A network dynamic nomogram for predicting overall survival and cancer-specific survival in patients with breast cancer liver metastases: an analysis based on the SEER database

Abstract

The liver stands out as one of the most frequent sites for distant metastasis in breast cancer cases. However, effective risk stratification tools for patients with breast cancer liver metastases (BCLM) are still lacking. We identified BCLM patients from the SEER database spanning from 2010 to 2016. After meticulously filtering out cases with incomplete data, a total of 3179 patients were enrolled and randomly divided into training and validation cohorts at a ratio of 2:1. Leveraging comprehensive patient data, we constructed a nomogram through rigorous evaluation of a Cox regression model. Validation of the nomogram was conducted using a range of statistical measures, including the concordance index (C-index), calibration curves, time-dependent receiver operating characteristic curves, and decision curve analysis (DCA). Both univariable and multivariable analyses revealed significant associations between OS and CSS in BCLM patients and 14 variables, including age, race, and tumor stage, among others. Utilizing these pertinent variables, we formulated nomograms for OS and CSS prediction. Subsequent validation involved rigorous assessment using time-dependent ROC curves, decision curve analysis, C-index evaluations, and calibration curves. Our web-based dynamic nomogram represents a valuable tool for efficiently analyzing the clinical profiles of BCLM patients, thereby aiding in informed clinical decision-making processes.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-024-01719-1.

Introduction

Cancer stands as one of the foremost medical challenges globally and ranks as the second leading cause of death in the United States. Among women, breast cancer comprises a significant portion, representing 31% of new cancer cases and ranking as the foremost cancer type, while also claiming the second spot in mortality at 15% [1]. Alarmingly, 90% of cancer-related deaths in breast cancer stem from metastases originating in the primary tumor. Reports indicate that the 5-year overall survival (OS) rate for patients without distant metastasis can exceed 80%, contrasting sharply with the meager 20% 5-year OS rate post-metastasis [2, 3]. This underscores the critical impact of distant metastasis on the prognosis of breast cancer patients, with approximately 10% experiencing distant metastasis during primary tumor diagnosis and treatment, with notable organs such as bone, lung, and liver being primary sites of metastasis [4].

Among these organs, the liver ranks as the third most common site of metastasis in breast cancer cases. Despite its lower incidence compared to bone and lung metastasis, the 5-year survival rate for patients with breast cancer liver metastasis (BCLM) plummets to a mere 8.5%, significantly lower than those with bone or lung metastases [5]. Current treatment modalities for BCLM encompass chemotherapy, endocrine therapy, immunotherapy, target therapies, surgery and radiotherapy [6]. However, the efficacy of these treatments for BCLM patients remains suboptimal due to factors such as chemotherapy and endocrine therapy resistance. Hence, the development of a precise BCLM risk assessment system holds paramount importance in guiding clinical decision-making processes.

Nomograms serve as invaluable tools in assessing tumor patient prognosis [7]. Overcoming the limitations of tumor lymph node metastasis (TNM) staging, nomograms provide a comprehensive numerical probability of individual prognosis by integrating diverse prognostic variables, aligning with the personalized medicine paradigm emphasized in contemporary healthcare [8]. By inputting variables, clinicians can swiftly and accurately calculate patient prognoses based on the cumulative score.

In this study, we have developed two efficient and precise web-based dynamic nomograms tailored for assessing the prognosis of BCLM patients. These nomograms respectively evaluate overall survival (OS) and cancer-specific survival (CSS) in BCLM patients, offering clinicians valuable insights to make informed clinical decisions tailored to individual patient profiles.

Methods

Patient selection, nomogram construction and validation

We utilized the SEER*Stat 8.3.8 program to retrieve and download patient information from the SEER 18 database. This publicly accessible database contains comprehensive cancer incidence data and detailed patient information across the United States.

Our study encompassed adult breast cancer patients diagnosed with liver metastases between 2010 and 2016. The collected patient data included race, age, marital status, detailed TNM stage, history of radical resection at the primary site, receipt of radiation and chemotherapy, ER, PR, and HER2 receptor status in postoperative pathology, as well as the presence of bone, brain, or lung metastases. Patients lacking complete pathological and survival information were excluded. Ultimately, 3197 patients were enrolled in the study, with random allocation into a training cohort (2131 cases) and a verification cohort (1066 cases) at a ratio of 2:1.

