The clinicopathological and imaging parameters contributing to the similarity network fusion subtypes and prognosis of hormone receptor-positive HER2-negative breast cancer: an exploratory study

Introduction

Breast cancer stands as one of the most prevalent malignancies among women, significantly contributing to cancer-related fatalities globally (1), especially among those under 45 years old (2). Breast cancer is a heterogeneous disease, which can be delineated and classified into four different intrinsic molecular subtypes, relying on the presence of hormone receptor (HR), which includes both estrogen receptor (ER) and progesterone receptor (PR), and the status of human epidermal growth factor receptor 2 (HER2). Specifically, these subtypes include HR positive (HR⁺)/HER2 positive (HER2⁺), HR⁺/HER2 negative (HER2⁻), HR negative (HR⁻)/HER2⁺, and HR⁻/HER2⁻ (3).

The subtype of HR⁺/HER2⁻ breast cancer stands as the most prevalent subtype, ranging from approximately 65% to 70% of all breast cancer cases, thereby posing a significant medical challenge on a global scale (4). In HR⁺/HER2⁻ breast tumors, estrogen binds to ER and stimulates receptor-regulated transcription, which promotes the growth and proliferation of tumor cells (5). Owing to the presence of the HR, endocrine therapy emerges as an effective adjuvant therapy for HR⁺/HER2⁻ breast cancers. Nevertheless, HR⁺/HER2⁻ breast cancer is marked by a persistent threat of distant recurrence throughout the prescribed 5-year course of endocrine therapy and even persists for 20–30 years post-diagnosis. Furthermore, developing resistance to endocrine therapy poses a formidable challenge in enhancing the patients’ prognosis (6,7).

HR⁺/HER2⁻ breast cancer is also characterized by its remarkable heterogeneity, encompassing a vast spectrum of variations, such as histological grade, distinct pathological subtype, varying Ki-67 index, unique gene expression profiles, and numerous other differentiating signatures (8). A recent comprehensive study successfully established a large multi-omics cohort, containing gene sequencing data, quantitative protein analysis via mass spectrometry, and metabolic profiles, among other features (9). This comprehensive approach has classified HR⁺/HER2⁻ breast cancer into four distinct molecular types: canonical [similarity network fusion (SNF)1], immunogenic (SNF2), proliferative (SNF3), and receptor tyrosine kinase (RTK)-driven (SNF4). Research has also pinpointed precise treatments strategies tailored to each subtype, thereby enhancing the precision medicine approach for HR⁺/HER2⁻ breast cancer (10).

The SNF classification primarily emphasizes the disparities in tumor genetic expression patterns and the intricate tumor microenvironment, offering a direct insight into the heterogeneity inherent in HR⁺/HER2⁻ breast cancer. Imaging reports provide invaluable insights into the intricate patterns of tumor growth, whereas tumor heterogeneity is also intricately intertwined with the clinical profiles and pathological characteristics of patients. This study will integrate routine clinical and pathological data as well as imaging assessments, aiming to validate the clinical significance of these common indicators in differentiating SNF subtypes and predicting the prognosis of HR⁺/HER2⁻ breast cancer. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2045/rc).

Methods

Patients

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The retrospective study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center (FUSCC) (No. 1612167-18). Owing to the retrospective nature of this study, written informed consent was not required.

Data were collected from patients who were diagnosed with HR⁺/HER2⁻ breast cancer after receiving surgical treatment at FUSCC between September 2009 and October 2015. Clinical data were collected, specifically including age, immunohistochemical findings, histological grade, status of lymph node metastasis, and lymphovascular invasion (LVI), alongside ultrasound (US), mammography (MG), and magnetic resonance imaging (MRI) reports. Additionally, follow-up data were meticulously documented for survival analysis. The inclusion criteria stipulated those patients must have undergone preoperative US examination at FUSCC, along with accessible US reports and a comprehensive compilation of both clinical and post-surgical pathological data.

