MiRNAs regulate gene expression by interacting with target mRNA, playing key roles in the cell cycle, development, apoptosis, and disease progression [21, 22]. A recent study achieved selective drug delivery by co-loading miR-34a and photosensitizers within cationic liposomes [23]. This strategy leads to the suppression of immune escape in triple-negative breast cancer (TNBC), offering a novel perspective for the enhancement of cancer immunotherapy. A separate study demonstrated that BCa cells hijack the neuron-astrocyte metabolic coupling to promote brain metastasis by secreting miR-199b-5p [24]. This suggests a potential target for therapeutic strategies against brain metastases in BCa. Therefore, microRNAs are a promising and safe biomarker for disease diagnosis and treatment, offering significant potential for personalized medicine. Currently, there is a lack of relevant studies both domestically and internationally confirming the associations of miR-1911-3p, miR-4694-5p, miR-548ao-5p, and miR-4804-3p with the development of BCa. This study validated significant reductions in the serum levels of miR-1911-3p, miR-4694-5p, miR-548ao-5p, and miR-4804-3p in BCa patients through three-phase testing, suggesting that these miRNAs may play an important role in the occurrence and development of breast cancer. Several studies indicate that a molecular signature, combining multiple biomarkers, provides greater specificity and accuracy for diagnostics than individual biomarkers [25]. Hong’s experiment established a panel of eight miRNAs (miR-324-5p, miR-10b-5p, miR-455-3p, miR-486-5p, miR-20a-5p, miR-107, miR-146b-5p and miR-139-5p) to predict recurrence in TNBC patients and offer personalized treatment [26]. This study identified miR-548ao-5p and miR-4804-3p as independent inhibitors of BCa through multivariate analysis. Their expression patterns were consistent in both serum and tissues of BCa patients. However, no significant differences were found in the levels of these microRNAs among hepatocellular carcinoma, gastric cancer, and lung cancer patients. These findings suggest that miR-548ao-5p and miR-4804-3p may serve as non-invasive biomarkers for BCa diagnosis, providing a promising alternative to traditional tissue biopsy methods.

Serum tumor marker testing is widely utilized in clinical laboratories as an indicator because of its convenience, low invasiveness, and cost-effectiveness in clinical development [27]. The most commonly used serum tumor markers in BCa screening and auxiliary diagnosis include AFP, CEA, CA153, CA152, and CA199. However, a single tumor marker is usually insufficient for accurate diagnosis. The use of multiple markers can increase the accuracy of diagnosis, and different combinations of markers yield varying diagnostic values [28]. Luo’s research confirmed that among the various combinations of three parallel detections of BCa tumor markers, the sensitivity of AFP + CEA + CA153 was the highest, whereas the AUC of CEA + CA125 + CA199 was the highest [29]. Our study also revealed that both CA125 and CA153 independently contribute to breast cancer incidence, but they both exhibited poor sensitivity. However, when these methods were combined, they had greater higher diagnostic effectiveness. Although numerous studies have demonstrated that molecular signatures composed of multiple miRNAs are valuable for the diagnosis and treatment of BCa, the diagnostic utility of miRNAs combined with serum tumor markers remains poorly understood. (miRNA + TM) Panel III, a novel diagnostic model of miRNA combined with tumor markers, was developed. This approach addresses the issue of low sensitivity associated with serum tumor markers and demonstrates excellent sensitivity, specificity, accuracy, stability, and overall diagnostic performance. The robustness of this model was validated through multi-phase design and comprehensive comparative analyses from various perspectives. This finding addresses a significant gap in BCa research regarding novel miRNAs, which not only has substantial scientific merit but also holds potential for breakthroughs in diagnosis and treatment.

Breast cancer is a highly heterogeneous malignancy, with diverse tumors in different patients in terms of growth rate, aggressiveness, hormone dependence, and treatment response [30, 31]. ER, PR and HER2 are widely used indicators for targeted BCa treatment and prognosis assessment [32, 33]. Invasive breast cancer can be categorized into four major molecular subtypes, namely, luminal A, luminal B, HER2-enriched, and triple-negative subtypes, which are based on hormone receptor status and gene expression patterns [34]. The management of BCa is complex and involves personalized treatment tailored to the molecular subtypes of patients. Scholars have proposed that miRNAs may play a role in regulating the pathogenicity of tumors by disrupting signaling pathways, potentially influencing the response to hormone therapy in the luminal A and luminal B subtypes of BCa [35]. For example, high expression levels of miR-221/222 promote the transition of BCa cells from ER-positive to ER-negative, while miR-221/222 sponge can effectively treat Tamoxifen-resistant BCa patients through G1 cell cycle arrest and restoration of ER-α expression [36]. Multifactorial linear regression revealed a strong correlation between miR-548ao-5p and miR-4804-3p concerning HER2 status. Patients with HER2-enriched breast cancer often receive targeted anti-HER2 therapy, which is linked to an aggressive clinical course and poor prognosis [37]. As clinical trials advance, an increasing amount of research is focusing on targeted delivery systems. These include the use of engineered cellular carriers [38], self-assembling nanoparticles [39] and photo-responsive cationic liposome platform [23] to deliver microRNAs to specific sites for therapeutic purposes. Our study shows that miR-548ao-5p and miR-4804-3p are linked to the onset and progression of BCa. In the future, combining targeted therapies with microRNAs is expected to improve treatment efficacy for HER2-positive breast cancer, leading to more precise personalized treatment strategies.