We employed univariable Cox proportional hazards regression analysis to identify independent prognostic factors, upon which prognostic factor models were constructed. Subsequently, based on the results of univariable analysis, we conducted multivariable Cox proportional risk regression analysis to formulate a nomogram incorporating significant variables from the training cohort. The nomogram was designed to analyze 1-year, 3-year, and 5-year overall survival (OS) and cancer-specific survival (CSS) rates. Calibration curves, C-index, and decision curve analyses (DCA) were utilized to assess the predictive performance of the nomograms. Furthermore, time-dependent ROC analysis was employed to evaluate the efficacy of the OS and CSS models developed in our study (Fig. 1).

Fig. 1.

Flow charts of selected patients in this study. We included a total of 4974 BCLM patients from the SEER database and excluded 1777 patients according to the exclusion criteria described earlier. For the remaining 3197 patients who met the requirements, we randomly divided them into the training cohort and the verification cohort on a 2:1 basis. The training queue is used to construct the Norman diagram, and the effectiveness of the validation queue is used to evaluate the architectural model. We validated the constructed Norman plots using calibration curves, C-index decision curve analysis (DAC), and time-varying ROC methods

Statistical analysis

All of these statistical analyses were conducted using R version 4.2.0. Univariable and multivariable Cox regression analyses were employed to assess the independent factors influencing overall survival (OS) and cancer-specific survival (CSS) in BCLM patients. Statistically significant differences were determined with a threshold of P < 0.1 in the univariable analysis and P < 0.05 in the multivariable analysis.

Web-based dynamic nomogram publication

Following validation, the web-based dynamic nomograms for BCLM OS and CSS were published online via the R package “DynNom” and hosted on shinyapps.io (https://www.shinyapps.io/) [9], accessed on 9 February 2024.

Ethical consideration

This is an observational study, therefore it does not necessitate review by the Xiangya Hospital Ethics Committee.

Result

Basic patient information

According to the process outlined in Fig. 1 above, we excluded patients lacking detailed pathological diagnosis, immunohistochemical results, and survival information, resulting in a total of 3197 patients being included in our study. The basic information of all patients is presented in Table 1. In summary, 41% of the patients in the study were aged between 51 and 65, 99.6% were female, 53.6% were unmarried, and 73.9% were white. Regarding pathological information, 100% of patients were diagnosed with de novo stage IV, 34.4% had T4 tumors, and 78.7% had lymph node metastasis. Concerning treatment, while 74.1% of patients received chemotherapy, the proportion of patients who underwent surgery was significantly lower, at only 32.1%. Regarding organ metastasis, bone remained the most common site for breast cancer patients with liver metastasis, affecting 56.6% of patients, followed by the lung at 32.7%. The percentages of patients testing positive for ER, PR, and HER-2 were 65.3%, 50.4%, and 40.9%, respectively. The mean OS and CSS for patients included in the study were 32.9 months and 35.1 months, respectively.

Table 1.

