MicroRNA-486: a dual-function biomarker for diagnosis and tumor immune microenvironment characterization in non-small cell lung cancer

Abstract

Background

This investigation evaluates the clinical significance and molecular mechanisms of microRNA-486 (miR-486) as a potential biomarker in non-small cell lung cancer (NSCLC) through an integrative analytical approach.

Methods

We conducted systematic search and meta-analysis of diagnostic studies from major biomedical databases from inception through April 04, 2025, followed by comprehensive bioinformatics interrogation. Protein–protein interaction (PPI) networks were constructed using STRING to identify key hub genes regulated by miR-486. Validation of hub genes employed TCGA datasets, while immune infiltration analysis utilized TIMER2.0 platform.

Results

The meta-analysis indicated that miR-486, both individually and in combination, could be effective biomarkers for NSCLC detection. Afterwards, functional enrichment analyses of miR-486 target genes highlighted significant ontology terms and pathways crucial to the initiation and progression of NSCLC. PPI networks revealed key proteins and modules that participate in multiple essential pathways associated with NSCLC pathogenesis. Furthermore, the identified hub genes were validated for differential expression in cancerous versus normal tissues, suggesting their potential diagnostic utility, while subsequent survival analyses confirmed their prognostic value through significant associations with overall survival. Notably, these hub genes were found to be significantly associated with immune infiltration levels, immune microenvironment scores, and immune-related proteins in NSCLC.

Conclusions

This dual-modality investigation establishes miR-486 as a multi-functional biomarker in NSCLC, demonstrating both diagnostic utility and immunoregulatory potential through tumor microenvironment modulation.

Introduction

Lung cancer persists as the leading cause of cancer-related mortality worldwide, with an estimated 2.2 million new cases and 1.8 million deaths annually [1]. Among histological subtypes, non-small cell lung cancer (NSCLC) represents approximately 85% of cases, with lung adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC) constituting the predominant subtypes [2]. While therapeutic innovations have expanded treatment options—including precision radiotherapy, molecularly targeted agents, and immune checkpoint inhibitors, the five-year survival rate for NSCLC remains suboptimal at 15–20% across most populations [3]. This poor prognosis is largely attributable to late-stage diagnosis, with over 70% of patients presenting with locally advanced or metastatic disease at initial detection [4]. Current diagnostic approaches, ranging from low-dose CT screening to invasive biopsy procedures, face substantial limitations in clinical implementation due to radiation exposure risks, procedural complications, and suboptimal predictive values. These diagnostic challenges underscore the urgent need for minimally invasive, highly reliable biomarkers that can facilitate early detection and improve risk stratification in NSCLC management.

MicroRNAs (miRNAs) represent a class of small non-coding RNA molecules, typically 18–25 nucleotides in length, that play crucial regulatory roles in cellular processes through sequence-specific interactions with target mRNAs [5]. These molecules primarily bind to complementary sequences within the 3’-untranslated regions (3’-UTRs) of target mRNAs, leading to either mRNA degradation or translational repression, thereby fine-tuning gene expression at the post-transcriptional level. Accumulating evidence has established the critical involvement of miRNAs in multiple hallmarks of cancer, including NSCLC pathogenesis, where they regulate key processes such as cell proliferation, apoptosis resistance, and metastatic potential [6]. The clinical utility of miRNAs is further enhanced by their remarkable stability in various biological fluids (serum, plasma, saliva) and tissues, even under extreme conditions such as prolonged storage, multiple freeze–thaw cycles, and varying pH environments [7]. These unique characteristics coupled with their tissue-specific expression patterns and early dysregulation in tumorigenesis-position miRNAs as promising candidates for the development of minimally invasive diagnostic and prognostic assays in NSCLC management.

MicroRNA-486 (miR-486), located at chromosome 8p11.21, has emerged as a crucial regulator in NSCLC pathophysiology [8]. Extensive research has demonstrated that miR-486 modulates multiple oncogenic pathways, exerting pleiotropic effects on cellular processes ranging from apoptosis regulation and cell cycle progression to epithelial-mesenchymal transition (EMT) and chemoresistance development [9]. Clinical investigations have revealed that circulating miR-486 levels show significant correlations with tumor staging and metastatic potential in NSCLC patients, suggesting its utility as a liquid biopsy biomarker [10]. However, the diagnostic performance of miR-486 exhibits considerable heterogeneity across studies, with reported sensitivity ranging from 68 to 89% and specificity varying between 74 and 92% in different patient cohorts. This variability, coupled with incomplete understanding of its molecular targets and regulatory networks in NSCLC pathogenesis, underscores the need for comprehensive mechanistic studies and standardized validation protocols to establish miR-486 as a reliable diagnostic marker.

In this study, through a multi-dimensional approach, we demonstrate that miR-486 exhibits robust diagnostic accuracy in NSCLC detection, with further improvement when combined with complementary biomarkers. Bioinformatics analyses reveal that miR-486 coordinately regulates ten hub genes, which are significantly enriched in immune-related pathways and show strong associations with tumor-infiltrating lymphocytes. These findings establish miR-486 as a multifunctional biomarker with potential clinical applications in NSCLC management.

Methods

Study design

In this comprehensive investigation, we employed a multi-dimensional approach to elucidate the clinical and mechanistic significance of miR-486 in NSCLC. First, we conducted a systematic meta-analysis of miR-486 expression profiles across multiple cohorts to establish its diagnostic potential. Second, we developed and validated a novel miRNA panel incorporating miR-486 to enhance diagnostic accuracy through combinatorial biomarker analysis. Finally, we implemented an integrative bioinformatics framework combining network pharmacology and machine learning algorithms to identify and characterize the molecular pathways through which miR-486 influences NSCLC pathogenesis and progression. The study workflow is depicted in Fig. 1.

