A systematic literature search was performed to identify relevant studies investigating the relationship between bladder cancer, tumoral immune infiltration, immunotherapy, and miRNA. The search strategy employed a combination of targeted keywords, including “bladder cancer,” “urothelial carcinoma,” “immunotherapy,” “checkpoint inhibitor,” “immune response,” “immune evasion,” “immune infiltration,” “miR,” “microRNA,” “miRNA,” and “non-coding RNA.”

The databases used for the identification of relevant studies were PubMed, SCOPUS, Web of Science, ClinicalTrials.gov, and Google Scholar. Literature retrieval from Google Scholar was done via the Publish or Perish software (8), which facilitates structured bibliographic data extraction. The last updated search was conducted in March 2025.

The inclusion criteria were peer-reviewed original research articles published in English that addressed the role of miRNAs in the context of immunotherapy and/or immune mechanisms in bladder cancer. We considered studies including patients and animal models, as well as those conducted on cell lines. Case reports, editorials, and book chapters were excluded. The results from all databases were imported into Rayyan (9). Duplicates were removed, and titles and abstracts were screened manually by three authors for relevance. Full texts of eligible studies were downloaded and read in detail. Disagreements were resolved by discussion with a fourth researcher. Relevant data from each study were extracted and summarized in tables.

Expression data (mRNA and miRNA) and clinical data for muscle-invasive urothelial carcinoma (pT2 or above) were obtained from The Cancer Genome Atlas (TCGA-BLCA) project using the FirebrowseR R package, via the Broad Institute’s Firehose pipeline. Only primary tumor samples with complete expression and clinical annotation were included in the subsequent analyses (10). Corresponding expression and clinical annotation files were integrated based on TCGA patient barcodes. Clinical data for survival analysis were obtained from the TCGA Pan-Cancer Clinical Data Resource (TCGA-CDR) (11).

Immune cell infiltration estimates were obtained from the Tumor Immune Estimation Resource (TIMER) (7), a computational framework that deconvolutes the composition of the tumor immune microenvironment using bulk RNA-sequencing data. TIMER provides robust quantification of six key immune cell populations: B cells, CD4⁺ T cells, CD8⁺ T cells, neutrophils, macrophages, and dendritic cells. To ensure a comprehensive evaluation of immune infiltration, we utilized multiple established deconvolution algorithms that estimate immune cell abundance from the transcriptome data. The R packages igraph and ggraph were used for the construction of the networks between TME populations and miRNAs (12, 13).

All statistical analyses were conducted in R version 4.4.3 (2025 - 02–28 ucrt) - “Trophy Case” (14). For correlation analyses, the Spearman rank correlation test was used. To enhance the strength of the findings and reduce the likelihood of false-positive results due to weak associations between variables of interest, a minimum correlation coefficient threshold of 0.3 was set to ensure a moderate to strong association. This cutoff was used to prioritize miRNAs with a potentially higher influence on the TME. All p-values were adjusted for multiple comparisons using the Benjamini–Hochberg procedure to control the false discovery rate (15).

For clinical analyses, one-way ANOVA tests were performed to compare miRNA expression levels between patient groups. To evaluate the prognostic impact of miRNAs, patients were stratified into high- and low-expression groups based on the median expression level of each miRNA. Progression-free interval (PFI) and overall survival (OS) were assessed using Kaplan–Meier survival curves and compared with the log-rank tests. Cox proportional hazards regression models were used to estimate the degree and significance of the association between miRNA expression and patient outcomes. A p-value of 0.05 was set for statistical significance.

We identified a robust set of positive correlations between miRNA expression and immune cell infiltration in bladder cancer, highlighting potential immunoregulatory roles for these miRNAs. Among the most significant results, miR-155 had the strongest associations, showing high positive correlations with CD4⁺ T cells (r = 0.568) and M1 macrophages (r = 0.537). Similarly, miR-142, miR-146b, and miR-511 were strongly associated with CD4⁺ memory-activated T cells (r = 0.483, r = 0.477, and r = 0.452, respectively). miR-146b and miR-142 also significantly correlated with M1 macrophage infiltration (r = 0.453 and r = 0.441, respectively) ( Figure 2 , Supplementary Table 3 ).