Basic information for all patients included in this study

Variables	Total Cohort (n = 3197)		Training cohort (n = 2131)		Validation cohort(n = 1066)
	No	%	No	%	No	%
Age
≤ 50	982	30.7	640	30	342	32.1
51–65	1313	41	907	42.6	406	38.1
> 65	902	28.3	584	27.4	318	29.8
Gender
Female	3183	99.6	2122	99.6	1061	99.5
Male	14	0.4	9	0.4	5	0.5
Marital status
Married	1482	46.4	977	45.8	505	47.4
Unmarried	1715	53.6	1154	54.2	561	52.6
Race
White	2361	73.9	1574	73.9	787	73.8
Black	570	17.8	379	17.8	191	17.9
Other	266	8.3	178	8.3	88	8.3
Grade
Grade I/II	1299	40.6	872	40.9	427	40
Grade III	1898	59.4	1259	59.1	639	60
T stage
T1	426	13.3	269	12.6	157	14.7
T2	1094	34.2	721	33.8	373	35
T3	577	18.1	408	19.2	169	15.9
T4	1100	34.4	733	34.4	367	34.4
N stage
N0	680	21.3	459	21.5	221	20.7
N+	2517	78.7	1672	78.5	845	79.3
Primary surgery
No	2170	67.9	1437	67.4	733	68.8
Yes	1027	32.1	694	32.6	333	31.2
Chemotherapy
No	828	25.9	560	26.3	268	25.1
Yes	2369	74.1	1571	73.7	798	74.9
Radiation
No	2335	73	1553	72.9	782	73.4
Yes	862	27	578	27.1	284	26.6
Bone metastatic
No	1386	43.4	932	43.7	454	42.6
Yes	1811	56.6	1199	56.3	612	57.4
Brain metastatic
No	2951	92.3	1969	92.4	982	92.1
Yes	246	7.7	162	7.6	84	7.9
Lung metastatic
No	2153	67.3	1432	67.2	721	67.6
Yes	1044	32.7	699	32.8	345	32.4
ER
Negative	1108	34.7	721	33.8	387	36.3
Positive	2089	65.3	1410	66.2	679	63.7
PR
Negative	1586	49.6	1055	49.5	531	49.8
Positive	1611	50.4	1076	50.5	535	50.2
HER2
Negative	1891	59.1	1264	59.3	627	58.8
Positive	1306	40.9	867	40.7	439	41.2
OS (months)	32.9		33.7		31.4
CSS (months)	35.1		35.9		33.7

ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2

Selection of variables in our nomogram

We conducted univariable and multivariable Cox regression analyses for all the variables listed in Table 1, determining their respective weights in OS and CSS. Initially, we performed a univariable analysis of the 16 variables, revealing that 14 of them displayed statistical significance. Subsequently, a multivariable analysis was conducted on these 14 variables to identify the final variables associated with OS and CSS, respectively (refer to Tables 2 and 3). The findings indicated that, with the exception of patient gender and whether radiotherapy was administered, all other variables demonstrated significant correlations with the OS and CSS of BCLM patients, thus warranting inclusion in the subsequent construction of the nomogram.

Table 2.

Univariable and multivariable Cox regression analysis of OS in BCLM patients

Characteristics	Univariable Analysis			Multivariable Analysis
	OR	95%CI	p	OR	95%CI	p
Age			< 0.001			< 0.001
< 50		Reference	1		Reference	1
51–65	1.273	1.106–1.466	0.001	1.173	1.016–1.354	0.03
> 65	2.071	1.785–2.403	< 0.001	1.7	1.446–1.998	< 0.001
Gender			0.903			NA
Female		Reference	1
Male	0.947	0.393–2.28	0.903
Marital status			< 0.001			< 0.001
Married		Reference	1		Reference	1
Unmarried	1.386	1.236–1.553	< 0.001	1.207	1.073–1.358	0.002
Race			< 0.001			0.001
White		Reference	1		Reference	1
Black	1.312	1.139–1.512	< 0.001	1.306	1.128–1.512	< 0.001
Other	0.887	0.71–1.109	0.293	0.953	0.761–1.193	0.672
Grade			< 0.001			< 0.001
I/II		Reference	1		Reference	1
III	1.263	1.125–1.419	< 0.001	1.441	1.272–1.634	< 0.001
T stage			< 0.001			0.019
T1		Reference	1		Reference	1
T2	1.005	0.832–1.215	0.955	1.087	0.898–1.317	0.391
T3	1.076	0.873–1.325	0.494	1.172	0.946–1.452	0.147
T4	1.375	1.142–1.657	0.001	1.305	1.074–1.586	0.007
N stage			0.023			0.001
N0		Reference	1		Reference	1
N+	0.857	0.751–0.979	0.023	0.821	0.714–0.944	0.006
Primary surgery			< 0.001			< 0.001
Yes		Reference	1		Reference	1
No	1.495	1.322–1.691	< 0.001	1.417	1.246–1.611	< 0.001
Chemotherapy			< 0.001			< 0.001
Yes		Reference	1		Reference	1
No	1.816	1.612–2.047	< 0.001	1.57	1.372–1.796	< 0.001
Radiation			0.441			NA
Yes		Reference	1
No	1.055	0.929–1.197	0.441
Bone metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	1.321	1.178–1.482	< 0.001	1.231	1.091–1.39	0.001
Brain metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	2.068	1.709–2.502	< 0.001	1.789	1.467–2.182	< 0.001
Lung metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	1.597	1.422–1.793	< 0.001	1.272	1.127–1.435	< 0.001
ER			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.43	1.273–1.607	< 0.001	1.457	1.243–1.707	< 0.001
PR			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.44	1.286–1.612	< 0.001	1.389	1.192–1.619	< 0.001
HER-2			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.773	1.571–2	< 0.001	1.931	1.695–2.199	< 0.001

ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2, OR odds ratio, CI confidence interval, NA not available

Table 3.