Fig. 1

Schematic overview of the study design and analytical workflow

Literature search methodology

A systematic literature retrieval was performed across three major biomedical databases (PubMed, Web of Science, and EMBASE) covering publications from inception through April 04, 2025. The search strategy employed a comprehensive combination of MeSH terms and free-text keywords, including: (“microRNA-486” OR “miR-486” OR “miRNA-486”) AND (“Lung cancer” OR “Pulmonary carcinoma” OR “Lung malignancy” OR “Carcinoma, Non-Small-Cell Lung” OR “NSCLC”). The detailed search terms were provided in Supplementary Table S1. To ensure literature saturation, we implemented a snowball search strategy by manually screening reference lists of included articles and relevant review papers. All retrieved records were managed using EndNote X8 software, with duplicate removal performed prior to screening.

Eligibility criteria

Studies were considered eligible if they met the following requirements: (1) studies were required to investigate the diagnostic performance of miR-486 in NSCLC using human clinical specimens, specifically peripheral blood derivatives (serum or plasma collected using standardized protocols) or properly preserved sputum samples; (2) all enrolled patients needed histopathological confirmation of NSCLC diagnosis according to current WHO classification standards, with no restrictions applied to disease stage or histological subtype; (3) included studies had to report complete diagnostic accuracy data, including sensitivity and specificity values along with the raw data necessary to reconstruct 2 × 2 contingency tables (true positives, false positives, true negatives, and false negatives).

Studies were excluded based on the following considerations: (1) non-primary research articles (reviews, meta-analyses, editorials, conference abstracts) or preclinical studies (cell lines, animal models); (2) publications with overlapping patient cohorts or duplicate datasets; (3) studies lacked essential methodological details about sample processing, analysis procedures; (4) studies provided insufficient data for meaningful quantitative synthesis; (5) non-English studies.

Data extraction and quality assessment

Two independent investigators (JY and YS) performed blinded data extraction and methodological quality evaluation using a standardized protocol. Any discrepancies were resolved through iterative discussion and, when necessary, adjudication by a senior researcher (YX). The extraction protocol encompassed four domains: (1) study identification: first author’s name, publication year, and country of origin; (2) population characteristics: ethnicity, cohort size, age distribution, and TNM staging according to AJCC guidelines; (3) methodological parameters: biospecimen type (serum/plasma/sputum), RNA isolation protocol, quantification method, and normalization strategy; (4) Diagnostic performance: area under the receiver operating characteristic curve, sensitivity, specificity, and derived 2 × 2 contingency table data (true positives, false positives, true negatives, false negatives).

Study quality was rigorously evaluated using the QUADAS-2 instrument, which assesses risk of bias across four domains: patient selection, index test, reference standard, and flow/timing. Each domain was rated as low, high, or unclear risk of bias based on predefined criteria [11].

Statistical analysis framework

The diagnostic accuracy of miR-486 was evaluated through comprehensive meta-analytic methods. We calculated pooled estimates of sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR), each accompanied by their 95% confidence intervals. Diagnostic performance was further characterized using hierarchical summary receiver operating characteristic (HSROC) modeling, with the area under the curve (AUC) serving as the primary metric of discriminative capacity [12]. Heterogeneity assessment incorporated both Cochran's Q statistic (significance threshold: p < 0.10) and the I² index, with I² values interpreted as follows: ≤ 25% (low heterogeneity), 25–50% (moderate), and > 50% (substantial) [13]. Model selection followed contemporary meta-analytic guidelines: fixed-effects models were applied for homogeneous datasets (I² ≤ 50%), while random-effects models were implemented for heterogeneous datasets (I² > 50%). Threshold effects were examined through Spearman’s rank correlation analysis between sensitivity and false positive rates. For non-threshold effect heterogeneity, we performed stratified subgroup analyses, mixed-effects meta-regression, and leave-one-out sensitivity analyses. Publication bias was assessed using Deeks’ regression asymmetry test, with p < 0.10 indicating significant bias [14]. All statistical tests were two-sided, with p < 0.05 considered statistically significant except where otherwise noted. Analyses were conducted using STATA version 14.0 software.

Identification of the regulated genes by miR-486

Experimentally validated targets of miR-486 were retrieved from miRTarBase, a curated repository of miRNA-mRNA interactions supported by rigorous experimental evidence including reporter assays, western blot, and next-generation sequencing data [15]. To ensure biological relevance, we specifically focused on human miRNA-target interactions that were validated through at least two independent experimental methodologies. The resulting interaction network was subsequently subjected to comprehensive functional annotation and pathway enrichment analysis.

Functional annotation and pathway enrichment analysis

To elucidate the biological significance of miR-486-regulated targets, we performed comprehensive functional annotation using Gene Ontology (GO) and pathway analysis through the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [16, 17]. The GO analysis systematically categorized target genes into three ontological domains: biological processes (focusing on cellular pathways and systems-level functions), cellular components (emphasizing subcellular localization), and molecular functions (highlighting biochemical activities). Pathway enrichment analysis was conducted to identify significantly perturbed signaling cascades and molecular networks. These analyses were implemented using the DAVID bioinformatics platform (version 6.8), with statistical significance determined by a modified Fisher’s exact test (EASE score) at a threshold of p < 0.05 after Benjamini–Hochberg correction for multiple testing [18]. A P-value of less than 0.05 was considered to indicate statistical significance.

Protein–protein interaction network construction and analysis

To systematically investigate the molecular mechanisms mediated by miR-486, we constructed a comprehensive protein–protein interaction (PPI) network using an integrative bioinformatics approach [19]. The network was generated through the STRING database, incorporating multiple evidence channels including experimental data, curated databases, co-expression patterns, and computational predictions [20]. A confidence threshold of 0.4 (medium confidence) was applied to ensure biologically relevant interactions while maintaining network specificity. The resulting PPI network was visualized and analyzed using Cytoscape [21]. Network topology analysis was performed using CytoNCA to identify hub proteins based on multiple centrality measures (degree, betweenness, and closeness centrality) while functional modules within the network were detected using MCODE [22, 23]. Statistical significance was determined at p < 0.05, with Benjamini–Hochberg correction for multiple testing.

