The diagnostic performance of micro-RNA and metabolites in lung cancer: A meta-analysis

Abstract

Background

The diagnosis of lung cancer is based on the microscopic exam of tissue or liquid. During the recent decade, many biomarkers have been pointed to have a potential diagnostic role. These biomarkers may be assessed in blood, pleural effusion or sputum and they could avoid biopsies or other risky procedures. The authors aimed to assess the diagnostic performances of biomarkers focusing on micro-RNA and metabolites.

Methods

This meta-analysis was conducted under the PRISMA guidelines during a nine-year-period (2013–2022). the Meta-Disc software 5.4 (free version) was used. Q test and I² statistics were carried out to explore the heterogeneity among studies. Meta-regression was performed in case of significant heterogeneity. Publication bias was assessed using the funnel plot test and the Egger's test (free version JASP).

Results

According to our inclusion criteria, 165 studies from 79 articles were included. The pooled SEN, SPE and dOR accounted, respectively, for 0.76, 0.79 and 13.927. The AUC was estimated to 0.859 suggesting a good diagnostic accuracy. The heterogeneity in the pooled SEN and SPE was statistically significant. The meta-regression analysis focusing on the technique used, the sample, the number of biomarkers, the biomarker subtype, the tumor stage and the ethnicity revealed the biomarker number (p = 0.009) and the tumor stage (p = 0.0241) as potential sources of heterogeneity.

Conclusion

Even if this meta-analysis highlighted the potential diagnostic utility of biomarkers, more prospective studies should be performed, especially to assess the biomarkers’ diagnostic potential in early-stage lung cancers.

Keywords

Biomarkers micro-RNA metabolites lung cancer specificity sensitivity diagnostic odds ratio

Introduction

The diagnosis of lung cancer is commonly made on biopsies, surgical specimen or cytologies (bronchial liquid, pleural effusion). Since a decade, biopsies have been considered as evils and risky procedures, and many researches were performed in order to use non-invasive procedures to make a diagnosis of lung cancer. In this context, biomarkers represented a possible surrogate for tissue in order to perform the diagnosis.^1–3

A biomarker is a physical, biological or molecular state defining a new entity, a prognostic group or predicting a response to a treatment. A good biomarker implies an improvement in the take-care process.^2,4 Many samples can be used in the assessment of biomarkers in lung cancer including bronchial liquid, bronchoalveolar lavage (BAL), sputum, urine, pleural or peritoneal fluid or blood. Liquid biopsy has changed the management of lung cancer and represents an alternative to tissue in order to highlight therapeutic targets.¹ It also gained a place in the diagnostic field of lung cancer.

The most cited diagnostic biomarkers of lung cancer are represented by auto-antibodies, circulating tumor cells, exosomes, micro-RNA, metabolomics and proteomics.

Autoantibodies are defined as antibodies produced against abnormal tumor cell antigens. They are present in little amounts and are assessed using common laboratory techniques (ELISA). Many false negative cases have been reported.^4,5 Circulating tumor cells are rare and are always surrounded by 10⁶ to 10⁷ leucocytes. Their presence and amount are independent from the tumor stage or the microscopic subtype.^6–10 The exosomes are vesicles measuring 40–100 nm. They are present in many biological liquids consisting of plasma, serum, sputum, bile, BAL or amniotic liquid. They play a key role in angiogenesis, vascular permeability, epithelial–mesenchymal transition and resistance to chemotherapy. The assessment of exosomes is based on extraction techniques which consist of precipitation, ultra-centrifugation, chip-microfluidics and detection techniques in order to assess their composition. The detection techniques consist of real-time PCR, ELISA, flow cytometry or western blot techniques. The major challenges of diagnosis are represented by the fact that the extracted exosomes aren’t necessarily from tumor cells.¹¹ Micro-RNA are tissue-specific and cellular type-specific. They have a role in post-transcriptional regulation and are implicated in the carcinogenesis, the cell apoptosis or some viral infections. In addition to the messenger RNA, many non-coding RNA molecules have been assessed including ribosomal RNA, transfer RNA, small nuclear RNA, small nucleolar RNA and other RNA including large RNA. Micro-RNAs have been reported to play a key role in the diagnosis of lung cancer.^10,11 Proteomics studies are defined as the analysis of protein and have represented an alternative to the DNA or RNA analysis because of the fact that the same genome can produce different proteomes. Metabolomic consists of the analysis of different metabolites and has been reported in the diagnosis of lung cancer. It consists of the assessment of different metabolites present in urine or sputum.¹¹ According to the literature, using diagnostic biomarkers is needed in three major circumstances consisting of the early diagnosis and the assessment of indeterminate nodules diagnosed by the low dose CT-scan, early diagnosis of recurrences and the diagnosis of multiple pulmonary nodules. The authors aimed to assess the diagnostic value of the micro-RNA, exosomal micro-RNA and metabolites biomarkers that were reported in lung cancer.

Methods

Data source and search

This meta-analysis was conducted under the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).¹² To retrieve all eligible articles, PubMed, Embase databases and Cochrane Library were comprehensively searched during a nine-year period, from 2013 to 2022 with limitation to French and English language. The search medical subject heading (MeSH) terms employed for literature retrieval included: biomarkers OR microRNA OR exosomes OR metabolomic OR metabolites OR extracellular vesicles AND diagnostic OR diagnosis OR sensitivity OR specificity OR Receiver operating characteristic curve AND lung cancer OR non-small cell lung carcinoma OR large cell lung carcinoma OR squamous cell carcinoma. The reference list of eligible articles was also independently searched to obtain other valuable sources. According to inclusion criteria, titles and abstracts of all retrieved publications were screened and judged.