A total of 347 patients fulfilling the inclusion criteria were enrolled in the study. Among them, 290 patients underwent MG examination at FUSCC. To mitigate the potential bias arising from substantial missing data, two cohorts were established. Cohort I contained all patients (n=347), whereas cohort II exclusively comprised those who had undergone MG examination (n=290).

Pathological data

The pathological types were categorized into three groups: invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), and others. Histological grades included the following: I, characterized by high differentiation; II, featuring moderate differentiation, and III, marked by poor differentiation. The assessment of Ki-67 expression was determined by immunohistochemical staining following standard procedures (11). The threshold for Ki-67 expression was established at 20% immunostaining positivity among malignant cells.

The immunohistochemical technique was utilized to detect the expression of ER across all HR⁺/HER2⁻ breast cancer cases. ER positivity was defined as ≥1% of tumor cells with positive nuclear staining, and further categorized by staining intensity as weak (+), moderate (++), or strong (+++). Based on a comprehensive study published by the breast cancer center at our institute, encompassing a vast multi-omics cohort comprising 579 HR⁺/HER2⁻ breast cancer patients from FUSCC, the four distinct molecular subtypes of SNF1, SNF2, SNF3, and SNF4 were categorized (9).

Follow-up

Follow-up information was gathered through inpatient or outpatient medical records, as well as via telephone communication. The commencement date was marked as the date of the surgical procedure, and the conclusion date spanned up to 31 December 2023. The termination point of the follow-up period was defined as the time of death or loss of tracking. Progression-free survival (PFS) referred to the period commencing from the time of surgery until the first local recurrence, distant metastasis, or death attributed to breast cancer.

Statistical analyses

Statistical analysis and plotting were performed utilizing the software SPSS 20.0 (IBM Corp., Armonk, NY, USA) and R version 4.3.3 (Alcatel-Lucent, Colombes, Paris, France). To ascertain independent attributes with the SNF classification of HR⁺/HER2⁻ breast cancer tumors, we employed both univariate and multivariate logistic regression analyses, incorporating a forward stepwise selection method. The relationships between covariates and dependent variables were quantitatively expressed through odds ratios (ORs), accompanied with 95% confidence intervals (CIs) for enhanced precision and interpretation. Survival analysis was performed using Cox regression, with loss to follow-up or death as the observed endpoint. The primary endpoints of interest in this analysis were overall survival (OS) and PFS. Survival curves for both OS and PFS were plotted using the Kaplan-Meier method and used to compare the prognosis of the four SNF subtypes.

The nomogram for predicting 3-, 5-, and 10-year OS/PFS in patients with HR⁺/HER2⁻ breast cancer was developed according to the multivariate Cox regression analysis. The discriminative capability of the established model was evaluated through the area under the receiver operating characteristic (ROC) curve (AUC). Calibration curves were plotted to compare the predicted OS/PFS with the observed OS/PFS using a bootstrap approach with 1,000 resamples. The calculations encompassed the period from diagnosis to either the occurrence of all-cause mortality or the last follow-up date on 31 December 2023. Data pertaining to patients who were still alive at the final follow-up were subject to censoring. Categorical data were presented as number (percentage). A P value less than 0.05 indicated statistical significance.

Results

After rigorously applying the inclusion criteria, we analyzed a cohort of 347 patients diagnosed with HR⁺/HER2⁻ breast cancer, comprising 85 canonical cases (SNF1, 24.5%), 89 immunogenic cases (SNF2, 25.6%), 115 proliferative cases (SNF3, 33.1%), and 58 RTK-driven cases (SNF4, 16.7%). Based on the established SNF classification system, subtypes for the 347 patients were derived from multi-omics clustering of RNA expression, somatic copy number alteration (SCNA), and metabolite abundance data (9). The clinicopathological and imaging characteristics of these patients are concisely presented in Table 1.

