Skip to main content
NCBI home page
BioMedical Engineering OnLine logoLink to BioMedical Engineering OnLine
. 2025 Aug 26;24:103. doi: 10.1186/s12938-025-01431-3

Unveiling the diagnostic power of lncRNAs in colorectal cancer: a meta-analysis

Wen Chen 1,#, Xinliang Liu 2,#, Zhenheng Wu 3,#, Haifen Tan 4,#, Fuqian Yu 5, Dongmei Wang 6,✉, Xiaodan Lin 7,✉, Zhigang Chen 6,✉
PMCID: PMC12379428  PMID: 40859296

Abstract

Background

Colorectal cancer (CRC) is a highly aggressive and extensive malignancy. Although long noncoding RNAs (lncRNAs) are often used as diagnostic biomarkers, their diagnostic effectiveness in CRC remains uncertain.

Methods

From January 1, 2015, to April 1, 2024, we conducted a comprehensive search of Embase, China National Knowledge Infrastructure (CNKI), Wanfang, PubMed, Cochrane Library, and Web of Science (WoS). The pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), area under the receiver operating characteristic curve (AUC) and Fagan plot analysis were used to assess the overall test performance of lncRNAs. Moreover, we evaluated the publication bias using the Deeks’ funnel plot asymmetry test.

Results

Twenty-eight publications were identified and incorporated into this meta-analysis. The aggregated diagnostic data were as follows: The pooled sensitivity was 0.79 (95% CI, 0.75–0.83). The pooled specificity was 0.81 (95% CI, 0.78–0.84). The PLR was 3.68 (95% CI, 3.18–4.26). The NLR was 0.28 (95% CI, 0.24–0.33). The DOR was 15.01 (95% CI, 11.85–19.00). The AUC was 0.87 (95% CI, 0.84–0.90). Deeks’ funnel plot asymmetry test indicated no significant evidence of publication bias (p > 0.05). The Fagan plot analysis showed that the post-test probability was 81% for positive results and 20% for negative results. Univariate meta-regression identified multiple sources of heterogeneity in the data, including year, sample size and specimen.

Conclusion

In summary, our findings demonstrate that lncRNAs have a promising diagnostic accuracy for CRC, underscoring their potential as effective non-invasive biomarkers.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12938-025-01431-3.

Keywords: LncRNAs, Colorectal cancer (CRC), Sensitivity, Specificity, Diagnostic performance

Introduction

Colorectal cancer (CRC) is one of the most common cancers of the gastrointestinal tract and a leading cause of cancer-related death [1–3]. The implementation of CRC screening programs has improved early detection rates, but many CRC patients are still diagnosed at a later stage, and they often miss the opportunity for curative excision [4]. Despite great improvements, current treatment options and survival rates for advanced colorectal cancer remain limited. Furthermore, CRC also exerts significant impacts on both physical and mental functioning, characterized by elevated stress levels, diminished self-worth and heightened rates of mood disorders [5]. Therefore, timely identification of CRC has the ability to avert these significant consequences.

The newly discovered non-coding RNAs are called long noncoding RNAs (lncRNAs), defined as > 200 nucleotides in length, with mRNA-like transcripts but no protein-coding capabilities [6, 7]. In fact, an increasing number of studies have demonstrated that lncRNAs are closely related to the occurrence, progression, prognosis, drug resistance and sensitivity of tumors. Therefore, studies are necessary to incorporate lncRNAs into preclinical models to develop diagnostic biomarkers.

In recent years, many studies have confirmed that lncRNAs can be used as biomarkers for the diagnosis of colorectal cancer [8–13]. A study conducted in Egypt confirmed that ASB16-AS1 has been employed to diagnose CRC in individuals with a diagnostic accuracy of up to 99.6% [12]. Futhermore, another Chinese study found the combined use of LINC01410 with traditional serum markers (CEA and CA199) improved sensitivity to 92.6%, outperforming each individual marker [13]. The study on the use of lncRNAs for CRC detection is growing. However, the accuracy of these lncRNAs varies greatly in different studies, mostly due to differences in algorithmic techniques and diagnostic criteria.

Imaging methods such as PET and MRI are widely used in public health services because they can simply and rapidly detect CRC lesions. However, these technologies are often costly, and more economical technologies will be needed in the future. Therefore, this study aims to systematically review and conduct a meta-analysis of the diagnostic accuracy of lncRNAs in detecting CRC, providing a theoretical basis for clinical diagnosis.

Materials and methods

This study adhered to the reporting elements recommended by the Preferred Reporting Items for Systematic Reviews and Meta—Analyses (PRISMA) standards [14]. The protocol was registered on the International Prospective Register of Ongoing Systematic Reviews (PROSPERO Number: CRD42024566443; https://www.crd.york.ac.uk/prospero).

Search strategy

We conducted an extensive review of the literature in several databases, starting from January 1, 2015, to April 1, 2024. The databases included in the search were the China National Knowledge Infrastructure (CNKI), PubMed, Cochrane Library, Embase, Wanfang, and Web of Science. Only studies published in Chinese or English were considered and included.

Search subject terms included “cancer”, “CRC”, “Colorectal cancer”, “ncRNA”, “non-coding RNA”, “lncRNA”, “Long non-coding RNA”, “True positive (TP)”, “False positive (FP)”, “Sensitivity”, “Specificity”, “AUC”, and “Area under the curve”. If the two researchers had differences on literature retrieval, discrepancies were resolved by consensus or through consultation with a third reviewer.

Inclusion criteria and exclusion criteria

Inclusion criteria: (1) Published articles evaluating the value of lncRNAs in the diagnosis of colorectal cancer; (2) Cohort studies with case-control studies or diagnostic experiments; (3) patients were diagnosed with CRC by histopathology; (4) Sensitivity, TP, FN or other estimable indices can be calculated.