Study selection criteria

To be qualified for inclusion in this meta-analysis, articles must comply with all the following criteria: The articles were original articles consisting of cohort studies or case–control studies, articles evaluated the diagnostic value of the different biomarkers (microRNA, exosomes, metabolites) in plasma/serum, sputum, BAL, pleural effusion or blood for lung cancer diagnosis, the diagnosis of lung cancer was confirmed by the gold standard test which is the microscopic exam performed on biopsy or cytology and articles provided sufficient data (true negative (TN), true positive (TP), false negative (FN) and false positive (FP)). The major exclusion criteria were as follows: Studies assessing biomarkers in tissue, studies without gold standard test, studies published before 2013, technical reports, case reports, comments or letters with invalid data. Studies with duplicate data reported in other studies and reviews weren’t included.

Data extraction and quality assessment

Two investigators independently reviewed all the articles and extracted data from the selected articles: first authors’ name, publication year, characteristics of the biomarkers (source of control, number of cases and controls, sample types), assay methods, assay indicators, microscopic subtype, TP, FP, TN, FN cases and quality assessment information.

Quality assessment was based on the revised quality assessment of diagnosis, accuracy studies-2 (QUADAS-2) criteria.¹³ The most relevant items were chosen to be assessed and two reviewers (MM and MMZ) used the quality of assessment checklist (Table 1). Any disagreements were resolved by discussion and consensus. The included articles were rated taking into account the four key domains: patient selection, index test, reference standard and flow and timing.¹³ Study quality was assessed using the QUADAS-list, with each item scored as 1 for yes, −1 for no and 0 for unclear. Patient selection items focused on the representativeness of the patients who received the test and the justification of the withdrawals (maximum score = 2), the flow and timing domain focused on the time period between the reference standard and the index test (maximum score = 1), the reference standard domain focused on the capacity of the reference test to classify the target condition, the use of the reference test for all the patients regardless of the index test, the independence between the reference and the index test (maximum score = 4) and the index test domain focused on the condition of interpretation of the index test and the details about the uninterruptible and intermediate results (maximum score = 4). According to the items of quality chosen from QUADAS checklist, the maximum score accounted for 11.

Table 1.

The detailed quality assessment of the different articles included.

	Patient selection		Flow and Timing	Reference standard				Index test				Total score
	Was the spectrum of patient representative of those who will receive the test in practice?	Were withdrawals from the study explained?	Is the time period between reference standard and index test short enough to be reasonably sure that the target condition did not change between the 2 tests?	Is the reference standard likely to correctly classify the target condition?	Did patients receive the same reference standard regardless of the index test?	Was the reference standard independent of the index test?	Were the reference standard results interpreted without knowledge of the results of the index test?	Did the whole sample or a random selection of the sample, receive verification using a reference standard of diagnosis?	Were the index test results interpreted without knowledge of the results of the reference test?	Were the same clinical data available when test results were interpreted as would be available when the test is used in practice?	Were uninterruptible/ intermediate test results reported?	Total score
Sanfiorenzo 2013	−1	1	0	1	1	1	1	1	0	1	1	7
Tang 2013 (1)	−1	−1	0	1	1	1	1	1	0	1	0	6
Ulivi 2013	−1	−1	0	1	1	1	1	1	0	1	0	6
Zhu 2016	−1	−1	0	1	1	1	1	1	1	1	0	5
Xing 2015	1	−1	0	1	1	1	1	1	1	1	0	7
Shen 2014	−1	1	0	1	1	1	1	1	0	1	0	6
Li 2014	−1	−1	0	1	1	1	1	1	0	1	0	4
Li 2013	1	−1	0	1	1	1	1	1	0	1	0	6
Anjuman 2013	−1	1	0	1	1	1	1	1	0	1	1	7
Gao 2013	−1	0	0	1	1	1	1	1	0	1	0	5
Geng 2014 (1)	−1	0	0	1	1	1	1	1	0	1	1	6
Wang 2015	1	1	1	1	1	1	1	1	1	1	1	11
Zhu 2016	−1	−1	0	1	1	1	1	1	0	1	0	4
Shang 2017	−1	−1	1	1	1	1	1	1	0	1	0	5
Bagheri 2018	−1	−1	1	1	1	1	1	1	0	1	0	5
Zhang 2016	−1	−1	1	1	1	1	1	1	0	0	0	4
Dong 2016	−1	−1	1	1	1	1	1	1	0	0	0	4
Leng 2018	1	−1	1	1	1	1	1	1	0	0	0	6
Abu-Duhier 2018	−1	1	1	1	1	1	1	1	0	0	0	6
Abdollahi 2019 (1)	−1	−1	1	1	1	1	1	1	0	0	0	4
Wang 2018	−1	−1	1	1	1	1	1	1	0	0	0	4
Wang 2018	−1	−1	1	1	1	1	1	1	0	0	0	4
Sun 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Abd-El-Fattah 2013	−1	−1	0	1	1	1	0	1	0	0	0	2
Mozzoni 2013	−1	−1	0	1	1	1	0	1	0	0	0	2
Zhang 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Li 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Yang 2014	−1	−1	0	1	1	1	0	1	0	0	0	2
Zhu 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Li 2018	−1	−1	0	1	1	1	0	1	0	0	0	2
Hang 2018	−1	−1	0	1	1	1	0	1	0	0	0	2
Liu 2019	−1	1	1	1	1	1	0	1	0	0	0	5
Su 2015	−1	−1	1	1	1	1	0	1	0	0	0	3
Zhao 2015	−1	−1	1	1	1	1	0	1	0	0	0	3
Cazzoli 2013	−1	−1	1	1	1	1	0	1	0	0	0	3
Xi 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Li 2013	−1	−1	0	1	1	1	0	1	0	0	0	2
Klupczynska 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Zhang 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Terlizzi 2018	−1	−1	0	1	1	1	0	1	0	0	0	2
Yu 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Sun 2018	−1	−1	0	1	1	1	0	1	0	0	0	2
Fahrmann 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Ros-Mazurczik 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Li 2014	−1	−1	0	1	1	1	0	1	0	0	0	2
Anjuman 2013	−1	1	0	1	1	1	0	1	0	0	0	4
Su 2016	−1	0	0	1	1	1	0	1	0	0	0	4
Razzak 2016 (1)	−1	0	0	1	1	1	0	1	0	0	0	3
Kim 2015	−1	0	0	1	1	1	0	1	0	0	0	3
Ma 2013	−1	0	0	1	1	1	0	1	0	0	0	3
Zhang 2017	−1	0	0	1	1	1	0	1	0	0	0	3
Xi 2018	−1	0	0	1	1	1	0	1	0	0	0	3
Aiso 2018	−1	0	0	1	1	1	0	1	0	0	0	3
Sui 2019	−1	0	0	1	1	1	0	1	0	0	0	3
Jin 2017(1)	−1	−1	0	1	1	1	0	1	0	0	0	2
Zhou 2017	−1	−1	0	1	1	1	0	1	0	0	0	2
Tamiya 2018	−1	−1	0	1	1	1	0	1	0	0	0	2
Czarnecka 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Zhang 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Yao 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Fang 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Xia 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Wu 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Liu 2019	−1	−1	0	1	1	1	0	1	0	0	0	2
Wang 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Liu 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Han 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Yang 2020	−1	−1	0	1	1	1	0	1	0	0	0	2
Havorsen 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Sun 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Gao 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Wei 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Leidinger 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Wang 2016	−1	−1	0	1	1	1	0	1	0	0	0	2
Nadal 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Ma 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Yang 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Dou 2015	−1	−1	0	1	1	1	0	1	0	0	0	2
Yang 2015	−1	−1	0	1	1	1	0	1	0	0	0	2

