kunming Medical University Joint Special Fund for Applied Basic Research-Key Projects (Nos. 202301AY070001-009 and 202401AY070001-042), Yunnan Provincial Department of Science and Technology-Kunming Medical University Applied Basic Research Joint Special Fund-General Project (Nos. 202401AY070001-268 and 202501AY070001-108). Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/coif). All authors declare that this study was supported by scientific research foundation of Yunnan Education Department (No. 2023Y0654), Yunnan Provincial Department of Education Youth Talent Basic Research Project (No. 2024J0248), Yunnan Provincial Department of Science and Technology-Kunming Medical University Joint Special Fund for Applied Basic Research-Key Projects (Nos. 202301AY070001-009 and 202401AY070001-042), Yunnan Provincial Department of Science and Technology-Kunming Medical University Applied Basic Research Joint Special Fund-General Project (Nos. 202401AY070001-268 and 202501AY070001-108). The authors have no other conflicts of interest to declare. Data Sharing Statement Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/dss References Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/prf Supplementary Materials The article’s supplementary files as Data Availability Statement Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1652/dss