Univariable and multivariable Cox regression analysis of CSS in BCLM patients

Characteristics	Univariable analysis			Multivariable analysis
	OR	95%CI	p	OR	95%CI	p
Age			< 0.001			< 0.001
< 50		Reference	1		Reference	1
51–65	1.261	1.08–1.473	0.003	1.16	0.991–1.359	0.065
> 65	1.935	1.637–2.287	< 0.001	1.535	1.279–1.843	< 0.001
Gender			0.889			NA
Female		Reference	1
Male	0.933	0.349–2.49	0.889
Marital status			< 0.001			< 0.001
Married		Reference	1		Reference	1
Unmarried	1.379	1.212–1.569	< 0.001	1.211	1.06–1.384	0.005
Race			0.001			0.018
White		Reference	1		Reference	1
Black	1.274	1.087–1.493	0.003	1.236	1.049–1.457	0.011
Other	0.808	0.626–1.044	0.103	0.883	0.682–1.144	0.346
Grade			< 0.001			< 0.001
I/II		Reference	1		Reference	1
III	1.347	1.18–1.538	< 0.001	1.553	1.347–1.791	< 0.001
T stage			< 0.001			0.05
T1		Reference	1		Reference	1
T2	0.961	0.764–1.21	0.737	1.041	0.825–1.314	0.733
T3	1.144	0.896–1.46	0.281	1.213	0.944–1.559	0.131
T4	1.368	1.095–1.71	0.006	1.265	1.003–1.595	0.047
N stage			0.023			0.008
N0		Reference	1		Reference	1
N+	0.837	0.717–0.976	0.023	0.803	0.683–0.943	0.008
Primary surgery			< 0.001			< 0.001
Yes		Reference	1		Reference	1
No	1.573	1.336–1.768	< 0.001	1.466	1.266–1.696	< 0.001
Chemotherapy			< 0.001			< 0.001
Yes		Reference	1		Reference	1
No	1.837	1.603–2.106	< 0.001	1.617	1.386–1.886	< 0.001
Radiation			0.397			NA
Yes		Reference	1
No	1.063	0.923–1.225	0.397
Bone metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	1.429	1.253–1.628	< 0.001	1.326	1.155–1.523	< 0.001
Brain metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	2.099	1.691–2.606	< 0.001	1.679	1.341–2.102	< 0.001
Lung metastatic			< 0.001			< 0.001
No		Reference	1		Reference	1
Yes	1.666	1.463–1.897	< 0.001	1.328	1.16–1.519	< 0.001
ER			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.396	1.225–1.592	< 0.001	1.462	1.225–1.745	< 0.001
PR			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.453	1.279–1.651	< 0.001	1.438	1.212–1.707	< 0.001
HER-2			< 0.001			< 0.001
Positive		Reference	1		Reference	1
Negative	1.868	1.631–2.14	< 0.001	2.097	1.81–2.43	< 0.001

ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2, OR odds ratio, CI confidence interval, NA not available

Nomogram construction and validation

Following univariable and multivariable Cox regression analyses, we identified a total of 14 variables suitable for constructing the nomogram. These variables were found to significantly correlate with both OS and CSS. Among them are age, marital status, race, tumor grade, T stage, lymph node metastasis, surgery, chemotherapy, bone metastasis, lung metastasis, brain metastasis, ER status, PR status, and HER2 status. Utilizing these variables, we developed 1-year, 3-year, and 5-year nomograms for OS and CSS, respectively. These nomograms include individual scores and total scores, with the total score determined based on the individual scores computed using the nomogram. These scores enable the evaluation of patients’ 1-year, 3-year, and 5-year OS and CSS outcomes (refer to Figs. 2A and 3A).