Statistical analysis

We used the Meta-Disc software 5.4 (free version) to conduct this meta-analysis. The pooled sensitivity (SEN) (TP/TP + FN), specificity (SPE) (TN/TN + FP), negative likelihood ratio (NLR), positive likelihood ratio (PLR) and diagnostic odds ratio (DOR) with the 95% confidence intervals were calculated.

At the same time, we constructed the summary receiver operator characteristic (sROC) curve and calculated the area under the SROC curve based on the SEN and SPE of each study.

Threshold effect

A threshold effect was assessed using the Moses model with calculation of the Spearman correlation coefficient.

Heterogeneity

Q test and I² statistics were carried out to explore the heterogeneity among studies. P value <0.1 for q test or I² value >50% represented substantial between-study heterogeneity. A random-effects model was used if significant heterogeneity existed (I² value >50% or p < 0.1), otherwise a fixed-effects model was chosen. Based on the characteristics of the included articles, meta-regressions and subgroup analyses were performed to explore the sources of heterogeneity if necessary.

Publication bias

Publication bias was assessed using the funnel plot test and the Egger's test (free version JASP).

Results

Studies characteristics

A total of 898 articles were identified based on the search strategy adopted and were retrieved from PubMed, MEDLINE and the Cochrane database (Figure 1). After reviewing the titles and abstracts, removal of reviews, letters, non-English, non-human studies and unrelated articles, 123 full-text articles were obtained. Due to the lack of the gold standard test, assessment of only prognostic or therapeutic performances or incomplete data to reassess sensitivity and specificity, 79 full-text articles were included. After screening the references, no additional study was included. Finally, 165 studies from 79 articles were included, involving 22,224 participants. Twenty-one articles were considered as 2–8 studies because of the different biomarkers assessed.^14–33 Mozzoni et al. assessed different biomarkers in different samples including plasma and exhaled breath.³⁴ Razzak et al., Tamiya et al., Fang et al., Yang et al. and Halvorsen et al. assessed different biomarkers with different controls including healthy participants, participants with inflammatory pulmonary diseases or participants with benign pulmonary nodules.^35–39 Kim assessed a panel of biomarkers in different samples, BAL and sputum.⁴⁰ Jin et al. assessed different cases including NSCLC, adenocarcinomas and squamous cell carcinomas.⁴¹ Liu et al. assessed different biomarkers using different techniques in different controls.⁴² The data related to the diagnostic performance of biomarkers in lung cancer events eligible for inclusion are represented in Table 2. One hundred and nine studies (66.1%) assessed Asian participants and 56 studies (33.9%) included Caucasian participants. The gold standard test consisted of the microscopic diagnosis performed on tissues or cytologies. The event studies were lung cancer and 120 studies (72.7%) focused on non-small cell lung carcinomas (NSCLC), 29 focused on adenocarcinomas (17.6%), 15 studies focused on lung cancers (9.1%) and 1 study focused on squamous cell carcinomas (0.6%).

Figure 1.

The flow chart shows the search strategy.

Table 2.

The diagnostic performance of the different studies included.