Hub gene expression profiling and clinical correlation analysis

Differential expression analysis of identified hub genes was performed using the Kruskal–Wallis test on TCGA-LUAD and TCGA-LUSC datasets, with Benjamini–Hochberg correction for multiple comparisons [24]. Results were visualized using customized boxplots generated with the ggplot2 package (3.4.4) in R. To assess diagnostic potential, we constructed ROC curves using RNA-seq expression data from TCGA, calculating AUC values with 95% confidence intervals. Prognostic significance was evaluated through the GEPIA2 platform (http://gepia2.cancer-pku.cn), employing Kaplan–Meier survival analysis with log-rank testing [25]. Hazard ratios (HR) with 95% confidence intervals were calculated for overall survival (OS) and disease-free survival (DFS) endpoints. To assess whether hub gene serves as an independent prognostic factor, we performed univariate and multivariate Cox regression analyses on hub gene expression and additional clinical variables using the “survival” package (3.3.1). Variables demonstrating statistical significance (P < 0.05) in the univariate analysis were subsequently incorporated into a multivariate Cox proportional hazards model. To facilitate prognostic prediction, a nomogram was developed utilizing the “RMS” R package (6.3–0). Calibration curves were then plotted to evaluate the concordance between the predicted probabilities generated by the model and the observed outcomes for 1-year, 3-year and 5-year OS under varying conditions. This approach provides a robust framework for validating the predictive accuracy of the nomogram in clinical settings.

Immune landscape characterization

The immune landscape associated with hub gene expression was interrogated through a multi-modal analytical framework. Single-sample gene set enrichment analysis (ssGSEA) was performed using the “GSVA” package (3.21) with 24 precisely defined immune cell gene signatures from Bindea et al.’s immunome database, applying 1000 permutations for significance testing [26]. The ESTIMATE algorithm was implemented with default parameters to deconvolute stromal (stromal score), immune (immune score), and combined tumor purity indices (ESTIMATE score), with batch effects corrected using ComBat. For immune modulator analysis, we curated five functional categories of molecular features: (1) immunostimulating genes, (2) immunosuppressive genes, (3) chemokines, (4) chemokine receptors, and (5) major histocompatibility complex (MHC) molecules. Spearman correlations were computed using the “stats” package, retaining interactions meeting p < 0.05 thresholds. All visualizations were generated using “ggplot2” (3.4.4) with consistent color mapping and clustering parameters.

Results

Study selection and characteristics

The systematic literature selection process is detailed in Fig. 2, following PRISMA guidelines. From an initial pool of 251 potentially relevant articles, 13 high-quality publications (comprising 14 independent studies) met our stringent inclusion criteria after rigorous screening [27,28,29,30,31,32,33,34,35,36,37,38,39].

Fig. 2

The flowchart of database search and study identification

Table 1 summarizes the key characteristics of included studies. The selected publications represent a decade of research (2010–2020) across diverse geographical regions: USA (n = 6), China (n = 5), Italy (n = 1), Poland (n = 1), and Egypt (n = 1). Study populations encompassed multiple ethnicities: Caucasian (n = 3), solely African (n = 1), mixed Caucasian/African (n = 5), and Asian (n = 5).

Table 1 The main features of the included studies for miR-486 in the diagnosis of NSCLC

Biospecimen types included plasma (n = 8), serum (n = 3), sputum (n = 2), and tissue (n = 1). All studies employed quantitative real-time PCR (qRT-PCR) for miR-486 quantification, with normalization to endogenous controls. Quality assessment using QUADAS-2 revealed that 85% of studies (n = 11) scored ≥ 6, indicating robust methodological quality.

Diagnostic accuracy of miR-486

Substantial heterogeneity was observed across studies, with I² values of 73.52% (95% CI: 59.47–87.57%) for sensitivity and 90.69% (86.99–94.39%) for specificity, necessitating the use of a random-effects model for meta-analysis.

Pooled analysis (Fig. 3A) of 14 studies revealed a diagnostic sensitivity of 0.78 (95% CI: 0.72–0.83) and specificity of 0.81 (95% CI: 0.70–0.89). The PLR was 4.1 (2.4–6.8), indicating moderate ability to rule in diagnosis, while the NLR was 0.27 (0.20–0.36), suggesting good ability to rule out disease. The DOR was 15 (7–32), reflecting overall diagnostic accuracy.

Fig. 3

Forest plots for pooled results in diagnosing NSCLC for sensitivity and specificity. A Single miR-486; B Combination marker containing miR-486

HSROC analysis (Fig. 4A) yielded an AUC of 0.85 (95% CI: 0.81–0.88), indicating good discriminative ability. The 95% confidence region and prediction region of the HSROC curve suggested consistent diagnostic performance across studies.

Fig. 4

SROC curve with confidence and prediction regions around mean operating sensitivity and specificity point. A SROC curve overall including the outliers for single miR-486; (B) SROC curve of circulating single miR-486; (C) SROC curve of single miR-486 based on large sample size; (D) SROC curve of single miR-486 based on small sample size; (E) SROC curve of outliers excluded for single miR-486; (F) SROC curve of combination biomarkers containing miR-486

Heterogeneity exploration and subgroup analysis

Threshold effect analysis revealed no significant correlation (Spearman’s ρ = 0.33, p = 0.11), prompting further investigation through subgroup stratification (Table 2) and mixed-effects meta-regression.

