Immune-related long noncoding RNAs in predicting the prognosis and immune landscape of intrahepatic cholangiocarcinoma: a bioinformatics analysis with experimental verification

Introduction

Intrahepatic cholangiocarcinoma (ICC) is a fatal tumor deriving from the biliary system and accounts approximately for 10–20% of all bile duct malignancies (1,2). The incidence of ICC has been gradually increasing in most parts of the world (2). ICC often occurs in patients with underlying liver disease such as hepatitis B or C and primary sclerosing cholangitis (3). ICC lacks distinct symptoms in the early stage and is usually detected at an advanced stage, making it difficult for clinicians to perform early intervention (4). A large number of clinical studies have intensively investigated the immunotherapy and targeted therapy for ICC, which has resulted in the appreciable prolongation of progression-free survival (PFS) and overall survival (OS) among these patients (5,6). Further development of immunotherapy based on immune check inhibitors is thus warranted.

Long noncoding RNAs (lncRNAs) are longer 200-bp RNAs that lack a protein-coding function and can participate in various disease progression by mediating gene expression (7,8). Several competing endogenous RNA (ceRNA) networks have been clarified, in which lncRNAs regulate certain key genes by modifying the expression of microRNAs (miRNAs) (9). Consequently, lncRNAs have gradually come to be considered a type of functional RNA similar to tumor proto-oncogenes or oncogenes. A myriad of lncRNAs have been identified through next-generation-sequencing technology, some of which exert crucial functions in hepatobiliary malignancies (10,11). This property of lncRNAs is of substantial relevance for the research on tumor development, invasion, and immune infiltration.

Recent studies have shown that lncRNAs are key molecules in determining the immune status of various cancers (12,13) and regulating messenger RNA (mRNA) expression to aid in the activation of tumor-infiltrating immune cells and the remodeling of the tumor immune microenvironment. Corresponding lncRNA signatures related to immune cell infiltration have demonstrated satisfactory value in prediction tasks related malignancies. Several studies have identified fractional lncRNAs that participate in tumor immune progression. Zhang et al. constructed a six-lncRNA signature for predicting the prognosis of patients with triple-negative breast cancer, providing insights into the immune status associated with lncRNAs (14). Cao et al. examined immune-related lncRNAs (irlncRNAs) and developed a five-irlncRNA signature for prognostic prediction in patients with bladder cancer (15). Although irlncRNA prognostic signatures have been developed for a number of malignancies, the irlncRNAs relevant to ICC have not been identified, and no similar signature has been designed; however, completing this work is essential in order to generate insights into the progression and immune microenvironment of ICC.

In this study, we developed an irlncRNA prognostic signature for ICC and evaluated the association of this signature with survival, immune cell infiltration, chemotherapeutic efficacy, and the expression of immune checkpoint genes. Our findings offer a novel perspective for the individualized treatment of patients with ICC. We present this article in accordance with the TRIPOD reporting checklist (16) (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-705/rc).

Methods

Data collection and preprocessing and analysis of differentially expressed irlncRNAs (DE-irlncRNAs)

The transcriptome profiling data of patients with ICC were downloaded from The Cancer Genome Atlas (TCGA) dataset (https://portal.gdc.cancer.gov/), and 2,483 immune-related genes (IRGs) were extracted from the ImmPort database (https://immport.niaid.nih.gov) via a webpage interface module (17). Another dataset, GSE107943, was downloaded from the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) for biomarkers verification. The separation of mRNAs and lncRNAs was performed based on the corresponding attributes via the Perl programming language, and the RNA-sequencing data were annotated based on gene transfer format (GTF) files from the Ensemble database. The correlation analysis was performed between IRGs and lncRNAs, and lncRNAs with P<0.001 and an absolute value of Pearson correlation coefficient >0.4 were considered to be irlncRNAs. Finally, we identified DE-irlncRNAs as those with a threshold log₂ fold change (FC) >2 and a false-discovery rate (FDR) <0.01 via the “limma” package in R software (18).

Construction of an irlncRNA prognostic signature

Clinicopathological and survival data were extracted from TCGA dataset, with OS as the outcome. Univariate regression analysis was carried out to screen for survival-related DE-irlncRNAs, which were further incorporated into the multivariable regression analysis to obtain DE-irlncRNAs with independent prognostic value for constructing the prognostic signature. P<0.05 was used as the threshold. Subsequently, the coefficients and expression levels of DE-irlncRNAs were obtained to calculate the risk score via the following formula: riskscore=∑i=1kβi×expi. The median risk score was used as the cutoff value to divided all patients into high- and low-risk groups.