Author	Study type	Stage	Biomarker	Sample	Technique	Ethnicity	TP	FP	FN	TN	QUADAS	Subtype
Sanfiorenzo 2013⁴³	Case–control	I, II/III	Panel miRNA	Plasma	qRT-PCR	Caucasian	42	3,5	10	16,5	7	NSCLC
Tang 2013 (1)¹⁴	Case–control	I/II/III	Panel miRNA(145, 155, 21, 25)	Plasma	qRT-PCR	Asian	37	15	25	45	6	NSCLC
Ulivi 2013⁴⁴	Case–control	I/II/III/IV	miR-328	Whole blood	qRT-PCR	Caucasian	60	4	26	20	6	NSCLC
Zhu 2016¹⁷	Case–control	I/IV	miR-210	Serum	QT-PCR (taqman)	Asian	38	0	74	40	5	NSCLC
Xing L 2015⁴⁵	Case–control	I/II	Panel miRNA	Sputum	QT-PCR (taqman)	Caucasian	45	9	15	53	7	NSCLC
Shen J 2014⁴⁶	Case–control	I/IV	Panel miRNA	Sputum	QT-PCR (taqman)	Caucasian	43	18	23	50	6	NSCLC
Li N 2014⁴⁷	Case–control	I	miR-210	Sputum	QT-PCR (taqman)	Caucasian	20	4	15	36	4	NSCLC
Li ZH 2013⁴⁸	Case–control	I/IV	miR-210	Serum	QT-PCR (taqman)	Asian	47	8	13	22	6	NSCLC
Anjuman N 2013⁴⁹	Case–control	I	Panel miRNA	Sputum	QT-PCR (taqman)	Caucasian	27	10	12	32	7	NSCLC
Gao F 2013⁵⁰	Case–control	I/II/III/IV	miR-155	Serum	RT-PCR	Asian	26	8	10	22	5	Adc
Geng Q 2014 (1)¹⁵	Case–control	I	panel miRNA(miR-20a, miR-223, miR-21, miR-221 and miR-145)	Plasma	RT-PCR	Asian	130	13	21	72	6	NSCLC
Wang P 2015 (1)¹⁶	Case–control	I/II	panel miRNA(miR-126, 125, 25	Serum	qRT-PCR	Asian	21	3	3	21	11	LC
Zhu 2016 (2)¹⁷	Case–control	0/III	miR-126	Serum	qRT-PCR	Asian	68	3	44	37	4	NSCLC
Shang 2017(1)¹⁸	Case–control	I/IV	miR-126	Serum	qRT-PCR	Asian	122	19	5	93	5	NSCLC
Bagheri 2018(1)¹⁹	Case–control	I/IV	miR-126	Sputum	QT-PCR (taqman)	Asian	29	7	1	23	5	NSCLC
Geng Q 2014 (2)¹⁵	Case–control	I	microRNA-21	Plasma	RT-PCR	Asian	84	19	42	41	6	NSCLC
Zhang H 2016²⁴	Case–control	I	microRNA-21	Plasma	RT-PCR	Asian	100	12	29	71	4	NSCLC
Tang 2013 (3)¹⁴	Case–control	I/II/III	mi-R145	Serum	qRT-PCR	Asian	19	27	15	35	6	NSCLC
Tang 2013 (4)¹⁴	Case–control	I/II/III	micro-R 155	Serum	qRT-PCR	Asian	23	21	11	41	6	NSCLC
Dong J 2016⁵¹	Case–control	I/II/III/IV	microRNA-21	Serum	qRT-PCR	Asian	63	25	5	39	4	LC
Leng Q 2018(1)⁵²	Case–control	I/II/III/IV	microRNA-21	Plasma	ddPCR	Caucasian	30	11	10	18	6	NSCLC
Abu-Duhier 2018⁵²	Case–control	I/II/III/IV	microRNA-21	Plasma	RT-PCR	Caucasian	64	16	16	64	6	NSCLC
Abdollahi 2019 (1)²¹	Case–control	I/II/III/IV	miR-148a-3p	Serum	qRT-PCR	Caucasian	35	17	8	26	4	NSCLC
Tang 2013 (2)¹⁴	Case–control	10I/8II/16III	microRNA-21	Serum	qRT-PCR	Asian	18	9	16	23	6	NSCLC
Wang S(1) 2018²²	Case–control	I/II/III/IV	microRNA-21	Serum	RT-PCR	Asian	30	8	20	16	4	LC
Wang X(1) 2018²³	Case–control	I/II/III/IV	microRNA-21	Serum	RT-PCR	Asian	108	46	59	82	4	NSCLC
Zhang H 2016 (2)²⁴	Case–control	I	Panel mi-R(15, 145, 201, 21, 2230	Plasma	RT-PCR	Asian	106	8	23	75	4	NSCLC
Wang 2015 (2)⁵³	Case–control	I/I/III/IV	panel miRNA(miR-126, 125, 25	Serum	qRT-PCR	Asian	125	1	17	110	11	LC
Wang 2015 (3)⁵³	Case–control	I/I	miR-126	Serum	qRT-PCR	Asian	78	41	16	70	11	LC
Leng Q 2018(2)²⁰	Case–control	I/II/III/IV	microRNA-270	Plasma	ddPCR	Caucasian	33	9	3	24	6	NSCLC
Leng Q 2018(3)²⁰	Case–control	I/II/III/IV	microRNA-486-5p	Plasma	ddPCR	Caucasian	48	12	18	21	6	NSCLC
Leng Q 2018(4)²⁰	Case–control	I/II/III/IV	long non coding RNA(SNHG1)	Plasma	ddPCR	Caucasian	50	6	16	27	6	NSCLC