Fig. 2.

Nomogram and model assessment for predicting 1-, 3-, and 5-year overall survival (OS) in BCLM patients. A The OS nomogram for BCLM patients. B The calibration plots in the training cohort for 1-year, 3-year, and 5-year OS. AUC values of ROCs of the nomograms for 1-year, 3-year, and 5-year OS. C The ROC curves of training cohort. D The decision curve analyses (DCA) of the nomogram and other factors’ overall survival for the training cohort. E The calibration plots in the verification cohort for 1-year, 3-year, and 5-year OS. F The ROC curves of validation cohort. G The decision curve analyses (DCA) of the nomogram and other factors’ overall survival for the validation cohort. AUC Area under Curve, ROC Receiver Operating Characteristic

Fig. 3.

Nomogram and model assessment for predicting 1-, 3-, and 5-year cancer-specific survival (CSS) in BCLM patients. A The CSS nomogram for BCLM patients. B The calibration plots in the training cohorts for 1-year, 3-year, and 5-year CSS. AUC values of ROCs of the nomograms for 1-year, 3-year, and 5-year CSS. C The ROC curves of training cohort. D The decision curve analyses (DCA) of the nomogram and other factors’ cancer-specific survival for the training cohort. E The calibration plots in the verification cohorts for 1-year, 3-year, and 5-year CSS. F The ROC curves of verification cohort. G The decision curve analyses (DCA) of the nomogram and other factors’ cancer-specific survival for the verification cohort. AUC Area under Curve, ROC Receiver Operating Characteristic

Following the construction of the nomogram, we assessed its accuracy using various methods, including the concordance index (C-index), calibration curves, time-dependent receiver operating characteristic [10] curves, and decision curve analysis (DCA). For OS, the C-index values were 0.709 (95% confidence interval (CI]) 0.693–0.725) in the training cohort and 0.726 (95% CI 0.705–0.747) in the verification cohort. For CSS, the C-index values were 0.710 (95% CI 0.693–0.728) in the training cohort and 0.721 (95% CI 0.697–0.745) in the verification cohort (refer to Table 4). The calibration curves of the nomogram (Figs. 2B, E, 3B, and E) demonstrate a high level of agreement between predicted and observed survival probabilities in both the training and verification cohorts for OS and CSS, respectively. Furthermore, the time-dependent area under the curve [11] values of OS and CSS predictions at 1, 3, and 5 years are all greater than 0.75 in both the training and verification cohorts (Figs. 2C, F, 3C, and F), indicating a favorable discriminatory ability of our developed nomogram.

Table 4.

The C-indices for predictions of OS and CSS

Group	OS		CSS
	C-Index	95%CI	C-Index	95%CI
Training group	0.709	0.693–0.725	0.710	0.693–0.728
Validation group	0.726	0.705–0.747	0.721	0.697–0.745

C-index index of concordance, CI confidence interval

DCA curves are now widely utilized to compare nomograms with other clinically relevant factors. Figures 2D, G, 3D, and G depict DCA curves for OS and CSS of the nomogram and other clinical factors in BCLM patients. Our findings reveal that, in comparison with other clinical factors, the DCA curve of our designed nomogram exhibits significant superiority, suggesting its enhanced suitability for clinical use in aiding physicians to make informed decisions.

Risk stratification and their subgroup survival analysis

We aggregated the scores for each variable of BCLM patients to derive the total score and subsequently determined the cutoff value using X-tile software (version 3.6.1; Yale University, New Haven, CT, USA) (refer to Fig. 4A and C). Based on the X-tile software analysis results, we categorized the OS and CSS of BCLM patients into three risk levels: high-risk, medium-risk, and low-risk. Specifically, for OS, the high-risk, medium-risk, and low-risk groups for BCLM patients correspond to scores > 450, 305–460, and < 305, respectively (see Fig. 4B). Similarly, for CSS, the high-risk, medium-risk, and low-risk categories for BCLM patients are defined as scores > 437, 289–437, and < 289, respectively (refer to Fig. 4D).

Fig. 4.

