Non-small cell lung cancer (NSCLC) accounts for approximately 85% of all lung tumor cases and is considered one of the leading causes of cancer-related mortality world-wide (Molina et al., 2008). Lung adenocarcinoma (LUAD) represents the most common histological sub-type of NSCLC, around 40%–50% of all lung cancer patients are diagnosed with it (Herbst et al., 2018). In addition, LUAD frequently gives rise to distant metastases, particularly in the brain. Approximately 25% of patients with LUAD develop brain metastases (BMs) during the course of their disease (Lukas et al., 2014). LUAD brain metastases (LUAD-BM) are often associated with severe neurological symptoms and poor prognosis, with a median overall survival of 4–15 months (Sperduto et al., 2020).

From a clinical perspective, LUAD-BMs are often diagnosed synchronously with the primary tumor. Treatment strategies include a multimodal approach combining local therapies—such as stereotactic radiosurgery or whole-brain radiotherapy–with systemic treatments, including tyrosine kinase inhibitors in case of epidermal growth factor receptor (EGFR) mutant NSCLC patients (Le Rhun et al., 2021).

The molecular mechanisms underlying brain-specific metastasis in LUAD remain incompletely understood. However, current evidence suggests that they involve a complex interplay between tumor-intrinsic factors, such as gene expression signatures and microRNA (miRNA) profiles (Najjary et al., 2023). Certain oncogenic drivers, particularly EGFR mutations and anaplastic lymphoma kinase rearrangements, have been associated with a higher incidence of BMs, suggesting that these molecular subtypes may possess intrinsic neurotropic properties (Remon and Besse, 2018).

MiRNAs are short non-coding RNAs (∼22 nucleotides) that regulate gene expression post-transcriptionally, primarily by binding to complementary sequences within the 3′untranslated region of target mRNAs. In cancer, dysregulated miRNA expression contributes to tumor initiation, progression, and metastasis by modulating oncogenes, tumor suppressors, and key signaling pathways (Smolarz et al., 2022). Given that miRNAs regulate a wide range of genes involved in migration, invasion, adhesion, colonization, and epithelial–mesenchymal transition (EMT), they can be considered key molecular regulators within the metastatic cascade. Moreover, several studies have reported that miRNAs may con-tribute to the disruption of the blood–brain barrier and the establishment of pro-metastatic microenvironment (Souza et al., 2023; Solé and Lawrie, 2019; Kim et al., 2018). In LUAD-BM, limited therapeutic options, delayed diagnosis, and difficulties in accurate detection contribute to poor clinical outcomes (Souza et al., 2023; Kim et al., 2018; Tominaga et al., 2015).

Only a limited number of studies to date have identified dysregulated miRNAs in LUAD-BM (Remon et al., 2016; Koh et al., 2024; Zhang L. et al., 2022; Daugaard et al., 2017). Furthermore, it should be considered that expression patterns can vary across racial, ethnic, and geographic groups (Dluzen et al., 2016; Nassar et al., 2017; Rawlings-Goss et al., 2014). Identifying biomarkers that predict early metastatic dissemination and the establishment of a pro-metastatic microenvironment is therefore essential, as such molecules may offer novel therapeutic opportunities to prevent or limit the development of LUAD-BM.

In this study, we focused on characterizing the miRNA landscape in LUAD-BM using intraoperative tissue samples collected at the Department of Neurosurgery, Faculty of Medicine, University of Debrecen. Our objective was to identify specific miRNAs whose altered expression may contribute to metastatic progression and brain colonization, providing insights into the molecular mechanisms underlying LUAD-BM. Differentially expressed (DE) miRNAs between LUAD-BM and normal brain tissues were first identified through high-throughput next-generation sequencing (NGS) and subsequently validated in an extended cohort using real time quantitative polymerase chain reaction (RT-qPCR). To further investigate their functional relevance, pathway enrichment analyses were con-ducted to explore their roles in tumor progression and invasion. Integrating these findings with publicly available LUAD and normal lung miRNA-seq data allowed us to examine expression patterns across the normal lung–primary LUAD–LUAD-BM axis, high-lighting miRNAs potentially involved in metastatic spread and revealing novel candidates for diagnostic or therapeutic targeting.