Exclusion criteria: (1) Systematic reviews, narrative reviews, case reports, and conference; (2) Studies with incomplete data (e.g., sensitivity and specificity could not be obtained); (3) Duplicate publications; (4) Full-text articles not available; (5) Retracted articles or publications with unresolved concerns regarding scientific integrity.

Data extraction

Two independent researchers collected the following information from each publication: (1) Basic information, including the first author, publication year, lncRNA type, test samples, test methods, and the number of patients; (2) Diagnostic information, including sensitivity, specificity, area under the curve (AUC), and summary receiver operating characteristic curve (SROC); (3) When the data not directly reported, the sensitivity and specificity, etc. would be calculated through formulas.

Quality evaluation

Publication quality was evaluated using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) quality scoring standard [15]. This tool comprises 4 domains: patient selection, index test, reference standard, and flow and timing. The risk of bias or concerns in applicability in each domain is rated as either “low”, “high”, or “unclear”.

Statistical analysis

Metadisc 1.4 was used to evaluate the threshold effect, analyze the diagnostic accuracy data (PLR, NLR, DOR), and generate forest plots [16–18]. Stata version 16.0 (StataCorp LP, USA) was used to analyze sensitivity and meta-regression to explore the sources of heterogeneity. The statistical I2 and Q tests were used to analyze the heterogeneity. Subgroup analysis was also performed to explore the sources of heterogeneity. According to Moses-Littenberg method, SROC with pooled sensitivity and specificity were estimated. A deek funnel plot was drawn to detect publication bias. P < 0.05 was considered statistically significant. The positive and negative likelihood ratios (PLR and NLR) were calculated from the pooled sensitivity and specificity using the standard formulas: PLR = sensitivity/(1-specificity) and NLR = (1-sensitivity)/specificity [19]. Fagan plot was drawn to calculate the pre-test probability and post-test probability to evaluate the clinical value.

Result

Search results

Following the initial screening of titles and abstracts from a total of 1831 papers, 1099 were identified as redundant and subsequently excluded. A full-text review led to the exclusion of an additional 61 entries that did not meet the predetermined inclusion criteria. Ultimately, 28 studies were deemed eligible for analysis after satisfying all eligibility requirements. Figure 1 offers a detailed depiction of the literature screening process.

Fig. 1.

Fig. 1

Open in a new tab

Flow diagram

Basic characteristics

Table 1 provides a brief overview of all the studies included. Among them, most studies used serum to detect lncRNA, while others used plasma, tissue or feces. It covers research from multiple countries such as China, Iran, Italy and Egypt. The main type of research is retrospective study. The sample types include serum, plasma, tissue, feces, etc. The detection methods are mainly qPCR and RT-qPCR. Each study evaluated the performance indicators of different lncRNAs (such as BANCR, CCAT1, HOTAIR, UCA1, etc.) in cancer diagnosis, including sensitivity (Recall), Specificity, Precision, negative predictive value (NPV), Accuracy, AUC, Balanced-accuracy, F1 score, Youden index, etc. Some studies have also explored the diagnostic efficacy of combined detection (such as the combination of multiple lncRNAs or the combination of lncRNAs with other markers).

Table 1.

