Incidence and risk factors of lymph node metastasis in breast cancer patients without preoperative chemoradiotherapy and neoadjuvant therapy: analysis of SEER data

Introduction

Breast cancer (BC) is the leading cause of death in the female reproductive system, accounting for 31% of cancer cases and 15% of cancer-related deaths in women. The incidence of BC in women is increasing by 0.5% annually (1). Based on statistical data, the overall incidence rate of positive sentinel lymph nodes in BC patients is approximately 33%. Patients who test positive for lymph node involvement often experience a higher mortality rate and an elevated risk of disease recurrence (2,3).

The lymphatic vessels within the breast form an open system within the surrounding matrix environment (4). As breast tumors develop, they have a significant chance of invading nearby lymph nodes or lymphatic vessels, using the lymphatic system within the breast for metastasis, resulting in multiple invasive tumor foci. The high heterogeneity of BC contributes to different disease progression and prognosis. Therefore, certain patient and tumor characteristics remain valuable in assessing the status of lymph node metastasis (5). Evaluating the burden of lymph node metastasis in BC is crucial for neoadjuvant therapy (NAT) planning, initial surgical procedures, and guiding axillary treatment (6). Although existing literature has described some risk factors for lymph node metastasis (2,7,8), a comprehensive and in-depth analysis of various clinicopathological factors influencing lymph node metastasis has not been conducted. Moreover, the sample sizes in previous investigations are often small or limited to data from a single medical center, which may lack representativeness. Therefore, based on a large sample of retrospective survey data from the period of 2010 to 2015, this study aims to construct a predictive model to examine the accuracy of previous experiences and provide a detailed analysis and summary of each predictive factor. This will lay the foundation for further refinement of future predictive models.

First, we used chi-square tests to screen for influencing factors with significant differences in these large sample data. Then, we conducted univariate and multivariate logistic regression analyses to analyze the impact of different subgroups within each factor on lymph node positivity. Additionally, the influence of different clinicopathological characteristics on lymph node metastasis in BC may yield inconsistent results due to uncontrolled potential confounding factors. Therefore, it is essential to employ propensity score analysis to assess the association between each independent clinicopathological factor of interest and the outcome (9). To create a balanced cohort, we utilized inverse probability weighting (IPW). Our plan is to evaluate the impact of confounding factors on lymph node metastasis by comparing univariate and multivariate logistic regression analyses before and after propensity score matching (PSM). Subsequently, we will construct a nomogram based on consistent influencing factors identified from the results of univariate logistic regression analysis, multivariate logistic regression analysis, and univariate logistic regression analysis after PSM. We will also investigate potential reasons for any discrepancies between PSM and univariate/multivariate logistic regression in specific clinicopathological characteristics. We present this article in accordance with the TRIPOD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-23-258/rc).

Methods

Patient selection

We searched and downloaded medical records of BC patients from the Surveillance, Epidemiology, and End Results (SEER) database (SEER Research Plus Data, 17 Registries, Nov 2021 Sub, 2000–2019), which covers cancer incidence and survival records of over one-third of the US population. We collected data from 256,504 patients based on the following criteria: (I) January 2010 to December 2015; (II) female patients aged over 20 years old; (III) no NAT given; (IV) diagnosed with primary BC; (V) no radiation and/or cancer-directed surgery & radiation after surgery; (VI) M0 stage; (VII) unilateral and unifocal tumor; (VIII) not classified as “T0”, “Ta”, “Tis”, “Tispu”, “Tispd”, “T4”, “T4a”, “T4a(s)”, “T4a(m)”, “T4b”, “T4b(m)”, “T4(c)”, “T4d”, “T4e”, “T4 NOS(s)”, “T4 NOS(m)”, “T4 NOS”, “NA”, “TX”, and “Blank(s)” in the T staging of the 7^th Cancer Staging Manual of the American Joint Committee on Cancer (AJCC 7^th T 1988–2015); and (IX) exclusion of multiple BC lesions. The flowchart of participant inclusion and exclusion is shown in Figure 1. Ethical approval was not required for this study as the clinical data of recruited BC patients were collected from publicly available and anonymized data in the SEER database. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Figure 1 The flow diagram of participant inclusion and exclusion. SEER, Surveillance, Epidemiology, and End Results; NOS, not otherwise specified; AJCC T, the T staging of Cancer Staging Manual of the American Joint Committee on Cancer; W, White; B, Black; AI, American Indian or Alaskan Native; API, Asian or Pacific Islander; ICD-O-3 hist/behav, 3rd edition of the International Classification of Diseases for Oncology histology code and behavior.