Validation of the signature

The Kaplan-Meier (KM) survival curve was plotted to determine the predictive power of the risk score via the “survival” and “survminer” packages. The area under curve (AUC) of the receiver operating characteristic (ROC) was calculated via the “timeROC” R package to verify the signature’s predictive accuracy. Principal component analysis (PCA) was implemented to efficiently downscale high-dimensional lncRNA data to determine the discriminative ability of the risk score among patients (19), and gene set enrichment analysis (GSEA) was applied to clarify the mechanisms underlying the efficacy of the prognostic signature (20).

Quantitative real-time polymerase chain reaction (qRT-PCR)

Eight ICC and paired normal tissues were prospectively collected. Tissues samples were obtained from the patients with ICC from the Department of General Surgery of No. 924 Hospital of PLA Joint Logistic Support Force between January 2023 and December 2024. The detailed methods of RNA extraction and purity assessment, reverse transcription, and real-time quantification can be found in previous study (21). We recorded the cycle threshold (Ct) and calculated the relative expression of genes using the 2^−ΔΔCt method. All the six irlncRNAs and actin primer sequences are shown in available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Ethics committee of the No. 924 Hospital of PLA Joint Logistic Support Force (No. GY2024-93). Written informed consent was obtained from all patients.

The clinical application of the signature and nomogram construction

To validate the independently predictive ability of signature, we combined the risk score with clinicopathological characteristics and incorporated them into Cox regression analyses. The clinicopathologic characteristics included age, gender, grade, American Joint Committee (AJCC) stage, T stage, N stage, M stage, and peripheral nerve infiltration (PNI). A forest map was used to visualize the results. Further, we assessed the association between the risk score and the clinicopathological characteristics, calculated the difference in the risk score and these clinicopathological characteristics between the two risk groups, and used band diagrams and box diagrams to visualize the results. The R packages “survival” and “ComplexHeatmap” were used for these analyses (22). Subsequently, based on the identified predictive variables, we constructed a prediction nomogram using the “rms” and “foreign” R packages. The prediction ability of the nomogram was validated via calibration curves, and multivariate ROC curves were drawn to demonstrate the predictive ability of the risk score.

Association of the risk score with tumor-infiltrating immune cells, chemotherapeutic efficacy, immune checkpoint genes, and tumor mutation burden (TMB)

To evaluate the immune infiltration status in ICC, we integrated several established methods, including TIMER (23,24), CIBERSORT (25), XCELL (26,27), QUANTISEQ (28,29), MCPcounter (30), and EPIC (31), and analyzed the correlation between the risk score and immune cell infiltration (32). Differential analysis was performed, the results of which were visualized in a box plot via the Wilcoxon signed-rank test. The relationship between the risk score and tumor-infiltrating immune cells was analyzed through use of Spearman correlation analysis via the “ggplot2” R package, with P<0.05 being considered significant; the corresponding results were displayed in a lollipop plot.

To evaluate the capacity of the signature to predict chemotherapeutic efficacy, we compared the half-maximal inhibitory concentration (IC₅₀) difference of several chemotherapeutic agents, including gemcitabine, cisplatin, and rapamycin, along with partly targeted therapy drugs, between the high- and low- risk groups via the Wilcoxon signed-rank test using the “pRRophetic” R package (33). The subsequent results were visualized in a box diagram through use of the “ggplot2” package in R software. We further examined the association between risk score and the crucial immune check points in order to forecast the immunotherapy response. The “ggpubr” R package was employed for violin plot visualization. Finally, we collected and processed the TMB data from TCGA database in order to clarify the relationship between the risk score and TMB score.

Statistical analysis

R software (version 4.0.2) was used for statistical analyses, and the related visualizations were created through use of its resource packages. The Chi-squared test was conducted to determine the correlation of the risk score with clinicopathological characteristics. Differences in clinicopathological characters, tumor-infiltrating cells, and IC₅₀ between the two risk groups were calculated via the Wilcoxon signed-rank test. In all statistical tests in this study, a two-sided P<0.05 was considered statistically significant.