Table 2 Summary table of the value of miR-486 in the diagnosis of NSCLC

Ethnicity-based subgroup analysis demonstrated superior diagnostic performance in Asian populations (sensitivity: 0.80 [0.73–0.85]; specificity: 0.85 [0.59–0.96]; AUC: 0.84 [0.80–0.87]) compared to Caucasian (sensitivity: 0.74 [0.67–0.80]; specificity: 0.75 [0.65–0.83]; AUC: 0.82 [0.78–0.86]) and mixed Caucasian/African populations (sensitivity: 0.71 [0.64–0.78]; specificity: 0.73 [0.65–0.79]; AUC: 0.76 [0.72–0.79]).

Biospecimen-specific analysis indicated enhanced diagnostic accuracy in plasma-based studies (sensitivity: 0.78 [0.73–0.83]; specificity: 0.88 [0.74–0.95]; AUC: 0.84 [0.81–0.87]) relative to serum-based investigations (sensitivity: 0.75 [0.67–0.82]; specificity: 0.60 [0.53–0.66]; AUC: 0.70 [0.62–0.78]). In addition, analysis of circulating miR-486 in 11 studies (Fig. 4B) revealed pooled sensitivity of 0.77 [0.73–0.81], specificity of 0.81 [0.67–0.90], and AUC of 0.82 [0.78–0.85].

Study size stratification showed improved specificity in larger cohorts (≥ 100 subjects: specificity = 0.87 [0.73–0.95]; AUC = 0.88 [0.84–0.90], Fig. 4C) compared to smaller studies (< 100 subjects: specificity = 0.67 [0.58–0.75]; AUC = 0.80 [0.76–0.83], Fig. 4D).

Meta-regression analysis identified no significant associations between heterogeneity and ethnicity (p > 0.05), sample size (p > 0.05), or biospecimen type (p > 0.05).

These findings highlight the necessity for large-scale, multi-ethnic validation studies to establish robust diagnostic thresholds for miR-486 in NSCLC detection.

Robustness assessment and publication bias evaluation

The stability of our meta-analytic findings was rigorously evaluated through leave-one-out sensitivity analysis. The random-effects bivariate model demonstrated excellent robustness, as evidenced by goodness-of-fit statistics and bivariate normality tests (Fig. 5).

Fig. 5

Sensitivity analysis results

Influence analysis identified two potential outliers through standardized residual diagnostics. Exclusion of these studies resulted in minimal changes to pooled estimates: sensitivity (0.78 → 0.74), specificity (0.81 → 0.74), PLR (4.1 → 2.9), NLR (0.27 → 0.35), DOR (15 → 8), and AUC (0.85 → 0.79). The marginal impact of outlier removal on diagnostic parameters confirmed the robustness of our primary findings. The corresponding HSROC curve for the reduced dataset is presented in Fig. 4E.

Publication bias assessment using Deek’s funnel plot asymmetry test revealed significant bias (p = 0.01), suggesting potential underrepresentation of negative or null findings in the literature. This observation underscores the need for prospective registration of diagnostic accuracy studies to mitigate publication bias in future meta-analyses.

Diagnostic performance of miR-486-based combination biomarkers

We delved further into assessing the diagnostic accuracy of combinations of biomarkers associated with miR-486. From an initial pool of 17 relevant studies [27,28,29, 32, 34, 36,37,38,39,40,41,42], we identified robust evidence supporting the enhanced diagnostic capability of combination biomarkers over miR-486 alone. Key study characteristics are summarized in Table 3, with diagnostic performance metrics including sensitivity and specificity illustrated in Fig. 3B.

Table 3 The main features of the enrolled studies regarding combination markers containing miR-486 in diagnosing NSCLC

Pooled analysis demonstrated superior diagnostic performance for combination biomarkers, with sensitivity of 0.83 (95% CI: 0.79–0.87), specificity of 0.91 (0.87–0.94), and AUC of 0.92 (0.90–0.94). The SROC curve was presented at Fig. 4F. The PLR was 9.4 (5.9–14.8), indicating strong ability to rule in diagnosis, while the NLR was 0.19 (0.14–0.24), suggesting excellent ability to rule out disease. The DOR was 51 (25–100), reflecting substantially improved diagnostic accuracy compared to single-marker approaches.

These findings suggest that miR-486-based biomarker panels offer enhanced diagnostic precision for NSCLC detection, potentially facilitating earlier and more accurate clinical decision-making.

Functional annotation and pathway enrichment of miR-486 targets

GO analysis of miR-486 target genes revealed distinct functional patterns across biological processes, cellular components, and molecular functions. At the biological process level, target genes were significantly enriched in transcriptional regulation, epithelial-mesenchymal transition, cellular proliferation, and TGF-β signaling (Fig. 6A). Cellular component analysis demonstrated predominant localization in nuclear compartments, and nuclear matrix (Fig. 6B). Molecular function analysis identified enrichment in protein and protein interactions, kinase activity, and chromatin remodeling (Fig. 6C).

Fig. 6

Functional enrichment analysis results. (A) Top 10 of the most significantly enriched GO items at the biological process level; (B) Top 10 of the most significantly enriched GO items at the cellular component level; (C) Top 10 of the most significantly enriched GO items at the molecular function level; (D) The significant pathways enriched by all the genes regulated by miR-486

KEGG pathway analysis identified 22 significantly enriched pathways, with particular relevance to cancer biology and metabolic regulation (Fig. 6D). Key pathways included transcriptional dysregulation in cancer, cancer metabolism, hypoxia response (HIF-1 signaling), tumor suppression (p53 signaling), and receptor tyrosine kinase signaling (ErbB pathway). Notably, the NSCLC pathway emerged as a central hub of miR-486 activity, suggesting direct involvement in lung cancer pathogenesis.