Variable description

The demographic and clinical characteristics of the participants are as follows: (I) the independent variables included age at diagnosis (20–44, 45–55, 56+ years), race (White, Black, Asian or Pacific Islander, others), primary site (upper-outer quadrant, upper-inner quadrant, lower-outer quadrant, lower-inner quadrant, central portion & nipple, others), histology (infiltrating duct carcinoma, lobular carcinoma, mucinous adenocarcinoma, others), breast subtype [hormone receptor (HR)⁺/human epidermal growth factor receptor 2 (HER2)⁻, HR⁻/HER2⁻, HR⁺/HER2⁺, HR⁻/HER2⁺, unknown; +, positive; −, negative], grade (I = well differentiated, II = moderately differentiated, III = poorly differentiated, IV = undifferentiated, unknown), T stage (derived AJCC, 7^th) (T1mic, T1a, T1b, T1c, T2, T3, unknown; mic, microinvasive carcinoma); and (II) the outcome variable is determined based on the information provided in the “Regional nodes positive (1988+)” column, where values greater than 1 are considered to be positive for lymph node involvement.

Statistical analysis

This study employed a retrospective cross-sectional survey for statistical analysis. Patient and tumor characteristics were presented as percentages, and data were analyzed using chi-squared tests or Fisher’s exact tests. A P value less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 25.0. In order to visualize the differences more intuitively, relevant analysis results were transformed into stacked bar charts using Microsoft Office Excel 2019. We conducted univariate and multivariate logistic regression analyses using SPSS version 25.0. Furthermore, R software (version 4.2.1) was utilized to perform propensity score reweighting analysis on selected data based on individual independent factors, followed by univariate binary logistic regression analysis. Nomograms based on regression models, calibration curves were generated using various functional packages such as RMS, Foreign, Cmprsk, and other software. A two-tailed P value less than 0.05 was considered statistically significant (*, P<0.05; **, P<0.01). We validated the developed predictive model using internally generated validation data through a 3:1 random allocation. The sensitivity and specificity of the constructed nomogram were reflected using receiver operating characteristic (ROC) curves, while its accuracy was evaluated and validated using calibration plots. Additionally, decision curve analysis (DCA) was utilized to calculate the net benefits for each risk threshold probability.

Results

Baseline characteristics of enrolled patients

Figure 2 presents the overall cohort, which included a total of 232,730 eligible BC patients. Among them, 25.57% (n=59,512) had positive lymph nodes (PLNs), and 74.43% (n=173,218) had negative lymph nodes (NLNs). Additional demographic and clinical characteristics of the recruited BC patients were also provided. Significant differences were observed in age, race, primary site, histology, breast subtype, grade, T stage (derived AJCC, 7^th), and other factors. In the age group, the 56+ subgroup had the highest number of patients (n=155,765). Among the three age groups, the 20–44 subgroup had the highest percentage of PLN (35.13%), indicating that PLN is more likely to occur in younger women. In the race group, the largest number of patients were white people (n=187,111), while the highest percentage of PLN was observed in black patients (29.94%). In the analysis and comparison within the group of primary sites, excluding “others”, the upper-outer quadrant had the highest number of patients (n=80,331), and the central portion & nipple had the highest percentage of PLN (34.47%). In the group of histology, infiltrating duct carcinoma had the highest number of patients (n=173,008), while lobular carcinoma had the highest percentage of PLN (29.72%). Among the group of breast subtype, the HR⁺/HER2⁻ subgroup had the highest number of patients, but the highest percentage of PLN was observed in HR⁻/HER2⁺ BC patients (31.16%). In the grade group, grade II had the highest number of patients (n=27,702), while grade III had the largest percentage of PLN (33.83%). In the group of T stage, the largest number of patients belonged to the T1c category (n=88,707), while the highest percentage of PLN was observed in the T3 category. Consistent with our expectations, larger tumors were associated with a higher likelihood of lymph node metastasis. Additionally, the baseline characteristics of the training and validation cohorts are provided in Tables S1-S7.

