Abstract
Background/Aim: Following the National Comprehensive Cancer Network guidelines, radiotherapy is administered after breast-conserving surgery (BCS) in patients with more than four positive lymph nodes. Four positive lymph nodes are typically considered an indicator to assess disease spread and patient prognosis. However, the subjective counting of positive axillary lymph nodes underscores the need for biomarkers to improve diagnostic precision and reduce the risk of unnecessary treatments. Loss of E-cadherin expression is associated with cancer metastasis, but its potential as a predictive marker for cancer treatment remains uncertain. This study aimed to investigate the validity of E-cadherin as a reference for adjuvant radiotherapy in breast cancer patients with positive lymph nodes post-mastectomy. Materials and Methods: Immunohistochemistry was performed on 60 clinical tissue specimens to assess these implications. Results: Although no significant result was found in a single E-cadherin subgroup (low, medium, and high subgroups according to the X-tile algorithm), the proposed multivariate model, including the E-cadherin category, breast cancer subtype, and tumor size, yielded satisfactory recurrence risk estimation results for patients undergoing BCS. Patients with a low E-cadherin category, triple-negative breast cancers, and tumor size over 5 cm could have an increased risk of recurrence. Conclusion: Our study proposed a multivariate model that serves as a candidate prognostic factor for recurrence-free survival in patients undergoing BCS and radiotherapy. Utilizing this model for patient stratification in high-risk diseases and as a standard for assessing postoperative intensified therapy can potentially improve patient outcomes.
Breast-conserving surgery (BCS), also known as lumpectomy or partial mastectomy, aims to excise a tumor or a section of the breast while preserving as much healthy tissue as possible (1, 2). Despite its benefits, BCS may carry a slightly higher risk of local recurrence compared to mastectomy, particularly when the tumor is large or margins are unclear (3, 4). Patients undergoing BCS require long-term monitoring and follow-up to detect recurrence (5). Additionally, radiation therapy is often recommended to minimize the risk of local recurrence, introducing another layer to the treatment regimen that may extend the duration and introduce potential side effects (6). Consequently, BCS may not be the optimal choice for every patient, with suitability varying based on tumor size, location, and patient preferences (7).
BCS is typically more appropriate for smaller tumors. Tumor size is a critical factor, with BCS often advised for tumors less than a specific size (e.g., T1 or T2). Additionally, the degree of lymph node involvement is a crucial consideration. BCS may be an option if the cancer has not extensively spread to adjacent lymph nodes. Achieving clear surgical margins is crucial, ensuring the complete excision of cancerous tissue along with a border of normal tissue. Inadequate removal heightens the risk of local recurrence (4). Research also suggests that the likelihood of recurrence varies with different breast cancer subtypes following BCS (8). The hormone receptor-positive subtype usually presents a lower recurrence risk after BCS, particularly when combined with adjuvant hormonal therapy, such as tamoxifen or aromatase inhibitors. Human epidermal growth factor receptor 2 (HER2)-positive tumors might carry a marginally increased recurrence risk. Nonetheless, the concurrent use of targeted therapies, like trastuzumab (Herceptin®), with BCS and other treatments can mitigate this risk. Triple-negative breast cancers (TNBC) are linked to a higher recurrence risk compared to hormone receptor-positive subtypes (9). Adjuvant chemotherapy is frequently recommended to manage this heightened risk. Consequently, it is important to recognize that recurrence risk is personalized, with treatment decisions based on an integration of these factors.
E-cadherin is involved in cell adhesion and the maintenance of tissue integrity (10). In many breast cancer cases, there is a loss or downregulation of E-cadherin expression. This loss is associated with a more malignant phenotype, contributing to increased invasiveness and metastatic potential of cancer cells and worse clinical outcomes, including decreased disease-free survival and overall survival (11). A study investigated the association between E-cadherin expression and specific molecular subtypes of breast cancer (12). For instance, in hormone receptor-negative and TNBC sub-types, low E-cadherin levels may contribute to a more aggressive phenotype and higher recurrence rates (13). Therefore, monitoring E-cadherin expression and understanding its molecular interactions can provide valuable insights for prognostic assessment and developing targeted therapeutic approaches in breast cancer treatment.
