A nomogram integrating clinical and multimodal ultrasound features for pretreatment prediction of HER2 status in breast cancer: a retrospective study

Introduction

Breast cancer is the most common cancer in women, accounting for 11.7% of all new cancer cases diagnosed annually. GLOBOCAN 2020 estimates indicate that breast cancer accounted for 6.9% of global cancer deaths and represents the leading cause of cancer mortality in 110 countries (1). As a heterogeneous disease, breast cancer is categorized into distinct molecular subtypes according to their specific molecular markers. The expression profiles of these markers and the resulting molecular classification are critical for guiding personalized treatment plans and improving patient outcomes. Human epidermal growth factor receptor 2 (HER2), a transmembrane tyrosine kinase receptor, is a key oncogenic driver, a prognostic marker, and a primary predictive biomarker for anti-HER2 therapy in breast cancer. Approximately 15–20% of breast cancers exhibit HER2 overexpression. Furthermore, increased angiogenesis and vascular endothelial growth factor (VEGF) expression are closely linked to HER2 activation in breast cancer cells (2). HER2-positive breast cancer is associated with a high proliferation rate, an elevated risk of metastasis, shorter disease-free survival, and consequently, a poorer overall prognosis (3,4). Advancements in molecular biology have established anti-HER2 targeted therapy as a cornerstone of the standard treatment regimen. Consequently, this therapeutic approach has led to significant improvements in recurrence rates, disease-free survival, and overall survival among patients with HER2-positive breast cancer (5,6). Clinical studies have demonstrated that trastuzumab-based therapy reduces the risk of disease recurrence by approximately 50% in patients with early-stage HER2-positive breast cancer. Despite the significant clinical efficacy of trastuzumab in combination with adjuvant chemotherapy, drug resistance remains a major challenge to its long-term success. Therefore, profiling the molecular, genetic, and immunologic characteristics of tumor cells is essential prior to initiating targeted therapy (7). The decision to incorporate anti-HER2 targeted therapy into a treatment regimen is primarily determined by the HER2 receptor status of the tumor. Current clinical practice for assessing HER2 status relies on immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) assays performed on biopsy samples (8). As invasive procedures, both IHC and FISH are associated with potential complications, including pain, infection, and in rare cases, tumor seeding. Because biopsy is invasive, repeated assessment is not always easy or feasible. Breast cancer is widely recognized as a heterogeneous disease, a characteristic further evidenced by intratumoral heterogeneity in HER2 expression, which is observed in approximately 40% of cases (9). Consequently, a pretreatment needle biopsy alone may not fully capture the heterogeneous nature of the entire tumor. These limitations highlight the urgent need for the development of convenient, non-invasive, and accurate diagnostic tools to assess HER2 status.

Ultrasonography is one of the most widely utilized imaging modalities for the pretreatment diagnosis and evaluation of breast cancer. Compared to other imaging techniques, it offers distinct advantages, including cost-effectiveness, procedural simplicity, and the absence of ionizing radiation. Accumulating evidence suggests that various sonographic features are potentially associated with specific molecular characteristics of breast cancer (10,11). Previous studies have identified associations between HER2 expression status and clinical parameters such as patient age, clinical T stage, and lymph node status (12). Given the persisting challenges in quantitatively extracting sonographic features, this study aims to develop a predictive model for HER2 status by integrating clinical data with multimodal ultrasound characteristics. This approach seeks to leverage fundamental, readily available information to inform treatment strategy formulation and prognosis assessment. We present this article in accordance with the TRIPOD reporting checklist (available at https://cco.amegroups.com/article/view/10.21037/cco-2026-1-0029/rc).

Methods

Data collection and patient selection

This study was conducted in accordance with the principles of the Declaration of Helsinki and its subsequent amendments. The study protocol was reviewed and approved by the Institutional Ethics Committee of Yunnan Cancer Hospital (approval No. KYLX2023-021), which granted a waiver of informed consent for this retrospective analysis. All data were anonymized prior to analysis.