Figure 2 Clinicopathological features of BC patients diagnosed with and without lymph node metastasis (chi-square test; P<0.001 for each independent factor). HR, hormone receptor; HER2, human epidermal growth factor receptor 2; AJCC T, the T staging of Cancer Staging Manual of the American Joint Committee on Cancer; mic, microinvasive carcinoma; BC, breast cancer.

Logistic regression analysis and propensity score validation of lymph node occurrence in BC patients

Single-factor and multiple-factor logistic regression analyses were conducted to assess the independent risk factors for lymph node occurrence (PLN) in newly diagnosed BC patients. The results showed that the diagnosis age, race, primary site, histology, breast subtype, grade, and T stage (derived AJCC, 7^th) was significantly associated with PLN occurrence (Table 1).

In the age group, the 45–55 subgroup [odds ratio (OR) =0.86; 95% confidence interval (CI): 0.83–0.892, P<0.001] and the 56+ subgroup (OR =0.652; 95% CI: 0.631–0.674, P<0.001) had significantly lower PLN occurrence rates compared to the 20–44 subgroup. In terms of race, Asian or Pacific Islander had a significantly lower risk of PLN occurrence (OR =0.864; 95% CI: 0.834–0.896; P<0.001) compared to black people (OR =1.152; 95% CI: 1.114–1.191; P<0.001). In the group of primary sites, there was no significant difference in PLN occurrence between the upper-outer quadrant and lower-outer quadrant (OR =1; 95% CI: 0.962–1.041; P=0.993). The upper-inner quadrant (OR =0.567; 95% CI: 0.547–0.588; P<0.001) and lower-inner quadrant (OR =0.754; 95% CI: 0.718–0.791; P<0.001) had a lower PLN risk compared to the upper-outer quadrant, while the central portion & nipple had the highest PLN risk (OR =1.387; 95% CI: 1.324–1.452; P<0.001). In terms of histology, mucinous adenocarcinoma had a significantly lower risk of PLN (OR =0.21; 95% CI: 0.187–0.236; P<0.001) compared to infiltrating duct carcinoma. Infiltrating lobular carcinoma showed inconsistent results in single-factor analysis (OR =1.234; 95% CI: 1.197–1.272; P<0.001) and multiple-factor analysis (OR =0.907; 95% CI: 0.876–0.939; P<0.001). Further evaluation using PSM would be considered. Triple-negative BC (TNBC) (OR =0.585; 95% CI: 0.563–0.608; P<0.001) had a lower risk of PLN compared to HR⁺/HER2⁻ patients. HR⁺/HER2⁺ (OR =1.018; 95% CI: 0.982–1.055; P=0.329) and HR⁻/HER2⁺ patients (OR =1.017; 95% CI: 0.961–1.076; P=0.572) did not show statistically significant differences in PLN risk compared to HR⁺/HER2⁻ patients. In terms of grade, grade II (OR =1.466; 95% CI: 1.425–1.509; P<0.001), grade III (OR =1.648; 95% CI: 1.594–1.704; P<0.001), and grade IV (OR =1.478; 95% CI: 1.213–1.801; P<0.001) were all identified as risk factors for PLN occurrence in BC patients. Furthermore, in the T stage (derived AJCC, 7^th), the risk of PLN increased with the size of the tumor, with T3 having the highest risk (OR =34.704; 95% CI: 30.106–40.004; P<0.001). This suggests a correlation between tumor size and the likelihood of lymph node involvement, possibly indicating the progression of the disease.

After PSM, a certain number of patients were successfully matched in each independent factor. The baseline characteristics between the two groups, including diagnostic age (n=113,000), race (n=112,160), primary site (n=113,164), histology (n=112,268), breast subtype (n=113,496), grade (n=112,530), and T stage (derived AJCC, 7^th) (n=118,970), achieved good balance [standardized mean difference (SMD) <0.2, see Tables 1-3 and Tables S1-S7]. In the PSM dataset, the results of the single-factor logistic regression analysis were consistent with the previous results of the multivariable regression analysis. This indicates that the analysis results have a strong level of credibility.