Additional miRNAs, including members of the miR-200 family (miR-200c, miR-200a, miR-200b, and miR-429), demonstrated moderate correlations with activated myeloid dendritic cells (r ranging from 0.338 to 0.42), suggesting a role in dendritic cell-mediated immune responses. Other miRNAs such as miR-223, miR-29a, miR-150, and miR-146a showed moderate associations with various immune subsets, including neutrophils, T cells, and macrophages ( Figure 2 , Supplementary Table 3 ).

In contrast, a substantial number of miRNAs demonstrated moderate to strong negative correlations with immune cell subsets. Namely, miR-146b, miR-125b, and miR-100 exhibited strong inverse correlations with activated myeloid dendritic cells (correlation coefficients ranging from −0.39 to −0.37), while miR-200b, miR-200a, and miR-429 were negatively associated with M1 macrophages and activated CD4⁺ memory T cells. miR-590 and miR-17 also showed consistent negative correlations with activated mast cells. Interestingly, several miRNAs such as miR-200c, miR-141, miR-155, and miR-142 were negatively correlated with multiple immune cell types, highlighting their possible involvement in broader immune regulatory networks within the TME. This multiple regulation was also seen when analyzing how many different miRNAs correlated with each immune population ( Figure 2 , Table 2 ). M1 macrophages were the most frequently targeted immune cells, with 10 miRNAs being positively correlated with their degree of infiltration in the tumors and 13 miRNAs showing a negative correlation. CD4⁺ T cells were also correlated with multiple miRNAs (10 positively and 4 negatively correlated) ( Figure 2 , Table 2 ).

Analysis of miRNA expression across different tumor stages yielded a total of 49 miRNAs with significant differences ( Supplementary Table 4 ). Of these, 13 were also significantly associated with patients’ prognosis ( Supplementary Tables 5 , 6 ). Notably, six miRNAs (miR-1278, miR-29c, miR-30c, miR-93, miR-191, and miR-141) exhibited a decreasing expression trend, with the highest levels observed in stage II and the lowest in stage IV. Conversely, six other miRNAs (miR-125b, miR-100, miR-99a, let-7c, miR-145, and miR-409) showed increasing expression levels from stage II to the more advanced stages ( Figure 3 , Supplementary Table 4 ).

Since only nine studies evaluated the impact of miRNAs on OS and only one on PFS, we sought to analyze these data on the TCGA-BLCA cohort to further evaluate the clinical relevance of these miRNAs ( Figures 4 , 5 , Supplementary Tables 5 , 6 ). Out of all miRNAs analyzed, 12 were significantly associated with PFI, and 23 were significantly associated with OS. Specifically, elevated expression levels of miR-200c, miR-141, miR-30c, miR-191, miR-125a, miR-200a, and miR-29c were associated with improved PFI and OS. Additionally, higher levels of miR-625, miR-3200, miR-590, miR-186, miR-93, miR-1287, miR-1307, miR-192, and miR-200b were significantly associated with better OS. Conversely, increased expression of let-7c, miR-99a, and miR-125b correlated with poorer PFI and OS. Higher levels of miR-204, miR-100, miR-145, and miR-409 were linked to worse OS. Notably, elevated miR-92b expression was associated with a shorter PFI, although its association with OS did not reach statistical significance ( Table 3 , Supplementary Tables 5 , 6 ).

The results of this systematic review and comprehensive analysis highlight a complex network of interactions between specific miRNAs and components of the bladder cancer TME, which is yet to be fully explored and understood. These findings, completed by the bioinformatic analyses aimed at filling the literature gaps, support their potential as modulators of immune infiltration in bladder cancer. Although different perspectives have been explored, there are still insufficient data to definitively demonstrate the value of miRNAs in the selection of patients undergoing immunotherapy.