Protein–protein interaction network analysis

The PPI network was constructed using confidence interactions (score > 0.4) from the STRING database, yielding 190 nodes and 392 edges. Network topology analysis using CytoNCA identified ten hub proteins based on multiple centrality measures: CDKN1 A, ERBB2, ITGB3, PIK3R1, PTEN, RAF1, SMAD2, SNAI1, SYK, and UBB (Fig. 7A-C). A focused subnetwork comprising these hub genes and their first-order interactors was reconstructed (Fig. 7D). miR-486 and its ten target genes along with their corresponding regulatory relationships are detailed in Supplementary Table S2.

Fig. 7

PPI network construction results. A Betweenness centrality distributions of nodes; (B) Closeness centrality distributions of nodes; (C) Degree distributions of nodes; (D) The sub-network reconstructed with the selected hub proteins and their first neighbor proteins; (E) Top 20 pathways enriched by the 10 hub genes of miR-486

Functional annotation of hub genes revealed significant enrichment in cancer-related pathways, including FoxO signaling, PI3K-Akt signaling, and NSCLC-specific pathways (Fig. 7E). Notably, the hub genes were prominently involved in receptor tyrosine kinase signaling (ErbB pathway), metabolic reprogramming (central carbon metabolism:) and immune-related pathways (B cell receptor signaling pathway).

Module analysis using MCODE identified a critical subnetwork (7 nodes, 20 edges) significantly enriched in cancer hallmark pathways, including cell cycle regulation, p53 signaling, and PI3K-Akt-mTOR signaling (Supplementary Table S3). This module analysis reinforced the central role of miR-486 targets in NSCLC pathogenesis through multiple interconnected signaling axes.

Expression and diagnosis value of hub genes in NSCLC

Differential expression analysis of hub genes in TCGA datasets revealed significant dysregulation in both LUAD and LUSC compared to unpaired adjacent tissues and pared adjacent tissues (Fig. 8). The majority of hub genes showed consistent expression patterns across histological subtypes, suggesting their fundamental role in NSCLC pathogenesis.

Fig. 8

The expression of hub genes in LUAD and LUSC compared with unpaired adjacent tissues and pared adjacent tissues

Diagnostic evaluation using ROC curve analysis demonstrated strong discriminative capacity of hub genes in distinguishing LUAD and LUSC patients from healthy controls (Fig. 9). The findings suggested that these hub genes possessed a relatively strong predictive ability for differentiating patients with LUAD and LUSC from healthy individuals.

Fig. 9

Receiver operating characteristic (ROC) curves for hub genes expression in LUAD and LUSC based on TCGA database

Survival analysis of hub genes in NSCLC patients

Figure 10 presents the survival analysis results of hub genes in NSCLC patients, focusing on both LUAD and LUSC. The analysis includes OS and DFS outcomes. The survival maps of hub gene expression in LUAD were illustrated, highlighting the association between gene expression levels and OS (Fig. 10A) and DFS (Fig. 10B). Similarly, the survival maps for LUSC were depicted, showing the correlation between hub gene expression and OS (Fig. 10C) and DFS (Fig. 10D). These maps provide a comprehensive overview of the prognostic significance of various hub genes in both subtypes of NSCLC.

Fig. 10

Survival analysis results of the hub genes in NSCLC patients. A-B Survival maps of hub genes expression in LUAD: OS (A), and DFS (B); (C-D) Survival maps of hub genes expression in LUSC: OS (C), and DFS (D); (C-D) Kaplan–Meier survival curves of high and low expression of SNAI1 in LUAD and LUSC: OS in LUAD (E), DFS in LUAD (F), OS in LUSC (G), and DFS in LUSC (H)

The Kaplan–Meier survival curves demonstrate the impact of SNAI1 expression on survival outcomes in LUAD and LUSC. In LUAD, the high expression of SNAI1 is associated with significantly worse OS (Fig. 10E) and DFS (Fig. 10F) compared to low expression. Similarly, in LUSC, high SNAI1 expression correlates with poorer OS (Fig. 10G) and DFS (Fig. 10H).

These results underscore the prognostic value of SNAI1 in both LUAD and LUSC, suggesting its potential role as a biomarker for survival outcomes in NSCLC patients. The findings from this survival analysis highlight the critical role of hub genes, particularly SNAI1, in predicting survival outcomes in NSCLC. The differential expression of these genes in LUAD and LUSC provides valuable insights into their prognostic significance and potential therapeutic targets.

Prognostic significance of SNAI1 in NSCLC

We further explored the prognostic significance of the pivotal hub gene, SNAI1, in patients with NSCLC, focusing on both LUAD and LUSC. The analysis includes univariate and multivariate Cox regression models, as well as predictive nomograms and calibration curves.

Figure 11A-D illustrated the prognostic significance of SNAI1 in LUAD. Univariate Cox regression analysis (Fig. 11A) reveals that high expression of SNAI1 is significantly associated with worse OS (HR = 1.440, 95% CI: 1.070–1.923, p = 0.013). Multivariate Cox regression analysis (Fig. 11B) further substantiates this relationship, adjusting for additional variables such as pathologic T stage, N stage, and M stage, confirming SNAI1 as an independent prognostic factor (HR = 1.439, 95% CI: 1.010–2.030, p = 0.039). A predictive nomogram (Fig. 11C) was constructed using TCGA-LUAD datasets, integrating SNAI1 expression levels with other prognostic indicators to estimate survival probabilities in LUAD. The calibration curve (Fig. 11D) associated with the nomogram demonstrates good concordance between predicted and observed survival outcomes, validating the accuracy of the model.