This study aimed to enhance our understanding of E-cadherin’s impact on clinical outcomes in breast cancer patients. It involved protein level assessments through immunohistochemistry (IHC) staining. The primary objective was to elucidate the combined role of E-cadherin, breast cancer subtypes, and tumor size in recurrence among patients undergoing BCS combined with RT. The aim was to assess the potential significance of E-cadherin in guiding therapeutic strategies for individuals with breast cancer.
Materials and Methods
Data source. Data were retrospectively collected from a single medical institute under approved protocols [IRB no. KMUHIRB-F(I)-20200107]. Inclusion criteria were patients with invasive ductal carcinoma (IDC), age over 20 years, and the availability of a primary tumor tissue specimen for IHC staining. Sixty women diagnosed with breast cancer who met these criteria were analyzed. Assessed clinical characteristics included age at diagnosis, breast cancer subtype, lymphovascular invasion (LVI), dermal invasion (DI), perineural invasion, tumor size (cm), lymph node (LN) invasion, and tumor stage. Treatment characteristics, such as breast mastectomy, chemotherapy (CT), and radiotherapy (RT) history were also recorded. Survival outcomes observed were recurrence-free survival (RFS) and all-cause mortality.
E-cadherin IHC staining. In brief, 5 μm sections of formalin-fixed paraffin-embedded tissues were heated at 65°C for 1 h. These sections were then subjected to deparaffinization and dehydration through a series of steps: twice incubated in xylene for 5 min each, washed in pure ethanol twice (the second wash lasting 5 min), washed in 95% ethanol twice (the second wash lasting 5 min), soaked in 75% ethanol for 5 min, and soaked in 50% ethanol for 5 min. Subsequently, the slides were washed twice in water, with the second wash lasting 5 min. Antigen retrieval involved heating the slides in a steamer for 10 min in 1× trilogy buffer (Merck KGaA, Darmstadt, Germany). After cooling, the slides were washed three times with phosphate-buffered saline (PBS), incubated in 3% hydrogen peroxide in double-distilled water for 10 min, and then washed three times with PBS. Tissues were blocked with 5% goat serum in PBS for 1 h at room temperature. The primary antibody (anti-E-cadherin; BD Transduction Laboratories, San Jose, CA, USA; Cat. No. 610182) was applied at a 1:2,000 dilution in blocking buffer and incubated at 4°C overnight. Following three PBS washes, the secondary antibody reagent, EnVision+ Peroxidase system (Dako, Glostrup, Denmark) was applied at room temperature for 1 h. Hematoxylin was used for counterstaining and incubated for 1 min. Slides were washed in running water for 10 min, dehydrated in a series of ethanol washes (75% for 3 min, 95% for 1 min, 95% for 3 min, 100% for 1 min, 100% for 3 min), and immersed twice in xylene for 10 min each. Finally, the slides were mounted for microscopic examination.
Scoring. The H-score was calculated by summing the products of the average percentage of E-cadherin-positive cells and their respective intensity scores across various staining intensities. The calculation formula was as follows: H-score=[1× (% of E-cadherin-positive cells with intensity category 1)] + [2× (% of E-cadherin-positive cells with intensity category 2)] + [3× (% of E-cadherin-positive cells with intensity category 3)]. To ascertain the average percentage of tumor cells at different intensities, ten fields were randomly selected and examined at 400× magnification. In each field, the total cell count and the number of cells in each intensity category were documented. The average percentage was then computed across these fields, yielding an H-score that ranged from 0 to 300.
E-cadherin expression subgroups. The estimated E-cadherin IHC scores for the study cohort were classified into high, medium, and low subgroups using the X-tile algorithm (14). The X-tile cutoff results, stratified by RFS status, are depicted in Figure 1. Representative IHC images for each E-cadherin subgroup, according to RFS status, are presented in Figure 2.
X-tile cutoff of E-cadherin IHC score according to recurrence-free survival. (A) The histogram represents the distribution of E-cadherin expression in high, medium, and low subgroups categorized using X-tile algorithm. (B) The panel graphically represents a right-triangular grid in which each point indicates the protein expression in the study cohort. The black dot denotes the optimal cut point estimated using X-tile algorithm.
The expression of E-cadherin by immunohistochemistry (IHC) according to the x-tile cutoff in recurrence patients and controls.