Lung cancer represents the most prevalent cause for cancer-related mortality world-wide; within this, approximately 85% of all cases are diagnosed with NSCLC (Zhou et al., 2021). A considerable proportion of patients with NSCLC develop BM that significantly worsen prognosis and pose unique therapeutic challenges, especially among patients with LUAD. Response rates to monotherapy with PD-1 or PD-L1 inhibitors among those with BMs are generally low (Kim et al., 2020). It can be explained by the unique immune microenvironment of the central nervous system, restricted drug penetrance across the blood–brain barrier, and the clonal selection and evolution of those groups of tumor cells that tend to migrate to the brain (Tsakonas et al., 2022). Recently, miRNAs have emerged as critical regulators of gene expression and key players in cancer biology. These small, non-coding RNA molecules can influence multiple oncogenic processes, including proliferation, invasion, metastasis, and therapy resistance (Hudson et al., 2024). In this study, we aimed to identify miRNAs that are DE in LUAD-BM tissue samples compared to control brain and primary LUAD samples, with the goal of uncovering candidates potentially involved in the development of BM originating from LUAD. Further-more, we explored their potential utility as biomarkers for predicting the risk of BM in LUAD patients.

KEGG pathway analysis was performed using the 50 most upregulated and 50 most downregulated miRNAs identified by miRNA NGS in the LUAD-BM versus control brain tissue comparison. The predicted target genes of these dysregulated miRNAs showed a strong association with cancer-related pathways, highlighting their potential involvement in tumor progression and metastasis. For example, aberrant expression of focal adhesion molecules that act as mechanosensors mediating bidirectional communication between the cell and its microenvironment could lead to enhanced invasion and migration capacity of the tumor cells (Yayan et al., 2024). It has been shown that the application of integrin inhibitors and E-cadherin up-regulators during therapy leads to a 70% reduction of invasion (Riley et al., 2019). Furthermore, neurotrophin signaling pathways play a significant role in tumor development. Specifically, nerve growth factor and brain-derived neurotrophic factor were reported to stimulate tumor cell proliferation, survival, migration, and/or invasion and favor tumor angiogenesis (Chopin et al., 2016).

Our RT-qPCR validation confirmed that the expression levels of miR-200c-3p, miR-210-3p, miR-10a-5p, miR-130b-3p, miR-146b-5p, miR-503-5p, miR-196b-5p, and miR-3934-5p were significantly upregulated, while miR-138-2-3p and miR-195-5p were significantly downregulated in LUAD-BM tissue samples compared to control brain tis-sue. In addition to identifying dysregulated miRNAs in LUAD-BM tissue samples compared to brain tissue controls, we performed differential expression analysis on 15 normal lung tissue samples, 20 primary LUAD samples–downloaded from the TCGA database and selected from European populations–and our six LUAD-BM samples (Table 4). Furthermore, based on literature data, we analyzed the functions of the target genes of the selected miRNAs, focusing on those previously shown to be associated with migratory processes in LUAD/NSCLC (Table 5; Figure 5).

Based on differential expression analysis conducted on the above-mentioned tissue samples miR-200c-3p expression was significantly decreased, whereas miR-146b-5p and miR-3934-5p expression was significantly increased in LUAD-BM samples compared to primary LUAD samples, while no detectable differences were observed between primary tumors and normal lung tissue. These findings suggest that the expression changes of these miRNAs are specifically associated with the development of secondary brain tumors. The upregulation of miR-146b-5p and miR-3934-5p observed in the LUAD-BM versus primary LUAD comparison is consistent with our findings in the LUAD-BM versus normal brain comparison, where we also detected upregulation.

According to several studies, miR-200c may exert contrasting regulatory effects depending on the cancer type, tumor microenvironment, and even the specific genetic background of the tumor cells (Guo et al., 2025; Klicka et al., 2022). However, in LUAD tissue samples compared to normal lung tissue, miR-200c has generally been identified as a tumor suppressor (Liu et al., 2017). It has been reported that all members of the miR-200 family are highly expressed in lung tissue but are barely detectable in the brain (Teplyuk et al., 2012; Liang et al., 2007). These data are consistent with our results, as our sequencing data indicated that normal brain expresses miR-200c-3p at significantly lower levels than normal lung. This provides an explanation for the observed relative increase of miR-200c-3p in LUAD-BM samples compared to normal brain, despite the downregulation detected in primary LUAD. MiR-200c proved to be involved in EMT in LUAD, which is a hallmark of metastasis that facilitates detachment from the primary tumor, enhanced motility, and invasion. (Gregory et al., 2008). Lei et al. demonstrated that miR-200c exhibits tumor-suppressive effects in NSCLC by inhibiting cell proliferation and migration through targeting lactate dehydrogenase A, representing a potential therapeutic target (Lei et al., 2018). Moreover, miR-200c-3p contributes to EGFR TKI sensitivity by regulating the EMT process (Wang et al., 2020).