While numerous studies have investigated the role of individual miRNAs in bladder cancer immunity, their focus has often been fragmented, with different studies examining different miRNAs and distinct immune cell populations, very few on checkpoint molecules, and with limited clinical endpoints. For instance, miR-100-5p and miR-138-5p were linked to PD-L1/PD-L2 regulation (17), while others, such as miR-21 and miR-31, were associated with BCG response, disease progression, and recurrence (18), but no validation studies to confirm these results were found. Additional studies explored the interplay between select miRNAs and specific immune subsets, including CD4⁺ IFN-γ⁺ T cells, NK cells, macrophages, and regulatory T cells (19, 22, 26, 28). However, these investigations frequently lacked integration across cell types, mechanisms, and clinical impact. Moreover, only a few studies examined the prognostic significance of immune-related miRNAs systematically and comparatively.

Our study addresses these gaps by providing a comprehensive synthesis of the roles of the 104 miRNAs implicated in shaping the immune microenvironment of bladder cancer. By integrating findings from the published experimental models and validating them with large-scale data (TCGA-BLCA), we connected the dots between immune modulation and clinical outcomes such as PFI and OS. This integrative approach summarizes the current state-of-the-art on the immunoregulatory role of miRNAs in bladder cancer and offers a more unified view for understanding their potential impact on immunotherapy response. Moreover, this work lays the ground for their future use as prognostic and potentially predictive biomarkers.

The differential expression of specific miRNAs and their association with immune cell infiltration patterns highlight the potential of miRNAs as both biomarkers and functional regulators of immune dynamics (52). These observations align with the broader framework of tumor immunoediting, a dynamic process encompassing the phases of elimination, equilibrium, and escape, through which cancer cells continuously evolve to evade immune surveillance (53).

In the escape phase, immune evasion becomes dominant, and tumors acquire multiple mechanisms to suppress antitumor immunity. Many of the miRNAs identified in this review are likely active participants in this phase. For instance, expression of miR-21 and miR-141-3p correlates with diminished cytotoxic immune cell infiltration (e.g., CD8⁺ T cells, NK cells), suggesting a role in establishing immune exclusion or immune suppression (18, 19, 34, 39, 43). Similarly, downregulation of miRNAs such as miR-29c-3p and miR-186-5p, which have been linked to immune activation, may further support the development of an immunosuppressive TME (23, 25) ( Figure 6 ).

The heterogeneity of miRNA expression patterns observed across patient cohorts also reflects the broader variability in immune infiltration and response observed in bladder tumors. The bladder cancer TME frequently harbors immunosuppressive components, including tumor-associated macrophages (TAMs), regulatory T cells (Tregs), and myeloid-derived suppressor cells (MDSCs), all of which can be recruited and sustained via miRNA-regulated signaling pathways (54). Moreover, the enrichment of immune checkpoint ligands such as PD-L1 and PD-L2 in miRNA-associated gene networks suggests a role for miRNAs in modulating immune checkpoint pathways, which are central to current immunotherapeutic strategies. High expression of PD-L1, often driven by oncogenic or hypoxia-induced signaling, is further modulated by miRNAs such as miR-138-5p and miR-100-5p (54). Loss or suppression of these miRNAs may contribute to the upregulation of immune inhibitory signals, allowing tumor cells to escape T-cell-mediated cytotoxicity ( Figure 6 ).

The immunosuppressive TME interferes with normal immune responses and, thus, can reduce the efficacy of immune checkpoint blockade, which has been recently approved for bladder cancer treatment (55). Our review suggests that miRNA expression profiles could potentially serve as predictive biomarkers for immunotherapy response, as evidenced, for example, by their association with BCG failure (e.g., miR-21, miR-31) (18) ( Figure 6 ).