Risk stratification by X-tile software, the subgroup and their survival analysis. A The cut-off values of BCLM patients for OS. The black dot means the OS distinction point. B Risk stratification of OS in BCLM patients. The blue color is the low-risk group, the gray color is the medium-risk group, and the purple color is the high-risk group. C The cut-off values of BCLM patients for CSS. The black dot means the CSS distinction point. D Risk stratification of CSS in BCLM patients. The blue color is the low-risk group, the gray color is the medium-risk group, and the purple color is the high-risk group. E The Kaplan–Meier survival curves of three risk subgroups of BCLM patients for OS. F The Kaplan–Meier survival curves of three risk subgroups of BCLM patients for CSS

Utilizing this risk stratification, we generated Kaplan–Meier survival curves for BCLM patients across different risk levels. The 5-year OS rates were 35.0%, 11.3%, and 3.02% for high-risk, medium-risk, and low-risk BCLM patients, respectively (see Fig. 4E). Likewise, the 5-year CSS rates were 39.9%, 14.1%, and 1.6% for high-risk, medium-risk, and low-risk BCLM patients, respectively (refer to Fig. 4F). Notably, the differences among the three risk subgroups for both OS and CSS were statistically significant.

Web-based dynamic nomogram publication

We have officially released our web-based nomograms for predicting OS in BCLM patients (Fig. 5A) and CSS (Fig. 5B). These nomograms are accessible at the following links: (https://osnomogram-xyhospital.shinyapps.io/BCLM-OSnomogram/) (accessed on 9 February 2024) and (https://cssnomogram-xyhospital.shinyapps.io/BCLM-CSSnomogram/) (accessed on 9 February 2024). You can utilize these platforms to perform data calculations and conduct risk analysis.

Fig. 5.

The web-based dynamic nomogram. A Web-based dynamic nomogram of OS of BCLM patients (https://osnomogram-xyhospital.shinyapps.io/BCLM-OSnomogram/) (accessed on 9 February 2024). B Web-based dynamic nomogram of CSS of BCLM patients (https://cssnomogram-xyhospital.shinyapps.io/BCLM-CSSnomogram/) (accessed on 9 February 2024)

Discussion

Breast cancer holds the highest incidence and ranks as the second leading cause of mortality among female malignancies. It’s estimated that over 150,000 breast cancer survivors in the United States are living with metastatic disease, with distant metastasis being a primary cause of mortality [12, 13]. Liver metastasis stands as the third most prevalent site for solid tumor metastasis in breast cancer, with the survival rates of breast cancer liver metastasis (BCLM) patients notably lower compared to those with bone or lung metastasis [14]. Despite the existence of prognostic nomograms for bone and lung metastasis of breast cancer established by Lu et al. [15] and Wang et al. [16], respectively, there has been a notable absence of prognostic nomograms specifically tailored for BCLM patients. Hence, we have developed a dynamic web-based nomogram leveraging fundamental patient information, pathological classifications, and treatment protocols of BCLM patients. This tool facilitates rapid and accurate evaluation of both overall survival (OS) and cancer-specific survival (CSS) statuses across diverse patient profiles. Clinicians can seamlessly input patient-specific information into our nomogram to assess prognosis and guide clinical treatment decisions effectively.

The typical cascade of tumor metastasis involves tumor cells invading blood or lymphatic vessels, transitioning into circulating tumor cells (CTCs), adapting to the microenvironment as disseminated tumor cells (DTCs), and eventually evolving into metastatic initiation cells upon reaching the site of metastasis to establish metastatic tumors [17, 18]. However, the molecular mechanism underlying liver metastasis in breast cancer remains a relatively unexplored area. The influence of liver-specific sinus structures and the liver microenvironment on breast cancer cells remains ambiguous [5]. Contrary to common belief, our findings challenge the notion that patients with breast cancer liver metastasis (BCLM) and positive lymph node metastasis exhibit better overall survival (OS) and cancer-specific survival (CSS) compared to those without metastasis. Instead, our results suggest that breast cancer cells may have a predilection for escaping into blood vessels rather than lymphatics to form CTCs, ultimately leading to liver metastases. Consequently, future research on the mechanism of liver metastasis in breast cancer may benefit from focusing on the mechanism of blood-derived tumor cell dissemination.