Fig. 11

Survival analysis outcomes for the pivotal hub gene, SNAI1, in patients with NSCLC. (A-D) The prognostic significance of SNAI1 in LUAD is depicted through various analytical approaches: A Univariate Cox regression analysis reveals the influence of SNAI1 expression on OS in LUAD; (B) Multivariate Cox regression further substantiates this relationship, adjusting for additional variables; (C) A predictive nomogram, constructed from TCGA datasets, combines SNAI1 levels with other prognostic indicators to estimate survival probabilities in LUAD; (D) The calibration curve associated with the nomogram evaluates the accuracy of survival predictions. (E–H) The prognostic relevance of SNAI1 in LUSC: E Univariate Cox regression analysis assesses the effect of SNAI1 expression on OS in LUSC; (F) Multivariate analysis corroborates these findings, accounting for other prognostic factors; (G) Another nomogram, derived from TCGA data, integrates SNAI1 expression with additional prognostic markers to forecast survival outcomes in LUSC; (H) The corresponding calibration curve assesses the precision of the nomogram’s predictive capability

Similar results also highlighted the prognostic relevance of SNAI1 in LUSC. Univariate Cox regression analysis (Fig. 11E) indicates that high SNAI1 expression is significantly associated with poorer OS (HR = 1.637, 95% CI: 1.245–2.152, p < 0.001). Multivariate analysis (Fig. 11F) corroborates these findings, adjusting for other prognostic factors, and confirms the independent prognostic value of SNAI1 in LUSC (HR = 1.659, 95% CI: 1.228–2.241, p < 0.001). Another nomogram (Fig. 11G) was developed using TCGA data, incorporating SNAI1 expression with additional prognostic markers to predict survival outcomes in LUSC. The corresponding calibration curve (Fig. 11H) also evaluates the precision of the nomogram, showing strong agreement between predicted and actual survival probabilities.

The findings from this comprehensive survival analysis underscore the critical role of SNAI1 as a prognostic marker in both LUAD and LUSC. The integration of SNAI1 expression into predictive nomograms provides a robust tool for estimating survival probabilities, offering valuable insights for personalized therapeutic strategies in NSCLC patients.

Immune cell infiltration patterns

Given the pivotal role of tumor-infiltrating immune cells in shaping immunotherapy responses, we employed ssGSEA to quantify immune cell infiltration levels and their association with hub gene expression. Significant correlations (p < 0.05) were observed between hub gene expression profiles and infiltration levels of multiple immune cell subsets in LUAD and LUSC (Fig. 12). These findings suggest that hub genes may modulate the immune landscape of LUAD and LUSC through distinct cellular mechanisms. These findings provide new insights into understanding the tumor immune microenvironment and may contribute to the development of more effective immunotherapy strategies.

Fig. 12

The correlation between the infiltration of immune cells and the expression of the hub genes in LUAD and LUSC

Tumor microenvironment characterization

The tumor microenvironment, comprising stromal and immune components, plays a crucial role in tumor progression. Using ESTIMATE algorithm-derived scores, we identified significant associations (p < 0.05) between hub gene expression and stromal, immune, and combined ESTIMATE scores (Fig. 13). These correlations indicate that hub genes may influence the composition and functional state of the LUAD and LUSC microenvironment.

Fig. 13

Correlations between the hub genes expression and immune microenvironment scores in LUAD and LUSC including stromal score, immune score, and estimate score

Immune modulator interactions

To elucidate potential mechanisms underlying immune modulation, we analyzed correlations between hub gene expression and five categories of immune-related proteins: (1) immunostimulating genes, (2) immunosuppressive genes, (3) chemokines, (4) chemokine receptors, and (5) MHC molecules. Hierarchical clustering analysis revealed distinct patterns of association (p < 0.05), suggesting that these hub genes may play a role in modulating the immune response within the tumor microenvironment (Figs. 14 and 15).

Fig. 14

Correlation of hub genes expression with immune regulator expression in LUAD and LUSC

Fig. 15

Correlation of hub genes expression with immune related genes in LUAD and LUSC

Discussion

The persistently high mortality rates and poor survival outcomes in NSCLC underscore the critical need for innovative, minimally invasive biomarkers to facilitate early detection. While molecular diagnostics have advanced significantly, current approaches for NSCLC screening remain limited by procedural invasiveness, patient compliance issues, and technical constraints in sample acquisition. Emerging evidence positions miR-486 as a promising diagnostic biomarker, yet its clinical translation has been hampered by methodological heterogeneity and inconsistent validation across studies. These discrepancies stem from variations in study design, sample collection protocols, and statistical power limitations. To address these challenges, we implemented a comprehensive two-phase investigation: (1) a systematic meta-analysis to establish the diagnostic performance of miR-486 across diverse clinical settings, and (2) an integrative bioinformatics approach to elucidate the molecular networks and biological pathways underlying miR-486’s biomarker potential in NSCLC.

Our systematic review incorporated data from 14 independent studies, revealing pooled diagnostic parameters for miR-486 in NSCLC detection: sensitivity = 0.78 (95% CI: 0.72–0.83), specificity = 0.81 (0.70–0.89), and AUC = 0.85 (0.81–0.88). Substantial heterogeneity was observed (I² = 73.52% for sensitivity; I² = 90.69% for specificity), prompting comprehensive subgroup analyses. While meta-regression failed to identify significant moderators, stratification by sample size revealed enhanced diagnostic accuracy in larger cohorts (AUC = 0.88 [0.84–0.90]) compared to smaller studies (AUC = 0.80 [0.76–0.83]). Ethnicity-based analysis demonstrated superior performance in Asian populations (AUC = 0.84 [0.80–0.87]) relative to Caucasian (AUC = 0.82 [0.78–0.86]) and mixed populations (AUC = 0.76 [0.72–0.79]), aligning with our prior findings that genetic ethnicity may influence the diagnostic performance of miR-486 [43]. Notably, plasma-based assays showed improved diagnostic accuracy (AUC = 0.84 [0.81–0.87]) compared to serum-based approaches, potentially reflecting differential miRNA stability across sample types (AUC = 0.70 [0.62–0.78]). These findings highlight the need for standardized protocols and large-scale, multi-ethnic validation studies to establish robust diagnostic thresholds for miR-486 in NSCLC detection.