Table 2 presents the results of a comprehensive univariate analysis of clinicopathological and imaging characteristics associated with HR⁺/HER2⁻ breast cancer. Notably, significant statistical correlation was found between the SNF categorization and key factors, including histological grade, Ki-67, ER grade, and Breast Imaging Reporting and Data System (BI-RADS) grade of MG (P<0.05). Conversely, age, the maximal diameter, menopausal status, personal/family histories, pathological type, LVI, axillary lymph node metastasis (ALNM), as well as BI-RADS grades of US and MRI did not exhibit notable differences among the four distinct SNF subtypes.

As our findings indicated, the distribution of MG BI-RADS grades differs among the four SNF subtypes of breast cancer, prompting us to investigate whether mammographic features also vary across these subtypes. Compared with the BI-RADS classification results of US and MRI, the composition of the MG BI-RADS grades based on the four SNF types seems to have a greater difference (Figure 1). Among the 290 patients with MG reports, 68 were excluded for the following reasons: examination performed outside our institution (n=46), loss of images due to system issues (n=14), or poor visualization of the lesion (n=8). The remaining 222 patients constituted the mammographic image-feature analysis cohort. The prevalence of each mammographic feature among the four SNF subtypes is summarized in Table 3.

Figure 1 The composition of four SNF subtypes in the two patient cohorts. SNF1, canonical; SNF2, immunogenic; SNF3, proliferative; SNF4, RTK-driven. RTK, receptor tyrosine kinase; SNF, similarity network fusion.

The evaluated mammographic features included American College of Radiology (ACR) breast density, mass shape, margin, internal density, calcification, presence of lymphadenopathy, and associated features. Some masses were not visible on MG due to factors such as high breast density or deep lesion location. Results showed that the ACR breast density and mass density have statistically significant differences among the four SNF classifications (P<0.05). Compared to other subtypes, SNF3 breast cancers were more frequently associated with suspicious malignant calcifications, whereas SNF1 tumors typically presented as masses without calcifications. SNF4 tumors were more likely to be mammographically occult. Both SNF4 and SNF1 subtypes were associated with denser breast tissue (Figure 2).

Figure 2 Representative mammographic images of different SNF subtypes in HR⁺/HER2⁻ breast cancer. (A) Right breast mammogram of an SNF1 subtype patient with dense fibroglandular breast tissue. Craniocaudal view demonstrates an irregular, slightly hyperdense mass with spiculated margins in upper outer quadrant. (B) Mediolateral oblique view of right breast MG of an SNF3 subtype breast cancer patient. The image demonstrates an irregular mass in the deep inner upper quadrant, accompanied by clustered fine pleomorphic calcifications and architectural distortion. (C) Right mammogram in a patient with SNF4 subtype breast cancer. Craniocaudal view shows an irregular high-density mass in the upper breast with spiculated margins, peritumoral edema, skin thickening, and nipple retraction. ER, estrogen receptor; HER2⁻, human epidermal growth factor receptor 2-negative; HR⁺, hormone receptor-positive; MG, mammography; SNF, similarity network fusion.

Notably, the distribution of SNF types among patients was found to be comparable across both cohorts (P<0.05), thereby precluding any anticipated analytical errors (Figure 3). For each cohort, an independent multifactorial logistic regression analysis was performed, systematically integrating variables that demonstrated statistically significant differences in the univariate analysis into the multivariate logistic regression model using a forward stepwise approach (Table 2). The analysis revealed that histological grade, Ki-67, ER grade, and BI-RADS grade of MG had a significant impact on the SNF subtype categorization of HR⁺/HER2⁻ breast cancer (P<0.05).

Figure 3 Kaplan-Meier curves for OS and PFS of the four SNF subtypes. SNF1, canonical; SNF2, immunogenic; SNF3, proliferative; SNF4, RTK-driven. OS, overall survival; PFS, progression-free survival; RTK, receptor tyrosine kinase; SNF, similarity network fusion.

