Abstract

Background

   Sarcopenia, an age-related syndrome characterized by a decline in
   muscle mass, not only affects patients’ quality of life but may also
   increase the risk of breast cancer recurrence and reduce survival
   rates. Therefore, investigating the genetic mechanisms shared between
   breast cancer and sarcopenia is significant for the prevention,
   diagnosis, and treatment of breast cancer.

Methods

   This study downloaded gene expression datasets and clinical data
   related to breast cancer and skeletal muscle aging from the GEO
   database. Data preprocessing, integration, differential gene
   identification, functional enrichment analysis, and construction of
   protein-protein interaction networks were performed using R language.
   Subsequently, COX proportional hazards model analysis and survival
   analysis were conducted, and survival curves and nomograms were
   generated. The expression levels of genes in tissues were detected
   using qRT-PCR, and the Radiant DICOM viewer software was used to
   delineate the pectoralis major muscle area in CT images.

Results

   We identified 152 differentially expressed genes (P < .05) and 226
   sarcopenia-related genes (r > .4) associated with skeletal muscle
   aging. The TCGA-BRCA dataset revealed 106 genes associated with breast
   cancer (P < .05, logFC = 1). Functional enrichment analysis indicated
   significant enrichment in cell proliferation and growth pathways. The
   PPI network identified critical molecules involved in muscle aging and
   tumor progression. After dimensionality reduction, a strong correlation
   was observed between the expression of the muscle aging-related gene
   set and the prognosis of breast cancer patients (P < .01). The
   expression of SLC38A1 identified through multivariate COX analysis was
   significantly associated with poor prognosis in breast cancer patients
   (P = .03). Incorporating SLC38A1 expression, the prognostic model
   precisely forecasted breast cancer survival (P < .01). External
   validation confirmed the higher expression of the SLC38A1 gene in
   breast cancer tissues compared to adjacent non-cancerous tissues
   (P < .01). The SLC38A1 index, calculated in combination with the
   patient’s age and BMI, can optimize the prognostic prediction model,
   providing a powerful tool for personalized treatment of breast cancer.

Conclusion

   High SLC38A1 gene expression was significantly associated with poor
   prognosis in breast cancer patients. The combination of SLC38A1
   expression and the pectoralis major muscle area provided an optimized
   prognostic prediction model, offering a potential tool for personalized
   breast cancer treatment.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12885-024-13326-y.

   Keywords: Breast cancer, Sarcopenia, Prognostic prediction model,
   SLC38A1, Multivariate analysis

Introduction

   In breast cancer development, fat tissue and lean tissue play different
   roles [[32]1, [33]2]. Research suggests that obesity and high body fat
   may increase the risk of breast cancer, while lean tissue, especially
   skeletal muscle, may be somewhat protective against breast cancer
   [[34]3, [35]4]. In sarcopenia, muscle mass, strength, and physical
   function are reduced as people age [[36]5]. With age, the number and
   size of skeletal muscle fibers in the human body gradually decrease,
   leading to a decline in muscle strength and function [[37]6].
   Sarcopenia is often accompanied by a chronic inflammatory state that
   may promote tumor growth and metastasis, thereby affecting the survival
   time of breast cancer patients [[38]7, [39]8]. Zhang’s study showed
   that people with sarcopenia younger than 55 years old had a lower risk
   of death than those with sarcopenia older than 55 years old, and both
   had an increased risk of death compared with non-sarcopenia patients
   [[40]5]. A retrospective study using L3 plane psoas density to assess
   breast cancer prognosis suggested that the median OS was significantly
   shorter in the group with low psoas density than in the group with high
   psoas density [[41]9]. In addition, sarcopenia may also lead to immune
   dysfunction in patients, reducing the body’s ability to clear tumor
   cells and further increasing the risk of breast cancer recurrence
   [[42]10]. On the other hand, breast cancer itself and its treatment
   process may also have adverse effects on skeletal muscle [[43]11].
   Patients with breast cancer often experience changes in body
   composition, including loss of skeletal muscle mass. Chemotherapy,
   radiotherapy, and surgery may lead to skeletal muscle damage and
   accelerated aging, further affecting the quality of life and prognosis
   of patients [[44]12, [45]13]. However, the mechanisms that breast
   cancer and sarcopenia share at the genetic level are unclear.