Our study revealed a correlation between the racial category of BCLM patients and their overall survival (OS) and cancer-specific survival (CSS), with black women demonstrating a tendency toward poorer prognoses. This disparity may stem from socioeconomic disparities, as black women often face economic challenges leading to lower levels of healthcare access and disparities in receiving standard treatment protocols. Reports indicate notable differences in treatment patterns between black and white breast cancer patients [19, 20]. For instance, for stage I and II breast cancer, a higher percentage of white patients receive breast-conserving surgery compared to black women. Similarly, for stage III breast cancer, black women are less likely to receive radiation and chemotherapy alone, with a higher propensity for mastectomy compared to white women. Furthermore, black women with hormone receptor (ER, PR) positive breast cancer are less likely to undergo endocrine therapy compared to their white counterparts [21]. Hence, it is imperative, based on our findings and similar reports, to ensure equitable access to high-quality treatment and health screening services for breast cancer patients, especially those with BCLM, across all racial backgrounds. Specifically, for black women, initiatives should focus on ensuring appropriate patients receive standardized treatments such as hormone therapy or mastectomy [22, 23]. Additionally, efforts to provide comprehensive, equitable, and affordable health insurance coverage for individuals of all racial backgrounds are crucial [24].

Our study also revealed that the receipt of radiotherapy did not significantly impact the overall survival (OS) and cancer-specific survival (CSS) of BCLM patients. Several randomized clinical trials have indicated that older patients with hormone-receptor-positive breast cancer who undergo breast-conserving surgery may not experience survival disadvantages without radiation therapy [25–27]. Therefore, drawing from these clinical trials and our own findings, we suggest that for BCLM patients who opt out of radiotherapy or have contraindications to it, omitting radiotherapy from the treatment regimen may not adversely affect distant recurrence or long-term overall survival. Furthermore, our findings suggest that surgical intervention is associated with improved survival outcomes in BCLM patients, aligning with conclusions drawn from multiple retrospective studies and randomized clinical trials [28–31]. Possible explanations for this phenomenon include: the potential for enhanced effectiveness of systemic treatments by reducing tumor burden through primary tumor removal; removal of necrotic and non-vascularized tumor regions, potentially inaccessible to drugs, thereby enhancing chemotherapy efficacy; elimination of the primary tumor as a source of systemic seeding; and alteration of the immunosuppressive tumor microenvironment, potentially rendering systemic therapy more effective [32, 33]. To date, the relationship between surgical removal of the primary tumor and prognosis in stage IV breast cancer patients has remained inconclusive. Nonetheless, our study findings may provide support for the notion that primary tumor removal significantly enhances both OS and CSS in BCLM patients.

Limitations

Our study also carries several limitations that warrant consideration. Firstly, certain variables that could potentially impact the prognosis of BCLM patients, such as nutritional status and the Ki67 index, were not included in our analysis. Incorporating these variables could provide further insights into patient outcomes. Secondly, our study relied on a retrospective analysis of data from the SEER database, which inherently introduces selection bias. Future studies employing prospective designs could help mitigate this limitation. Lastly, our verification of the nomogram was conducted using cases from within the SEER database rather than external cases, which may affect the generalizability of our findings. Utilizing external cases for validation would enhance the accuracy and reliability of our nomogram.

Abbreviations

BCLM

Breast cancer liver metastases

OS

Overall survival

CSS

Cancer-specific survival

DCA

Decision curve analysis

TNM

Tumor lymph node metastasis

ER

Estrogen receptor

PR

Progesterone receptor

HER2

Human epidermal growth factor receptor 2

OR

Odds ratio

CI

Confidence interval

CTC

Circulating tumor cells

DTC

Disseminated tumor cells

Author contributions

Conception and design: M.T., K.W. and M.L.; administrative support: M.T. and K.W.; provision of study materials or patients: M.T.; collection and assembly of data: M.T., K.W. and M.L.; data analysis and Interpretation: K.W., and M.L.; writing—original draft: M.T., K.W. and M.L.: writing—review and editing: all authors final approval of the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by funds from the Natural Scientific Foundation of China (No. 30771122) to Ming Li.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Informed consent

Not applicable.

Competing interests

The authors declare no competing interests.