This study enrolled 173 female patients with breast cancer who were managed at the Department of Breast Surgery, Yunnan Cancer Hospital, between March 2022 and August 2024. A total of 31 patients were excluded due to not meeting the inclusion criteria, such as a history of prior treatment, absence of a definitive diagnosis, male sex, or incomplete patient information. All remaining patients underwent pretreatment multimodal ultrasound examinations and had complete clinical and ultrasound data, enabling a comprehensive evaluation of the lesion characteristics.

The inclusion criteria were as follows: (I) complete clinical and imaging data; (II) no history of prior treatment; (III) a definitive pathological diagnosis; and (IV) high-quality ultrasound images permitting clear feature extraction. The exclusion criteria included: (I) poor-quality ultrasound images; (II) an unclear pathological diagnosis; (III) a history of prior treatment; (IV) incomplete clinical or imaging data; and (V) male breast cancer patients.

Ultrasound protocol

Ultrasonography was performed using an Aixplorer V system (Supersonic Imagine, France) equipped with an SL15-4 linear transducer (4–15 MHz). Patients were positioned in supine and contralateral decubitus positions during a standardized scanning protocol. The candidate ultrasound features were selected based on a combination of prior literature and our clinical experience. Specifically, features including hyperechoic halo, shape, calcifications, Adler grade, and lymph node status have been previously reported to be associated with HER2 expression or molecular subtypes of breast cancer (13-16). Features such as maximum elasticity (Emax) and elasticity ratio (Eratio) (shear-wave elastography parameters) were included based on our preliminary data exploration and clinical observation, as their association with HER2 status remains less established but potentially valuable. All candidate features were entered into the least absolute shrinkage and selection operator (LASSO) regression model without pre-filtering to avoid subjective bias.

The following definitions were used for each ultrasound feature, in accordance with the Breast Imaging Reporting and Data System (BI-RADS) and published criteria. Maximum lesion diameter: the longest dimension of the lesion measured on grayscale ultrasound. Shape: classified as regular (round, oval, or gently lobulated) or irregular (non-lobulated, angular, or spiculated). Hyperechoic halo: a thick, echogenic rim partially or completely surrounding the lesion; classified as present or absent. Calcifications: presence of punctate hyperechoic foci within the lesion, with or without acoustic shadowing; classified as present or absent. Adler grade: assessment of intralesional blood flow using the Adler grading system: Grade I = no or minimal flow (1–2 vessels); Grade II = moderate flow (3–4 vessels); Grade III = marked flow (≥5 vessels). Lymph node status was evaluated via pretreatment ultrasonography and was considered suspicious (positive) if any of the following criteria were met: diffuse or focal cortical thickening (>3 mm), abnormal morphological structure, loss of the fatty hilum, or the presence of microcalcifications; otherwise, it was classified as negative (17). Emax and Eratio: maximum shear-wave elasticity value (kPa) within the lesion, and the ratio of lesion elasticity to adjacent normal tissue elasticity, measured using shear-wave elastography. Two board-certified ultrasound physicians (>5 years breast imaging experience) independently performed and interpreted all examinations.

Clinical data collection

Clinical information, including patient age, menopausal status, body mass index (BMI), and parity, was collected through a retrospective review of medical records and pathology reports. The pathological confirmation for all patients was based on the results from pretreatment core needle biopsy. The HER2 expression status was quantified using IHC or FISH. HER2 negativity was defined as an IHC score of 0 or 1+. HER2 positivity was defined as an IHC score of 3+. For cases with an IHC score of 2+, further FISH analysis was performed to determine HER2 gene amplification status; these cases were classified as HER2-positive if FISH was positive and HER2-negative if FISH was negative (18).