Statistical analysis. The clinical characteristics of the study cohort are summarized using the median and range or frequency and percentage. Differences in clinical characteristics between E-cadherin IHC subgroups were estimated using Kruskal–Wallis, chi-squared, and Fisher’s exact tests. The follow-up duration for RFS was measured from the date of initial diagnosis to the date of first recurrence. The RFS rate of breast cancer patients was computed using the Kaplan– Meier method, and overall and pairwise comparisons between subgroups were tested using the log-rank test. Cox proportional hazards regression was performed to investigate the impact of E-cadherin and associated clinical characteristics on RFS. The multivariate model retained only variables with a p-value <0.2 in univariate analyses. LN invasion is the most commonly used factor to assess recurrence risk in clinical settings. Therefore, Harrell’s C-indices for the proposed model, E-cadherin subgroup, and LN invasion are reported to compare the risk assessment performance for RFS. All p-values were two-tailed, and p<0.05 was considered statistically significant. All statistical analyses were conducted using R 4.2.3 (R Core Team, 2023, Vienna, Austria).
Results
Determination of E-cadherin cutoff value. To assess the predictive value of E-cadherin for RFS in patients with breast cancer, we utilized the X-tile software. Figure 1 demonstrates the classification of patients based on their E-cadherin IHC scores: patients with scores <11.6 were assigned in the low subgroup, those with scores between 11.6 and 87.3 in the medium subgroup, and those with scores ≥87.3 in the high subgroup. The low, medium, and high E-cadherin subgroups comprised 11, 28, and 21 patients, respectively. Figure 2 presents the distribution of low, medium, and high E-cadherin expression levels in recurrence patients and controls, as determined by the X-tile cutoff.
Correlation between E-cadherin and clinicopathological characteristics. Table I summarizes the baseline characteristics of the study cohort by E-cadherin subgroups. The low E-cadherin subgroup had a younger age at diagnosis, larger median tumor size, slightly higher LVI proportion, and more advanced tumor stage. This subgroup also had a higher likelihood of receiving CT+RT and undergoing BCS, and they exhibited a greater proportion of recurrence and all-cause mortality. Despite these differences, the clinical characteristics between the E-cadherin subgroups were not significantly different overall.
Baseline characteristics of study cohort (n=60).
Clinical outcomes according to E-cadherin status combined with high-risk factors in patients with breast cancer. Subsequently, we performed Cox proportional hazards regression analysis to evaluate the influence of E-cadherin and related clinical characteristics on recurrence risk within the study cohort (Table II). Univariate analysis revealed that the low E-cadherin IHC subgroup had a higher recurrence risk compared to the medium and high subgroups, although the differences were not statistically significant. However, after adjusting for tumor subtype and size, a significant elevation in recurrence risk was noted in the low E-cadherin IHC subgroup relative to the high E-cadherin IHC subgroup [adjusted hazard ratio (HR)=4.38, 95% confidence interval (CI)=1.13-17.0, p=0.033]. Additionally, larger tumor size (≥5 cm vs. <2 cm: crude-HR=6.04, 95%CI=1.36-26.80, p=0.018) was associated with a significant increase in recurrence risk in the univariate analysis; however, this significance was not maintained after adjusting for E-cadherin IHC and tumor subtype.
Cox proportional hazard regression results for recurrence risk in study cohort.
Predictive value of E-cadherin expression for determining recurrence risk Figure 3 depicts the RFS curve for the study cohort. The low E-cadherin IHC subgroup exhibited a reduced RFS rate compared to the medium and high E-cadherin IHC subgroups. Specifically, Figure 3A reveals an overall RFS rate of 20.2% (95%CI=3.6-100) for the low E-cadherin IHC subgroup, while the medium and high subgroups had RFS rates of 51.2% (95%CI=32.0-82.0) and 68.8% (95%CI=49.2-96.2), respectively. Figure 3B-D presents the RFS for the CT+RT, BCS, and radical mastectomy subcohorts. A similarly poor RFS threshold was observed in the low E-Cadherin subgroup in both CT+RT and BCS subcohorts. These findings suggest that patients with low E-cadherin IHC may have a higher risk of breast cancer recurrence.
Kaplan–Meier plot for recurrence-free survival in the study cohort. (A) Overall. (B) Chemotherapy (CT)+ radiotherapy (RT) subcohort. (C) Breast conserving surgery (BCS) subcohort. (D) Radical subcohort.