The Kaplan-Meier curves distinctly illustrated substantial variations in survival duration across these four SNF subtypes (Figure 4). Specifically, the average survival time was 152.6 months for SNF 1, 143.4 months for SNF2, 138.2 months for SNF3, and 117.5 months for SNF 4. The difference of OS/PFS among these four subgroups was statistically significant (OS: P=0.014, PFS: P=0.012). Among the total 347 patients, distant organ metastases were observed in 115 cases, with notable occurrences in the lung (n=26, 17.9%), liver (n=24, 16.5%), bone (n=69, 47.6%), distant lymphatic system (n=15, 10.3%), pleura (n=3, 2.1%), and contralateral breast (n=4, 2.8%).

Figure 4 Nomogram of predicting 3-, 5-, and 10-year OS (left) and PFS (right) for HR⁺/HER2⁻ breast cancer patients. *, P<0.05; **, P<0.01; ***, P<0.001. SNF1, canonical; SNF2, immunogenic; SNF3, proliferative; SNF4, RTK-driven. HER2⁻, human epidermal growth factor receptor 2-negative; HR⁺, hormone receptor-positive; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; LVI, lymphovascular invasion; OS, overall survival; PFS, progression-free survival; RTK, receptor tyrosine kinase.

Using OS as a pivotal variable, regression models were meticulously constructed through univariate and multivariate Cox regression analyses. It is found that the pathological maximal diameter, SNF subtype, pathological type, LVI, and ALNM had statistically independent impacts on the prognosis of HR⁺/HER2⁻ breast cancer (P<0.05) as shown in Table 4. Treating these variables as independent measures, R software was employed to further analyze the OS and PFS of patients with HR⁺/HER2⁻ breast cancer. Leveraging the 5- and 10-year survival rates as dependent variables, a comprehensive visual nomogram model was established (Figure 5). The cumulative score for each patient was derived by integrating individual scores calculated via the nomogram. Notably, the majority of patients in the study exhibiting total risk points ranging from 12 to 145, offering a robust framework for assessment.

Figure 5 Time-dependent AUC and calibration curves of the nomogram. AUC, area under the receiver operating characteristic curve; OS, overall survival; PFS, progression-free survival.

The application of the time-dependent AUC analysis, incorporating the nomogram, exhibited robust predictive capabilities for both OS and PFS over 5 and 10 years, yielding remarkable AUC values (5-year OS: 0.681, 10-year OS: 0.766; 5-year PFS: 0.744, 10-year PFS: 0.760) (Figure 6). According to the total points calculated using the nomogram, patients with HR⁺/HER2⁻ breast cancer were divided into two risk groups: a low-risk cohort (total points <87) and a high-risk cohort (total points ≥87). The Kaplan-Meier OS curves underscored a pronounced divergence in survival outcomes between these two risk groups (P<0.05) (Figure 7).

Figure 6 The Kaplan-Meier curves of OS: low-risk group (total points <87) and high-risk group (total points ≥87) and PFS: low-risk group (total points <79) and high-risk group (total points ≥79). OS, overall survival; PFS, progression-free survival.

Figure 7 Distribution of Ki-67 index across four SNF subtypes. SNF1, canonical; SNF2, immunogenic; SNF3, proliferative; SNF4, RTK-driven. RTK, receptor tyrosine kinase; SNF, similarity network fusion.

Discussion

Globally, HR⁺/HER2⁻ breast cancer constitutes the most prevalent immunohistochemical subtype (12). Despite its favorable response to endocrine therapy, this subtype exhibits notable heterogeneity, evidenced by varying patterns in ER expression, histological grade, Ki-67 index, and so on (13), which makes it of great importance to delve into the genetic underpinning of the tumor heterogeneity.