Results

Identification of DE-irlncRNAs

We extracted the lncRNA transcriptome data of ICC from TCGA dataset, including 33 tumor and 8 normal samples, as well as 2,483 IRGs from the ImmPort database. Based on the co-expression analysis results, a total of 1,394 irlncRNAs with log₂FC >2 and an FDR <0.01 met the aforementioned criteria. Subsequently, 723 irlncRNAs were identified as DE-irlncRNAs, of which 682 were upregulated and 41 were downregulated in ICC (Figure 1A,1B and available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf).

Figure 1 Identification of survival-related and differentially expressed irlncRNAs from TCGA. (A) Heatmap plot of the differentially expressed irlncRNAs. (B) Volcano plot of up- or downregulated differentially expressed irlncRNAs. The red dots represent upregulation, while the green dots represent downregulation. (C) Univariate regression analysis for screening survival-related and differentially expressed irlncRNAs. (D) Multivariate regression analysis for screening independently prognostic irlncRNAs. The green dots indicate an HR value <1, while the red dots indicated an HR value >1. (E) Association of mRNAs with irlncRNAs and risk types. (F) Boxplot of the differential expression levels of six irlncRNAs. **, P<0.01; ***, P<0.001. FC, fold change; FDR, false-discovery rate; HR, hazard ratio; irlncRNAs, immune-related long noncoding RNAs; TCGA, The Cancer Genome Atlas.

Construction of the irlncRNA prognostic signature

Ten survival-related DE-irlncRNAs were screened in the univariate Cox regression analysis (P<0.05; Figure 1C). A prognostic signature involving six irlncRNAs (APCDD1L-DT, WAC-AS1, LINC01615, AL391056.1, AC090114.2, and LINC01711) was then constructed via multivariate Cox regression analysis (Figure 1D). Based on the coexpression results, a total of 81 IRGs were found to be positively correlated with six irlncRNAs, and an alluvial plot of the top 10 IRGs in terms of associations with irlncRNAs and survival state was drawn (Figure 1E). The risk coefficients indicated that all of the hazard ratios of the six irlncRNAs exceeded 0 and were considered to be prognostic risk factors (available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf); their expression levels are shown in Figure 1F. The risk score of each sample was calculated as follows:

riskscore=(−2.65)×APCDD1L-DT+(−1.15)×WAC-AS1+(−1.06)×LINC01615+(2.63)×AL391056.1+(−2.68)×AC090114.2+(1.44)×LINC01711[1]

Finally, using the median risk score, we divided all patients into a high-risk group (n=16) and a low-risk group (n=17). The corresponding survival condition plots are shown in Figure 2A,2B. These results suggested that the high-risk patients had higher mortality compared with the low-risk patients. The related expression patterns of the irlncRNAs are also shown in Figure 2C.

Figure 2 Validation of the irlncRNA-based prognostic signature. (A-C) Survival condition plots and heatmap of 6 differentially expressed irlncRNAs. (D) Kaplan-Meier survival curve. (E) Time-dependent ROC curves predicting OS at 1, 2, and 3 years. AUC, area under the curve; irlncRNA, immune-related long noncoding RNA; OS, overall survival; ROC, receiver operating characteristic.

Validation of the irlncRNA prognostic signature

Next, to evaluate the signature’s prognostic discrimination of patients with ICC, KM survival curves were plotted. The results revealed that high-risk patients had a significantly shorter OS (Figure 2D). We also performed ROC analyses to calculate the AUCs, which were 0.993, 0.901, and 0.904 at 1, 2, and 3 years, demonstrating this signature performed well as a predictor of prognosis (Figure 2E). Moreover, the PCA indicated that the risk model more rationally distributed the high- and low-risk groups (available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf). Importantly, the results of qRT-PCR indicated that all irlncRNAs were significantly highly expressed in tumor tissues compared with normal controls (all P<0.05; Figure 3A-3F).

Figure 3 The relative expression levels of irlncRNAs between normal and tumor tissues. (A) APCDD1L-DT. (B) WAC-AS1. (C) LINC01615. (D) AL391056.1. (E) AC090114.2. (F) LINC01711. *, P<0.05; **, P<0.01. irlncRNAs, immune-related long noncoding RNAs.