   In the study of breast cancer, using public data to screen differential
   genes has become an essential means to reveal the mechanism of tumor
   development. Qiu et al. used 33 genes related to scorch death to build
   a risk model to analyze the correlation between scorch death risk
   score, immune checkpoint-related gene expression, and anticancer drug
   sensitivity [[46]14]. Based on 15 immune genes, Chen et al. developed a
   breast cancer prognostic risk-scoring model [[47]15]. The screening of
   differential genes provides the basis for the follow-up research, but
   the massive gene expression data needs to be effectively reduced
   dimensionality processing for better analysis. Non-negative matrix
   decomposition (NMF), as a data dimensionality reduction method, has
   been widely used in bioinformatics [[48]16, [49]17]. researchers have
   also begun combining genetic information with clinical imaging features
   to predict breast cancer prognosis more fully. Bismeijer. et al. found
   a relationship between breast cancer MRI phenotype and its underlying
   molecular biology from gene expression data, suggesting that the
   enhancement and clarity of tumor margins are associated with ribosomes
   [[50]18]. The results of the study by Christina et al. showed that
   combining imaging and genetic hypoxia biomarkers to classify hypoxia
   could improve the prediction of the effect of radiotherapy and
   chemotherapy in cervical cancer [[51]19]. The application of this
   comprehensive approach provides a new perspective for the accurate
   treatment and prognosis assessment of breast cancer.

   In conclusion, the combination of genetic information and clinical
   imaging features can be used to evaluate breast cancer’s biological
   characteristics and clinical phenotypes comprehensively. First, we
   downloaded skeletal muscle age-related datasets from public databases
   and performed data integration and differential gene screening. We
   further conducted dimension reduction processing and survival analysis
   of breast cancer data. Using COX proportional risk models, we analyzed
   the relationship between breast cancer prognosis and muscle aging
   genes. We constructed a prognostic risk prediction model based on these
   findings and verified its effectiveness and practicability. Finally, we
   conducted external data verification of key gene SLC38A1 and explored
   the application of SLC38A1 gene expression and imaging data in breast
   cancer prognosis assessment. This study provides an important basis for
   understanding the role of SLC38A1 in the occurrence and development of
   breast cancer. It provides a powerful tool for developing personalized
   treatment and prognosis assessment. Figure [52]1 shows the study’s flow
   chart.

Fig. 1.

   [53]Fig. 1
   [54]Open in a new tab

   Our research workflow began with extracting gene expression profiles
   and clinical datasets linked to breast cancer and sarcopenia from the
   GEO repository. Subsequent steps included data preprocessing using R,
   differential gene detection with the limma package, and functional
   enrichment analysis via ClusterProfiler. We mapped out protein
   interactions using STRING and assessed patient survival rates with
   Survminer, applying the COX proportional hazards model. The prognostic
   risk prediction model was constructed using the RMS package, while
   subgroup analysis was conducted with the NMF package. Tumor
   characteristics and immune cell infiltration were evaluated with
   ESTIMATE and IOBR tools. Gene expression was validated through qRT-PCR,
   and the Radiant DICOM viewer software was utilized to outline the
   pectoralis major muscle area. GraphPad handled statistical analysis,
   and the research was conducted with ethical approval

Method

Data download and preprocessing

   Gene expression datasets and clinical data related to breast cancer and
   skeletal muscle aging were downloaded from the GEO database.
   [55]GSE25941, [56]GSE28392, and [57]GSE28422 were used to identify
   differential genes through data integration. The TCGA-BRCA dataset was
   used to identify muscle aging genes and their association with
   prognosis. External validation was conducted using the skeletal muscle
   aging-related gene sets [58]GSE117525, [59]GSE47881, [60]GSE4667,
   [61]GSE674, [62]GSE362, [63]GSE167186, [64]GSE38718, [65]GSE5086,
   [66]GSE8479, and [67]GSE9103. Additionally, breast cancer datasets
   [68]GSE42568, [69]GSE5364, [70]GSE20711, [71]GSE21653, [72]GSE25066,
   [73]GSE88770, and [74]GSE97342 were also used for external validation.
   To mitigate batch effects and technical variations, we utilized the Sva
   package in R. Principal Component Analysis (PCA) was performed using
   the prcomp function in R, demonstrating the main variations among
   samples through dimensionality reduction. Certain gene sets undergo
   correction via data normalization, ensuring that expression levels are
   consistently represented in the TPKM format. Differential gene analysis
   was conducted using the limma package in R. From the Genecard website,
   226 sarcopenia-related genes were identified with a correlation
   coefficient of at least 0.40.

