Breast-conserving surgery is an appropriate procedure for centrally located breast cancer: a population-based retrospective cohort study

Background The evidence of breast-conserving therapy (BCT) applied in centrally located breast cancer (CLBC) is absent. This study aims to investigate the long-term survival of breast-conserving therapy (BCT) in centrally located breast cancer (CLBC) compared with mastectomy in CLBC and BCT in non-CLBC. Methods Two hundred ten thousand four hundred nine women with unilateral T1-2 breast cancer undergoing BCT or mastectomy were identified from the Surveillance, Epidemiology, and End Results database. Kaplan–Meier survival curves were assessed via log-rank test. Propensity score matching (PSM) was used to balance baseline features, and the multivariable Cox model was used to estimate the adjusted hazard ratio [HR] and its 95% confidence interval [CI] for breast cancer-specific survival (BCSS) and overall survival (OS). Results With a median follow-up of 91 months, the BCSS and OS rates in patients who received BCT were greater than those patients treated with mastectomy in the entire CLBC set. Multivariable Cox analyses showed that CLBC patients who received BCT had better BCSS (HR = 0.67, 95%CI: 0.55–0.80, p < 0.001) and OS (HR = 0.78, 95%CI: 0.68–0.90, p = 0.001) than patients who received a mastectomy, but there were no significant differences of BCSS (HR = 0.65, 95%CI: 0.47–0.90, p = 0.009) and OS (HR = 0.82, 95%CI: 0.65–1.04, p = 0.110) after PSM. In patients treated with BCT, CLBC patients had a similar BCSS (HR = 0.99, 95%CI: 0.87–1.12, p = 0.850) but a worse OS (HR = 1.09, 95%CI: 1.01–1.18, p = 0.040) compared to that of the non-CLBC patient, but there was no significant difference both BCSS (HR = 1.05, 95%CI: 0.88–1.24, p = 0.614) and OS (HR = 1.08, 95%CI: 0.97–1.20, p = 0.168) after PSM. Conclusion Our findings revealed that BCT should be an acceptable and preferable alternative to mastectomy for well-selected patients with CLBC. Supplementary Information The online version contains supplementary material available at 10.1186/s12893-023-02181-6.


Introduction
Centrally located breast cancer (CLBC) refers to the location of cancer in the nipple areola or the central region of the breast.Although breast-conserving treatment (BCT) has become the standard care for early-stage breast cancer, direct evidence of BCT for CLBC patients is lacking [1,2].
BCT consists of breast-conserving surgery (BCS) and radiotherapy, which has been proven to be at least equivalent or even superior to mastectomy concerning survival outcomes [3][4][5].The choice between BCS and mastectomy also depends on the tumor's location, marginal status, adjuvant therapy, and cosmetic appearance [6,7].The patient's willingness and the recurrence risk are important factors influencing the choice of BCS [6,7].
CLBC had different clinical features compared with non-CLBC.Patients with CLBC had a higher axillary lymph node metastasis rate, higher possibility of positive margin and invasion of the NAC, lower satisfying cosmetic outcome, and increased local recurrence rate [8,9].The published data regarding BCS in CLBC are scarce; only limited studies with a small sample size support the safety of BCS in CLBC [10][11][12][13].Several researches based on real-world data concluded that BCS may be an alternative for CLBC patients.However, these studies had limitations in the study design, including advanced breast cancer, absence of post-surgical radiotherapy, small sample, and insufficient follow-up, which cannot provide evidence for the utility of BCT in patients with early-stage CLBC [14][15][16][17][18][19].Further study is needed to evaluate the long-term oncological safety of BCT in CLBC.
Limited evidence supports BCS as a safe oncological alternative to mastectomy in CLBC patients.Hence, we hypothesized that the prognosis of BCT in CLBC is like that of mastectomy in CLBC, and like that of BCT in non-CLBC patients.We conducted a retrospective cohort study using the data extracted from the US Surveillance, Epidemiology, and End Results (SEER) database, to analyze the overall and cancer-specific survival between BCT and mastectomy in CLBC patients, and between CLBC and non-CLBC patients treated with BCT.