We subsequently estimated Harrel’s C-index for the proposed multivariate model, single E-cadherin according to IHC, and LN invasion for recurrence risk assessment in the overall study cohort, as well as in the CT+RT and BCS subcohorts. The results of the recurrence risk assessment comparison were visualized using a receiver operating characteristic curve. Figure 4A demonstrates that the proposed model (C-index=0.706, 95%CI=0.562-0.850) outperformed single E-cadherin IHC (C-index=0.618, 95%CI=0.474-0.762) and LN invasion (C-index=0.538, 95%CI=0.466-0.609) in the overall study cohort in terms of recurrence risk. Figure 4B and C show that the proposed model achieved even better performance in recurrence risk assessment for the CT+RT (C-index=0.725, 95%CI=0.568-0.882) and BCS (C-index=0.799, 95%CI=0.638-0.960) subcohorts.
Proposed model performance on recurrence risk assessment in the study cohort. (A) Overall. (B) Chemotherapy (CT)+ radiotherapy (RT) subcohort. (C) Breast conserving surgery (BCS) subcohort.
Although no significant results were found in the single E-cadherin IHC subgroup, the proposed model, which includes E-cadherin IHC, tumor subtype, and tumor size, yielded satisfactory recurrence risk estimation with Harrel’s C-index (>0.70) in the overall study cohort, as well as in the CT+RT and BCS subcohorts. In summary, this model could potentially assess the recurrence risk for breast cancer patients, particularly for those who have undergone BCS or CT+RT treatments.
Discussion
The high heterogeneity of breast cancer presents a challenge for treatment (15). Consequently, this study aimed to integrate clinical parameters with predictive biomarker analyses to aid in assessing the risk of recurrence after breast and lymph node surgery in pre-breast cancer patients. In this retrospective cohort study of patients with breast cancer, we report the following findings: low E-cadherin expression, as indicated by the corresponding IHC subgroup, was correlated with an increased risk of recurrence. However, this association did not reach statistical significance in the univariate analysis. Significance emerged only after adjusting for the E-cadherin IHC subgroup, tumor subtype, and tumor size, underscoring the necessity of considering these factors collectively. Notably, our proposed multivariate model exhibited superior performance in evaluating recurrence risk, with a concordance index (C-index) of 0.725 (95%CI=0.568-0.882) in subcohorts receiving both chemotherapy and radiotherapy (CT+RT) and a C-index of 0.799 (95%CI=0.638-0.960) in those undergoing BCS. To our knowledge, this study is the first to ascertain the predictive value of pretreatment E-cadherin according to IHC in patients with breast cancer receiving BCS and chemoradio-therapy.
Adjuvant radiation therapy is commonly used for patients at high risk of recurrence following partial mastectomy. However, it can cause irreversible damage at the lesion site, affecting postoperative quality of life and care (16-18). Thus, it is crucial to address the challenge of avoiding excessive medical interventions. Tumor size, lymphatic invasion, and breast subtypes are closely associated with postoperative recurrence (19-21). Previous studies have shown that E-cadherin expression is regulated by various mechanisms (22-24). E-cadherin plays a role in mediating cell adhesion and maintaining tissue integrity (25, 26). A reduction or loss of E-cadherin expression is associated with epithelial-mesenchymal transition, a process in which cancer cells acquire invasive properties, often correlated with a poorer prognosis. E-cadherin also plays a pivotal role in cancer progression and in regulating the response to chemotherapy in several cancer types. Most studies suggest that reduced E-cadherin expression may be a predictor of poorer prognosis for patients with breast cancer, making it a potential prognostic biomarker for the disease (11, 27, 28). However, its role in modulating the effects of radiation therapy is not well understood.
In our previous study, we established a correlation between the RNF8–TWIST signaling pathway and the downstream E-cadherin (CDH1) gene in breast cancer (29). Our findings suggested that the RNF8high/CDH1low index is not only a prognostic and therapeutic marker but also represents a potential target for developing anti-cancer agents for patients with TNBC. Individuals with an RNF8high/CDH1low index exhibited lower survival rates when treated solely with chemotherapy compared to those with an RNF8low-medium/CDH1medium-high index. Furthermore, CT+RT significantly improved survival rates over CT alone, particularly for patients with an RNF8high/CDH1low index, suggesting that E-cadherin modulates the response to radiotherapy in TNBC (29). Consequently, E-cadherin expression appears to be associated with the efficacy of RT in patients with breast cancer. In this study, we evaluated the correlation between combined E-cadherin protein expression and clinical parameters in relation to postoperative recurrence.