We then performed GSEA to clarify the mechanisms related to the prognostic signature’s efficacy. The results revealed that primary bile acid biosynthesis, pantothenate and co-enzyme A biosynthesis, and complement and coagulation cascades pathways were significantly enriched in the high-risk group (Figure 4A). The detailed GSEA results are provided in available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf. Only irlncRNAs WAC-AS1 and AL391056.1 were identified in the GSE107943 dataset; thus, further transcriptome sequencing analyses are needed validate all the DE-irlncRNAs.

Figure 4 Gene set enrichment analysis and clinical relevance analysis. (A) Gene set enrichment analysis in the high-risk group. (B) Forest plot for the univariate multivariate Cox regression analyses. (C) Forest plot for multivariate Cox regression analysis. (D) Band diagram showing the relationship between risk and clinicopathologic characteristics. (E) Box diagram showing the correlation of risk with T stage. (F) Box diagram showing the correlation of risk with AJCC stage. *, P<0.05. AJCC, American Joint Committee on Cancer; PNI, peripheral nerve infiltration.

Clinical application and nomogram construction

To analyze the clinical utility of prognostic signature, we performed univariate and multivariate Cox regression analysis based on all the clinicopathologic characteristics. The results indicated that the risk score (P=0.02) and N stage (P=0.049) were independent prognostic indicators (Figure 4B,4C and available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf). The detailed clinical and follow-up information of 33 patients with ICC are shown in available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf. We further examined the relationship between the risk score and clinicopathological characteristics. The band diagram and box diagram suggested that T stage (T1 vs. T3 and T2 vs. T3) and AJCC stage (stage II vs. stage IV) were significantly associated with the risk score (Figure 4D-4F).

Next, we developed a predictive nomogram integrating the above-mentioned predictors for clinical application (Figure 5A). The multivariate ROC curve demonstrated that the risk score had the best predictive capability among the factors examined (Figure 5B). The calibration curves indicated that the nomogram had the ability to predict the actual survival rate at 1, 2, and 3 years (Figure 5C-5E). These results suggest that the nomogram has good prospects for clinical application.

Figure 5 Construction and validation of nomogram. (A) Nomogram for predicting OS at 1, 2, and 3 years. (B) The AUC for the risk score and clinical features. (C-E) Calibration curves comparing the probability of 1-, 2-, and 3-year OS between the nomogram prediction and actual outcomes. AJCC, American Joint Committee on Cancer; AUC, area under curve; OS, overall survival; PNI, peripheral nerve infiltration.

The association of the signature with tumor-infiltrating immune cells, immunosuppressed biomarkers, chemotherapeutics efficacy, and TMB

First, we investigated the correlation of irlncRNAs with the tumor microenvironment, and the results indicated that a high risk score was positively correlated with enrichment of resting natural killer (NK) and macrophage M1 cells, but not of CD4⁺ T cells (Figure 6A-6D and available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf).

Figure 6 Correlation analysis of the signature with tumor-infiltrating immune cells, immunosuppressed biomarkers, and the IC₅₀ of commonly used chemotherapeutics drugs. (A) A high risk score was positively correlated with the enrichment of resting NK cells and M1 macrophages and negatively correlated with the enrichment of CD4+ T cells. (B-D) Box diagram showing the enrichment of immune cell infiltration in the high- and low- risk groups. (E) The correlation of risk score with CD244. (F) The correlation between risk and IC₅₀ for dasatinib. *, P<0.05. IC₅₀, half-maximal inhibitory concentration.

We further examined the differential expression of crucial immune checkpoint genes in the different risk groups. We discovered that a high risk score was significantly associated with a high expression of CD244 (P<0.05, Figure 6E) but not with that of CD274 or CTLA4 (available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf). These outcomes suggest that the signature substantively reflects the immune process of the tumor microenvironment and, to a certain extent, the tumor response to immunotherapy.

We investigated the relationship between the risk score and chemotherapeutics efficacy in ICC. The results showed that a low risk score was positively related with a higher IC₅₀ of the chemotherapeutic agents dasatinib (P=0.03), docetaxel (P=0.25), gemcitabine (P=0.87), rapamycin (P=0.19), and cisplatin (P=0.28) (Figure 6F, available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf); however, the other agents exhibited no significant relationship.

Finally, we found that patients with a high-risk score had a higher TMB score, but this did not represent a significant difference (available online: https://cdn.amegroups.cn/static/public/tcr-2025-705-1.pdf). Overall, the risk score may be somewhat capable of predicting chemotherapeutic efficacy and TMB.