Methods
We used the data identified from the SEER database, which represents approximately 30% of the American population (https:// seer.cancer.gov/).Informed consent was not required because personal identifying information was not accessed and no intervention was conducted.This study protocol was approved by the Clinical Research Ethics Committee of the Suining Central Hospital (No. LLSLH20220013).This study was conducted and reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement [20].

Patients selection
Data were retrieved using the SEER*Stat version 8.3.9 on May 23, 2022 (user name: 10143-Nov2021).The case list of breast cancer who met the following inclusion criteria was generated: aged 18 to 70-year-old; female, year of diagnosis between 2004 and 2015, histologically diagnosed as breast carcinoma, breast cancer was the first primary carcinoma, tumor stage was T1-2, N0-3 and M0, local surgery was performed.Patients with one of the following exclusion criteria were excluded: tumor accumulation in the whole breast or unknown tumor location, surgery procedure unknown, without radiotherapy after BCS, without follow-up data, or tumor stage.
Patients were classified as CLBC group if the case with a tumor located in the nipple (code: C500) or central portion of the breast (code: C501).In contrast, the remaining patients were classified as the non-CLBC group.Then, patients were divided into BCT (including BCS and radiotherapy) and mastectomy cohorts according to the breast surgical procedure (Supplementary Table 1).

Variables
The sociodemographic features, clinicopathological characteristics, and survival data were extracted from the database.According to the age at diagnosis, patients were divided into groups of 18-40 years, 41-50 years, 51-60 years, and 61-70 years.Marital status was classified as unmarred (including single, separated, unmarried or domestic partner, widowed, and divorced), married, and unknown.The histological type was classified into three subtypes: invasive ducal carcinoma (IDC, code:8500/3), invasive lobular carcinoma (ILC, code: 8520/3), and others according to the 3 rd edition of the International Classification of Diseases for Oncology (ICD-O-3).Tumor TNM stage was classified referenced to the standard of the 6 th edition of the breast cancer anatomical staging system of the American Joint Committee on Cancer (AJCC) [21].Estrogen receptor (ER) status and progesterone receptor (PR) status were divided into negative, positive, and unknown.The human epidermal growth factor receptor-2 (HER-2) status was available from 2010, which was classified as negative, positive, and unknown.The breast surgery was classified into the BCS group (code: 19 and 20) and the mastectomy group (code: 80).
The primary outcomes of this study were breast cancerspecific survival (BCSS) and overall survival (OS).BCSS was calculated from the date of breast cancer diagnosis to the date of death from breast cancer or the last follow-up for patients still alive.OS was computed from the time of diagnosis until the time of death from any cause, or the last follow-up for patients still alive.

Statistical analysis
All variables were converted to categorical variables and presented with percentages.The Chi-square test was used to compare patient-specific variables between BCT and mastectomy in the CLBC cohort and between CLBC and non-CLBC in the BCT cohort.To overcome the effects of baseline differences on survival outcomes, the propensity score match (PSM) method was used to balance baseline features between BCT and mastectomy patients in the CLBC cohort, and between CLBC and non-CLBC patients in the BCT cohort at a ratio of 1:1, respectively.Survival outcomes were estimated using the Kaplan-Meier plot and compared across groups using the log-rank test.Univariable and multivariable Cox proportion risk regression models were used to identify the prognostic factors for BCSS and OS, and hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated.Variables with p < 0.1 in univariable analysis or with a clinical consideration were enrolled in the multivariable model.Statistical analyses were performed by R software (version 4.0.3 for Windows) using the packages "survival" and "survminer".All statistical tests were two-sided, and p < 0.05 was considered statistically significant.