Data set analysis

   ClusterProfiler in R was used to conduct enrichment analyses on GO and
   KEGG. The STRING website constructed the PPI network. Survival curves
   were plotted following the COX proportional hazards model analysis and
   survival analysis. The rms package was employed to construct nomograms,
   while the cor package in R was utilized for correlation analysis.
   Non-negative matrix factorization was carried out using the NMF package
   to explore potential subgroups and survival differences. The R package
   ESTIMATE calculated stromal, immune, and ESTIMATE scores for each
   patient based on gene expression. The R package IOBR reassessed the
   infiltration scores of immune cells for each patient in every tumor
   based on gene expression. The ggplot2 package was used to enhance
   graphics.

Clinical data acquisition and processing

   Tissue samples from breast cancer patients were collected at the First
   Affiliated Hospital of Guangxi Medical University in 2022. Inclusion
   criteria: pathologically confirmed primary breast cancer; survival ≥ 3
   months; patients had their hands raised above their heads during CT
   scanning. Exclusion criteria: The previous treatment of breast cancer
   (such as surgery, radiotherapy, chemotherapy, endocrine therapy, etc.);
   the presence of other malignant tumors or severe comorbidities;
   inability to cooperate with the required examinations, treatments, and
   follow-ups for the study. Finally, 34 patients met the requirements for
   subsequent research. Clinical information collected from patients,
   including age, height, weight, platelet count, neutrophil count,
   lymphocyte count, and monocyte count.

   In the analysis of CT images of breast cancer patients, the T4 thoracic
   vertebral level was selected based on published studies [[75]20,
   [76]21] to measure the pectoralis major muscle area (Figure [77]S1A).
   The outcome of the patient was recorded during follow-up due to breast
   cancer recurrence, lymph node metastasis, re-admission, or death.
   Referring to previous studies, NLR = neutrophils/lymphocytes;
   MLR = monocytes/lymphocytes; SII = neutrophils × platelets/lymphocytes;
   SIRI = neutrophils × monocytes/lymphocytes; PIV = neutrophils ×
   monocytes × platelets/lymphocytes [[78]22]. To further optimize the
   prognostic prediction model, we calculated three different SLC38A1
   indices, namely the ratio of SLC38A1 gene CT value to age (SLC38A1
   index 1), the ratio of SLC38A1 gene CT value to BMI (SLC38A1 index 2),
   and the product of SLC38A1 gene CT value and pectoralis major muscle
   index (SLC38A1 index 3). This research protocol underwent rigorous
   review and approval by the hospital’s ethics committee, ensuring
   ethical compliance and the protection of patient’s rights and interests
   throughout the research process.

qRT-PCR

   The expression difference of the SLC38A1 gene between breast cancer
   tissue and adjacent normal tissue was compared, and GAPDH was used as
   the internal reference gene for standardization. qRT-PCR was performed
   as previously described. The primer sequence is as follows: SLC38A1-F:
   5′- GCATTTGTTTGCCACCCGTC − 3′; SLC38A1-R: 5′- GTCGGACTGCACGTTGTCAT
   − 3′; GAPDH-F: 5′- ATGGGGAAGGTGAAGGTCG − 3′; GAPDH-R: 5′-
   CTCCACGACGTACTCAGCG − 3′.

Statistical analysis

   Data were summarized using descriptive statistics, including mean,
   standard deviation, median, and range. We employed independent sample
   t-tests or Mann-Whitney U tests to compare differences between
   different groups, and the results of correlation analysis were
   described using the Pearson correlation coefficient. The qRT-PCR
   results were completed using GraphPad software. The delineation of the
   pectoralis major muscle in CT images of breast cancer patients was done
   using Radiant DICOM viewer software. We defined |logFC| > 1 as the
   threshold for gene differential analysis. Statistical significance was
   defined as a p-value less than 0.05.