Leng Q 2018(5)²⁰	Case–control	I/II/III/IV	long non coding RNA(RMRP)	Plasma	ddPCR	Caucasian	41	3	25	30	6	NSCLC
Leng Q 2018(6)²⁰	Case–control	I/II/III/IV	FLUCOSYLTRANSFERASE FUT2	Plasma	ddPCR	Caucasian	47	4	19	29	6	NSCLC
Leng Q 2018(7)²⁰	Case–control	I/II/III/IV	FLUCOSYLTRANSFERASE POFUT 1	Plasma	ddPCR	Caucasian	32	3	34	30	6	NSCLC
Sun Y(1) 2017²⁵	Case–control	I/II/III	microRNA-21	Plasma	RT-PCR	Asian	23	1	5	27	2	Adc
Sun Y(2) 2017²⁵	Case–control	I/II/III	microRNA-339-5P	Plasma	RT-PCR	Asian	23	2	5	26	2	Adc
Abd-El-Fattah (1) 2013²⁶	Case–control	I/II/III/IV	microRNA-21	Serum	RT-PCR	Caucasian	56	5	9	32	2	NSCLC
Mozzoni (1) 2013³⁴	Case–control	I/II/III	microRNA-21	Plasma	RT-PCR	Caucasian	27	4	27	42	2	NSCLC
Zhang YL 2019⁵⁴	Case–control	I/IV	microRNA-25	Serum	RT-PCR	Asian	85	20	29	60	2	NSCLC
Li J 2019⁵⁵	Case–control	I/IV	microRNA-25	Serum	RT-PCR	Asian	98	20	30	108	2	NSCLC
Li 2015 (1)²⁷	Case–control	I/II/III/IV	miR-148a	Serum	RT-PCR	Asian	28	4	8	16	2	NSCLC
Li 2015(2)²⁷	Case–control	I/II/III/IV	miR-148b	Serum	RT-PCR	Asian	25	4	11	16	2	NSCLC
Li 2015(3)²⁷	Case–control	I/II/III/IV	miR-152	Serum	RT-PCR	Asian	26	2	10	18	2	NSCLC
Yang 2014 (1)⁵⁶	Case–control	I/II/III/IV	miR-148a	Serum	qRT-PCR	Asian	255	26	45	126	2	NSLC
Yang 2014 (2)⁵⁶	Case–control	I/II/III/IV	miR-148b	Serum	qRT-PCR	Asian	249	26	51	126	2	NSLC
Yang 2014 (3)⁵⁶	Case–control	I/II/III/IV	miR-152	Serum	qRT-PCR	Asian	225	35	75	117	2	NSLC
Zhu X 2017⁵⁷	Case–control	I/II/III/IV	circular RNA	Plasma	RT-PCR	Asian	20	2	10	28	2	ADC
Li 2018⁵⁸	Case–control	I/II/III/IV	circular RNA	Serum	RT-PCR	Asian	32	9	13	36	2	NSCLC
Hang 2018⁵⁹	Case–control	I/II/III/IV	circular RNA	Plasma	RT-PCR	Asian	32	17	18	33	2	NSCLC
Liu 2019⁶⁰	Case–control	I/II/III/IV	circular RNA	Plasma	RT-PCR	Asian	228	33	78	75	5	ADC
Zhu 2016 (3)¹⁷	Case–control	I/II	panel miRNA(182, 183, 210, 126) + CEA	Serum	qRT-PCR	Asian	91	0	21	104	4	NSCLC
Zhu 2016 (4)¹⁷	Case–control	I/II	mi-R182	Serum	qRT-PCR	Asian	71	20	92	84	4	NSCLC
Zhu 2016 (5)¹⁷	Case–control	I/II	mi-R183	Serum	qRT-PCR	Asian	46	18	66	86	4	NSCLC
Shang 2017 (2)¹⁸	Case–control	I/IV	miR-22	Serum	qRT-PCR	Asian	113	20	14	92	5	NSCLC
Bagheri 2018(2)¹⁹	Case–control	I/IV	panel miRNA	Sputum	QT-PCR (taqman)	Asian	27	3	3	27	5	NSCLC
Bagheri 2018(3)¹⁹	Case–control	I/IV	mi-R145	Sputum	QT-PCR (taqman)	Asian	30	10	0	20	5	NSCLC
Bagheri 2018(1)¹⁹	Case–control	I/IV	mi-R7	Sputum	QT-PCR (taqman)	Asian	22	5	8	25	5	NSCLC
Geng Q 2014 (3)¹⁵	Case–control	I	mi-R20	Plasma	RT-PCR	Asian	21	5	4	21	6	NSCLC
Geng Q 2014 (4)¹⁵	Case–control	I	mi-R 223	Plasma	RT-PCR	Asian	22	4	3	21	6	NSCLC
Geng Q 2014 (5)¹⁵	Case–control	I	mi-R155	Plasma	RT-PCR	Asian	22	4	3	21	6	NSCLC
Geng Q 2014 (6)¹⁵	Case–control	I	mi-R145	Plasma	RT-PCR	Asian	18	8	7	17	6	NSCLC
Tang 2013 (5)¹⁴	Case–control	I/II/III	panel (mi-R21/145)	Serum	qRT-PCR	Asian	25	11	9	50	6	NSCLC
Tang 2013 (6)¹⁴	Case–control	I/II/III	panel( mi-R 21/155)	Serum	qRT-PCR	Asian	19	21	15	41	6	NSCLC
Tang 2013 (7)¹⁴	Case–control	I/II/III	Panel (mi-R155/145)	Serum	qRT-PCR	Asian	27	14	7	48	6	NSCLC
Leng Q 2018 (8)²⁰	Case–control	I/II/III/IV	panel (miR-210, SNHG1, FUT8)	Plasma	ddPCR	Caucasian	63	1	3	32	6	NSCLC
Abdollahi 2019 (2)²¹	Case–control	I/II/III/IV	panel (21-5p/638/152-3p)	Serum	qRT-PCR	Caucasian	41	6	2	37	4	NSCLC