Importantly, the functional links between miRNAs and immune regulation open up new opportunities for therapeutic intervention. Targeting specific miRNAs could complement existing immunotherapies by reversing immunosuppressive programming or restoring antitumor immunity. For example, miRNA mimics or antagomirs could be employed to modulate key pathways associated with T-cell recruitment, macrophage polarization, or checkpoint expression, thus enhancing the depth and durability of immune responses (54).

With the highly positive results obtained by immunotherapy in other cancers (54) and the approval of novel therapies for bladder cancer, research focusing on identifying the most promising candidates who would benefit from it is highly needed. Significant progress has been made in the use of miRNAs as biomarkers in bladder cancer, including the validation of a panel capable of diagnosing and stratifying patients (7). However, whether these panels or similar ones can also identify responders to treatment remains an open question. Of note, several of the miRNAs (miR-221-3p, miR-93-5p, miR-191-5p, miR-200c-3p, miR-192-5p, miR-21-5p) included in this panel have already been investigated for their roles in tumor immunity, suggesting that the development of miRNA-based tools to predict immunotherapy response is a feasible and promising avenue.

Emerging evidence highlights the critical role of miRNAs (often transported via tumor-derived exosomes) in preparing the pre-metastatic niche, particularly by modulating immune landscapes at distant sites (56). These miRNA-loaded exosomes act as messengers that condition the future metastatic environment, promoting immune evasion, suppressing cytotoxic T-cell activity, and recruiting immunosuppressive cells such as regulatory T cells, M2 macrophages, and myeloid-derived suppressor cells (56, 57). For example, certain miRNAs can inhibit antigen presentation by downregulating MHC class I molecules or modulate cytokine production to create a tolerogenic or “immune-silent” microenvironment that favors metastatic colonization. This immune reprogramming not only facilitates tumor cell seeding and outgrowth at distant sites but also contributes to resistance against immunotherapies such as immune checkpoint blockade (56, 57). By converting immunologically “hot” environments into “cold” niches, these miRNAs help shield disseminated tumor cells from T-cell-mediated clearance. Understanding and targeting these miRNA-mediated pathways offer a novel opportunity for disrupting the metastatic cascade and enhancing immunotherapeutic efficacy through restoration of immune surveillance in both primary and secondary tumor sites (53, 58).

miRNAs play multifaceted roles in modulating immune infiltration in bladder cancer and are increasingly recognized as promising biomarkers. Although increasing evidence shows their potential clinical relevance in immune-oncology, no studies to date have directly assessed miRNA profiles in patients receiving immunotherapy. Prospective studies and robust clinical validation are urgently needed to assess their utility as reliable biomarkers of immunotherapy response in bladder cancer patients.

While this systematic review and integrated TCGA analysis provide a comprehensive overview of the miRNA–immune axis in bladder cancer, several limitations should be mentioned to contextualize our findings and guide future research. The primary studies included in this systematic review had considerable heterogeneity. This was evident in the diversity of miRNAs investigated, the variety of experimental models used (ranging from in vitro cell lines to animal models and human tumor tissues), the different immune cell populations assessed, and the clinical or biological endpoints reported. Such heterogeneity, while reflective of the exploratory nature of much of the research on this topic, prevented any meaningful meta-analysis and therefore requires caution when concluding from this combined literature.

Moreover, our TCGA-BLCA data analysis, while robust, has some limitations as well. The use of bulk RNA-sequencing data for immune cell deconvolution provides estimates of immune cell abundance but cannot fully show the spatial organization or functional state of these cells within the tumor microenvironment. The associations identified between miRNA expression, immune infiltration, and clinical outcomes are also correlational and do not imply causality. While these findings provide valuable hypotheses, they require experimental validation on other cohorts.

Despite the promising associations identified, a significant translational gap remains. The gap from identifying significant miRNA expression patterns or in vitro/in vivo mechanistic links to clinically validated biomarkers predictive of immunotherapy response is quite wide. The current body of evidence, as synthesized here, highlights the need for prospective, well-designed clinical trials to validate the utility of these miRNAs in guiding therapeutic decisions in bladder cancer.