Results

Data Integration and Preliminary Analysis

   The PCA algorithm was employed to illustrate the effect of data
   integration. The results suggest that the batch effect has been
   eliminated(Figs. [79]2A and B). We further utilized the limma package
   to filter out those significantly associated with skeletal muscle aging
   and visualized 152 of these genes through a volcano plot (Fig. [80]2C).
   Additionally, 226 genes related to sarcopenia were obtained from the
   Genecard website (correlation coefficient > 0.40). After removing
   duplicates, 374 genes were included in subsequent studies. Among these,
   106 genes were differentially expressed in the TCGA-BRCA dataset. The
   results of GO enrichment analysis revealed a significant enrichment of
   functional pathways related to cell proliferation and growth
   (Fig. [81]2D). Meanwhile, KEGG pathway enrichment analysis further
   highlighted the importance of membrane protein signaling, kinase
   signaling, and longevity-associated signaling pathways (Fig. [82]2E).
   Furthermore, we constructed a protein-protein interaction network,
   identifying interacting molecules primarily comprising growth factors,
   cytokines, and growth regulators (Fig. [83]2F), which may play crucial
   roles in regulating muscle aging and tumor progression. To explore the
   potential role of muscle aging-related gene sets in breast cancer, we
   utilized these genes as features to reduce dimensionality on TCGA
   breast cancer tumor data. The results of NMF data reduction indicated
   an optimal K value of 2 (Fig. [84]2G), enabling us to segment breast
   cancer patients into two distinct subgroups (Fig. [85]2H). Survival
   analysis revealed a higher survival rate in group 2 compared to group 1
   (Fig. [86]2I, P = .49). When focusing on tumor patients over 60 years
   old, we observed a similar trend, with group 2 exhibiting a higher
   survival rate than group 1 (Fig. [87]2J, P = .25). These findings
   suggest that the expression of muscle aging-related gene sets in breast
   cancer patients may be closely associated with patient prognosis.

Fig. 2.

   [88]Fig. 2
   [89]Open in a new tab

   Molecular Mechanisms Associated with Skeletal Muscle Aging and Their
   Potential Role in Breast Cancer. (A) and (B) Data integration and PCA
   results. The integrated data exhibit good clustering in the PCA plot,
   effectively eliminating batch-to-batch variations. (C) Volcano plot of
   differentially expressed genes. (D) and (E) show the results of GO and
   KEGG enrichment analysis. (F) Protein-protein interaction network. (G)
   and (H) NMF analysis results indicate an optimal K value of 2, leading
   to the classification of breast cancer patients into two distinct
   subgroups. (I) and (J) Survival analysis results

Genetics of muscle aging associated with breast cancer outcomes

   We employed the COX proportional hazards model for analysis to further
   investigate the association between breast cancer prognosis and muscle
   aging-related genes. Through univariate analysis, we identified 12
   genes significantly associated with breast cancer patients’ prognosis
   (Fig. [90]3A). Subsequently, multivariate COX analysis was conducted,
   successfully identifying four independent prognostic genes: CHD9,
   LDLRAD3, GREB1, and SLC38A1 (Fig. [91]3B, P < .05). Patients with high
   expression of SLC38A1 had significantly fewer survival rates than those
   with low expression (Fig. [92]3C, P = .03), indicating that high
   SLC38A1 expression may be closely associated with poor prognosis in
   breast cancer patients. There were no significant differences in
   survival between high and low expression groups for CHD9, GREB1, and
   LDLRAD3(Fig. [93]3D, E and F; P = .81, P = .07, P = .08). We examined
   SLC38A1’s prognostic value in various breast cancer subtypes based on
   the heterogeneity of breast cancer. The results showed that patients in
   the low SLC38A1 expression group exhibited better survival advantages
   in triple-negative breast cancer (Fig. [94]3G, P = .64), HER2-positive
   breast cancer (Fig. [95]3H, P = .019), Luminal A breast cancer
   (Fig. [96]3I, P = .57), and Luminal B breast cancer (Fig. [97]3J,
   P = .42).

Fig. 3.