The establishment of the BC lymph node metastasis prediction model

In the original cohort of 232,730 patients, we allocated 174,548 to the training set (75.0%) and 58,182 patients to the validation set (25.0%) (Tables 2,3). The flowchart of participant inclusion and exclusion is shown in Figure 1. We established a nomogram to visually display the score distribution and the predicted probabilities of risk factors (Figure 3). How do we utilize this model? For example, a 46-year-old (score: 7.5) African American (score: 8) female with BC, presenting with a tumor located in the lower-inner quadrant (score: 7.5), histologically diagnosed as invasive ductal carcinoma (score: 46), molecular subtype HR⁻/HER2⁺ (score: 16), unknown grade (score: 5), and T stage classified as T1c (score: 50). The total score is 7.5+8+7.5+46+16+5+50=140, corresponding to an estimated probability of lymph node metastasis of approximately 25%. In clinical practice, it is easy for us to estimate the probability of lymph node metastasis in patients using this nomogram. In order to better utilize this nomogram, we also employed the following objective evaluation methods to analyze the sensitivity, specificity, and accuracy of the model. The calibration curves with similar area under the curve (AUC) values demonstrate the good predictability of our nomogram model (Figure 4A,4B). DCA suggests that the threshold probability of 0–0.6 is the most favorable predictive factor for lymph node metastasis (Figure 4C,4D). The calibration curve with a concordance index (C-index) of 0.749 (95% CI: 0.747–0.751) indicates a strong consistency between the observed values and the predicted probabilities (Figure 4E,4F).

Figure 3 Seven independent factors, including age, race, primary site, histology, breast subtype, grade, derived AJCC T, 7^th were included in the nomogram. HR, hormone receptor; HER2, human epidermal growth factor receptor 2; AJCC T, the T staging of Cancer Staging Manual of the American Joint Committee on Cancer; mic, microinvasive carcinoma.

Figure 4 Nomogram of lymph node metastasis risk prediction in BC patients who did not receive any treatment before surgery. (A) The ROC curve of training set. (B) The ROC curve of validation data. (C) The DCA curve of training set. (D) The DCA curve of validation data. (E) The calibration curve of training cohort. (F) The calibration curve of validation data. ROC, receiver operating characteristic; AUC, area under the curve; DCA, decision curve analysis; BC, breast cancer.

Discussion

The topic of lymph node metastasis in BC has been extensively researched in the past. Previous studies by Van Zee et al. constructed a nomogram for predicting the likelihood of additional nodal metastases in BC patients with a positive sentinel node biopsy, but it had limitations as it only focused on preoperative biopsy-positive patients (10). Bevilacqua et al. included 3,786 cases of invasive BC between 1996 and 2002, who did not receive NAT, to develop an effective nomogram for predicting the probability of positive sentinel lymph nodes. However, the included clinical features in the nomogram were not sufficiently detailed, for example, the location was categorized only as upper-inner quadrant and other. Additionally, the predictive factors of the nomogram did not include molecular subtypes, which resulted in a lack of precision (2). Li et al. constructed a nomogram for lymph node metastasis in T1–2 and non-metastatic (M0) BC patients using the SEER database, but their model did not exclude patients with multiple tumors or those who did not undergo lymph node dissection (11). Gao et al. (8) developed nomogram models for stratified prediction of axillary lymph node metastasis in cN0 BC patients using SEER data from 2010 to 2015. However, it’s important to note an anomaly in the SEER database during the years 2010 to 2015—it lacks the capability to selectively identify cN0 BC patients. The only columns where cN0 stage patients can be reliably identified are “Derived SEER Combined N [2016–2017]” and “Derived SEER Combined N Src [2016–2017]”. In contrast, other columns can solely screen BC patients who have been clinically or pathologically determined as N0 stage. Consequently, the screening method provided in this study lacks the necessary specificity, which may impact the accuracy of the results. In our study, we specifically excluded patients who received NAT or radiation therapy before surgery and excluded patients with M1, T0, Tis, and T4 stages to accurately study the impact of independent factors on lymph node metastasis in female BC. Since traditional level I and II axillary lymph node dissection (ALND) requires at least 10 lymph nodes for pathological evaluation (12), we performed statistical analysis on the number of lymph node biopsies greater than 5, greater than 10, and greater than 12 before establishing the model (Figure 5). We found that the lymph node positivity rate in the downloaded data was higher than our expectations, so we reviewed the literature on lymph node positivity rates in different T stages in the past. In previous literature surveys, Iwasaki et al. conducted a retrospective study on 823 T1N0M0 invasive BC patients, with a lymph node metastasis rate of 25% (208/823) for T1N0M0 invasive BC (13). Tan et al. conducted a retrospective study on 380 early-stage BC patients, with lymph node positivity rates of 4.3% for T1a, 18.9% for T1b, and 27.6% for T1c (14).