Our investigation extended to evaluating the diagnostic potential of miR-486-containing biomarker panels for NSCLC detection. The results demonstrated superior diagnostic accuracy for combination panels (AUC = 0.92 [0.90–0.94]) compared to miR-486 alone (AUC = 0.85 [0.81–0.88]), consistent with previous reports of enhanced performance in multi-analyte miRNA profiling [44, 45]. This finding suggests that miR-486 functions synergistically with other miRNAs within diagnostic panels, potentially through complementary molecular pathways. The observed improvement in diagnostic performance underscores the clinical potential of multi-miRNA panels for NSCLC screening. However, the optimal composition of such panels requires further validation through large-scale, prospective studies to establish standardized miRNA combinations and diagnostic thresholds.

When contextualized against canonical NSCLC biomarkers, miR-486 demonstrates some advantages and limitations. Compared to miR-21, the most extensively validated NSCLC miRNA biomarker, miR-486 exhibits similar diagnostic power (sensitivity: 0.78 vs. 0.77; specificity: 0.81 vs. 0.76; AUC: 0.85 vs. 0.87) for early-stage detection in our meta-analysis, likely due to its more restricted expression pattern in normal tissues [46]. However, miR-486 lags behind these biomarkers in clinical translation, with fewer large-scale validation studies and more complex normalization requirements due to lower basal expression levels. Crucially, our combinatorial analyses suggest miR-486 synergizes with these biomarkers; when paired with miR-21 in diagnostic panels, AUC improves from 0.85 (single) to 0.92 (combined), outperforming either marker alone. This positions miR-486 as a complementary rather than competing biomarker, potentially enhancing existing miRNA-based diagnostic frameworks while providing unique immune microenvironment insights.

Since miRNAs primarily function by regulating downstream target genes, the biological roles of these target genes can, to some extent, reflect the functions of the miRNAs themselves. To elucidate the molecular mechanisms underlying miR-486’s biological functions, we conducted comprehensive functional enrichment analysis of its predicted target genes. GO analysis revealed significant enrichment in critical biological processes, including transcriptional regulation, epithelial-mesenchymal transition, cellular proliferation, and TGF-β signaling pathways. Cellular component analysis demonstrated predominant localization of miR-486 targets in fundamental cellular structures: nucleus, nucleoplasm, cytoplasm, and nuclear matrix. Molecular function annotation identified enrichment in protein interaction domains, particularly protein binding, enzyme regulation, kinase activity, and chromatin remodeling. These findings suggest that miR-486 orchestrates a complex regulatory network influencing multiple aspects of cellular function and architecture, potentially contributing to its role in NSCLC pathogenesis.

KEGG pathway enrichment analysis identified 22 signaling pathways significantly associated with miR-486 target genes, many of which are central to NSCLC pathogenesis. The direct inclusion of the NSCLC pathway in these findings strongly implicates miR-486 in lung cancer development. Among these pathways, transcriptional regulation emerged as a key process, with its dysregulation linked to cancer stem cell emergence and tumorigenesis through enhanced self-renewal and impaired differentiation [47]. Metabolic reprogramming, particularly through central carbon metabolism, has become a critical area in cancer research, with clinical applications in monitoring disease progression and therapeutic response [48]. The HIF-1 signaling pathway, a major regulator of cancer cell metabolism and metastasis, has been consistently associated with tumor development through its dysregulated expression [49]. Similarly, the p53 pathway, one of the most extensively studied tumor suppressor networks, is nearly universally altered in cancer and represents a promising therapeutic target in NSCLC [50]. Numerous studies have established that alterations in the p53 pathway are pivotal to the varying clinical manifestations of NSCLC. Targeting this signaling cascade holds the potential to emerge as a promising therapeutic approach for the treatment of NSCLC [51]. Proteoglycans have emerged as key macromolecules that play critical roles in cellular development, human diseases, and the pathogenesis of malignancies. These molecules are known to influence the biological activities of cancer cells and their microenvironment, significantly impacting the initiation and progression of both solid tumors and hematopoietic malignancies [52]. The FoxO signaling pathway, crucial for regulating genes involved in diverse biological processes, has been implicated in NSCLC development through its dysregulation [53]. Pathways in cancer encompass array of renowned signaling cascades that significantly contribute to processes such as apoptosis, proliferation, differentiation, invasion, and metastasis [54]. Additionally, the ErbB signaling pathway, a well-established driver of solid tumors, has been successfully targeted in several cancers, including NSCLC [55, 56]. These findings collectively position miR-486 as a central regulator of multiple oncogenic pathways, influencing key aspects of NSCLC development and progression through diverse molecular mechanisms.

To elucidate the molecular mechanisms of miR-486, we performed PPI network analysis, identifying ten hub genes with the highest interaction scores. Pathway enrichment analysis of these genes revealed significant associations with NSCLC pathogenesis, consistent with previous experimental findings. While many of these pathways have been extensively studied, it is noteworthy that all ten hub genes are directly implicated in NSCLC development. Among these, the PI3 K-Akt signaling pathway emerged as a central regulator of multiple oncogenic processes, including cell proliferation, apoptosis, metastasis, stemness, immune modulation, and drug resistance [57]. Furthermore, the involvement of miR-486 in cancer-related microRNA networks highlights its role in tumor initiation and progression. The B-cell receptor signaling pathway, crucial for B-cell survival and activation, has been shown to influence adaptive immune responses and tumor-infiltrating B-cell populations, potentially impacting recurrence rates in early-stage lung adenocarcinoma following surgical intervention [58]. Earlier research has demonstrated that the B-cell receptor signaling pathway affects the intensity of human adaptive immune responses, as well as the number of tumor-infiltrating B cells. This, in turn, may enhance the prevention of recurrence of early-stage lung adenocarcinoma following aggressive surgical resection [59]. These findings collectively validate the ten identified hub genes as critical components of the NSCLC-associated PPI network, playing pivotal roles in disease pathogenesis and offering potential targets for therapeutic intervention.