Abdollahi 2019 (3)²¹	Case–control	I/II/III/IV	miR-21-5p	Serum	qRT-PCR	Caucasian	38	17	5	26	4	NSCLC
Abdollahi 2019 (4)²¹	Case–control	I/II/III/IV	miR-63	Serum	qRT-PCR	Caucasian	29	14	14	29	4	NSCLC
Abdollahi 2019 (5)²¹	Case–control	I/II/III/IV	miR-152-3p	Serum	qRT-PCR	Caucasian	33	12	10	31	4	NSCLC
Wang S(2) 2018²²	Case–control	I/II/III/IV	microRNA-31	Serum	RT-PCR	Asian	30	6	20	18	4	LC
Wang S(3) 2018²²	Case–control	I/II/III/IV	microRNA-LET7	Serum	RT-PCR	Asian	27	10	23	14	4	LC
Wang S(4) 2018²²	Case–control	I/II/III/IV	Panel (microRNA-21/31)	Serum	RT-PCR	Asian	35	7	15	17	4	LC
Wang S(5) 2018²²	Case–control	I/II/III/IV	Panel (microRNA-31/LET7)	Serum	RT-PCR	Asian	30	6	20	18	4	LC
Wang S(1) 2018²²	Case–control	I/II/III/IV	Panel (microRNA-21/31/LET-7)	Serum	RT-PCR	Asian	36	6	14	18	4	LC
Wang X(2) 2018²²	Case–control	I/II/III/IV	Panel (microRNA-21/Let-7a)	Serum	RT-PCR	Asian	132	32	35	96	4	NSCLC
Wang X(3) 2018²²	Case–control	I/II/III/IV	microRNA-Let7a	Serum	RT-PCR	Asian	109	36	58	92	4	NSCLC
Sun Y(3) 2017²⁵	Case–control	I/II/III	Panel (microRNA-339-5P/21)	Plasma	RT-PCR	Asian	26	2	2	26	2	Adc
Abd-El-Fattah (2) 2013²⁶	Case–control	I/II/III/IV	microRNA-155	Serum	RT-PCR	Caucasian	62	2	3	35	2	NSCLC
Abd-El-Fattah (3) 2013²⁶	Case–control	I/II/III/IV	microRNA-182	Serum	RT-PCR	Caucasian	65	4	0	33	2	NSCLC
Abd-El-Fattah (4) 2013²⁶	Case–control	I/II/III/IV	microRNA-197	Serum	RT-PCR	Caucasian	47	14	18	23	2	NSCLC
Mozzoni (2) 2013³⁴	Case–control	I/II/III	microRNA-486-5p	Plasma	RT-PCR	Caucasian	38	5	16	41	2	NSCLC
Mozzoni (3) 2013³⁴	Case–control	I/II/III	Panel (microRNA-21/486)	Plasma	RT-PCR	Caucasian	47	6	7	40	2	NSCLC
Mozzoni (4) 2013³⁴	Case–control	I/II/III	microRNA-21	Exhaled breath condensate	RT-PCR	Caucasian	38	17	16	29	2	NSCLC
Mozzoni (5) 2013³⁴	Case–control	I/II/III	microRNA-486	Exhaled breath condensate	RT-PCR	Caucasian	34	17	20	29	2	NSCLC
Su (1) 2015²⁹	Case–control	I	micro-RNA21	Sputum	RT-PCR	Caucasian	42	14	14	59	3	NSCLC
Su (2) 2015²⁹	Case–control	I	micro-RNA31	Sputum	RT-PCR	Caucasian	36	12	20	61	3	NSCLC
Su (3) 2015²⁹	Case–control	I	micro-RNA210	Sputum	RT-PCR	Caucasian	46	15	10	58	3	NSCLC
Su (4) 2015²⁹	Case–control	I	panel(micro-RNA21/31/210)	Sputum	RT-PCR	Caucasian	47	9	9	64	3	NSCLC
Zhao W 2015⁶¹	Case–contorl	NC	micro-RNA21	Serum	RT-PCR	Asian	59	17	21	43	3	NSCLC
Cazzoli R 2013⁶²	Case–control	I	exosomal microRNA	Plasma	RT-PCR	Caucasian	48	22	2	33	3	Adc
Xie Y 2020⁶³	Case–control	I/II/III	panel metabolites	Plasma	lipid chromatography mass spectromatry (LC-MS)	Asian	108	0	2	43	2	LC
Li X 2013⁶⁴	Case–control	I/II	panel metabolites	Plasma	MRM-MS	Asian	37	29	15	23	2	NSCLC
Klupczynska 2017⁶⁵	Case–control	I/II	1 metabolite	Plasma	LC-MS	Asian	28	5	22	20	2	NSCLC
Zhang L 2020⁶⁶	Case–control	I/II	panel metabolites	Plasma	LC-MS	Asian	44	1	9	19	2	NSCLC
Terlizzi 2018⁶⁷	Case–control	I/II/III/IV	1 metabolite	Plasma	ELISA	Caucasian	125	10	0	69	2	NSCLC
Yu 2017⁶⁸	Case–control	I/II	panel metabolites	Plasma	ESI-MS	Caucasian	163	143	36	104	2	NSCLC
Sun Q 2018⁶⁹	Case–control	I/II/III/IV	panel metabolites	Serum	GC-MS	Asian	29	0	3	29	2	NSCLC
Fahrmann JF 2015⁷⁰	Case–control	I/II/III/IV	Cassifier metabolites	Serum	GC-MS	Caucasian	30	11	13	31	2	Adc