The main characteristics of the included studies in the meta-analysis

Authors Country Study type Specimen Assay method LncRNA Sample Recall Specificity Precision NPV Accuracy AUC Balanced − accuracy F1 Youden index
Rui Wang et al. 2016 China Retrospective Serum qPCR NR_029373 240 0.76 0.73 0.74 0.75 0.75 0.812 0.75 0.75 0.49
Rui Wang et al. 2016 China Retrospective Serum qPCR NR_034119 240 0.62 0.71 0.68 0.65 0.66 0.724 0.67 0.65 0.33
Rui Wang et al. 2016 China Retrospective Serum qPCR BANCR 240 0.63 0.62 0.62 0.63 0.63 0.638 0.63 0.63 0.25
Rui Wang et al. 2016 China Retrospective Serum qPCR NR_026817 240 0.68 0.84 0.66 0.67 0.66 0.708 0.76 0.67 0.52
Rui Wang et al. 2016 China Retrospective Serum qPCR Combined 240 0.82 0.80 0.80 0.81 0.81 0.891 0.81 0.81 0.62
Weimin Zhao et al. 2015 China Retrospective Plasma qPCR CCAT1 64 0.76 0.85 0.83 0.77 0.80 0.836 0.81 0.79 0.61
Weimin Zhao et al. 2015 China Retrospective Plasma qPCR HOTAIR 64 0.68 0.90 0.88 0.74 0.80 0.777 0.79 0.76 0.57
Weimin Zhao et al. 2015 China Retrospective Plasma qPCR LincRNA- p21 64 0.82 0.85 0.84 0.82 0.83 0.886 0.84 0.83 0.67
Jian Shi et al. 2015 China Retrospective Plasma qPCR XLOC_006844 328 0.78 0.81 0.89 0.64 0.79 0.804 0.80 0.83 0.59
Jian Shi et al. 2015 China Retrospective Plasma qPCR LOC152578 328 0.83 0.81 0.90 0.70 0.82 0.826 0.82 0.86 0.64
Jian Shi et al. 2015 China Retrospective Plasma qPCR XLOC_000303 328 0.78 1.00 1.00 0.69 0.85 0.891 0.89 0.88 0.78
Changyi Fang et al. 2017 China Retrospective Plasma qPCR ZFAS1 200 0.92 0.77 0.82 0.90 0.85 0.88 0.85 0.87 0.69
Meiyu Dai et al. 2017 China Retrospective Serum qPCR BLACAT1 60 0.83 0.77 0.78 0.82 0.80 0.858 0.80 0.81 0.60
Wanjun Gong et al. 2017 China Retrospective Serum qPCR HIF1A- AS1 311 0.87 0.93 0.92 0.88 0.90 0.96 0.90 0.89 0.79
Jing Li et al. 2019 China Retrospective Tumor tissue and paracancer tissue qPCR IncRNA-MVIH 134 0.70 0.83 0.81 0.74 0.77 0.799 0.77 0.75 0.53
Wei Wang et al. 2019 China Retrospective Tumor tissue and paracancer tissue RT-qPCR MEG3 84 0.67 0.88 0.85 0.73 0.77 0.798 0.77 0.75 0.54
Cristina Barbagallo et al. 2018 Italy Retrospective Serum RT-qPCR UCA1 40 1.00 0.43 0.65 1.00 0.73 0.719 0.72 0.78 0.43
Cristina Barbagallo et al. 2018 Italy Retrospective Serum RT-qPCR TUG1:UCA1 Combined 40 0.93 0.64 0.73 0.93 0.80 0.814 0.79 0.82 0.57
Ehsan Lotf et al. 2024 Iran Case–control study Plasma RT-qPCR UICLM 60 0.85 0.93 0.93 0.88 0.90 0.883 0.89 0.89 0.78
Ehsan Lotf et al. 2024 Iran Case–control study Plasma RT-qPCR BBOX1-AS1 60 0.68 0.67 0.69 0.71 0.70 0.719 0.68 0.68 0.35
Ehsan Lotf et al. 2024 Iran Case–control study Plasma RT-qPCR FEZF1-AS1 60 0.69 0.63 0.65 0.66 0.65 0.672 0.66 0.67 0.32
Ehsan Lotf et al. 2024 Iran Case–control study Plasma RT-qPCR CCAT1 60 0.67 0.62 0.67 0.70 0.68 0.731 0.65 0.67 0.29
Ehsan Lotf et al. 2024 Iran Case–control study Plasma RT-qPCR LINC00698 60 0.67 0.60 0.64 0.00 0.50 0.701 0.64 0.65 0.27
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR ASB16-AS1 96 0.94 0.90 0.90 0.94 0.92 0.967 0.92 0.92 0.83
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR ASB16-AS1 96 0.89 0.86 0.86 0.89 0.88 0.934 0.88 0.87 0.75
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR AFAP1-AS1 96 0.91 0.84 0.84 0.91 0.88 0.944 0.88 0.88 0.75
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR AFAP1-AS1 96 0.85 0.80 0.80 0.85 0.82 0.923 0.82 0.82 0.65
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR ASB16-AS1 97 0.96 0.92 0.92 0.96 0.94 0.996 0.94 0.94 0.88
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR ASB16-AS1 97 0.91 0.88 0.88 0.92 0.90 0.974 0.90 0.90 0.79
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR AFAP1-AS1 97 0.91 0.90 0.90 0.92 0.91 0.984 0.91 0.91 0.81
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR AFAP1-AS1 97 0.87 0.84 0.84 0.88 0.86 0.965 0.86 0.85 0.71
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR ASB16-AS1 60 0.88 0.82 0.53 0.98 0.83 0.909 0.85 0.66 0.70
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR ASB16-AS1 60 0.82 0.92 0.69 0.96 0.90 0.888 0.87 0.75 0.74
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR AFAP1-AS1 60 0.76 0.86 0.53 0.93 0.83 0.900 0.81 0.63 0.62
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR AFAP1-AS1 60 0.71 0.82 0.47 0.93 0.80 0.842 0.76 0.56 0.52
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR ASB16-AS1 67 0.94 0.92 0.80 0.98 0.93 0.988 0.93 0.86 0.86
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR ASB16-AS1 67 0.88 0.62 0.44 0.94 0.69 0.949 0.75 0.59 0.50
Naglaa S Elabd et al. 2022 Egypt Retrospective Tissue qPCR AFAP1-AS1 67 0.82 0.90 0.74 0.94 0.88 0.973 0.86 0.78 0.72
Naglaa S Elabd et al. 2022 Egypt Retrospective Plasma qPCR AFAP1-AS1 67 0.71 0.84 0.60 0.89 0.81 0.925 0.77 0.65 0.55
Xianjuan Shen et al. 2020 China Retrospective Serum RT-PCR DANCR 50 0.68 0.88 0.58 0.92 0.84 0.745 0.78 0.63 0.55
Xianjuan Shen et al. 2020 China Retrospective Serum RT-PCR CEA 50 0.40 0.85 0.40 0.85 0.76 0.623 0.63 0.40 0.25
Xianjuan Shen et al. 2020 China Retrospective Serum RT-PCR CA199 50 0.33 0.80 0.27 0.82 0.70 0.573 0.56 0.30 0.13
Q-G Li et al. 2019 China Retrospective Tissue RT-PCR lnc-DILC 174 0.78 0.71 0.73 0.77 0.75 0.8264 0.75 0.75 0.49
Shadi Sadri et al. 2021 Iran Retrospective Stool qRT-PCR AUROC 40 0.95 0.80 0.83 0.94 0.88 0.95 0.88 0.88 0.75
Shadi Sadri et al. 2021 Iran Retrospective Stool qRT-PCR BANCK 40 0.59 0.61 0.60 0.60 0.60 0.66 0.60 0.59 0.20
Nehal Samir et al. 2017 Egypt Retrospective Serum RT-PCR lnc-RNA-RPH 60 0.80 0.93 0.97 0.70 0.85 0.867 0.87 0.88 0.73