Figure 5 Statistical description of LN status at regional nodes examined >5, >10, and >12. LN, lymph node; AJCC T, the T staging of Cancer Staging Manual of the American Joint Committee on Cancer; mic, microinvasive carcinoma.

In our analysis, we found that the lymph node positivity rates in T1mic BC patients who underwent lymph node dissection were significantly higher than those reported in previous literature. The positivity rate was 20.56% for more than five lymph nodes, 41.71% for more than 10 lymph nodes, and as high as 50.00% for more than 12 lymph nodes. This discrepancy suggests that there may be some bias in the data entry process. One possible explanation is that many patients who undergo lymph node dissection are suspected of having lymph node positivity, prompting physicians to consciously perform the procedure. Based on previous clinical experience data, we excluded patients who did not undergo lymph node dissection and established a more reasonable predictive model.

Our study is the first to utilize the SEER database to construct a nomogram for lymph node metastasis in BC patients who have not received any treatment. The nomogram was well validated. With the risk prediction model we established, we can promptly and scientifically assess and provide personalized treatment for high-risk patients, which has significant implications for their treatment outcomes.

In our present study, we initially conducted comparisons among various subgroups based on clinical and pathological characteristics through univariate and multivariate regression analyses. Following PSM, we noted that there were no substantial alterations in the distinctions between these subgroups when compared to the reference group. This implies that there is no significant interaction among the various clinical-pathological characteristics, and some hints regarding the relationships between individual subgroups can be gleaned from the results of univariate, multivariate, and PSM analyses. Therefore, it is not difficult to conclude that the risk factors for regional lymph node metastasis include young age (<45 years old), black ethnicity, involvement of the central portion and nipple, infiltrating duct carcinoma, HER2 positivity, grade III, and large tumor size. On the other hand, Asian or Pacific Islander ethnicity, upper-inner and lower-inner quadrants, mucinous adenocarcinoma, triple-negative subtype, low grade, and small tumor size is considered protective factors against lymph node metastasis.

BC accounts for a relatively small proportion in young women (Figure 2), approximately 9.4%. However, the positive lymph node ratio (PLNR) is highest among young women and tends to decrease with increasing age. The definition of young women varies among articles (15-17), therefore, this study used the range of perimenopausal women (45–55 years) as the dividing line (18), classifying women into young, perimenopausal, and elderly groups to investigate the impact of age on lymph node metastasis. It has been reported that young women are associated with poor prognosis in BC (19). Some retrospective studies also suggest a higher risk of recurrence after breast-conserving surgery in young women compared to older women (16,20). Patients younger than 40 years old more commonly exhibit lymphocytic stromal reaction, histologic grade 3, and extensive ductal carcinoma in situ (DCIS) (21), which is consistent with the findings of this data analysis, indicating that young women with BC have more aggressive biological behavior compared to older women.

Among the racial categories, Black women are more likely to experience lymph node metastasis. Even after adjusting for independent factors such as age, tumor location, histology, BC subtype, grade, and T stage (Table 1), Black women still show a statistically significant difference in lymph node positivity compared to White women, with an OR of 1.09 times higher for Black women (95% CI: 1.049–1.133, P<0.001). This conclusion has been widely validated in previous studies (7,22,23).

Regarding the primary tumor location, we found that the highest risk of lymph node metastasis is in the central portion and nipple, followed by the outer-upper and outer-lower quadrants (Table 1). There were no statistically significant differences in lymph node metastasis risk between the outer-upper and outer-lower quadrants in both univariate and multivariate logistic regression analyses and PSM, which aligns with previous research (24).

In terms of molecular subtypes, our analysis through multivariate regression analysis, PSM, and column line chart analysis all showed no significant difference in lymph node metastasis risk among HR⁺/HER2⁻, HR⁺/HER2⁺, and HR⁻/HER2⁺ subtypes (P=0.089; P=0.066). Interestingly, the lymph node metastasis rate of TNBC is lower than that of hormone receptor-positive, HER2⁻ patients. Although this observation may seem counterintuitive. It has also been reflected in previous studies. In a retrospective study by Si et al., luminal HER2⁺ was associated with the highest lymph node positivity (49.0%), followed by luminal HER2⁻ (46.8%), HER2⁺ (44.4%), luminal A (36.5%), and TNBC (34.7%). The occurrence of LN metastases was lowest in the TNBC subtype (25).