Statistical analysis

Statistical analyses were performed using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA) and R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria). First, the LASSO regression (19) was applied to all candidate predictor variables to mitigate overfitting and to identify the features most strongly associated with HER2 status. The variables selected by the LASSO regression were then entered into a multivariable logistic regression model to confirm their status as independent predictors and to calculate their odds ratios (ORs) with 95% confidence intervals (CIs). A nomogram was constructed based on this final multivariable model. The model’s discriminative ability was assessed by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). Calibration, which refers to the agreement between predicted probabilities and observed outcomes, was evaluated using a calibration curve. Finally, internal validation of the nomogram was conducted using the bootstrap resampling method with 1,000 iterations. This validation process yielded a bias-corrected (or optimism-corrected) calibration curve and a corrected concordance index (C-index), which quantified the degree of overfitting and provided a more reliable estimate of the model’s predictive performance.

Results

This study enrolled 142 breast cancer patients, who were categorized into HER2-positive and HER2-negative groups based on pathological results. Demographic, clinical, and ultrasonographic characteristics of all patients are summarized in Table 1. Feature selection was initially performed using the LASSO regression model, which identified six non-zero coefficient features (Figure 1). These features included maximum lesion diameter, the presence of a hyperechoic halo, Adler grade, shape, lymph node status, and microcalcifications. These six features were subsequently incorporated into a multivariable logistic regression analysis. The multivariable analysis confirmed that the presence of a hyperechoic halo (P=0.01), Adler grade (P<0.001), and lymph node status (P=0.01) were independent influencing factors for HER2 status (P<0.05). Multivariable analysis identified three independent predictors of HER2 status (P<0.05): hyperechoic halo presence (OR =2.655; 95% CI: 1.204–5.990), lymph node status (OR =0.325; 95% CI: 0.133–0.745), and Adler grade (OR =3.093; 95% CI: 1.672–6.110) (Table 2). Based on these three independent predictors, a nomogram was constructed to predict HER2 status (Figure 2). The predictive performance of the nomogram was subsequently evaluated and validated. The discriminative ability and calibration of the model were assessed using the ROC curve and calibration curve, respectively (Figure 3). The AUC was 0.728 (95% CI: 0.639–0.817), indicating moderate discriminative ability. The calibration curve demonstrated good agreement between the model’s predictions and actual observations, as supported by a non-significant Hosmer-Lemeshow goodness-of-fit test result (P=0.86). Furthermore, decision curve analysis (DCA) showed that the nomogram provided a higher net benefit across a wide range of risk thresholds for predicting HER2-positive status (Figure 4), indicating its potential clinical utility and value. Finally, internal validation via bootstrap resampling (1,000 iterations) yielded a corrected C-index (AUC) of 0.708 (95% CI: 0.620–0.796) and a calibration slope of 0.928, further confirming the model’s robustness.

Figure 1 Key results of LASSO regression for selecting HER2 status predictors. (A) Coefficient path plot: x-axis shows −log(λ), y-axis shows regression coefficients. Each trajectory represents coefficient changes of individual predictors with increasing penalty. (B) Ten-fold cross-validation curve: x-axis shows −log(λ), y-axis indicates AUC values. Vertical dashed lines mark optimal λ values (λmin and λ1se), with λmin selected for final model construction.

Figure 2 Nomogram for predicting the risk of HER2 positivity. This prognostic tool integrates three predictors selected through LASSO regression: lymph node status, hyperechoic halo, and Adler grade. Summing the points allocated to each predictor value on the top axis yields a total point score, which can be converted to the corresponding probability of HER2 positivity on the bottom axis. HER2, human epidermal growth factor receptor 2; LASSO, least absolute shrinkage and selection operator.

Figure 3 Performance validation of the HER2 positivity prediction model. (A) ROC curve demonstrating the model’s discriminative ability. (B) Calibration plot assessing the agreement between predicted probabilities and observed outcomes, based on 1,000 bootstrap internal validations. The diagonal dashed line represents perfect calibration. AUC, area under the curve; CI, confidence interval; HER2, human epidermal growth factor receptor 2; ROC, receiver operating characteristic.