Baseline characteristics
Finally, 210,409 patients with stage T 1-2 non-metastatic breast cancer as their first malignancy were retrieved (Fig. 1).Among them, 15,013 (7.14%) were CLBC patients, and 195,396 (92.86%) were non-CLBC patients.From 2004 to 2015, the proportion of BCS was increased slightly in CLBC patients and relatively stable in non-CLBC patients (Fig. 2).The proportion of BCS in CLBC patients was lower than that of the non-CLBC group (42.7% vs. 60.5%,p < 0.001).The distribution of characteristics between CLBC and non-CLBC patients was significantly different except for chemotherapy (Supplementary Table 2).
The median age was 56 [interquartile range (IQR): 48 ~ 63 years].The clinical characteristics of CLBC patients undergoing BCT and mastectomy are summarized in Table 1.Compared with patients who received a mastectomy, elderly patients, patients with a Fig. 1 Study flow chart and patient selection histological type of IDC, and the white race were more likely to receive BCS.In addition, CLBC patients with less aggressive characteristics such as lower histological grade, smaller tumor size, none or limited lymph node metastasis, or positive ER or PR status tended to be treated with BCS.CLBC patients treated with BCS were less likely to receive chemotherapy than mastectomy.For CLBC patients, the baseline clinicopathological characteristics of 4865 paired patients were balanced after the PSM (Table 1).
The clinical characteristics of CLBC patients and non-CLBC cases treated with BCT were also compared (Table 2).Compared with non-CLBC patients, younger people with less aggressive characteristics such as lower histologic grade, smaller tumor size, none or limited lymph node metastasis, and positive ER or PR status, or negative HER-2 in CLBC cohorts were more likely to receive mastectomy, instead of BCS.However, the baseline clinicopathological characteristics of 6416 paired patients were balanced after the PSM (Table 2).

Survival analyses before PSM
Overall, 15,013 CLBC cases and 195,396 non-CLBC cases were enrolled in the analysis (Supplementary Table 2).During a median follow-up of 91 months (IQR: 58 ~ 130 months), 24,687 patients died (11.73%) and 12,833 (6.10%) patients died of breast cancer in the total population.The Kaplan-Meier survival curves showed that CLBC patients who received BCS had a superior BCSS and OS than those patients who were treated with mastectomy (Fig. 3A, B).Meanwhile, there was no difference in BCSS and OS between CLBC and non-CLBC groups in patients treated with BCT (Fig. 3C, D).The estimated 3-, 5-, 7-, and 10-year BCSS rates and OS rates were summarized in Supplementary Table 3.
Among CLBC patients, the univariate Cox analysis found that year of diagnosis, age at diagnosis, race, marital status, grade, tumor stage, node stage, ER status, PR status, surgery procedure, and chemotherapy were associated with the BCSS and OS, while HER-2 status and radiotherapy were only associated with OS but not BCSS   (Supplementary Table 4).After adjusting the potential confounding factors, the multivariable Cox regression model showed that patients who received BCS had a better BCSS (HR = 0.67, 95%CI: 0.55-0.80,p < 0.001) and OS (HR = 0.78, 95%CI:0.68-0.89,p = 0.0005) compared with patients who received mastectomy (Supplementary Table 5).Among patients who received BCS, the univariate Cox analysis showed that the tumor location was not associated with BCSS but with OS (Supplementary Table 6).The multivariable Cox model showed that patients with CLBC had a similar BCSS (HR = 0.99, 95%CI: 0.87-1.12,p = 0.850) and a worse OS (HR = 1.09, 95%CI: 1.01-1.18,p = 0.030) compared with the non-CLBC patients (Supplementary Table 7).

Survival analyses after PSM
The survival curves showed that CLBC patients treated with BCS had a higher BCSS rate (log-rank test: p = 0.0005) and OS rate (log-rank test: p < 0.0001) compared to those who received mastectomy (Fig. 4A, B).