Based on the seminal article published in a prestigious journal, where researchers discerned that HR⁺/HER2⁻ breast cancer can be stratified into four distinct SNF subtypes (9), each characterized by unique biological and clinical attributes, we have undertaken a comprehensive analysis to further elucidate the clinicopathological characteristics, imaging manifestations, and their corresponding prognostic implications in HR⁺/HER2⁻ breast cancer. Our investigation revealed a significant correlation between histopathological patterns, the Ki-67 proliferation index, ER grade, and BI-RADS grade of MG, all being factors intimately linked to the subclassification of HR⁺/HER2⁻ breast cancer under the SNF subtype framework. We also investigated whether the four distinct SNF subtypes of breast cancer exhibit divergent features on MG. The results indicated that ACR breast density and mass density provided significant discriminatory value for differentiating among the four SNF subtypes. Furthermore, an integrated model that encapsulates pathological maximal diameter, SNF subtype, histopathological grading, presence of LVI, and BI-RADS grade of MG can predict the prognosis of HR⁺/HER2⁻ breast cancer.

According to our findings, the SNF1 subtype was distinguished by its minimal Ki-67 expression, the lowest percentage of high histological grade, and the most favorable survival rate. The SNF2 subtype exhibited a notably lowered expression of ER. The SNF3 subtype demonstrated the highest level of Ki-67 expression, and the SNF4 subtype had the shortest survival duration. This is in alignment with the finding shown by Jin et al. that the four subtypes of HR⁺/HER2⁻ breast cancers exhibited evident differences in gene expression, immunohistochemical features, tumor microenvironment, and immune microenvironment (9). The upregulation of the cell cycle pathway was discerned in the SNF3 subtype, whereas SNF2 tumors manifested an increased prevalence of immune cells, subsequently correlating with heightened expression of functional markers indicative of immune activation. The SNF4 subtype displayed the most adverse prognosis, likely attributed to the enrichment of the RTK pathway within this subtype.

In the context of HR⁺/HER2⁻ breast cancers, patients exhibit a consistent pattern of ER positivity coupled with HER2 negativity. However, they vary notably in the extent of ER expression levels. Our results constantly found that the SNF subtype was associated with the ER and Ki-67. ER expression plays an important role in the metabolism in breast cancer cells and is one of the key points to distinguish intrinsic breast cancer subtypes (14). Ki-67 in breast cancer is related to high risk of recurrence and favorable response to neoadjuvant chemotherapy. The Ki-67 index also serves as a discriminator between luminal A and luminal B tumor subtypes in HR⁺/HER2⁻ breast cancer (15,16). Upon the statistical analysis of the expression of Ki-67 across 347 patients, it can be seen that the majority of patients with the SNF1 subtype display a Ki-67 index of <20%, in contrast with SNF3 subtypes, for which the Ki-67 indices are notably higher (Figure 8). This finding aligns with previous research, which indicates that the SNF1 subtype is predominantly associated with the PAM50 luminal A subtype, characterized by low Ki-67 expression, whereas the SNF3 subtype primarily comprises the PAM50 luminal B subtype, known for its higher Ki-67 expression levels.

Figure 8 The categorization of imaging BI-RADS according to SNF subtypes. SNF1, canonical; SNF2, immunogenic; SNF3, proliferative; SNF4, RTK-driven. BI-RADS, Breast Imaging Reporting and Data System; MG, mammography; MRI, magnetic resonance imaging; RTK, receptor tyrosine kinase; SNF, similarity network fusion; US, ultrasound.

Accurately predicting the prognosis of HR⁺/HER2⁻ breast cancer holds paramount significance, given the persistent threat of long-term recurrence, which underscores its continued predominance as a leading cause of annual breast cancer mortality. We postulated that the underlying reason was the inherent diversity characteristic of HR⁺/HER2⁻ breast cancer. Our investigation revealed that prolonged pathological maximal diameter, advanced SNF stage, higher histopathologic grade, presence of LVI, and a higher BI-RADS grade of MG all contribute to a worse prognosis for HR⁺/HER2⁻ breast cancer patients. The contribution of maximal diameter, histological grade, and LVI is constantly with the literature, suggesting its potential to affect subsequent treatment strategies (17). The observation that there exists a notable absence of substantial contributions made by the SNF subtype classification to the prognosis of HR⁺/HER2⁻ breast cancer underscores a critical gap in our current understanding of this complex disease. Given the biological heterogeneity inherent in HR⁺/HER2⁻ breast cancer, where tumor behavior and response to treatment can vary significantly, it becomes imperative to delve deeper into the potential role of SNF subtypes in influencing the course and outcome of the disease (18,19). Further investigation into the contribution of SNF subtypes to the prognosis of HR⁺/HER2⁻ breast cancer is not only warranted but also urgent, given its potential to significantly advance our ability to manage and treat this disease.