   [98]Fig. 3
   [99]Open in a new tab

   Investigating the Relationship between Breast Cancer Prognosis and
   Muscle Aging-Related Genes through COX Analysis. (A) Results of
   univariate analysis. (B) Four independent prognostic genes were
   confirmed by multivariate COX analysis: CHD9, LDLRAD3, GREB1, and
   SLC38A1. (C) Survival analysis results for SLC38A1. (D), (E) and (F)
   Survival analysis results for CHD9, GREB1, and LDLRAD3, respectively.
   (G), (H), (I), and (J) Survival analysis results for SLC38A1 in
   triple-negative breast cancer, HER2-positive breast cancer, Luminal A
   breast cancer, and Luminal B breast cancer, respectively

Correlation analysis and construction of a prognostic risk prediction model

   We conducted a correlation analysis to investigate further the
   intrinsic relationship between muscle aging genes associated with
   breast cancer prognosis. The results revealed a strong correlation
   between the expression levels of HSD11B1 and HLA-DQA1 (Fig. [100]4A,
   R = .52, P < .001), while the correlations between other genes were
   relatively weak. Based on these discoveries, we further utilized seven
   muscle aging genes associated with breast cancer prognosis to construct
   a prognostic risk prediction model. Based on the mean risk index, we
   classified breast cancer patients into high-risk and low-risk
   categories (Fig. [101]4B). Heatmaps illustrate how muscle aging genes
   are expressed differently between high-risk and low-risk individuals
   (Fig. [102]4C). Figure [103]4D depicts the distribution of patients
   with varying risk scores, where red dots represent deceased patients
   and green dots represent surviving patients. Survival analysis revealed
   that low-risk patients’ survival rate was significantly lower than
   high-risk patients‘(Fig. [104]4E, P < .01). The risk model was used to
   develop a nomogram (Fig. [105]4F). This nomogram can assist clinicians
   in rapidly predicting the 3-year, 5-year, and 10-year survival rates of
   breast cancer patients. These results validate the effectiveness and
   practicality of the prognostic risk prediction model and provide a
   powerful tool for developing personalized treatment plans and
   prognostic evaluations.

Fig. 4.

   [106]Fig. 4
   [107]Open in a new tab

   Construction of a prognostic risk prediction model based on muscle
   aging genes associated with breast cancer prognosis. (A) Correlation
   analysis demonstrates the relationships between muscle aging genes and
   breast cancer prognosis. (B) Breast cancer patients were stratified
   into high-risk (red) and low-risk (green) groups using the mean risk
   index as the cutoff. (C) Heatmap illustrating the differential
   expression of prognosis-related muscle aging genes between the high-
   and low-risk groups. (D) Scatter plot showing patient risk distribution
   and survival status. Each point represents an individual patient, with
   red dots indicating deceased patients and green dots indicating
   surviving patients. The x-axis represents the patient’s risk score. (E)
   Nomogram assists clinicians in predicting breast cancer patients’
   3-year, 5-year, and 10-year survival rates

External data validation of SLC38A1 gene

   SLC38A1 gene expression decreased significantly in older participants
   compared to younger participants in age-related datasets (Fig. [108]5A,
   P < .05). Considering that breast cancer predominantly occurs in women,
   we further investigated whether there are gender differences in SLC38A1
   gene expression. Analysis of the [109]GSE117525 and [110]GSE8479
   datasets revealed no significant difference in SLC38A1 gene expression
   between males and females (Fig. [111]5B, P > .5), suggesting the
   functional universality of this gene across genders. Next, we focused
   on the relationship between the SLC38A1 gene and physical health
   status. The analysis showed that SLC38A1 expression was lower in
   overweight patients and those with sarcopenia compared to healthy
   individuals (Fig. [112]5C, P < .01). Notably, SLC38A1 gene expression
   exhibited an increasing trend when patients began physical training
   (Fig. [113]5D), implying a potential regulatory role of physical
   training on SLC38A1 expression. The [114]GSE42568 and [115]GSE5364
   breast cancer datasets were used to validate the expression of the
   SLC38A1 gene. According to consistent results, breast cancer tissues
   expressed significantly more SLC38A1 than normal tissues (Fig. [116]5E,
   P < .01), further supporting the crucial role of SLC38A1 in
   tumorigenesis and development. Finally, we conducted external survival
   validation to assess the impact of SLC38A1 gene expression on survival
   outcomes in breast cancer patients. In the [117]GSE20711 and
   [118]GSE88770 datasets, patients with low SLC38A1 expression appeared
   to have a better survival advantage (Fig. [119]5F, P > .1). However, in
   the [120]GSE42568 dataset, we observed opposite results (Fig. [121]5F,
   P < .01). Additionally, the results suggested a weak negative
   correlation between the SLC38A1 gene and breast cancer immunity (Figure
   [122]S1B). These inconsistent findings may reflect the heterogeneity of
   breast cancer and differences in patient characteristics across
   datasets.