Discussion
The choice between mastectomy and BCS refers to the research evidence, cosmetic appearance, and patient's desire.In our study, only 42.7% of CLBC patients underwent BCS, which is significantly less than that of non-CLBC patients (60.5%).However, the survival analysis results showed that there was no difference in OS and BCSS between BCS and mastectomy in the CLBC populations, and between CLBC patients and non-CLBC patients who received BCS.Therefore, there is rising confidence in advocating BCS as a surgical option for women with early-stage CLBC from a safety perspective.Some differences in clinicopathological features were observed between CLBC and non-CLBC patients.In the CLBC population, these patients who exhibited less aggressive histological traits tended to receive BCS.This may be one reason for the therapeutic difference between the two populations.Unfortunately, we were unable to access the detailed information on adjuvant therapy or other pathological features (Ki-67 et al.) in the SEER database, which limited us to determine whether these factors would influence the decision of breast surgery [22,23].Compared with non-CLBC, patients with less aggressive characteristics such as lower histologic grade, smaller tumor size, none or limited lymph node metastasis, positive hormone receptor status, or negative HER-2 in CLBC cohorts were more likely to receive mastectomy, instead of BCS.The tumor characteristics of CLBC patients who received BCS are not completely superior to those of non-CLBC patients.Interestingly, our survival data supported that BCS for CLBC resulted in non-inferior outcomes compared with mastectomy in CLBC or with BCS in non-CLBC, even in N2-N3 disease.These indicated that tumor biological behavior features should not be a barrier to BCS in CLBC.The improved survival for BCS was likely attributed to the advances in early diagnosis, surgery procedures, and adjuvant therapy over the last decades [24,25].
The standard implementation of BCS and adjuvant therapy is still the basic guarantee of BCT for patients with CLBC.There is no special requirement for a negative margin for BCS in CLBC patients.However, we need to be aware that central breast cancer may have the potential to invade the wider breast ducts [8,9].Considering the cosmetic appearance and local control, tumor size is an important factor associated with the selection of BCS in clinical practice, which is supported by several guidelines and consensuses [1,2,26].Several small sample, short-term follow-up studies based on the SEER database have enrolled some T3-4 or stage III/IV disease which would influence the application of research results   [14,15,17,18].Adjuvant radiotherapy was necessary for reducing local recurrence and improving long-term disease-free survival and overall survival for whole breast cancer populations treated with BCS [27][28][29].Without radiotherapy, the nipple-areola complex involvement was related to a higher risk of local recurrence after BCS [30].In previous reported studies, many patients treated with BCS but without radiotherapy, which would induce a select bias [15][16][17][18].Based on the consideration of clinical practice, the tumor stage was limited as T1-2 and radiotherapy after BCS was required in our study, which makes the study results suitable for clinical work.CLBC has a four times higher risk of involvement of nipple-areolar complex (NAC) than that of non-CLBC [31].Removal of NAC is usually recommended for BCS in CLBC patients.However, traditional incisions with NAC resection usually led to poor aesthetic outcomes [32,33].Preserving NAC is generally considered essential for maintaining the aesthetic appearance of the breast.Oncoplastic techniques for breast-conserving surgery could improve cosmetic results without jeopardizing oncological outcomes [34].A good cosmetic outcome improves psychosocial adjustment after breast cancer treatment.Fortunately, many articles have discussed various approaches to achieve optimal results following oncoplastic surgery, ranging from classic reduction mammaplasty to different flap reconstruction techniques [35][36][37].Additionally, restoration of the soft tissue defects by various oncoplastic procedures, with or without immediate reconstruction of the NAC, has been proven to be both oncological safe and cosmetically effective [10,35].So, the aesthetic factor should not be the main obstacle to the performance of BCS in CLBC patients.
In this study, many clinicopathological features were identified as independent prognostic factors for BCSS and OS.Patients with older age, married, lower grade, smaller tumor size, lymph node-negative, hormone receptor-positive, HER-2 negative, and BCS were the greater independent prognostic factor of BCSS and OS for CLBC.Meanwhile, chemotherapy was a superior prognostic indicator for OS.Contrastingly, radiotherapy did not improve survival in the whole CLBC population.Similar results were observed in CLBC patients who received BCS.Tumor stage (tumor size, lymph node stage), biological features (grade, hormone receptor status, HER2 status), and chemotherapy have been investigated well to predict survival for breast cancer patients [24,25,38,39].Marital status reflects the social, psychological, and economic conditions, which could affect the diagnosis, treatment, and prognosis of a female patient indirectly [40,41].The results of these prognostic analyses were well consistent with those of previous reports, which indicated the results of our study have high reliability.Identifying these risk factors is instrumental to accurately assess prognosis and develop individualized management strategies [42,43].
There are still some limitations in our research.First, although we have set inclusion and exclusion criteria strictly, and taken the PSM method to balance the baseline features, there may still exist some selective bias.Second, local recurrence data are unavailable from the SEER database, so we cannot evaluate the recurrence between each group.Third, we could not evaluate the influence of neoadjuvant chemotherapy on surgical choice and survival outcomes.Fourth, we did not have access to detailed information on adjuvant therapy, which may influence the survival outcomes.Fifth, the lack of information on comorbidities, performance status, treatment toxicities, tumor biological features (Ki-67 rate), education level, and socioeconomic status which may introduce bias into    • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year