Ros-Mazurczik 2017⁷¹	Case–control	I/II	Cassifier metabolites	Serum	LC-MS	Caucasian	73	66	27	234	2	NSCLC
Li N 2014⁴⁷	Case–control	I	Panel miR31/210	Sputum	ddPCR	Caucasian	23	6	12	34	2	NSCLC
Anjuman N 2013⁴⁹	Case–control	I	Panel miR	Sputum	RT-PCR	Caucasian	24	4	15	38	4	NSCLC
Su Y 2016(1)³⁰	Case–control	I	miR31	Sputum	RT-PCR	Caucasian	73	31	44	143	4	NSCLC
Su Y 2016(2)³⁰	Case–control	I	miR21	Sputum	RT-PCR	Caucasian	94	52	23	122	4	NSCLC
Su Y 2016(3)³⁰	Case–control	I	miR210	Sputum	RT-PCR	Caucasian	90	33	27	141	4	NSCLC
Su Y 2016(4)³⁰	Case–control	I	Panel miR	Sputum	RT-PCR	Caucasian	95	24	22	150	4	NSCLC
Razzak R 2016 (1)³⁵	Case–control	I/II	Panel miR21/210/372	Sputum	rT-PCR	Caucasian	14	1	7	9	3	NSCLC
Razzak R 2016 (1)³⁵	Case–control	III/IV	Panel miR21/210/372	Sputum	rT-PCR	Caucasian	13	0	8	10	3	NSCLC
Kim JO 2015 (1)⁴⁰	Case–control	I/II	Panel miR21, 143, 155, 210, 372	Bal	RT-PCR	Caucasian	18	1	3	9	3	NSCLC
Kim JO 2015 (2)⁴⁰	Case–control	I/II	Panel miR21, 143, 155, 210, 372	Sputum	RT-PCR	Caucasian	14	0	7	10	3	NSCLC
Ma J 2013⁷²	Case–control	I/II	Panel miR21-5, 335	Plasma	ddPCR	Caucasian	26	7	10	31	3	NSCLC
Zhang H 2017⁷³	Case–control	I/II	Panel miR	Plasma	rT-PCR	Asian	88	7	21	56	3	NSCLC
Xi KX 2018⁷⁴	Case–control	I/II	Panel miR	Plasma	RT-PCR	Asian	33	4	9	11	3	NSCLC
Aiso 2018⁷⁵	Case–control	I/II/III/IV	Panel miR	Serum	RT-PCR	Asian	40	3	16	23	3	NSCLC
Sui A 2019 (1)³¹	Case–control	I/II/III/IV	miR197	Serum	RT-PCR	Asian	56	9	20	51	3	NSCLC
Sui A 2019 (2)³¹	Case–control	I/II/III/IV	miR145	Serum	RT-PCR	Asian	64	17	10	43	3	NSCLC
Sui A 2019 (3)³¹	Case–control	I/II/III/IV	Panel miR	Serum	RT-PCR	Asian	70	13	6	47	3	NSCLC
Jin X 2017 (1)⁴¹	Case–control	I	panel extracellular vesicle miR(let-7b-5p, let-7e-5p,miR23a-3p,mi-R486-5p)	Plasma	RT-PCR	Asian	28	2	19	11	2	NSCLC
Jin X 2017 (2)⁴¹	Case–control	I	panel extracellular vesicle miR(181-b-5p, 361b-5p)	Plasma	RT-PCR	Asian	25	1	6	11	2	Adc
Jin X 2017 (3)⁴¹	Case–control	I	panel extracellular vesicle miR(l10b-5p, 320b)	Plasma	RT-PCR	Asian	9	3	2	29	2	sqcc
Zhou X(1) 2017⁷⁶	Case–control	I/II/III/IV	panel extracellular vesicle miR(19b-3p, 21-5p,221-3p,409-3p,425-5p, 584-5p)	Plasma	RT-PCR	Asian	35	3	7	29	2	Adc
Zhou X(2) 2017⁷⁶	Case–control	I/II/III/IV	panel extracellular vesicle miR(19b-3p, 21-5p,221-3p,409-3p,425-5p, 584-6p)	Plasma	RT-PCR	Asian	46	21	20	41	2	Adc
Zhou X(3) 2017⁷⁶	Case–control	I/II/III/IV	panel extracellular vesicle miR(19b-3p, 21-5p,221-3p,409-3p,425-5p, 584-7p)	Plasma	RT-PCR	Asian	72	27	36	67	2	Adc
Zhou X(4) 2017⁷⁶	Case–control	I/II/III/IV	panel extracellular vesicle miR(19b-3p, 21-5p,221-3p,409-3p,425-5p, 584-8p)	Plasma	RT-PCR	Asian	20	9	10	21	2	Adc
Tamiya H(1) 2018³⁶	Case–control	I/II/III/IV	exosomal microRNA 182	Pleural effusion	RT-PCR	Asian	33	8	8	7	2	Adc
Tamiya H(2) 2018³⁶	Case–control	I/II/III/IV	exosomal microRNA 210	Pleural effusion	RT-PCR	Asian	38	4	3	11	2	Adc
Tamiya H(3) 2018³⁶	Case–control	I/II/III/IV	exosomal microRNA 21	Pleural effusion	RT-PCR	Asian	24	1	17	14	2	Adc
Tamiya H(4) 2018³⁶	Case–control	I/II/III/IV	exosomal microRNA 31	Pleural effusion	RT-PCR	Asian	15	1	26	14	2	Adc
Czarnecka KH 2019⁷⁷	Case–control	I/II/III/IV	exosomal microRNA17	Serum	RT-PCR	Caucasian	19	6	2	15	2	Adc