Figure 4 Decision curve analysis for HER2 prediction. The model (red line) shows net benefit compared to treating all (grey line) or no patients (horizontal axis) across 0–80% probability thresholds. HER2, human epidermal growth factor receptor 2.

Discussion

The heterogeneity of HER2 expression in breast cancer has been well-documented. A primary contributor to this heterogeneity is the inherent limitation of standard detection methods, namely IHC and FISH. The samples for these tests are typically obtained via imaging-guided core needle biopsy. However, the diagnostic accuracy of this procedure is influenced by factors such as lesion size and morphology. Moreover, as a biopsy specimen represents only a small portion of the tumor, it may not fully capture the heterogeneous expression of HER2 across the entire lesion (20). Furthermore, a growing body of evidence suggests that the HER2 status of breast cancer patients can evolve during neoadjuvant therapy. Notably, studies have reported that comparable proportions of patients may experience either a loss or a gain in HER2 expression (21). Consequently, a single biopsy is often insufficient for comprehensively assessing the biological characteristics of breast cancer, while repeated biopsies would impose significant additional inconvenience and burden on patients. This underscores a pressing clinical need for a more convenient method to guide treatment decisions. Ultrasonography, a widely used imaging technique for breast cancer screening and diagnosis, provides features such as tumor margins, hyperechoic halos, and microcalcifications, which have been associated with biomarker status and molecular subtypes. Therefore, this study aims to develop an integrated model based on clinical and multimodal ultrasonographic features to predict HER2 status, thereby providing a practical tool to inform treatment planning and prognosis assessment.

This study demonstrates that HER2-positive breast cancers are characterized by several ultrasonographic features, including larger tumor size, irregular shape, the presence of a hyperechoic halo, rich internal vascularity (as assessed by Adler grade), microcalcifications, and ipsilateral axillary lymph node metastasis. Among these, the presence of a hyperechoic halo, high Adler grade (indicating rich vascularity), and positive ipsilateral lymph node status were identified as independent predictors of HER2 positivity. We incorporated these three variables into a predictive model that is both parsimonious and stable, demonstrating moderate discriminative ability (bootstrap-corrected AUC =0.708). The clinical utility of this model was further substantiated by DCA, which showed a positive net benefit across a wide range of risk thresholds.

The hyperechoic halo is recognized as an important feature in breast cancer diagnosis (22). Sonographically, it manifests as a thick, echogenic rim that partially or completely surrounds the lesion. This phenomenon corresponds to the disruption of the layered structure in the surrounding normal tissue. Histopathologically, it corresponds to a peripheral zone of active cancer cell proliferation, neovascularization, lymphocytic infiltration, and fibro-connective tissue hyperplasia (23). These underlying pathological features reflect a heightened degree of tumor invasiveness and serve as an important indicator of an unfavorable prognosis. Previous literature suggests that for accurate radiological measurement of breast malignancies, the hyperechoic halo should be included to avoid underestimation of tumor size (14). Furthermore, studies have confirmed an inverse correlation between the size of the hyperechoic halo and the degree of tumor differentiation. These pathological features reflect a heightened degree of cellular invasiveness and serve as an important indicator of an unfavorable prognosis. As a key oncogenic driver and prognostic marker in breast cancer, HER2 positivity is associated with higher proliferative rates, an increased risk of metastasis, and poorer overall outcomes. This pathological evidence establishes a close relationship between the presence of a hyperechoic halo and HER2 expression, which is consistent with the findings of our present study.