It is possible that the HR⁺/HER2⁻ subtype of BC tumors may have a higher proportion of luminal B subtype. In a retrospective study by Xiong et al., the LNM rates were 72.0% for luminal B HER2⁺, 51.9% for luminal B, 50.0% for TNBC, and 37.4% for luminal A, with statistically significant differences observed (26). Similarly, in a retrospective study by Cheng et al. investigating the relationship between molecular subtypes and clinicopathological features of BC in Chinese women, the basal-like subtype had a significantly lower risk of lymph node metastasis compared to luminal A, luminal B, and HER2⁺ subtypes (27). Min et al. also conducted a retrospective analysis of clinical and pathological features of 16,552 patients who underwent breast surgery at Samsung Medical Center from 2000 to 2015. The results showed a higher incidence of lymph node metastasis in the luminal subtype compared to HER2 and TNBC, with the lowest lymph node metastasis rate observed in the TNBC subtype (28). These clinical studies have reported a lower lymph node metastasis risk in TNBC.

In fact, it is well recognized that TNBC is prone to recurrence and distant metastasis, resulting in a relatively poorer prognosis compared with other BC subtypes (29). It is widely accepted in the field that lymph node metastasis is a precursor event to distant metastasis. Based on this premise, most researchers have traditionally believed that TNBC has a higher likelihood of lymph node involvement. The conclusion drawn from our statistical analysis seems contradictory to previous experience. However, we believe that although TNBC has the lowest risk of lymph node metastasis, this does not exclude its “aggressiveness”. Under the same treatment, TNBC has a worse prognosis (29), and it is often more likely to develop visceral metastases, including lung, liver, and brain metastases (30). Indeed, there is a substantial body of research that supports the notion that BC can progress to distant metastasis without necessarily involving axillary lymph node metastasis as an intermediate step (31). This study suggests that the rate of axillary lymph node metastasis in TNBC is relatively lower. However, it is important to note that this result should not be construed as evidence for a better prognosis among TNBC patients.

In the analysis of T-stage and lymph node positivity, it is easy to observe that the risk of lymph node metastasis increases with tumor size, as also demonstrated in the study by Wu et al. (32). However, there are clinical examples of small primary tumors with extensive lymph node metastasis (33) or distant metastasis (34), indicating that a small primary tumor does not always indicate a better prognosis. Although lymph node metastasis is not as immediately life-threatening as lung, liver, or brain metastasis, it still requires our attention. In our analysis, we observed that the incidence of lymph node metastasis in T1c stage BC patients was 22.64%, while it increased to 43.4% in T2 stage patients. Our nomogram further indicates that T1c stage BC receives a higher score, approximately 50 points, compared to the highest scores observed in other subgroups. This suggests that patients in T1c and T2 stages, particularly those facing delayed surgery, may consider NAT as a potential strategy to reduce the risk of lymph node metastasis. However, it’s important to note that the decision for adjuvant therapy should be based on additional clinical research findings to ensure its appropriateness.

There are several limitations of our study. The database does not include additional information such as imaging features, Ki-67, vascular embolism, vascular invasion, nerve invasion, etc. (35,36). Including these biomarkers would enhance the accuracy of the developed model. Additionally, The SEER database lacks detailed information on NAT regimens, and this study did not provide a detailed comparison. Further clinical research is needed to validate the efficacy of different NAT regimens for different tumor types.

Conclusions

This retrospective study revealed the lymph node positivity status of female BC patients with distinct clinicopathological characteristics who did not receive any treatment preoperatively. Valuable nomograms were constructed based on these patient populations. The study results demonstrate strong reliability through various calibration and discrimination statistical methods, making them potential tools for guiding clinical diagnosis and individualized treatment.

Risk factors for regional lymph node metastasis include young age (<35 years), Black race, central portion & nipple involvement, infiltrating duct carcinoma, HER2 positivity, grade III, and large tumor size.

Although TNBC has the lowest risk of lymph node metastasis, this does not exclude the “aggressiveness” of TNBC, as it still has a poorer prognosis under the same treatment.

Patients in T1c and T2 stages, particularly those with delayed surgery, may consider NAT to mitigate the potential threat of lymph node metastasis.