Discussion

ICC is a lethal hepatobiliary malignancy associated with a low early diagnosis rate and poor outcomes (1). In recent years, the interest in identifying novel tumor prognostic biomarkers for tumor treatment has intensified, and with the development of sequencing technology, lncRNAs have been recognized as elements capable of affecting tumorigenesis, tumor invasion, and the tumor immune microenvironment (34,35).

In our study, we extracted transcriptome data of lncRNAs and mRNAs from TCGA, along with IRGs from the ImmPort database, and performed coexpression, differential expression, and Cox regression analyses to develop an irlncRNA prognostic signature. Subsequently, we implemented the KM survival curves and calculated the AUC values. Our findings suggested that the signature has excellent predictive and discriminative ability. It has been reported that patients with primary hepatobiliary neoplasm—including ICC—have higher serum levels of primary bile acid (36), which is considered a metabolic biomarker. In our study, GSEA revealed that our signature was significantly enriched in primary bile acid biosynthesis pathway, indicating that bile acid synthesis is increased in these high-risk patients.

irlncRNA prognostic signatures have been constructed and validated in a variety of malignant tumors. Hong et al. constructed a novel irlncRNA pair signature based on TCGA database for predicting the prognosis of patients with hepatocellular carcinoma (32). Xie et al. developed a five-lncRNA signature for predicting the OS of patients with ICC and validated it on a prospective cohort (37). However, thus far, no irlncRNA-based prognostic signatures for ICC have been devised; thus, we attempted to develop an irlncRNA signature for predicting prognosis and characterizing the immune status of patients with ICC.

A portion of the irlncRNAs included in the signature have also been confirmed to play a role in various malignancies and immune signaling pathways. Although AL391056.1 has not been mentioned in the literature, Ju et al. proposed APCDD1L-DT as a prognostic indicator in lung squamous cell carcinoma (38). Pan et al. also indicated that APCDD1L-DT can contribute to the prediction of prognosis and immune response evaluation in patients with clear-cell renal cell carcinoma (39). Wang et al. performed a pan-cancer analysis and found that intensive WAC-AS1 figures prominently in the regulation of immune response, immune cell infiltration, and malignant properties (40). In Ji et al.’s study, LINC01615 was found to participate in metastasis-related process and was associated with extracellular matrix organization in hepatocellular carcinoma (41). Moreover, and Xi et al. identified LINC01615 to be a progression-related lncRNA in colorectal cancer, exerting meaningful effects in the immune microenvironment (42). Meanwhile, Xu et al. revealed that exosomal LINC01711 promotes cancer proliferation, migration, and invasion by upregulating FSCN1 and downregulating miR-326 to promote tumor emergence and progression (43). Hu et al. constructed and validated a prognostic signature including LINC01711 and found significant differences in the immune cell composition between the related risk groups, suggesting that LINC01711 contributes to the modulation of immune signaling pathways (44). Zhao et al. developed a risk model that included AC090114.2 for the prediction of prognosis and the immune microenvironment of patients with pancreatic cancer (45). Consistent with previous research, our study has preliminarily elucidated the key role of six irlncRNAs in regulating immune signaling pathways. It is hoped that the findings related to the signature’s irlncRNAs can support future research into the mechanisms of ICC.

Immunotherapy is a relatively a novel treatment strategy, but the nature of the tumor immune microenvironment may influence the response to this treatment (46). Ino et al. demonstrated that the patients with the better prognosis had a higher degree of CD4⁺ T-cell and CD8⁺ T-cell infiltration (47). Previous studies have reported a significantly high expression of irlncRNAs in human immune cells, with several irlncRNAs in prognostic signatures being highly correlated with immune cell infiltration (48). Thus, our study evaluated the relationship of the risk score with the abundance of tumor-infiltrating immune cells. The results indicated that the resting NK cells and M1 macrophages had a high degree of infiltration in the high-risk group, while CD4⁺ T cells were more enriched in the low-risk group, indicating a greater infiltration of CD4⁺ T cells portends a better prognosis.

Immune checkpoint inhibitors can modulate the interaction between immune and tumor cells. The expression level of immune checkpoint genes is a critical factor in immunotherapy response. Recently, Kim et al. found that CD244 could be used as an immunotherapy biomarker in combination with immune checkpoint inhibitors for patients with cancer (49). In our study, a high risk score was significantly correlated with a high expression of CD244. Thus, our signature could help identify patients who may benefit from CD244-dependent immunotherapy. However, assessments for the correlation between irlncRNAs and CD244 expression are generally lacking. To our knowledge, our study is the first to evaluate the impact of the six lncRNAs included in the signature on CD244 level, and further prospective and experimental studies are needed to confirm the efficacy of our signature.