An intriguing discovery underscores the correlation between the BI-RADS grade at MG and the SNF subtype, along with the prognosis of HR⁺/HER2⁻ breast cancer. Specifically, the SNF2 subtype exhibited a reduced prevalence of characteristic MG findings, including BI-RADS 4C and BI-RADS 5 classifications. There is a lack of similar research in the literature. This discovery merits further exploration and investigation.

There are several shortcomings in our research. Extensive studies showed that, in contrast to postmenopausal women, premenopausal individuals exhibit lower ER levels, coupled with a heightened histological grade, ultimately culminating in less favorable prognosis attributed to these factors (20). Nevertheless, in this study, age and menstrual history, factors traditionally considered pivotal in breast cancer risk, failed to exhibit notable effects on patients’ prognosis. This discrepancy might stem from the limited sample size of our analysis. Furthermore, the data originated from as far back as 10–15 years ago. US images were not available for image feature assessment, thereby precluding the in-depth exploration of the relationship between image features, biological behaviors, and clinical variables. Meanwhile, because of the limited number of patients with MRI, the BI-RADS grade of MRI in this cohort may not be representative.

Conclusions

The clinicopathological, immunohistochemical, and imaging features of the breast lesion exhibited considerable clinical significance in differentiating the SNF molecular subtype within HR⁺/HER2⁻ breast cancer. The prognosis of HR⁺/HER2⁻ breast cancer can be accurately anticipated utilizing an integrated model that incorporated clinicopathological parameters, imaging variables, and the classification of SNF molecular subtypes.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2045/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2045/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2045/coif). The 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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The retrospective study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center (FUSCC) (No. 1612167-18). Owing to the retrospective nature of this study, the written informed consent was not needed.