In contrast to genetic testing, IHC staining is currently the most convenient and cost-effective diagnostic method in clinical practice. Therefore, we primarily analyzed the optimal cutoff value of E-cadherin IHC staining in breast cancer using the X-tile analytic software. Patients were categorized into different subgroups based on their E-cadherin IHC scores. These values were used for the first time for patient grouping.
E-cadherin status can impact the response to certain treatments (30, 31). For instance, in hormone receptor-positive breast cancers, E-cadherin expression may influence the response to endocrine therapies such as tamoxifen. In this study, we demonstrated that the RFS rate in the subgroup with low E-cadherin expression, as determined by IHC, was lower than that in those with medium and high E-cadherin expression. This pattern was consistently observed in both subcohorts receiving CT+RT and those undergoing BCS. These findings suggest that breast cancer patients with low E-cadherin IHC levels may have an elevated risk of recurrence. Therefore, monitoring E-cadherin expression can help identify patients at higher risk of recurrence, potentially influencing the intensity of follow-up care and the consideration of additional therapies. These examples underscore the multifaceted role of E-cadherin in breast cancer outcomes, from influencing tumor behavior to serving as a valuable marker for prognosis and treatment response.
E-cadherin has emerged as a promising therapeutic target for the design of antitumor drugs in recent years, particularly in digestive system cancers. Numerous small bioactive compounds targeting E-cadherin expression have been identified, enhancing cell-cell adhesion in these cancers (32). However, research in breast cancer has not been as progressive. Micalizzi et al. discovered a dual antibody targeting epithelial E-cadherin (CDH1) and mesenchymal OB-cadherin (CDH11) with anti-metastatic activity in breast and pancreatic cancer mouse models, resulting in a decrease in circulating tumor cells in the bloodstream (33). Our study suggests a potential avenue for research into targeted therapies and the refinement of existing treatment regimens.
Study limitations. First, although the retrospective analysis of 60 cases showed that the proposed multivariate model was superior in assessing the risk of recurrence, particularly within subcohorts undergoing both CT+RT, the inclusion of individuals who only underwent partial mastectomy without any postoperative adjuvant treatment was limited. This insufficient representation hampers the evaluation of whether E-cadherin protein expression can be used for risk stratification to predict the need for additional postoperative adjuvant therapy. Therefore, larger prospective studies are necessary to confirm these preliminary findings. Second, this study did not explore varying treatment strategies for different clinical subtypes. Third, while E-cadherin expression levels change in response to specific treatments, leading to improved therapeutic outcomes, the mechanisms underlying these changes remain largely unknown.
Conclusion
In conclusion, our study is the first to show that the proposed multivariate model is a potential prognostic factor for RFS in patients undergoing BCS and RT. Particularly, this model accurately predicts the efficacy of CT+RT in high-risk patients. Thus, given its accessibility, the proposed multivariate model could be employed to identify patients with high-risk diseases, assess intensified therapy, and facilitate the development of new agents to enhance patient outcomes and minimize unnecessary medical interventions.
Acknowledgements
This work was supported by the following grants: 1) 112-2314-B-037-045-, 111-2314-B-650-006-MY3, 112-2320-B-037-002-, 112-2628-B-037-002- and 112-2314-B-037-044 from the Ministry of Science and Technology, Taiwan. 2) KMUH-DK(C)111003 from the Kaohsiung Medical University, Kaohsiung, Taiwan. 3) KMUH110-0M37 and KMUH108-8R36 from the Kaohsiung Medical University Hospital, Taiwan.
Footnotes
Authors’ Contributions
MRP and MFH participated in concept and study design, data compilation and integration. CNK performed the laboratory work and prepared the draft of the manuscript. CCC and WLC performed experiments and data analysis, CWL and WLH performed pathologic analysis and statistical analysis. SHM participated in data compilation and integration. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors report no conflicts of interest in relation to this study.
- Received February 12, 2024.
- Revision received March 10, 2024.
- Accepted March 11, 2024.
- Copyright © 2024, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