We further examined the correlation of signature with the sensitivity to common chemotherapeutic agents. The low-risk group had a higher IC₅₀ for dasatinib as compared to the high-risk group. Dasatinib, a targeted drug, has produced good results in a variety of tumors, especially hematologic tumors, and a recent study found ICC cells with the isocitrate dehydrogenase (IDH) mutants had a high sensitivity and meaningful response to dasatinib (50). Therefore, the efficacy of dasatinib for patients with ICC needs be further clarified in clinical trials. Gemcitabine, cisplatin, and rapamycin also had a higher IC₅₀ in the low-risk group, but not significantly so. Conventional chemotherapeutics have a certain ability to influence immunostimulatory effects, and thus our signature may help clarify the interaction between immunotherapy and chemotherapy to inform the systematic therapy of patients with ICC.

Certain limitations to this study should be addressed. First, all the data for the retrospective analysis were obtained from public databases, and thus selection bias was inevitable. Second, several irlncRNAs were not identified in the GEO database and thus lacked external validation. Therefore, the efficacy of our prognostic signature remains to be validated in a large, prospective cohort.

Conclusions

We constructed and validated a novel irlncRNA-based prognostic signature, which could predict the prognosis of patients with ICC and help identify those who could benefit from immunotherapy. Hopefully, our findings can contribute to further clarifying the tumor immune microenvironment of ICC and to identifying relevant biomarkers.

Acknowledgments

We would like to thank TCGA, GEO and ImmPort databases for providing their platforms and contributors for uploading their meaningful datasets.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-705/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-705/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-705/prf