•
At BMC, research is always in progress.

Learn more biomedcentral.com/submissions
Ready to submit your research Ready to submit your research ?Choose BMC and benefit from: ?Choose BMC and benefit from:

Fig. 2
Fig. 2 Trends of BCS rate from 2004 to 2015 in patients with CLBC and non-CLBC

Fig. 3
Fig. 3 Survival curves of BCSS and OS stratified by surgery procedure and tumor location before PSM (A BCSS of BCS versus mastectomy; B OS of BCS versus mastectomy; C BCSS of CLBC versus non-CLBC; D OS of CLBC versus non-CLBC)

Fig. 4
Fig. 4 Survival curves of BCSS and OS stratified by surgery procedure and tumor location after PSM (A BCSS of BCS versus mastectomy; B OS of BCS versus mastectomy; C BCSS of CLBC versus non-CLBC; D OS of CLBC versus non-CLBC)

Table 1
Comparison of baseline features between BCS and mastectomy in CLBC patients before and after PSM [n (%)] Abbreviations: AJCC American Joint Committee on Cancer, BCS Breast conserving surgery, BCSS Breast-cancer specific survival, CLBC Centrally located breast cancer, ER Estrogen receptor, HER2 Human epidermal growth factor receptor-2, IDC Invasive ductal carcinoma, ILC Invasive lobular carcinoma, OS Overall survival, PR Progesterone receptor, PSM Propensity score matching, CI Confidence interval, HR Hazard ratio

Table 2
Comparison of baseline features between CLBC and non-CLBC patients who received BCT before and after PSM Abbreviations: AJCC American Joint Committee on Cancer, BCS Breast conserving surgery, ER Estrogen receptor, HER2 Human epidermal growth factor receptor-2, IDC Invasive ductal carcinoma, ILC Invasive lobular carcinoma, PR Progesterone receptor, PSM Propensity score matching

Table 3
Multivariable Cox analysis for BCSS and OS in centrally located breast cancer patients after PSM Abbreviations: AJCC American Joint Committee on Cancer, BCS Breast conserving surgery, BCSS Breast-cancer specific survival, CI Confidence interval, ER Estrogen receptor, HER2 Human epidermal growth factor receptor-2, HR Hazard ratio, IDC Invasive ductal carcinoma, ILC Invasive lobular carcinoma, OS Overall survival, PR Progesterone receptor, PSM Propensity score matching

Table 3
(continued) ConclusionOur study demonstrated the long-term oncological safety of BCS in CLBC compared with mastectomy and non-CLBC disease.As a result, BCS should be an acceptable and preferable alternative to mastectomy for wellselected, early-stage CLBC patients.This study was supported by the Suining Talent Lifting Project and Foundation of State Key Laboratory of Ultrasound in Medicine and Engineering (No. 2021KFKT015).

Table 4
Multivariable Cox analysis for BCSS and OS in patients who received breast-conserving treatment after PSM Abbreviations: AJCC American Joint Committee on Cancer, BCS Breast conserving surgery, BCSS Breast-cancer specific survival, CI Confidence interval, ER Estrogen receptor, HER2 Human epidermal growth factor receptor-2, HR Hazard ratio, IDC Invasive ductal carcinoma, ILC Invasive lobular carcinoma, OS Overall survival, PR Progesterone receptor, PSM Propensity score matching