Nehal Samir et al. 2017 Egypt Retrospective Serum RT-PCR Combined miRNA595, lncRNARP11, LAMPT1 gene exp, CEA OR CA19.9 60 0.57 0.73 0.82 0.47 0.63 Not mentioned 0.65 0.67 0.30
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR lncRNA XIST 12 0.88 0.90 0.83 0.83 0.83 0.864 0.89 0.86 0.79
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR linc02037 12 0.56 0.83 0.75 0.63 0.67 0.698 0.70 0.64 0.39
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR linc01987 12 0.93 0.78 0.86 1.00 0.92 0.856 0.85 0.89 0.71
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR linc02041 12 0.62 0.90 0.80 0.71 0.75 0.774 0.76 0.70 0.52
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR TET2-AS1 12 0.96 0.71 0.75 1.00 0.83 0.828 0.83 0.84 0.66
Jinfeng Yu et al. 2020 China Retrospective Serum RT-qPCR linc174 12 0.85 0.78 0.83 0.83 0.83 0.849 0.82 0.84 0.63
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 89 0.86 0.50 0.70 0.73 0.71 0.894 0.68 0.77 0.36
Jin Xu et al. 2020 China Retrospective Serum RT-PCR CEA 89 0.26 0.92 0.81 0.48 0.54 0.701 0.59 0.39 0.18
Jin Xu et al. 2020 China Retrospective Serum RT-PCR CA199 89 0.28 0.92 0.82 0.49 0.55 0.640 0.60 0.41 0.20
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 + CEA 89 0.90 0.46 0.70 0.78 0.72 Not reported 0.68 0.78 0.36
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 + CA199 89 0.90 0.46 0.70 0.78 0.72 Not reported 0.68 0.79 0.36
Jin Xu et al. 2020 China Retrospective Serum RT-PCR Combined detection 89 0.93 0.34 0.65 0.76 0.67 Not reported 0.63 0.77 0.27
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 89 0.86 0.84 0.85 0.86 0.85 0.656 0.85 0.85 0.70
Jin Xu et al. 2020 China Retrospective Serum RT-PCR CEA 89 0.26 0.92 0.76 0.55 0.58 0.651 0.59 0.38 0.18
Jin Xu et al. 2020 China Retrospective Serum RT-PCR CA199 89 0.26 0.92 0.76 0.55 0.58 0.652 0.59 0.38 0.18
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 + CEA 89 0.90 0.77 0.81 0.89 0.84 Not reported 0.84 0.85 0.67
Jin Xu et al. 2020 China Retrospective Serum RT-PCR LINC01410 + CA199 89 0.90 0.76 0.79 0.88 0.83 Not reported 0.83 0.84 0.66
Jin Xu et al. 2020 China Retrospective Serum RT-PCR Combined 89 0.93 0.70 0.76 0.90 0.81 Not reported 0.81 0.83 0.63
Ledong Wan et al. 2016 China Retrospective Plasma RT-PCR HOTAIRM1 251 0.64 0.77 0.80 0.59 0.69 0.78 0.70 0.71 0.41
Ledong Wan et al. 2016 China Retrospective Plasma RT-PCR CEA 251 0.56 0.91 0.90 0.58 0.70 Not reported 0.74 0.69 0.47
Ledong Wan et al. 2016 China Retrospective Plasma RT-PCR HOTAIRM1 + CEA 251 0.84 0.77 0.84 0.76 0.81 Not reported 0.80 0.84 0.61
Ehsan Gharib et al. 2020 Iran Retrospective Stool (I-IV TNM stages) qRT-PCR A predictive panel of 10 significantly dysregulated lncRNAs (CCAT1, CCAT2, H19, HOTAIR, HULC, MALAT1, PCAT1, MEG3, PTENP1 and TUSC7) 60 0.75 0.94 0.92 0.78 0.83 0.8465 0.85 0.82 0.69
Ehsan Gharib et al. 2020 Iran Retrospective Stool (I-II TNM stages) qRT-PCR A predictive panel of 10 significantly dysregulated lncRNAs (CCAT1, CCAT2, H19, HOTAIR, HULC, MALAT1, PCAT1, MEG3, PTENP1 and TUSC7) 60 0.68 0.84 0.80 0.71 0.75 0.8121 0.76 0.74 0.52
Ehsan Gharib et al. 2020 Iran Retrospective Stool (III-IV TNM stages) qRT-PCR A predictive panel of 10 significantly dysregulated lncRNAs (CCAT1, CCAT2, H19, HOTAIR, HULC, MALAT1, PCAT1, MEG3, PTENP1 and TUSC7) 60 0.78 0.96 0.96 0.85 0.90 0.9236 0.87 0.86 0.74
Yanli Zhang et al. 2020 China Retrospective Serum RT-qPCR 5-exolncRNAs (AF079515, CCAT1, UCA1, RP11-434B12.1 and HOTTIP) 43 0.89 0.86 0.93 0.79 0.88 0.947 0.87 0.91 0.75
Rui Wang et al.2016 China Retrospective Serum RT-qPCR BANCR, NR_026817, NR_029373, and NR_034119 240 0.82 0.80 0.80 0.81 0.81 0.891 0.81 0.81 0.62
Zhaosheng Li et al. 2023 China Retrospective Serum qPCR lnc-PDZD8–1:5 + NEAT1:11 + LINC00910:16 186 0.75 0.81 0.82 0.72 0.77 0.85 0.78 0.78 0.55
Meng Xu et al.2020 China Retrospective Tissue qRT-PCR HANR 330 0.60 0.82 0.77 0.67 0.71 0.82 0.71 0.67 0.42
Yinghui Zhao et al.2019 China Retrospective Serum qPCR Exosomal LINC02418 250 0.95 0.66 0.74 0.93 0.81 0.8978 0.81 0.83 0.62
Xiangwei Sun et al.2016 China Retrospective Tissue qRT-PCR AK098081 235 0.79 0.78 0.86 0.67 0.78 0.721 0.78 0.82 0.57
Hossein Sadeghi et al. 2015 Iran Retrospective Tissue qRT-PCR CYP24A1 150 0.91 0.91 0.91 0.91 0.91 0.94 0.91 0.91 0.82
Hossein Sadeghi et al. 2015 Iran Retrospective Tissue qRT-PCR PFDN4 150 0.74 0.59 0.64 0.69 0.66 0.66 0.67 0.69 0.33
Hossein Sadeghi et al. 2015 Iran Retrospective Tissue qRT-PCR lnc-CYP24A1-3:1 150 0.85 0.55 0.65 0.79 0.70 0.7 0.70 0.74 0.40
Hossein Sadeghi et al. 2015 Iran Retrospective Tissue qRT-PCR lnc-TSHZ2-19:1 150 0.52 0.79 0.71 0.62 0.65 0.6 0.66 0.60 0.31
Yuchen Wu et al. 2015 China Retrospective Serum PCR NEAT1_v1 60 0.57 0.87 0.81 0.67 0.72 0.732 0.72 0.67 0.43
Yuchen Wu et al. 2015 China Retrospective Serum PCR NEAT1_v2 60 0.83 0.83 0.83 0.83 0.83 0.845 0.83 0.83 0.67
Yuchen Wu et al. 2015 China Retrospective Serum PCR NEAT1_v1 (variants) 60 0.96 0.79 0.82 0.95 0.88 0.787 0.88 0.88 0.75
Yuchen Wu et al. 2015 China Retrospective Serum PCR NEAT1_v2 (variants) 60 0.70 0.96 0.95 0.76 0.83 0.871 0.83 0.80 0.66
Zhexu Guo et al. 2019 China Retrospective Tissue PCR lncRNACTA-941F9.9 (cohort 1) 148 0.74 0.84 0.82 0.77 0.79 0.803 0.79 0.78 0.58
Zhexu Guo et al. 2019 China Retrospective Tissue PCR lncRNACTA-941F9.9 (cohort 2) 118 0.78 0.85 0.84 0.79 0.81 0.876 0.82 0.81 0.63
Negin Sadi Khosroshahi et al. 2024 Iran Retrospective Tissue RT-PCR PCAT-1 200 0.65 0.65 0.65 0.65 0.65 0.6395 0.65 0.65 0.30
Open in a new tab