Funding: This study was supported by Guilin Science and Technology Plan Project (No. 20220139-5-3), The Technology Project within the No. 924 Hospital of PLA Joint Logistic Support Force (No. S2022FH01), Guangxi Healthcare Science and Technology Plan Project (No. Z-C20241574), and Natural Science Foundation of Guangxi Zhuang Autonomous Region (No. 2025GXNSFBA069028).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-705/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the ethics committee of the No. 924 Hospital of PLA Joint Logistic Support Force (No. GY2024-93). Written informed consent was obtained from all patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Moris D, Palta M, Kim C, et al. Advances in the treatment of intrahepatic cholangiocarcinoma: An overview of the current and future therapeutic landscape for clinicians. CA Cancer J Clin 2023;73:198-222. [Crossref] [PubMed]
2. Qurashi M, Vithayathil M, Khan SA. Epidemiology of cholangiocarcinoma. Eur J Surg Oncol 2025;51:107064. [Crossref] [PubMed]
3. Bragazzi MC, Venere R, Ribichini E, et al. Intrahepatic cholangiocarcinoma: Evolving strategies in management and treatment. Dig Liver Dis 2024;56:383-93. [Crossref] [PubMed]
4. Ruff SM, Pawlik TM. Clinical management of intrahepatic cholangiocarcinoma: surgical approaches and systemic therapies. Front Oncol 2024;14:1321683. [Crossref] [PubMed]
5. Zheng Z, Wang J, Wu T, et al. Hepatic arterial infusion chemotherapy plus targeted therapy and immunotherapy versus systemic chemotherapy for advanced intrahepatic cholangiocarcinoma: a retrospective cohort study. Int J Surg 2025;111:1552-7. [Crossref] [PubMed]
6. Qu X, Wang Q, Zhu F, et al. Research hotspots and trends in immunotherapy for cholangiocarcinoma: a bibliometric analysis (2014-2023). Front Immunol 2024;15:1436315. [Crossref] [PubMed]
7. Ferrer J, Dimitrova N. Transcription regulation by long non-coding RNAs: mechanisms and disease relevance. Nat Rev Mol Cell Biol 2024;25:396-415. [Crossref] [PubMed]
8. Chen LL, Kim VN. Small and long non-coding RNAs: Past, present, and future. Cell 2024;187:6451-85. [Crossref] [PubMed]
9. Nejadi Orang F, Abdoli Shadbad M. Competing endogenous RNA networks and ferroptosis in cancer: novel therapeutic targets. Cell Death Dis 2024;15:357. [Crossref] [PubMed]
10. Gao Y, Zhang Z, Huang X, et al. HBV-associated hepatocellular carcinomas inhibit antitumor CD8(+) T cell via the long noncoding RNA HDAC2-AS2. Nat Commun 2025;16:2055. [Crossref] [PubMed]
11. Wang Q, Sun Z, Guo J, et al. Tumor-derived exosomal LINC01812 induces M2 macrophage polarization to promote perineural invasion in cholangiocarcinoma. Cancer Lett 2025;617:217596. [Crossref] [PubMed]
12. Park EG, Pyo SJ, Cui Y, et al. Tumor immune microenvironment lncRNAs. Brief Bioinform 2022;23:bbab504. [Crossref] [PubMed]
13. Yao H, Huang C, Zou J, et al. Extracellular vesicle-packaged lncRNA from cancer-associated fibroblasts promotes immune evasion by downregulating HLA-A in pancreatic cancer. J Extracell Vesicles 2024;13:e12484. [Crossref] [PubMed]
14. Zhang M, Zhang F, Wang J, et al. Comprehensive characterization of stemness-related lncRNAs in triple-negative breast cancer identified a novel prognostic signature related to treatment outcomes, immune landscape analysis and therapeutic guidance: a silico analysis with in vivo experiments. J Transl Med 2024;22:423. [Crossref] [PubMed]
15. Cao R, Yuan L, Ma B, et al. Immune-related long non-coding RNA signature identified prognosis and immunotherapeutic efficiency in bladder cancer (BLCA). Cancer Cell Int 2020;20:276. [Crossref] [PubMed]
16. Collins GS, Moons KGM, Dhiman P, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ 2024;385:e078378. [Crossref] [PubMed]
17. Bhattacharya S, Andorf S, Gomes L, et al. ImmPort: disseminating data to the public for the future of immunology. Immunol Res 2014;58:234-9. [Crossref] [PubMed]
18. Ritchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. [Crossref] [PubMed]
19. Ringnér M. What is principal component analysis? Nat Biotechnol 2008;26:303-4. [Crossref] [PubMed]
20. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102:15545-50. [Crossref] [PubMed]
21. Zou W, Wang H, Wu D, et al. ncRNA-mediated upregulation of FAM83A is associated with poor prognosis and immune infiltration in pancreatic cancer. Front Endocrinol (Lausanne) 2023;14:1093042. [Crossref] [PubMed]
22. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 2016;32:2847-9. [Crossref] [PubMed]
23. Li T, Fan J, Wang B, et al. TIMER: A Web Server for Comprehensive Analysis of Tumor-Infiltrating Immune Cells. Cancer Res 2017;77:e108-10. [Crossref] [PubMed]
24. Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res 2020;48:W509-14. [Crossref] [PubMed]
25. Chen B, Khodadoust MS, Liu CL, et al. Profiling Tumor Infiltrating Immune Cells with CIBERSORT. Methods Mol Biol 2018;1711:243-59. [Crossref] [PubMed]