Fig. 5.

   [123]Fig. 5
   [124]Open in a new tab

   The validation results of the SLC38A1 gene across multiple external
   datasets demonstrate consistent representation of expression levels in
   the TPKM format. (A) Comparison of SLC38A1 gene expression levels in
   age-related datasets. (B) Analysis of SLC38A1 gene expression
   differences between genders. (C) Relationship between the SLC38A1 gene
   and physical health status. (D) Impact of physical training on SLC38A1
   gene expression. (E) SLC38A1 gene expression in breast cancer datasets.
   (F) External survival validation of the effect of SLC38A1 gene
   expression on survival outcomes in breast cancer patients

SLC38A1 gene prognostic relationship in breast Cancer

   To understand the role of the SLC38A1 gene in the prognosis of breast
   cancer patients, we analyzed multiple datasets and validated our
   results using clinical samples. COX analysis revealed significant
   heterogeneity among the [125]GSE20711, [126]GSE21653, [127]GSE25066,
   [128]GSE88770, and [129]GSE97342 datasets. Specifically, in the
   [130]GSE97342 dataset, estrogen receptor (ER) and progesterone receptor
   (PR) positivity were unexpectedly identified as protective factors for
   breast cancer prognosis, while in the [131]GSE21653 dataset, a higher
   T-stage was considered a protective factor (Figure [132]S1C). These
   findings contradicted conventional clinical knowledge. Although the
   SLC38A1 gene was considered a risk factor in the [133]GSE88770,
   [134]GSE20711, and [135]GSE25066 datasets, these associations were not
   statistically significant (Figure [136]S1C, P > .1). To explore the
   clinical significance of SLC38A1 more directly, we included an
   independent sample set comprising 34 breast cancer patients. These
   patients were pathologically diagnosed with non-specific invasive
   ductal carcinoma grade II/III, with a mean age of 54.59 ± 12.85 years
   (range 37–81 years). Through qRT-PCR analysis, we again confirmed that
   SLC38A1 gene expression was significantly higher in tumors than in
   adjacent tissues (Fig. [137]6A, P = .001). As a result of univariate
   COX analysis, it was determined that SLC38A1 gene expression has a
   prognostic significance for breast cancer (Fig. [138]6B, P < .01).
   Although the SLC38A1 index had some reference value in breast cancer
   prognosis assessment, its statistical significance was not strong
   (Fig. [139]6B, P = .093, P = .063, P = .272). We then stratified
   patients into two groups based on the mean SLC38A1 expression. Survival
   analysis suggested that differences in SLC38A1 gene expression were
   helpful in prognosis assessment (Fig. [140]6D, P < .01). However, the
   pectoralis major index did not significantly contribute to prognosis
   (Fig. [141]6C, P = .5). The SLC38A1 index showed some promise in
   prognosis assessment (Fig. [142]6E, F, and G, P = .29, P = .22,
   P = .19). These findings suggest that gene expression levels and
   nutritional indicators may aid in the prognosis assessment of breast
   cancer patients.

Fig. 6.

   [143]Fig. 6
   [144]Open in a new tab

   Comprehensive exploration of the prognostic relationship between the
   SLC38A1 gene and breast cancer patients. (A) PCR results confirm the
   high expression pattern of SLC38A1 in breast cancer. (B) COX analysis
   results from external datasets. (C) The pectoralis major index did not
   significantly contribute to prognosis. (D) Survival analysis suggested
   that differences in SLC38A1 gene expression were helpful in prognosis
   assessment. (E), (F) and (G) Optimized prognostic prediction model
   results. This model was achieved by calculating three different SLC38A1
   indices and conducting a COX analysis