During the progression of breast cancer, soluble factors within the tumor microenvironment directly or indirectly stimulate the expression of pro-angiogenic factors such as VEGF, connective tissue growth factor (CTGF), and fibroblast growth factor (FGF), thereby promoting fibrosis and tumor angiogenesis. Research has demonstrated a close association between increased angiogenesis/VEGF expression and the activation of HER2 in cancer cells. VEGF facilitates tumor growth primarily by promoting angiogenesis and enhancing vascular permeability, which collectively lead to increased microvessel density. Furthermore, a multitude of other angiogenic growth factors—including epidermal growth factor (EGF), angiopoietin, platelet-derived growth factor (PDGF), transforming growth factor β (TGFβ), and tumor necrosis factor-α (TNF-α)—have been shown to play significant roles in this process (24). Consequently, in clinical practice, HER2-positive breast cancers, characterized by high proliferative activity and demanding substantial blood supply, induce the formation of more tortuous and dilated tumor vessels. This underlying vascular pathology is sonographically reflected as increased blood flow signals and a higher Adler grade, which aligns with the findings of our study.

Multivariable analysis identified lymph node status as an independent negative predictor of HER2-positive expression (OR =0.325, 95% CI: 0.133–0.745, P<0.05). This OR indicates that the odds of HER2 positivity were 67.5% lower in patients with axillary lymph node metastasis compared to those without, corresponding to an odds of only 30.2% relative to the lymph node-negative group. The 95% CI, which does not include 1, underscores that this inverse association is statistically significant and unlikely to be attributable to sampling variability. While previous studies have established that HER2 overexpression enhances tumor cell invasiveness by activating the PI3K-AKT and RAS-MAPK pathways (25,26) and is a key driver of metastasis in breast cancer (27,28), it is important to note that clinically defined HER2-positive breast cancer exhibits considerable heterogeneity at the molecular level. Diec et al. (29) identified advanced disease stage and a high HER2DX risk score as key determinants of increased metastatic risk. Their work also highlighted an association between the intrinsic molecular subtypes of HER2-positive breast cancer and distinct patterns of metastatic spread. These findings collectively underscore the heterogeneity in metastatic risk among HER2-positive tumors, indicating that not all exhibit equally high metastatic potential. In our cohort, 130 of the 142 patients (91.5%) were classified with clinical T1–T2 stage, a finding that aligns with the demographic profile reported by Weiss et al. (16). Their findings indicated that the probability of axillary lymph node metastasis increases with larger tumor size. Therefore, we postulate that in some early-stage, small HER2-positive tumors, the likelihood of axillary lymph node metastasis might be lower due to a lesser degree of epithelial-mesenchymal transition (EMT) or reduced secretion of matrix metalloproteinases (MMPs), which could prevent the breach of the basement membrane. This hypothesis, however, requires validation in future studies.

This study has several limitations. Firstly, the interpretation of breast ultrasound features is operator-dependent, and the reliance on stored static images may lead to the omission of significant features visible in other imaging planes. Secondly, while HER2-positive breast cancer encompasses diverse intrinsic molecular subtypes, our study did not comprehensively analyze the interactions among these molecular characteristics, which may limit the generalizability of our findings. Finally, this was a single-center, retrospective study with a relatively small sample size. Moreover, the lack of external validation limits the generalizability of our findings. Furthermore, lymph node status was determined by pretreatment ultrasound, which is susceptible to false-negative results and may have introduced understaging bias. Future multi-center, prospective studies that incorporate detailed pathological and molecular profiling are warranted to validate the generalizability of our findings.

Conclusions

In conclusion, this validated nomogram offers a clinically valuable adjunct for the pretreatment prediction of HER2 status. By integrating key ultrasonographic features—the presence of a hyperechoic halo, lymph node status, and a high Adler grade—it provides a non-invasive, whole-tumor perspective that complements pathological biopsy. Given that core needle biopsy is invasive and repeated assessments are not always easy or feasible, our nomogram may serve as a non-invasive alternative for risk refinement or dynamic monitoring. Although it cannot replace tissue-based diagnosis, it serves as a promising tool for refining pre-biopsy risk assessment, guiding biopsy strategies in heterogeneous tumors, and potentially facilitating dynamic monitoring during therapy.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://cco.amegroups.com/article/view/10.21037/cco-2026-1-0029/rc