AUC: Area Under the Curve; NPV: Negative Predictive Value; qPCR: quantitative polymerase chain reaction; RT-PCR: real-time reverse-transcription polymerase chain reaction; RT-qPCR: real-time reverse-transcription polymerase chain reaction

Quality assessment of included studies

Table 2 presents the findings of the QUADAS-2 quality evaluation. The majority of the articles included in the current meta-analysis met most of the criteria outlined in the QUADAS-2 assessment, suggesting that the overall quality of the studies included was moderate to high. After a detailed review of the articles and its evidence, the articles were included in the final analysis. Although some research designs adopted case–control, the sample sources were relatively new, the groupings were clear, the designs were reasonable, and they met the inclusion criteria, etc.

Table 2.

Quality assessments of included studies by using the QUADAS-2 tool

Study Risk of bias Concerns regarding applicability
Patient selection Index test Reference standard Flow and timing Patient selection Index test Reference standard
Sabrina M. Heman-Ackah et al. 2024 Low Unclear Unclear Low Unclear Unclear Low
Youssef M. Zohdy et al. 2023 Unclear Low Low Low Unclear Low Low
Chih-Yu Kuo et al. 2023 Unclear Low Low Low High Low Unclear
Yang Su et al. 2023 Unclear Low Low Low Low Unclear Low
Kunyue Wang et al. 2023 Low Unclear Low Low Low Low Unclear
Christiaan H.B. van Niftrik et al. 2019 Unclear Low Unclear Low High Unclear Low
Nidan Qiao et al. 2023 Low Unclear Low Low Unclear Low High
J. Lo¨tsch et al. 2017 Low Unclear Low Low Unclear Low High
Jiali Du et al. 2023 Low Low Low Low High Low Unclear
Chibueze A. Nwaiwu et al. 2024 Low Unclear Low Low High Unclear High
Todd C. Hollon et al. 2018 Low Low Low High High Unclear Unclear
Matthew C. Hernandez et al. 2024 Low Low Low High High Unclear Low
Jingwen Zhang et al. 2022 High Unclear Low Unclear Low Low Low
Dongsong Wu et al. 2023 Low Low Unclear Low Unclear Low Low
Qinxian Zhao et al. 2020 Low Low Unclear Low Low Unclear Unclear
Jiangying Li et al. 2024 Low Unclear Unclear Low Low Unclear Low
Wenhua Li et al. 2022 Low Low Unclear Unclear High Unclear Low
Yang Gao et al. 2023 Low Low Low Low Unclear High Unclear
Chuanli Liu et al. 2022 Low Low Low High High Unclear Low
Yangda Song et al. 2023 Low Low Low Low Unclear Low Low
Rong-yun Mai et al. 2021 Low Unclear Low Low High Unclear Unclear
Hui Qu et al. 2024 Low Unclear Low Low Low Low Low
Wenjuan Zhang et al. 2020 Low Low Unclear Low Low Unclear Low
Laura Alaimo et al. 2023 Low Low Low Unclear Unclear Low Low
Yubo Zhang et al. 2023 Low Low Unclear Low Unclear Low Low
Jihwan Park et al. 2024 Low Low Unclear Unclear High Unclear Unclear
Ying Zhao et al. 2024 High Low Low Low High Low Low
Giovanni Catalano et al. 2024 Low Unclear Unclear Low Low Unclear Low
Joshua S. Jolissaint et al. 2022 Low Low Low Unclear High Unclear Low
Qian Li et al. 2023 Low Low Low Low High Unclear Low
Jeong Hyun Lee et al. 2024 Unclear Low Unclear Low Low Low Low
Open in a new tab

QUADAS-2 Quality Assessment of Diagnostic Accuracy Studies-2

LncRNAs can accurately diagnose CRC.