Discuss

   Breast cancer prognosis is not only related to the characteristics of
   the tumor itself but also closely associated with the overall
   physiological status of the patient. Among various physiological
   indicators, muscle mass has received widespread attention in recent
   years as an essential factor. In this study, we conducted an integrated
   analysis of multiple datasets to delve deeper into the potential role
   of skeletal muscle aging-related molecules in breast cancer. Based on
   this, we further explored the application value of these molecules in
   the prognostic assessment of breast cancer. In addition, we paid
   particular attention to the importance and clinical application value
   of the SLC38A1 gene in breast cancer prognosis evaluation. Using the
   COX proportional hazards model, we identified four independent
   prognostic genes: CHD9, LDLRAD3, GREB1, and SLC38A1, among which the
   high expression of SLC38A1 is closely associated with poor prognosis in
   breast cancer patients. Furthermore, we constructed a prognostic risk
   prediction model that effectively predicts the survival rate of breast
   cancer patients, providing a powerful tool for personalized treatment.
   Subsequently, we delved into the relationship between SLC38A1 gene
   expression and age, gender, physical health status, as well as breast
   cancer. To further optimize the prognostic prediction model, we
   calculated three different SLC38A1 indices: the ratio of SLC38A1 gene
   CT value to age (SLC38A1 Index 1), the ratio of SLC38A1 gene CT value
   to BMI (SLC38A1 Index 2), and the product of SLC38A1 gene CT value and
   pectoralis major muscle index (SLC38A1 Index 3). Therefore, this study
   aims to deeply explore the application value of muscle aging genes in
   breast cancer prognosis assessment and seeks to develop more accurate
   prognostic prediction models to provide personalized treatment guidance
   and improve patients’ quality of life.

   Enrichment analysis of genes significantly associated with skeletal
   muscle aging revealed that they are primarily involved in cell
   proliferation and growth, membrane protein signaling, kinase signaling,
   and longevity-related signaling pathways. This result suggests that
   these pathways may play crucial roles in muscle aging. Analysis of
   protein-protein interaction networks has uncovered the central roles of
   molecules such as growth factors, cytokines, and growth regulatory
   factors in regulating muscle aging and tumor progression. These
   molecules have far-reaching effects on cell signaling, proliferation
   regulation, and microenvironment shaping. Growth factors and cytokines
   are particularly affected by muscle aging [[145]23, [146]24], which
   affect the muscle’s regenerative capacity and metabolic status
   [[147]25]. Notably, their role in breast cancer has been confirmed by
   multiple studies. EGF binds to EGFR on the cell surface, triggering the
   classical EGF/EGFR signaling pathway, leading to upregulation of ERK1/2
   and AKT phosphorylation levels and promoting tumor cell proliferation
   and survival [[148]26, [149]27]. The tumor necrosis factor mediates
   muscle atrophy in cancer cachexia or inflammation [[150]28]. The
   IL6/STAT3 pathway has been identified as a possible contributor to
   obesity-driven ER + breast cancer metastasis, according to studies
   [[151]29]. Abnormal activation of the IGF signaling pathway is closely
   associated with poor prognosis in breast cancer [[152]30, [153]31]. The
   combined effect of these factors forms a complex regulatory network
   that together drives the progression of breast cancer and sarcopenia.
   The results of our survival analysis show significant differences in
   survival among groups of breast cancer patients with different
   molecular characteristics of muscle aging. There is mounting evidence
   that sarcopenia reduces patients’ quality of life and substantially
   increases their death risk for breast cancer patients [[154]12,
   [155]32]. In addition, the study found that patients with sarcopenia
   were less tolerant during chemotherapy and were more likely to
   experience dose reductions or treatment interruptions, which affected
   the effectiveness of chemotherapy and patient outcomes [[156]33,
   [157]34]. These studies provide strong evidence for our results that
   skeletal muscle age-related molecules have important applications in
   evaluating breast cancer prognosis.