Acknowledgments

Funding: The work was supported by the National Natural Science Foundation of China (Nos. 82272855, 81972453, and 81972597), the Natural Science Foundation of Zhejiang Province (Nos. LR22H160011, LY19H160055, LY19H160059, LY18H160005, and LY20H160026), and the Medical and Health Science and Technology of Zhejiang Province (Youth Talent Program) Project (No. 2021RC016). The work was sponsored by the Zheng Shu Medical Elite Scholarship Fund.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-23-258/rc

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-23-258/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-23-258/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 (as revised in 2013).

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, Fuchs HE, et al. Cancer statistics, 2022. CA Cancer J Clin 2022;72:7-33. [Crossref] [PubMed]
2. Bevilacqua JL, Kattan MW, Fey JV, et al. Doctor, what are my chances of having a positive sentinel node? A validated nomogram for risk estimation. J Clin Oncol 2007;25:3670-9. [Crossref] [PubMed]
3. Nottegar A, Veronese N, Senthil M, et al. Extra-nodal extension of sentinel lymph node metastasis is a marker of poor prognosis in breast cancer patients: A systematic review and an exploratory meta-analysis. Eur J Surg Oncol 2016;42:919-25. [Crossref] [PubMed]
4. Asioli S, Eusebi V, Gaetano L, et al. The pre-lymphatic pathway, the rooths of the lymphatic system in breast tissue: a 3D study. Virchows Arch 2008;453:401-6. [Crossref] [PubMed]
5. Voogd AC, Coebergh JW, Repelaer van Driel OJ, et al. The risk of nodal metastases in breast cancer patients with clinically negative lymph nodes: a population-based analysis. Breast Cancer Res Treat 2000;62:63-9. [Crossref] [PubMed]
6. Chang JM, Leung JWT, Moy L, et al. Axillary Nodal Evaluation in Breast Cancer: State of the Art. Radiology 2020;295:500-15. [Crossref] [PubMed]
7. Zhao YX, Liu YR, Xie S, et al. A Nomogram Predicting Lymph Node Metastasis in T1 Breast Cancer based on the Surveillance, Epidemiology, and End Results Program. J Cancer 2019;10:2443-9. [Crossref] [PubMed]
8. Gao X, Luo W, He L, et al. Nomogram models for stratified prediction of axillary lymph node metastasis in breast cancer patients (cN0). Front Endocrinol (Lausanne) 2022;13:967062. [Crossref] [PubMed]
9. Park AR, Chae EY, Cha JH, et al. Preoperative Breast MRI in Women 35 Years of Age and Younger with Breast Cancer: Benefits in Surgical Outcomes by Using Propensity Score Analysis. Radiology 2021;300:39-45. [Crossref] [PubMed]
10. Van Zee KJ, Manasseh DM, Bevilacqua JL, et al. A nomogram for predicting the likelihood of additional nodal metastases in breast cancer patients with a positive sentinel node biopsy. Ann Surg Oncol 2003;10:1140-51. [Crossref] [PubMed]
11. Li H, Tang L, Chen Y, et al. Development and validation of a nomogram for prediction of lymph node metastasis in early-stage breast cancer. Gland Surg 2021;10:901-13. [Crossref] [PubMed]
12. Axelsson CK, Mouridsen HT, Zedeler K. Axillary dissection of level I and II lymph nodes is important in breast cancer classification. The Danish Breast Cancer Cooperative Group (DBCG). Eur J Cancer 1992;28A:1415-8. [Crossref] [PubMed]
13. Iwasaki Y, Fukutomi T, Akashi-Tanaka S, et al. Axillary node metastasis from T1N0M0 breast cancer: possible avoidance of dissection in a subgroup. Jpn J Clin Oncol 1998;28:601-3. [Crossref] [PubMed]
14. Tan LG, Tan YY, Heng D, et al. Predictors of axillary lymph node metastases in women with early breast cancer in Singapore. Singapore Med J 2005;46:693-7.
15. Rossi L, Mazzara C, Pagani O. Diagnosis and Treatment of Breast Cancer in Young Women. Curr Treat Options Oncol 2019;20:86. [Crossref] [PubMed]