The pooled sensitivity was 0.79 (95% CI, 0.75–0.83). The pooled specificity was 0.81 (95% CI, 0.78–0.84) (Fig. 2a, b). These findings indicate that lncRNAs are significantly more proficient in accurately detecting diseases than in accurately categorizing non-illness instances. In addition, an evaluation was performed on the combined diagnostic accuracy of the lncRNAs in diagnosing CRC. The pooled analysis revealed a Spearman’s correlation coefficient with a P value of less than 0.05. The sensitivity and specificity I2 values were 87.91 and 82.91%, respectively. The chi-square test P-values were all less than 0.05, indicating a considerable level of heterogeneity in the studies. The pooled results of the combined diagnostic data were as follows: The PLR was 3.68 (95% CI, 3.18–4.26). The NLR was 0.28 (95% CI, 0.24–0.33) (Fig. 3a, b). The pooled DOR was 15.01 (95% CI, 11.85–19.00) (Fig. 4a). Figure 4b shows the SROC curve that corresponds to the given data. The AUC was 0.87, demonstrating a substantial level of accuracy in diagnosing CRC utilizing lncRNAs.

Fig. 2.

Fig. 2

Open in a new tab

The sensitivity and specificity of lncRNA in diagnosing CRC. a Sensitivity forest plot. b Specificity forest plot

Fig. 3.

Fig. 3

Open in a new tab

PLR and NLR in the diagnosis of CRC by lncRNA. a PLR forest plot. b NLR forest plot

Fig. 4.

Fig. 4

Open in a new tab

The DOR and SROC curve of lncRNA in diagnosing CRC, and its clinical predictive value. a DOR forest plot. b SROC curve. c Fagan nomogram

Clinical value analysis

The clinical value disparities across lncRNAs for diagnosing CRC were assessed by Fagan plot analysis. The probability increased from 50 to 81% when the lncRNAs tested yielded good results and decreased to 20% when the results were negative (Fig. 4c).

Univariate meta-regression and subgroup analysis

To further analyze the source of heterogeneity, subgroup analysis and univariate meta-regression (Fig. 5, Supplementary Table 1) were performed. Covariates included country (China vs. others), sample size (< 150 vs. ≥ 150), publication year (Before 2019 and from 2020 onward), and specimen type (serum vs. plasma vs. tissue vs. stool). Meta-regression of all studies showed that lower sensitivity was associated with publication year from 2020 onward (0.79; 95% CI 0.74–0.83; P < 0.001), sample ≥ 150 (0.78; 95% CI 0.72–0.83; P < 0.001), studies from China (0.78; 95% CI 0.73–0.83; P < 0.001), the specimen was derived from serum (0.77; 95% CI 0.72–0.82; P < 0.001), from plasma (0.78; 95% CI 0.73–0.84; P < 0.001), and from tissue (0.78; 95% CI 0.71–0.85; P < 0.001). Higher specificity was associated with publication year < 2020 onward (0.84; 95% CI 0.80–0.88; P < 0.001), sample ≥ 150 (0.81; 95% CI 0.77–0.86; P < 0.001), studies from China (0.83; 95% CI 0.79–0.86; P < 0.001), the specimen was derived from tissue (0.83; 95% CI 0.78–0.88; P < 0.001), and from stool (0.86; 95% CI 0.75–0.96; P < 0.001).

Fig. 5.

Fig. 5

Open in a new tab

Univariate meta-regression and Subgroup Analysis

Publication bias

To evaluate the possibility of publication bias, we conducted the Deeks’ funnel plot asymmetry test in our meta-analysis (Fig. 6). A p-value of 0.83 suggests that there is a low probability of publication bias in the research [20, 21].

Fig. 6.

Fig. 6

Open in a new tab

The asymmetry test of Deeks’ funnel plot for publication

Discussion

CRC remains one of the most prevalent and lethal malignancies worldwide, with a significant proportion of patients diagnosed at advanced stages due to the lack of effective and accessible early detection methods [22]. The exceptional stability, conservation, and abundance of lncRNAs in body fluids, have made them potential biomarkers for a wide range of diseases [23]. Nevertheless, there is a scarcity of research investigating the precision of lncRNAs in the diagnosis of CRC.

The meta-analysis included 28 studies and used a diagnostic meta-analysis technique. The combined sensitivity and specificity of CRC were 0.79 (95% CI, 0.75–0.83) and 0.81 (95% CI, 0.78–0.84), respectively. LncRNAs showed moderate sensitivity and specificity for CRC diagnosis. In addition, the AUC in our meta-analysis results was 0.87 (95% CI, 0.84–0.90), suggesting that lncRNAs exhibited a promising accuracy in the diagnosis of CRC.

The studies from China (sensitivity 0.78, specificity 0.83) differed little from those from other regions (sensitivity 0.81, specificity 0.78), but were statistically significant. This phenomenon, which may be attributed to the larger sample sizes and greater homogeneity of Chinese cohorts, has raised concerns regarding the generalizability of these findings to non-Asian populations. In addition, genetic and environmental changes (such as diet and gut microbiota) may affect lncRNA expression [24]. A previous study reported that when LINC01410 was combined with CEA and CA199, the sensitivity was increased by 93%, outperforming by a single marker [13]. These findings emphasize the necessity of validating lncRNA biomarkers in different racial and environmental contexts to ensure their clinical applicability.