26. Aran D. Cell-Type Enrichment Analysis of Bulk Transcriptomes Using xCell. Methods Mol Biol 2020;2120:263-76. [Crossref] [PubMed]
27. Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol 2017;18:220. [Crossref] [PubMed]
28. Finotello F, Mayer C, Plattner C, et al. Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data. Genome Med 2019;11:34. [Crossref] [PubMed]
29. Plattner C, Finotello F, Rieder D. Deconvoluting tumor-infiltrating immune cells from RNA-seq data using quanTIseq. Methods Enzymol 2020;636:261-85. [Crossref] [PubMed]
30. Dienstmann R, Villacampa G, Sveen A, et al. Relative contribution of clinicopathological variables, genomic markers, transcriptomic subtyping and microenvironment features for outcome prediction in stage II/III colorectal cancer. Ann Oncol 2019;30:1622-9. [Crossref] [PubMed]
31. Racle J, de Jonge K, Baumgaertner P, et al. Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data. Elife 2017;6:e26476. [Crossref] [PubMed]
32. Hong W, Liang L, Gu Y, et al. Immune-Related lncRNA to Construct Novel Signature and Predict the Immune Landscape of Human Hepatocellular Carcinoma. Mol Ther Nucleic Acids 2020;22:937-47. [Crossref] [PubMed]
33. Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS One 2014;9:e107468. [Crossref] [PubMed]
34. Nemeth K, Bayraktar R, Ferracin M, et al. Non-coding RNAs in disease: from mechanisms to therapeutics. Nat Rev Genet 2024;25:211-32. [Crossref] [PubMed]
35. Chodurska B, Kunej T. Long non-coding RNAs in humans: Classification, genomic organization and function. Noncoding RNA Res 2025;11:313-27. [Crossref] [PubMed]
36. Cai D, Zhong GC, Dai X, et al. Targeting FDFT1 Reduces Cholesterol and Bile Acid Production and Delays Hepatocellular Carcinoma Progression Through the HNF4A/ALDOB/AKT1 Axis. Adv Sci (Weinh) 2025;12:e2411719. [Crossref] [PubMed]
37. Xie X, Wang Y, Zhang S, et al. A novel five-lncRNA signature panel improves high-risk survival prediction in patients with cholangiocarcinoma. Aging (Albany NY) 2021;13:2959-81. [Crossref] [PubMed]
38. Ju Q, Zhao YJ, Ma S, et al. Genome-wide analysis of prognostic-related lncRNAs, miRNAs and mRNAs forming a competing endogenous RNA network in lung squamous cell carcinoma. J Cancer Res Clin Oncol 2020;146:1711-23. [Crossref] [PubMed]
39. Pan J, Hu D, Huang X, et al. Identification of a cancer driver gene-associated lncRNA signature for prognostic prediction and immune response evaluation in clear cell renal cell carcinoma. Transl Cancer Res 2024;13:3418-36. [Crossref] [PubMed]
40. Wang Y, Gong H, Cao Y. LncRNA WAC-AS1 expression in human tumors correlates with immune infiltration and affects prognosis. Hereditas 2023;160:26. [Crossref] [PubMed]
41. Ji D, Chen GF, Liu X, et al. Identification of LINC01615 as potential metastasis-related long noncoding RNA in hepatocellular carcinoma. J Cell Physiol 2019;234:12964-70. [Crossref] [PubMed]
42. Xi W, Sun X, Wang M, et al. Identification of progression related LncRNAs in colorectal cancer aggressiveness. Sci Rep 2025;15:17258. [Crossref] [PubMed]
43. Xu ML, Liu TC, Dong FX, et al. Exosomal lncRNA LINC01711 facilitates metastasis of esophageal squamous cell carcinoma via the miR-326/FSCN1 axis. Aging (Albany NY) 2021;13:19776-88. [Crossref] [PubMed]
44. Hu J, Tian S, Liu Q, et al. A prognostic signature of Glutathione metabolism-associated long non-coding RNAs for lung adenocarcinoma with immune microenvironment insights. Front Immunol 2025;16:1477437. [Crossref] [PubMed]
45. Zhao K, Li X, Shi Y, et al. A Comprehensive Analysis of Pyroptosis-Related lncRNAs Signature Associated With Prognosis and Tumor Immune Microenvironment of Pancreatic Adenocarcinoma. Front Genet 2022;13:899496. [Crossref] [PubMed]
46. Luri-Rey C, Teijeira Á, Wculek SK, et al. Cross-priming in cancer immunology and immunotherapy. Nat Rev Cancer 2025;25:249-73. [Crossref] [PubMed]
47. Ino Y, Yamazaki-Itoh R, Shimada K, et al. Immune cell infiltration as an indicator of the immune microenvironment of pancreatic cancer. Br J Cancer 2013;108:914-23. [Crossref] [PubMed]
48. Luan J, Zhang D, Liu B, et al. Immune-related lncRNAs signature and radiomics signature predict the prognosis and immune microenvironment of glioblastoma multiforme. J Transl Med 2024;22:107. [Crossref] [PubMed]
49. Kim J, Kim TJ, Chae S, et al. Targeted deletion of CD244 on monocytes promotes differentiation into anti-tumorigenic macrophages and potentiates PD-L1 blockade in melanoma. Mol Cancer 2024;23:45. [Crossref] [PubMed]
50. Luk IS, Bridgwater CM, Yu A, et al. SRC inhibition enables formation of a growth suppressive MAGI1-PP2A complex in isocitrate dehydrogenase-mutant cholangiocarcinoma. Sci Transl Med 2024;16:eadj7685. [Crossref] [PubMed]

(English Language Editor: J. Gray)