Interestingly, both tumor-suppressive and oncogenic roles of miR-146b-5p have been reported in NSCLC, highlighting its context-dependent regulatory function. For example, Sang et al. demonstrated its significantly elevated level in NSCLC tissue samples coupled with worse prognosis. Furthermore, they proved that methyltransferase 16 mediated miR-146b-5p m6A modification led to the induction of cell proliferation and osimertinib resistance via activation of the phosphatidylinositol 3-kinase/AKT signaling pathway, and modification of miR-146b-5p may promote osimertinib resistance (Sang et al., 2025). These data, consistent with our results, support the notion that inhibition of miR-146b-5p overexpression could serve as a potential therapeutic strategy against LUAD-BM and the suppression of chemotherapy resistance.

MiR-3934-5p remains poorly characterized, as it has been rarely investigated in previous studies. Ren et al. highlighted the possible therapeutic potential of miR-3934-5p in A549 cell line, showing that inhibiting miR-3934-5p can restore tumor protein P53 inducible nuclear protein 1 expression and make LUAD cells more responsive to chemotherapeutic agents such as cisplatin. MiR-3934-5p upregulation enhances proliferation, inhibits apoptosis, and confers DDP resistance in lung cancer (Ren et al., 2019). We propose that inhibition of miR-3934-5p could possess possible therapeutic potential in overcoming therapy resistance.

The analysis revealed a significant downregulation of miR-10a-5p in primary LUAD samples compared to normal lung controls, with further decreased expression observed in LUAD-BM relative to primary LUAD. In contrast, miR-210-3p and miR-130b-3p were significantly upregulated in primary LUAD compared to normal lung, and their expression levels increased further when LUAD-BM was compared to primary LUAD. These results suggest that the expression changes of these miRNAs may be closely associated with the progression of metastasis.

The dual role of miR-10a-5p as both a tumor suppressor and an oncogene has been reported in several cancer types, including breast, bladder, ovarian, and gastric cancers (Hu et al., 2019; Nakayama et al., 2013; Gao et al., 2024; Zaravinos et al., 2012; Xiao et al., 2014; Ma et al., 2007; Moriarty et al., 2010; Wang et al., 2015; Singh et al., 2025). Additionally, previous studies have reported that miR-10a-5p acts as a tumor suppressor in gliomas by inhibiting glioma cell proliferation and migration via downregulation of tumor suppressor candidate 7 (Wang et al., 2023). However, only a few studies have investigated the role of miR-10a-5p in NSCLC. Yu et al. proved that upregulation of miR-10a tends to induce metastasis formation of lung cancer through the regulation of the phosphatase and tensin homolog/AKT/mitogen-activated protein kinase signaling pathway (Yan et al., 2015). Furthermore, silencing miR-10a reverses cisplatin resistance in human lung cancer cell lines through the transforming growth factor-beta/Smad2/signal transducer and activator of transcription 3/signal transducer and activator of transcription 5 path-way, thereby indicating a possible therapeutic relevance (Sun et al., 2015). Contradictory results have been reported regarding miR-10a-5p in studies of LUAD. These discrepancies may be attributed to differences in expression across racial, ethnic, and geographic populations. Contradictory results have been reported regarding miR-10a-5p in studies of LUAD. These discrepancies may be attributed to differences in expression across racial, ethnic, and geographic populations.