The heterogeneity among studies may also stem from the differences in detection techniques (such as qRT-PCR, RNA-seq, etc.) and diagnostic criteria. For example, some studies used qRT-PCR to quantitatively detect the expression levels of specific lncRNAs (such as HIF1A- AS1, MEG3), and took the preset Ct value or fold change as the diagnostic threshold [25, 26]. Other studies, however, combine multiple lncRNA combinations or traditional tumor markers (such as CEA and CA199) for comprehensive judgment [27].

In addition, the processing methods of sample types (serum, tissue or stool) and the standardized procedures of lncRNA (such as internal reference gene selection, RNA extraction kits) may also affect the consistency of the results [28, 29]. The performance of serum-based detection (with a comprehensive sensitivity of 0.77 and specificity of 0.78) is slightly lower than that of tissue samples (with a sensitivity of 0.78 and specificity of 0.83) or fecal samples (with a sensitivity of 0.78 and specificity of 0.86). This difference may reflect the variations in the stability and abundance of lncRNA in different biological body fluids [30]. Previous studies have also found that lncRNAs can prevent degradation by being packaged into microparticles (such as foreign bodies, microbubbles, apoptotic bodies) [31].

This study was limited in a few ways. First of all, we only included English studies published after 2015 and excluded other studies that might meet the conditions. Second, most of the studies included in our analyses originated from China, and the results obtained may include geographical bias. Third, most of the included studies were retrospective studies, which might have selection bias and affect the accuracy of diagnostic efficacy. Fourth, some studies use single lncRNA detection, while others use combined detection (such as lncRNA combination or combination with traditional tumor markers). In addition, the lncRNA expression profiles of different sample types may vary, affecting the diagnostic accuracy.

Although lncRNA shows a relatively high AUC in the diagnosis of CRC, its sensitivity and specificity still need to be further improved. In the future, multi-center, large-sample, and prospective studies should be carried out to improve the reliability and universality of the results.

Supplementary Information

Additional file 1. (14.7KB, docx)

Acknowledgements

We thank all authors and reviewers who participated in this work.

Author contributions

It was written by Wen Chen and Zhigang Chen. The manuscript was revised and subjected to critical discussion by Xinliang Liu, Zhenheng Wu, Haifen Tan and Fuqian Yu. Ideas and themes were created by Dongmei Wang, Xiaodan Lin, and Zhigang Chen. All authors have read and consented to publish this manuscript.

Funding

This work was supported by grants from the Key Clinical Specialty Discipline Construction Program of Fuzhou, Fujian, P.R.C (Grant No. 20220301), the Fuzhou Science and Technology talent training program (Grant No. 2024-R-009), the Fuzhou Science and Technology Plan Project (Grant No. 2024-S-056), and the Science and Technology Plan Project of Fujian Provincial Health Commission (Grant No. 2024QNA076).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Wen Chen, Xinliang Liu, Zhenheng Wu and Haifen Tan have contributed equally to this work.

Contributor Information

Dongmei Wang, Email: fautywang@163.com.

Xiaodan Lin, Email: linxiaodan4163@163.com.

Zhigang Chen, Email: czg99888@163.com.