   Our study reveals the potential role of muscle aging-related genes in
   breast cancer prognosis and specifically highlights the SLC38A1 gene as
   an important prognostic marker. SLC38A1 is a crucial transporter and
   belongs to the SLC38 family. It plays a variety of key functions in the
   human body, especially in amino acid transport and metabolism
   processes, which play an important role. The expression of SLC38A1 is
   abnormal in many types of cancer, including breast, lung, and colon
   cancers [[158]35–[159]37]. In the context of breast cancer, SLC38A1
   expression was significantly higher in tumors than in normal tissues,
   consistent with previous reports [[160]38]. Furthermore, research
   indicates that SLC38A1 may regulate the polarization of
   tumor-associated macrophages [[161]38]. Our results suggest a negative
   correlation between SLC38A1 expression and immune score in breast
   cancer. It is worth noting that this study observed some interesting
   but inconsistent results in some analyses. In exploring the
   relationship of the SLC38A1 gene to the prognosis of breast cancer
   patients, we found strong heterogeneity among different data sets.
   These results may be related to the heterogeneity of breast cancer,
   differences in patient characteristics between different data sets, and
   limitations of statistical methods. In addition to cancer, SLC38A1 may
   also be involved in the occurrence and development of other diseases.
   Some studies have found that SLC38A1 also plays a vital role in nervous
   and metabolic diseases [[162]39]. This study found that SLC38A1
   expression in skeletal muscle may decline with age, and SLC38A1
   expression is lower in overweight patients and patients with sarcopenia
   than in healthy people. It is worth noting that when patients began to
   receive physical training, the expression of the SLC38A1 gene in
   skeletal muscle showed an upward trend. These results suggest that
   physical training may have a regulatory effect on the expression of
   SLC38A1 in muscles. These results indicate that SLC38A1 expression is
   heterogeneous in different tissues, and high expression of SLC38A1
   positively affects the development of skeletal muscle cells and breast
   cancer cells. However, the specific mechanism of action of SLC38A1 in
   these diseases is not fully understood at present, and further studies
   are needed to reveal it.

   In calculating muscle mass, the area of the pectoralis major muscle at
   different thoracic levels is often used as an indicator [[163]21,
   [164]40]. This measurement provides us with an effective means of
   quantifying muscle mass and underscores its importance in the
   assessment of breast cancer prognosis. Our results show that the
   expression of the SLC38A1 gene in tumors is identified as a risk factor
   for breast cancer recurrence. However, the pectoral major index showed
   no significant effect. We then combined genetic information and imaging
   data to assess the risk of breast cancer recurrence. Although HR showed
   a clear advantage, the statistical significance was not strong, but
   they still provided us with valuable information for prognostic
   prediction. There are still many limitations to this study. Although
   the prognosis prediction model constructed in this study has achieved
   certain predictive effects, it still needs to be further verified and
   optimized by expanding the number of patients. At the same time, the
   selection of T4 level pectoral major muscle area as a reference for
   patient muscle mass is still worth considering. In future studies, we
   will continue to explore the relationship between SLC38A1 expression in
   skeletal muscle and muscle mass and the prognosis of breast cancer
   patients, further expand the sample size, and include more prognostic
   factors to improve the prognostic value of the SLC38A1 index.

   In conclusion, this study provides a new perspective and biomarker for
   the prognosis assessment of breast cancer. Future studies should
   further explore the mechanism of muscle mass loss and its role in the
   occurrence and development of breast cancer, to provide new ideas and
   methods for prognosis prediction and treatment of breast cancer.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [165]Supplementary Material 1^ (302.8KB, pdf)

Author contributions

   Y.W. and H.Y. are responsible for study conceptualization. Y.W.
   generated most of the data, assisted by P.Z. CJ.W. checked the
   statistical method. WJ.H., P.Z., and Y.W. prepared the figures. Y.W.
   wrote the manuscript. H.Y. provided resources, supervision, and
   editing. All authors contributed to the article and approved the
   submitted version.

Funding

   This study was funded by the Youth Science Foundation of Guangxi
   Medical University(GXMUYSF202327), the National Natural Science
   Foundation of China (82160336), the Natural Science Foundation of
   Guangxi (2023GXNSFDA026013, 2020GXNSFDA238005).

Data availability

   Data sources and handling of the publicly available datasets used in
   this study are described in Sect. 2. The raw data that support the
   findings of this study are available from the corresponding author upon
   request.

Declarations

Ethics approval and consent to participate

   The study was conducted in accordance with the principles of the
   Declaration of Helsinki and approved by the Medical Ethics Committee of
   the First Affiliated Hospital of Guangxi Medical University, with
   approval number 2021 (KY-E-034). All patients were fully informed about
   the study and provided written consent before participating.

Conflict of interest

   The authors declare no conflicts of interest.

Footnotes

   Publisher’s note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

References