The mortality rate of BCa has decreased in recent years. Approximately 20–30% of patients are estimated to experience metastasis after diagnosis or treatment [40]. The prognosis of invasive breast cancer is significantly impacted by the tumor-node-metastasis (TNM) staging system [41]. TNM anatomic staging predicts the overall stage of the disease by combining the quantitative classification of primary tumors (T), regional lymph nodes (N), and distant metastases (M) [42, 43]. A lower histologic grade, positive PR, and fewer positive lymph nodes are independently correlated with a low-risk recurrence score [44]. Currently, anatomic staging is combined with prognostic staging to provide a more comprehensive diagnosis, prognosis, and cancer management. Studies have indicated that patients with a larger tumor diameter or lymphatic proliferation are associated with an increased rate of early mortality [45]. Du et al. reported that the expression of miR-92b-3p increases in BCa patients with larger tumor diameters, low tumor differentiation, high TNM stage, and lymphatic metastasis, which suggests that miR-92b-3p could serve as a potential biomarker for both the diagnosis and prognosis of BCa [46]. Additionally, the downregulation of miR-150-5p is closely linked to larger tumors, higher p53 expression, increased tumor recurrence, the presence of distant metastasis, and poor survival outcomes [47]. The 5-year survival rate for patients with stage I significantly exceeds that of patients with distant metastases [48]. This study revealed a correlation between the expression of miR-548ao-5p and miR-4804-3p and the TNM anatomic staging of BCa. These was a significant decrease in the serum levels of these microRNAs in patients with stage III/IV disease as well as those who had undergone lymph node transplantation, with a decreasing trend noted as the T segmentation level increased. These findings clearly indicate that miR-548ao-5p and miR-4804-3p can serve as indicators for the progression of BCa and are closely associated with clinical prognosis.

In this study, we used the TargetScan, miRDB, and miRWalk databases to predict the target genes of miR-548ao-5p and miR-4804-3p. Target gene identification was performed using the Gene Expression Profiling Interactive Analysis (GEPIA) database, The University of ALabama at Birmingham CANcer data analysis Portal (UALCAN) database, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and Gene Ontology (GO) enrichment analysis. Our findings indicate that miR-548ao-5p may regulate cyclinB1 expression by targeting Heterogeneous Nuclear Ribonucleoprotein F (HNRNPF), thus inhibiting cell cycle progression and acting as a negative regulator in breast cancer (data not shown). In addition, the candidate target gene functions of miR-4804-3p were primarily shown to regulate DNA-binding transcription factors and ATF4-activated genes in response to endoplasmic reticulum stress (data not shown). Although this study’s robustness is enhanced by a multi-stage approach, it has several limitations. First, subjects were exclusively drawn from the First Affiliated Hospital of Guangxi Medical University and the Cancer Hospital of Guangxi Medical University, introducing regional constraints. Second, there is a lack of independent external validation and long-term survival data. In the future, we will conduct multi-center studies with hospitals from other regions to establish independent external validation, demonstrating the robustness of our diagnostic model across diverse populations. Additionally, we will incorporate more clinical variables for analysis and investigate the mechanisms of miRNAs to explore the potential of miR-548ao-5p and miR-4804-3p as biomarkers of BCa. In recent years, the investigation of miRNA combination diagnosis has transitioned from reliance on a single biomarker to an integrative approach that encompasses multiple omics disciplines, including radiology [49]. This field is advancing from laboratory settings into clinical applications. However, several challenges remain: first, there is an urgent need for large-scale clinical trial validation; second, standardizing data from various research platforms and sample sources is still unresolved. With advancements in artificial intelligence, nanotechnology, and body fluid biopsy techniques, miRNAs are expected to play a crucial role in early disease screening and personalized medicine in the future.

In summary, this study evaluated the diagnostic value of miR-1911-3p, miR-4694-5p, miR-548ao-5p, and miR-4804-3p in serum for BCa. A combined diagnostic model (including miR-548ao-5p, miR-4804-3p, CA125 and CA153) with high sensitivity, specificity, accuracy, and stability was successfully developed. This model overcomes the limitations of the low specificity of miRNAs and the low sensitivity of serum markers. The combined model shows very promising diagnostic performance and improved sensitivity for the early detection of BCa. Additionally, quantitative changes in miR-548ao-5p and miR-4804-3p may be associated with the progression and prognosis of BCa.