References

  • 1.Yin H, Li H, Xu J, et al. Primary tumor location impacts survival in colorectal cancer patients after primary resection: a population-based propensity score matching cohort study. J Gastrointest Oncol. 2023;14:886–99. 10.21037/jgo-23-71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Geng Z, Chen M, Yu Q, et al. Histone modification of colorectal cancer by natural products. Pharmaceuticals (Basel). 2023. 10.3390/ph16081095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Xu M, Xi S, Li H, et al. Prognosis significance and potential association between aldoa and akt expression in colorectal cancer. Sci Rep. 2024;14:6488. 10.1038/s41598-024-57209-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Wu J, Chen Q, Wang Y, et al. Linc01977 promotes colorectal cancer growth and metastasis by enhancing aerobic glycolysis via the erk/c-myc axis. J Gastrointest Oncol. 2024;15:271–85. 10.21037/jgo-24-52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Xia J, Li D, Yu G, et al. Effects of hypovitaminosis d on preoperative pain threshold and perioperative opioid use in colorectal cancer surgery: a cohort study. Pain Physician. 2022;25:E1009–19. [PubMed] [Google Scholar]
  • 6.Liu Z, Liu L, Weng S, et al. Machine learning-based integration develops an immune-derived lncrna signature for improving outcomes in colorectal cancer. Nat Commun. 2022;13:816. 10.1038/s41467-022-28421-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Golam RM, Khalil M, Shaker OG, et al. The clinical significance of long non-coding rnas malat1 and casc2 in the diagnosis of hcv-related hepatocellular carcinoma. PLoS ONE. 2024;19: e303314. 10.1371/journal.pone.0303314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Lotfi E, Kholghi A, Golab F, et al. Circulating mirnas and lncrnas serve as biomarkers for early colorectal cancer diagnosis. Pathol Res Pract. 2024;255: 155187. 10.1016/j.prp.2024.155187. [DOI] [PubMed] [Google Scholar]
  • 9.Raza A, Khan AQ, Inchakalody VP, et al. Dynamic liquid biopsy components as predictive and prognostic biomarkers in colorectal cancer. J Exp Clin Cancer Res. 2022;41:99. 10.1186/s13046-022-02318-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Luo XJ, Zhao Q, Liu J, et al. Novel genetic and epigenetic biomarkers of prognostic and predictive significance in stage ii/iii colorectal cancer. Mol Ther. 2021;29:587–96. 10.1016/j.ymthe.2020.12.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hong J, Guo F, Lu SY, et al. F. Nucleatum targets lncrna eno1-it1 to promote glycolysis and oncogenesis in colorectal cancer. Gut. 2021;70:2123–37. 10.1136/gutjnl-2020-322780. [DOI] [PubMed] [Google Scholar]
  • 12.Elabd NS, Soliman SE, Elhamouly MS, et al. Long non-coding rnas asb16-as1 and afap1-as1: diagnostic, prognostic impact and survival analysis in colorectal cancer. Appl Clin Genet. 2022;15:97–109. 10.2147/TACG.S370242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Xu J, Wang L, Wang Q. High expression of long noncoding rna 01410 serves as a potential diagnostic and prognostic marker in patients with colorectal cancer. Clin Lab. 2021. 10.7754/Clin.Lab.2020.200805. [DOI] [PubMed] [Google Scholar]
  • 14.Acharya P, Amin A, Nallamotu S, et al. Prehospital tranexamic acid in trauma patients: a systematic review and meta-analysis of randomized controlled trials. Front Med (Lausanne). 2023;10:1284016. 10.3389/fmed.2023.1284016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Whiting PF, Rutjes AW, Westwood ME, et al. Quadas-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155:529–36. 10.7326/0003-4819-155-8-201110180-00009. [DOI] [PubMed] [Google Scholar]
  • 16.Zheng X, Li W, Yan Y, et al. Association between the dietary inflammatory index and fracture risk in older adults: a systematic review and meta-analysis. J Int Med Res. 2024;52:645658329. 10.1177/03000605241248039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Hsu CY, Saver JL, Ovbiagele B, et al. Association between magnitude of differential blood pressure reduction and secondary stroke prevention: a meta-analysis and meta-regression. Jama Neurol. 2023;80:506–15. 10.1001/jamaneurol.2023.0218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fu X, Qi H, Qiu Z, et al. Outcomes of two types of iodine-125 seed delivery with metal stents in treating malignant biliary obstruction: a systematic review and meta-analysis. Diagn Interv Radiol. 2023;29:509–19. 10.5152/dir.2022.211277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kodama S, Fujihara K, Horikawa C, et al. Predictive ability of current machine learning algorithms for type 2 diabetes mellitus: a meta-analysis. J Diabetes Investig. 2022;13:900–8. 10.1111/jdi.13736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Chen W, Chen Y, Wu C, et al. The accuracy of fiber-optic raman spectroscopy in the detection and diagnosis of head and neck neoplasm in vivo: a systematic review and meta-analysis. PeerJ. 2023;11: e16536. 10.7717/peerj.16536. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Dehghanbanadaki H, Asili P, Haji GA, et al. Diagnostic accuracy of circular rna for diabetes mellitus: a systematic review and diagnostic meta-analysis. Bmc Med Genomics. 2023;16:48. 10.1186/s12920-023-01476-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Chen J, Hu Q, Zhang C, et al. Tendomodulin in pan-cancer analysis: exploring its impact on immune modulation and uncovering functional insights in colorectal cancer. BMC Cancer. 2025;25:239. 10.1186/s12885-025-13608-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Liu J, Ali MK, Mao Y. Emerging role of long non-coding rna malat1 related signaling pathways in the pathogenesis of lung disease. Front Cell Dev Biol. 2023;11:1149499. 10.3389/fcell.2023.1149499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Marangoni K, Dorneles G, Da SD, et al. Diet as an epigenetic factor in inflammatory bowel disease. World J Gastroenterol. 2023;29:5618–29. 10.3748/wjg.v29.i41.5618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Gong W, Tian M, Qiu H, et al. Elevated serum level of lncrna-hif1a-as1 as a novel diagnostic predictor for worse prognosis in colorectal carcinoma. Cancer Biomark. 2017;20:417–24. 10.3233/CBM-170179. [DOI] [PubMed] [Google Scholar]
  • 26.Wang W, Xie Y, Chen F, et al. Lncrna meg3 acts a biomarker and regulates cell functions by targeting adar1 in colorectal cancer. World J Gastroenterol. 2019;25:3972–84. 10.3748/wjg.v25.i29.3972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Samir N, Matboli M, El-Tayeb H, et al. Competing endogenous rna network crosstalk reveals novel molecular markers in colorectal cancer. J Cell Biochem. 2018;119:6869–81. 10.1002/jcb.26884. [DOI] [PubMed] [Google Scholar]
  • 28.Gharib E, Nazemalhosseini-Mojarad E, Baghdar K, et al. Identification of a stool long non-coding rnas panel as a potential biomarker for early detection of colorectal cancer. J Clin Lab Anal. 2021;35: e23601. 10.1002/jcla.23601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Wu Y, Yang L, Zhao J, et al. Nuclear-enriched abundant transcript 1 as a diagnostic and prognostic biomarker in colorectal cancer. Mol Cancer. 2015;14:191. 10.1186/s12943-015-0455-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Heidari R, Assadollahi V, Marashi SN, et al. Identification of novel lncrnas related to colorectal cancer through bioinformatics analysis. Biomed Res Int. 2025;2025:5538575. 10.1155/bmri/5538575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zhang Y, Sun L, Xuan L, et al. Reciprocal changes of circulating long non-coding rnas zfas1 and cdr1as predict acute myocardial infarction. Sci Rep. 2016;6:22384. 10.1038/srep22384. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional file 1. (14.7KB, docx)

Data Availability Statement

No datasets were generated or analysed during the current study.

Articles from BioMedical Engineering OnLine are provided here courtesy of BMC

RESOURCES