Data Sharing Statement: Available at https://cco.amegroups.com/article/view/10.21037/cco-2026-1-0029/dss

Peer Review File: Available at https://cco.amegroups.com/article/view/10.21037/cco-2026-1-0029/prf

Funding: This study was supported by the Scientific Research Foundation of Yunnan Education Department (Nos. 2023Y0654 and 2025J0309), 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), and 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://cco.amegroups.com/article/view/10.21037/cco-2026-1-0029/coif). X.K.L., J.D.L., D.C., N.Q.L., H.T.C., and X.M.L. report that this study was supported by the Scientific Research Foundation of Yunnan Education Department (Nos. 2023Y0654 and 2025J0309), 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), and Yunnan Provincial Department of Science and Technology-Kunming Medical University Applied Basic Research Joint Special Fund-General Project (Nos. 202401AY070001-268 and 202501AY070001-108). They also received consulting fees from AMCA. The other authors have no conflicts of interest to declare.

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. This study was conducted in accordance with the principles of the Declaration of Helsinki and its subsequent amendments. The study protocol was reviewed and approved by the Institutional Ethics Committee of Yunnan Cancer Hospital (approval No. KYLX2023-021), which granted a waiver of informed consent for this retrospective analysis. All data were anonymized prior to analysis.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Chen S, Cui W, Dong J, et al. A Novel Signature Based on Angiogenesis-Related Genes Predicts the Prognosis and Immunotherapy Response in HER2-Positive Breast Cancer. J Cancer 2024;15:4731-48. [Crossref] [PubMed]
3. Li Q, Liu J, Jiang Z, et al. CSCO breast cancer guideline: precise, economical and oriental. Sci China Life Sci 2020;63:1410-2. [Crossref] [PubMed]
4. Tesch ME, Gelmon KA. Targeting HER2 in Breast Cancer: Latest Developments on Treatment Sequencing and the Introduction of Biosimilars. Drugs 2020;80:1811-30. [Crossref] [PubMed]
5. Hu M, Zhang L, Wang X, et al. Enhanced HER-2 prediction in breast cancer through synergistic integration of deep learning, ultrasound radiomics, and clinical data. Sci Rep 2025;15:26992. [Crossref] [PubMed]
6. Li C, Yin J. Radiomics Nomogram Based on Radiomics Score from Multiregional Diffusion-Weighted MRI and Clinical Factors for Evaluating HER-2 2+ Status of Breast Cancer. Diagnostics (Basel) 2021;11:1491. [Crossref] [PubMed]
7. Ye F, Dewanjee S, Li Y, et al. Advancements in clinical aspects of targeted therapy and immunotherapy in breast cancer. Mol Cancer 2023;22:105. [Crossref] [PubMed]
8. Bonacho T, Rodrigues F, Liberal J. Immunohistochemistry for diagnosis and prognosis of breast cancer: a review. Biotech Histochem 2020;95:71-91. [Crossref] [PubMed]
9. Muller KE, Marotti JD, Tafe LJ. Pathologic Features and Clinical Implications of Breast Cancer With HER2 Intratumoral Genetic Heterogeneity. Am J Clin Pathol 2019;152:7-16. [Crossref] [PubMed]
10. Zhu JY, He HL, Jiang XC, et al. Multimodal ultrasound features of breast cancers: correlation with molecular subtypes. BMC Med Imaging 2023;23:57. [Crossref] [PubMed]
11. Huang Z, Chen L, Wang Y, et al. Molecular markers, pathology, and ultrasound features of invasive breast cancer. Clin Imaging 2021;79:85-93. [Crossref] [PubMed]