16. Collins LC, Marotti JD, Gelber S, et al. Pathologic features and molecular phenotype by patient age in a large cohort of young women with breast cancer. Breast Cancer Res Treat 2012;131:1061-6. [Crossref] [PubMed]
17. Gnerlich JL, Deshpande AD, Jeffe DB, et al. Elevated breast cancer mortality in women younger than age 40 years compared with older women is attributed to poorer survival in early-stage disease. J Am Coll Surg 2009;208:341-7. [Crossref] [PubMed]
18. Takahashi TA, Johnson KM. Menopause. Med Clin North Am 2015;99:521-34. [Crossref] [PubMed]
19. Kroman N, Jensen MB, Wohlfahrt J, et al. Factors influencing the effect of age on prognosis in breast cancer: population based study. BMJ 2000;320:474-8. [Crossref] [PubMed]
20. Arvold ND, Taghian AG, Niemierko A, et al. Age, breast cancer subtype approximation, and local recurrence after breast-conserving therapy. J Clin Oncol 2011;29:3885-91. [Crossref] [PubMed]
21. Kurtz JM, Jacquemier J, Amalric R, et al. Why are local recurrences after breast-conserving therapy more frequent in younger patients? J Clin Oncol 1990;8:591-8. [Crossref] [PubMed]
22. Iqbal J, Ginsburg O, Rochon PA, et al. Differences in breast cancer stage at diagnosis and cancer-specific survival by race and ethnicity in the United States. JAMA 2015;313:165-73. [Crossref] [PubMed]
23. McBride R, Hershman D, Tsai WY, et al. Within-stage racial differences in tumor size and number of positive lymph nodes in women with breast cancer. Cancer 2007;110:1201-8. [Crossref] [PubMed]
24. Rinaldi RM, Sapra A, Bellin LS. Breast Lymphatics. In: StatPearls. Treasure Island: StatPearls Publishing; 2023.
25. Si C, Jin Y, Wang H, et al. Association between molecular subtypes and lymph node status in invasive breast cancer. Int J Clin Exp Pathol 2014;7:6800-6.
26. Xiong J, Zuo W, Wu Y, et al. Ultrasonography and clinicopathological features of breast cancer in predicting axillary lymph node metastases. BMC Cancer 2022;22:1155. [Crossref] [PubMed]
27. Cheng HT, Huang T, Wang W, et al. Clinicopathological features of breast cancer with different molecular subtypes in Chinese women. J Huazhong Univ Sci Technolog Med Sci 2013;33:117-21. [Crossref] [PubMed]
28. Min SK, Lee SK, Woo J, et al. Relation Between Tumor Size and Lymph Node Metastasis According to Subtypes of Breast Cancer. J Breast Cancer 2021;24:75-84. [Crossref] [PubMed]
29. Bianchini G, De Angelis C, Licata L, Gianni L. Treatment landscape of triple-negative breast cancer - expanded options, evolving needs. Nat Rev Clin Oncol 2022;19:91-113. [Crossref] [PubMed]
30. Rakha EA, Chan S. Metastatic triple-negative breast cancer. Clin Oncol 2011;23:587-600. [Crossref] [PubMed]
31. Venet D, Fimereli D, Rothé F, et al. Phylogenetic reconstruction of breast cancer reveals two routes of metastatic dissemination associated with distinct clinical outcome. EBioMedicine 2020;56:102793. [Crossref] [PubMed]
32. Wu SL, Gai JD, Yu XM, et al. A novel nomogram and risk classification system for predicting lymph node metastasis of breast mucinous carcinoma: A SEER-based study. Cancer Med 2022;11:4767-83. [Crossref] [PubMed]
33. Wo JY, Chen K, Neville BA, et al. Effect of very small tumor size on cancer-specific mortality in node-positive breast cancer. J Clin Oncol 2011;29:2619-27. [Crossref] [PubMed]
34. Zheng YZ, Wang XM, Fan L, et al. Breast Cancer-Specific Mortality in Small-Sized Tumor with Stage IV Breast Cancer: A Population-Based Study. Oncologist 2021;26:e241-50. [Crossref] [PubMed]
35. Thangarajah F, Malter W, Hamacher S, et al. Predictors of sentinel lymph node metastases in breast cancer-radioactivity and Ki-67. Breast 2016;30:87-91. [Crossref] [PubMed]
36. Chen H, Meng X, Hao X, et al. Correlation Analysis of Pathological Features and Axillary Lymph Node Metastasis in Patients with Invasive Breast Cancer. J Immunol Res 2022;2022:7150304. [Crossref] [PubMed]