In the development and growth of metastatic brain tumors, the brain microenvironment, composed mainly of the astrocytes, plays an important role (Alsidawi et al., 2014). According to the results of several studies, miR-210 directly induces a change in the brain microenvironment. For example, according to the analysis of Camacho et al., miR-210 isolated from exosomes of metastatic brain cells with breast cancer origin was the only significantly over-expressed miRNA when compared to parental breast cancer cells (Camacho et al., 2013). In LUAD, elevated miR-210 supports survival in the brain niche by modulating mitochondrial metabolism and reducing oxidative stress (Puisségur et al., 2011). In the context of lung cancer, inhibition of miR-210-3p in A549 cells resulted in reduced migration, invasion capacity, and cell viability, while significantly increasing the rate of apoptosis (Chen et al., 2021). According to another study, a three-miRNA (miR-210, miR-214, and miR-15a) based signature was able to predict BM in LUAD patients with 91.4% accuracy (Zhao et al., 2018). It has also been reported that high levels of miR-210-3p promote tumor progression by modulating C-C motif chemokine ligand 2 mediated macrophage recruitment and polarization under hypoxic conditions (Arora et al., 2024). Furthermore, Chen et al. demonstrated that silencing miR-210-3p may represent a promising approach for lung cancer treatment, as it can inhibit tumor growth, invasion, and metastasis by modulating the polycomb group ping finger 3/upstream transcription factor 1 axis, while also affecting apoptosis, angiogenesis, and EMT (Chen et al., 2021). Daugaard et al. also reported the upregulation of miR-210-3p and found that it was significantly associated with the development of distant metastases (Daugaard et al., 2017). These results, consistent with our findings, support the notion that miR-210-3p may serve as a potential therapeutic target to prevent the development of LUAD-BM.

Previous clinical studies have demonstrated that overexpression of miR-130b is associated with aggressive tumor phenotypes and poor prognosis across several cancer types, including glioma (Alsidawi et al., 2014). Moreover, elevated miR-130b expression has been significantly correlated with the formation of distant metastases in both colon and lung cancers (Alsidawi et al., 2014). In LUAD, Kim et al. reported that miR-130b overexpression was linked to higher histological grade, lymph node metastasis, and lymphovascular invasion, further supporting its role in tumor progression and metastatic potential (Kim et al., 2021). It has also been reported that upregulation of miR-130b-3p promotes lung cancer progression by regulating serine/threonine kinase 11 mediated proliferation, migration, invasion, and immune escape, as well as FOXO3-dependent kelch like ECH associated protein 1/NFE2 like BZIP transcription factor 2/thioredoxin reductase 1 signaling (Chen et al., 2023; Guo et al., 2021). Moreover, miR-130b-3p is considered a possible therapeutic target in NSCLC, as exosomal miR-130b-3p has been shown to serve as a potential predictive marker and therapeutic target by suppressing NSCLC growth, migration, and EMT, and promoting apoptosis through inhibition of DEP Domain Containing 1 (Lv et al., 2024). Furthermore, inhibition of miR-130b or restoration of TIMP metallopeptidase inhibitor 2 may reduce NSCLC invasion by decreasing Matrix Metallopeptidase 2 activity (Hirono et al., 2019; Frydrychowicz et al., 2023). These data, consistent with our findings–showing that changes in miR-130b-3p expression can be detected even in primary tumor–suggest that targeting this miRNA may play a significant role in inhibiting the metastatic process from its initial phases.

Furthermore, no significant differences in expression of miR-503-5p and miR-195-5p were observed between LUAD-BM and primary LUAD. However, miR-503-5p was significantly upregulated, whereas miR-195-5p was significantly downregulated in primary LUAD compared to normal lung controls.

MiR-503-5p shows association with regulation of the development of different cancer types like LUAD and additionally, it accelerates metastasis and angiogenesis formation in tumors, too (Fei et al., 2020). Recent studies have reported that miR-503-5p plays a role in mediating chemotherapy resistance in various tumor types (Park and Kim, 2019). Han et al. demonstrated that elevated levels of it were associated with cisplatin resistance and angiogenesis of LUAD cells via regulation of CTD small phosphatase like gene expression, suggesting that miR-503-5p may serve as a potential therapeutic target (Han and Wang, 2023). Moreover, Zhang et al. demonstrated that inhibition of the JMJD2C/metastasis associated lung adenocarcinoma transcript 1/miR-503-5p/Septin 2 axis may suppress NSCLC progression by modulating histone methylation (Zhang J. et al., 2022).

Deregulation of miR-195-5p has been reported in multiple cancers. Depending on the specific cancer type, it can play an oncogenic or tumor-suppressive role as a posttranscriptional regulator (Mohan et al., 2024). Long et al. detected the significant downregulation of miR-195-5p both in LUAD tissue samples and in lung cancer cell lines (H1299, A549) in comparison with the control samples. Additionally, they demonstrated that elevated expression of miR-195-5p can induce inhibition of lung cancer cell proliferation, migration, and invasion, leading to reduced tumor growth and metastasis formation of lung cancer via targeting forkhead box K1 (Long and Wang, 2020). Other studies proved the tumor suppressor activity of miR-195-5p in NSCLC through the direct regulation of MYB, insulin-like growth factor 1 receptor, hepatoma-derived growth factor or checkpoint kinase 1 (Liu et al., 2015; Guo et al., 2014; Wang et al., 2014; Yongchun et al., 2014). It was further demonstrated that miR-195 is associated with a better prognosis in lung cancer, while targeting Fibroblast Growth Factor 2 can enhance chemosensitivity in A549/DDP cells (Liu et al., 2015; Wang et al., 2021).