12. Zhang L, Li J, Xiao Y, et al. Identifying ultrasound and clinical features of breast cancer molecular subtypes by ensemble decision. Sci Rep 2015;5:11085. [Crossref] [PubMed]
13. Joekel J, Eggemann H, Costa SD, et al. Should the hyperechogenic halo around malignant breast lesions be included in the measurement of tumor size? Breast Cancer Res Treat 2016;156:311-7. [Crossref] [PubMed]
14. Nguyen DL, Grimm LJ, Wells VA, et al. Echogenic Rind Associated With Suspicious Masses on Ultrasound: Performance for Breast Cancer Diagnosis and Associations With Pathologic Tumor Characteristics. AJR Am J Roentgenol 2025;225:e2533732. [Crossref] [PubMed]
15. Wang K, Yang X, Yang S, et al. A diagnostic test of two-dimensional ultrasonic feature extraction based on artificial intelligence combined with blood flow Adler classification and contrast-enhanced ultrasound for predicting HER-2-positive breast cancer. Transl Cancer Res 2025;14:640-50. [Crossref] [PubMed]
16. Weiss A, Martínez-Sáez O, Waks AG, et al. Nodal positivity and systemic therapy among patients with clinical T1-T2N0 human epidermal growth factor receptor-positive breast cancer: Results from two international cohorts. Cancer 2023;129:1836-45. [Crossref] [PubMed]
17. Magny SJ, Shikhman R, Keppke AL. Breast Imaging Reporting and Data System. 2026.
18. Allison KH, Hammond MEH, Dowsett M, et al. Estrogen and Progesterone Receptor Testing in Breast Cancer: ASCO/CAP Guideline Update. J Clin Oncol 2020;38:1346-66. [Crossref] [PubMed]
19. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1-22.
20. Zhang Y, Huang H, Yin L, et al. Preoperative prediction of HER-2 expression status in breast cancer based on MRI radiomics model. Zhonghua Zhong Liu Za Zhi 2024;46:428-37. [Crossref] [PubMed]
21. Lee HJ, Lee JE, Jeong WG, et al. HER2-Positive Breast Cancer: Association of MRI and Clinicopathologic Features With Tumor-Infiltrating Lymphocytes. AJR Am J Roentgenol 2022;218:258-69. [Crossref] [PubMed]
22. Durmus T, Stöckel J, Slowinski T, et al. The hyperechoic zone around breast lesions - an indirect parameter of malignancy. Ultraschall Med 2014;35:547-53. [Crossref] [PubMed]
23. Zhang XD, Kou M, Zou WY, et al. HER-2 expression is correlated with multimodal imaging features in breast cancer: a pilot study. Ann Med 2024;56:2434182. [Crossref] [PubMed]
24. Zhong L, Li Y, Xiong L, et al. Small molecules in targeted cancer therapy: advances, challenges, and future perspectives. Signal Transduct Target Ther 2021;6:201. [Crossref] [PubMed]
25. Ruiz-Saenz A, Dreyer C, Campbell MR, et al. HER2 Amplification in Tumors Activates PI3K/Akt Signaling Independent of HER3. Cancer Res 2018;78:3645-58. [Crossref] [PubMed]
26. Rubin E, Shan KS, Dalal S, et al. Molecular Targeting of the Human Epidermal Growth Factor Receptor-2 (HER2) Genes across Various Cancers. Int J Mol Sci 2024;25:1064. [Crossref] [PubMed]
27. Ahmed AR. HER2 expression is a strong independent predictor of nodal metastasis in breast cancer. J Egypt Natl Canc Inst 2016;28:219-27. [Crossref] [PubMed]
28. He ZY, Wu SG, Yang Q, et al. Breast Cancer Subtype is Associated With Axillary Lymph Node Metastasis: A Retrospective Cohort Study. Medicine (Baltimore) 2015;94:e2213. [Crossref] [PubMed]
29. Dieci MV, Conte P, Bisagni G, et al. Metastatic site patterns by intrinsic subtype and HER2DX in early HER2-positive breast cancer. J Natl Cancer Inst 2024;116:69-80. [Crossref] [PubMed]