The enrichment of miR-486-regulated genes in glioma-related pathways underscores both the multifaceted nature of oncogenic processes and the complexity of gene regulatory networks. This cross-cancer pattern likely reflects: (1) the conservation of core oncogenic pathways (e.g., PI3 K-AKT, HIF-1 signaling) across malignancies, where identical genetic circuits are co-opted in different cellular contexts; and (2) the pleiotropic nature of miR-486’s regulation, which may coordinate fundamental processes like metabolic reprogramming and microenvironment remodeling that transcend organ-specific boundaries. Notably, while these shared pathways highlight potential pan-cancer applications of miR-486 modulation, they also emphasize the critical importance of cellular context—the same molecular pathway may drive gliomagenesis through glial cell-specific mechanisms while promoting NSCLC progression via epithelial-mesenchymal crosstalk. This biological nuance reinforces the need for tissue-specific therapeutic development even when targeting apparently “universal” cancer pathways.

Module analysis identified a highly interconnected functional module consisting of seven nodes, five of which were previously characterized as hub genes. Pathway enrichment analysis of this module revealed significant associations with critical oncogenic pathways, including cell cycle regulation and NSCLC-specific signaling cascades, which are central to tumor initiation and progression. Notably, the cell cycle pathway, a well-established regulator of cellular proliferation, differentiation, and apoptosis, emerged as a key component of this module. Dysregulation of this pathway contributes to uncontrolled cell division in NSCLC, making it a promising target for therapeutic intervention [60]. These findings not only deepen our understanding of miR-486's role in NSCLC pathogenesis but also highlight potential therapeutic targets for the development of novel treatment strategies.

Notably, the identified hub genes exhibited significant differential expression between tumor and normal tissues in LUAD and LUSC, suggesting their potential diagnostic utility, while subsequent survival analyses confirmed their prognostic value through significant associations with overall survival. Notably, our study identifies SNAI1 as a pivotal prognostic hub gene within the miR-486 regulatory network. The strong inverse correlation between miR-486 and SNAI1 expression, coupled with SNAI1’s independent prognostic value in multivariate analysis, suggests its dual role as both a mediator of miR-486’s effects and a standalone prognostic indicator.

Our integrated analysis reveals that miR-486 orchestrates immune infiltration in NSCLC through multi-layered regulatory mechanisms. The identified hub genes (e.g., SNAI1, PIK3R1, and ERBB2) form a coordinated network that intersects with critical immune-modulatory pathways. Specifically, miR-486 likely influences the tumor immune microenvironment by regulating the hub genes, which showed strong correlations with immune infiltration levels, tumor microenvironment scores, and immune-related protein expression in LUAD and LUSC, suggesting their involvement in modulating the tumor immune landscape. The correlation analysis between hub genes expression and immune-related genes in LUAD and LUSC underscores the complex relationship between tumor genetics and the immune microenvironment. The significant associations observed suggest that these hub genes regulated by miR-486 may influence immune regulation and response, providing potential targets for immunotherapy in NSCLC. Further studies are warranted to explore the mechanistic roles of these genes regulated by miR-486 in immune modulation and their potential as biomarkers for immune-based therapies. These findings also position miR-486 as a key regulator of tumor immunity, with potential applications as a predictive biomarker for immunotherapy response. Our results provide new insights into the molecular mechanisms underlying immune regulation of miR-486 in NSCLC and may inform the development of innovative immunotherapeutic strategies targeting miR-486-mediated pathways.

While our study provides valuable insights into the diagnostic potential of miR-486 in NSCLC, several limitations should be acknowledged. First, the heterogeneity in cut-off values across included studies highlights the need for standardized protocols, including consensus normalization controls and diagnostic thresholds, to minimize technical variability and enhance reproducibility. Second, although we conducted subgroup analyses based on available variables, the lack of comprehensive clinical data, particularly regarding histological subtypes and tumor staging, limited our ability to perform stratified analyses across these critical parameters. Moreover, our findings are based on retrospective analyses, underscoring the urgent need for large-scale, prospective, and well-designed clinical trials to validate the diagnostic performance of miR-486 and its combinatorial panels. Finally, while TCGA data provide robust computational validation, the lack of independent experimental validation (e.g., RT-qPCR or IHC in patient samples) represents a limitation of this study. Future studies incorporating such validation would be valuable to confirm our findings.

Despite these limitations, our study establishes miR-486 as a promising non-invasive biomarker for NSCLC detection. The enhanced diagnostic accuracy observed with miR-486-containing panels suggests significant clinical utility in risk stratification and early detection. Furthermore, our integrated approach not only demonstrates the diagnostic potential of miR-486 but also provides mechanistic insights into its role in NSCLC pathogenesis, offering a foundation for future therapeutic development and biomarker optimization. These findings collectively advance our understanding of miR-486 as a multifaceted biomarker with potential applications in both NSCLC diagnosis and molecularly targeted therapies.

Conclusions

Our study establishes miR-486 as a promising non-invasive biomarker for NSCLC, with diagnostic performance enhanced in combinatorial panels. Integrated bioinformatics revealed its regulatory network involving key pathways and immune-modulatory roles, suggesting potential for immunotherapy prediction. Clinical translation warrants further validation through standardized, large-scale prospective studies.