A limited number of studies have focused on the miRNA expression differences in LU-AD-BM compared to primary LUAD. Zhang et al. compared the miRNA expression pro-files of LUAD samples without BM to those with BM to identify potential biomarkers and mechanisms involved in metastasis. They identified 20 dysregulated miRNAs using next-generation RNA sequencing in an Asian population (Zhang L. et al., 2022). In another study, the authors identified a 25-miRNA signature based on the mRNA profiles of LUAD patients with and without BMs, based on their tissue mRNA profiles using machine learning–based prediction (Koh et al., 2024). Remon et al. compared expression patterns between EGFR-mutant NSCLC samples without BMs and those with confirmed BMs (Remon et al., 2016).

Tsakonas et al. performed matched analyses to reveal a unique miRNA signature by comparing primary NSCLC samples to brain metastatic samples. The study reported downregulation of miR-129-2-3p, miR-124-3p, miR-219a-2-3p, miR-219a-5p, and miR-9-5p, and upregulation of miR-142-3p, miR-150-5p, miR-199b-5p, miR-199a-3p, and miR-199a-5p (Tsakonas et al., 2022). The heterogeneity observed in miRNA expression patterns is likely influenced by population-specific genetic factors that affect baseline miRNA levels. Consequently, expression patterns may vary across racial, ethnic, and geographic groups (Dluzen et al., 2016; Nassar et al., 2017; Rawlings-Goss et al., 2014). Therefore, geographic and ethnic context should be considered a key factor when assessing the clinical utility of miRNAs.

To the best of our knowledge, this is the first study to simultaneously compare the miRNA expression profiles of LUAD-BM tissues with normal brain tissue, primary LU-AD, and normal lung. This approach allows mapping of expression changes along the transition from normal lung to primary LUAD and LUAD-BM. Additionally, this is the first study specifically focusing on LUAD that utilized brain tissue samples to investigate the expression profiles of LUAD-BM. Furthermore, we identified for the first time the dysregulation of miR-10a-5p, miR-200c-3p, miR-3934-5p, miR-130b-3p, miR-146b-5p, miR-503-5p, and miR-195-5p in LUAD-BM human samples that could play an important role in the development of BM with LUAD origin. Our findings provide initial insights in-to the role of these miRNAs; however, they should be regarded as preliminary, and additional research is required to determine their diagnostic and therapeutic implications.

The identification of DE miRNAs holds promising potential for improving the diagnosis and treatment of various cancers. In our study, we identified a panel of ten miRNAs (miR-200c-3p, miR-210-3p, miR-10a-5p, miR-130b-3p, miR-146b-5p, miR-503-5p, miR-196b-5p, and miR-3934-5p) using NGS and RT-qPCR methods. This panel was able to distinguish LUAD-BM from normal brain tissue samples with excellent sensitivity and specificity in the Hungarian population. Differential expression analysis of 20 LUAD and 15 normal lung samples (selected from the TCGA database representing European populations), together with LUAD-BM samples, confirmed the dysregulation of six of the ten validated miRNAs. Among these, miR-200c-3p, miR-146b-5p, and miR-3934-5p differed significantly between primary LUAD and LUAD-BM, while miR-10a-5p, miR-210-3p, and miR-130b-3p showed progressive dysregulation along the normal lung–LUAD–LU-AD-BM axis. This study is the first to integrate brain tissue samples in the analysis of miRNA expression in LUAD-BM while simultaneously comparing LUAD-BM with nor-mal brain, primary LUAD, and normal lung tissues. Our findings are consistent with previously published data and support the crucial role of these miRNAs in the development of BM.

A limitation of this study is the relatively small sample size, highlighting the need for validation in larger independent cohorts. In addition, the lack of primary lung tumor samples restricted direct comparison between primary LUAD and LUAD brain metastases, which should be addressed in future studies using matched tumor pairs.