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. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin 2020;70:7-30. [Crossref] [PubMed]
2. Anastasiadi Z, Lianos GD, Ignatiadou E, Harissis HV, Mitsis M. Breast cancer in young women: an overview. Updates Surg 2017;69:313-7. [Crossref] [PubMed]
3. Kalinsky K, Barlow WE, Gralow JR, Meric-Bernstam F, Albain KS, Hayes DF, et al. 21-Gene Assay to Inform Chemotherapy Benefit in Node-Positive Breast Cancer. N Engl J Med 2021;385:2336-47. [Crossref] [PubMed]
4. Sparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS, Hayes DF, et al. Adjuvant Chemotherapy Guided by a 21-Gene Expression Assay in Breast Cancer. N Engl J Med 2018;379:111-21. [Crossref] [PubMed]
5. Huppert LA, Gumusay O, Idossa D, Rugo HS. Systemic therapy for hormone receptor-positive/human epidermal growth factor receptor 2-negative early stage and metastatic breast cancer. CA Cancer J Clin 2023;73:480-515. [Crossref] [PubMed]
6. Pan H, Gray R, Braybrooke J, Davies C, Taylor C, McGale P, Peto R, Pritchard KI, Bergh J, Dowsett M, Hayes DF. EBCTCG. 20-Year Risks of Breast-Cancer Recurrence after Stopping Endocrine Therapy at 5 Years. N Engl J Med 2017;377:1836-46. [Crossref] [PubMed]
7. Dowsett M, Cuzick J, Ingle J, Coates A, Forbes J, Bliss J, Buyse M, Baum M, Buzdar A, Colleoni M, Coombes C, Snowdon C, Gnant M, Jakesz R, Kaufmann M, Boccardo F, Godwin J, Davies C, Peto R. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010;28:509-18. [Crossref] [PubMed]
8. Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, Eisen MB, van de Rijn M, Jeffrey SS, Thorsen T, Quist H, Matese JC, Brown PO, Botstein D, Lønning PE, Børresen-Dale AL. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 2001;98:10869-74. [Crossref] [PubMed]
9. Jin X, Zhou YF, Ma D, Zhao S, Lin CJ, Xiao Y, Fu T, Liu CL, Chen YY, Xiao WX, Liu YQ, Chen QW, Yu Y, Shi LM, Shi JX, Huang W, Robertson JFR, Jiang YZ, Shao ZM. Molecular classification of hormone receptor-positive HER2-negative breast cancer. Nat Genet 2023;55:1696-708. [Crossref] [PubMed]
10. Pan JW, Ragu M, Chan WQ, Hasan SN, Islam T, Teoh LY, Jamaris S, See MH, Yip CH, Rajadurai P, Looi LM, Taib NAM, Rueda OM, Caldas C, Chin SF, Lim J, Teo SH. Clustering of HR + /HER2- breast cancer in an Asian cohort is driven by immune phenotypes. Breast Cancer Res 2024;26:67. [Crossref] [PubMed]
11. Li JW, Tong YY, Zhou J, Shi ZT, Sun PX, Chang C. Tumor Proliferation and Invasiveness Derived From Ultrasound Appearances of Invasive Breast Cancers: Moving Beyond the Routine Differential Diagnosis. J Ultrasound Med 2020;39:1589-99. [Crossref] [PubMed]
12. SEER. SEER Cancer Statistics Review (CSR) 1975-2016. 2020 [cited 2024 Sep 9]. Available online: https://seer.cancer.gov/csr/1975_2016/index.html
13. Giuliano AE, Connolly JL, Edge SB, Mittendorf EA, Rugo HS, Solin LJ, Weaver DL, Winchester DJ, Hortobagyi GN. Breast Cancer-Major changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin 2017;67:290-303.
14. Kulkoyluoglu-Cotul E, Arca A, Madak-Erdogan Z. Crosstalk between Estrogen Signaling and Breast Cancer Metabolism. Trends Endocrinol Metab 2019;30:25-38. [Crossref] [PubMed]
15. Ma Q, Liu YB, She T, Liu XL. The Role of Ki-67 in HR+/HER2- Breast Cancer: A Real-World Study of 956 Patients. Breast Cancer (Dove Med Press) 2024;16:117-26. [Crossref] [PubMed]
16. Andre F, Arnedos M, Goubar A, Ghouadni A, Delaloge S. Ki67--no evidence for its use in node-positive breast cancer. Nat Rev Clin Oncol 2015;12:296-301. [Crossref] [PubMed]
17. Zhong YM, Tong F, Shen J. Lympho-vascular invasion impacts the prognosis in breast-conserving surgery: a systematic review and meta-analysis. BMC Cancer 2022;22:102. [Crossref] [PubMed]
18. Gluz O, Nitz UA, Christgen M, Kates RE, Shak S, Clemens M, et al. West German Study Group Phase III PlanB Trial: First Prospective Outcome Data for the 21-Gene Recurrence Score Assay and Concordance of Prognostic Markers by Central and Local Pathology Assessment. J Clin Oncol 2016;34:2341-9. [Crossref] [PubMed]
19. Gluz O, Hofmann D, Würstlein R, Liedtke C, Nitz U, Harbeck N. Genomic profiling in luminal breast cancer. Breast Care (Basel) 2013;8:414-22. [Crossref] [PubMed]
20. Partridge AH, Hughes ME, Warner ET, Ottesen RA, Wong YN, Edge SB, Theriault RL, Blayney DW, Niland JC, Winer EP, Weeks JC, Tamimi RM. Subtype-Dependent Relationship Between Young Age at Diagnosis and Breast Cancer Survival. J Clin Oncol 2016;34:3308-14. [Crossref] [PubMed]