Zhang Y 2019⁵⁴	Case–control	I/II/III	exosomal panel microRNA(17-5p,CEA,CYFRA,SCCA)	Serum	rT-PCR	Asian	55	11	17	36	2	NSCLC
Yao B 2019⁷⁸	Case–control	I	exosomal panel microRNA(miR451a, 194-5p,486-5p)	Plasma	RT-PCR	Asian	37	1	2	38	2	Adc
Fang H (1) 2019³⁷	Case–control	I/II	exosomal microRNA 505-5P	Plasma	RT-PCR	Asian	90	3	18	47	2	Adc
Fang H (2) 2019³⁷	Case–control	I/II	exosomal microRNA 382-3P	Plasma	RT-PCR	Asian	89	14	19	36	2	Adc
Fang H (3) 2019³⁷	Case–control	I/II	Panel exosomal microRNA 505-5P/382-3p)	Plasma	RT-PCR	Asian	93	2	15	48	2	Adc
Xia Y 2020⁷⁹	Case–control	NC	exosomal microRNA1260b	Plasma	RT-PCR	Asian	36	8	24	42	2	Adc
Wu Q 2020(1)⁸⁰	Case–control	I/II	exosomal microRNA146a-5p	Serum	RT-PCR	Asian	33	8	15	72	2	Adc
Wu Q 2020 (2)⁸⁰	Case–control	I/II	exosomal microRNA486-5p	Serum	RT-PCR	Asian	34	4	14	76	2	Adc
Wu Q 2020(3)⁸⁰	Case–control	I/II	exosomal microRNA146a-5p/486-5p	Serum	RT-PCR	Asian	40	8	8	72	2	Adc
Liu W 2019⁶⁰	Case–control	I/II/III/IV	exosomal microRNA216-B	Serum	RT-PCR	Asian	91	15	14	45	2	NSCLC
Wang N 2020 (1)³³	Case–control	I/II	exosomal microRNA LET-7f-5p	Plasma	RT-PCR	Asian	51	6	29	24	2	NSCLC
Wang N 2020 (2)³³	Case–control	III/IV	exosomal microRNA 320a	Plasma	RT-PCR	Asian	29	15	11	25	2	NSCLC
Wang N 2020 (3)³³	Case–control	III/IV	exosomal microRNA 622	Plasma	RT-PCR	Asian	35	12	5	28	2	NSCLC
Wang N 2020 (4)³³	Case–control	III/IV	exosomal microRNA LET-7f-5p	Plasma	RT-PCR	Asian	19	4	21	36	2	NSCLC
Wang N 2020 (5)³³	Case–control	III/IV	PANEL exosomal microRNA 320,620	Plasma	RT-PCR	Asian	30	10	10	30	2	NSCLC
Liu C 2020 (1)⁴²	Case–control	I/II/III/IV	PANEL exosomal microRNA 21,25,155,210,486	Serum	Tethered cationic lipoplex nanoparticles (tCLN)	Asian	46	5	18	10	2	NSCLC
Liu C 2020 (2)⁴²	Case–control	I/II	PANEL exosomal microRNA 21,25,155,210,486	Serum	Tethered cationic lipoplex nanoparticles (tCLN)	Asian	22	5	10	10	2	NSCLC
Liu C 2020 (3)⁴²	Case–control	I/II/III/IV	PANEL exosomal microRNA 21,25,155,210,486	Serum	RT-PCR	Asian	37	2	27	13	2	NSCLC
Liu C 2020 (4)⁴²	Case–control	I/II	PANEL exosomal microRNA 21,25,155,210,486	Serum	RT-PCR	Asian	25	5	7	10	2	NSCLC
Liu C 2020 (5)⁴²	Case–control	III/IV	PANEL exosomal microRNA 21,25,155,210,486	Serum	RT-PCR	Asian	20	2	12	13	2	NSCLC
Han Z 2020⁸¹	Case–control	I/II	PANEL exosomal microRNA 342-5p,574-5p	Plasma	RT-PCR	Asian	45	11	12	29	2	Adc
Yang G 2020 (1)³⁸	Case–control	I/II/III	PANEL exosomal microRNA ratio miR21/let-7a	Serum	RT-PCR	Asian	47	6	28	18	2	NSCLC
Yang G 2020 (2)³⁸	Case–control	I/II/III	PANEL exosomal microRNA ratio miR21/let-7a	Serum	RT-PCR	Asian	52	3	25	15	2	NSCLC
Yang G 2020 (3)³⁸	Case–control	I/II/III	PANEL exosomal microRNA ratio miR21/let-7a	Serum	RT-PCR	Asian	60	3	15	15	2	NSCLC
Halvorsen A 2016 (1)³⁹	Case–control	I/II/III/IV	Panel miR429, 205, 200b, 203, 12, 34b	Serum	RT-PCR	Caucasian	88	17	12	41	2	Adc
Halvorsen A 2016 (2)³⁹	Case–control	I/II	Panel miR429, 205, 200b, 203, 12, 34b	Serum	RT-PCR	Caucasian	67	15	12	43	2	Adc
Sun L 2016⁸²	Case–control	I/II/III/IV	miR30a	Plasma	RT-PCR	Caucasian	53	15	34	81	2	NSCLC
Gao X 2016⁸³	Case–control	I	panel miR 324,1285	Plasma	RT-PCR	Asian	28	3	2	27	2	LC
Wei J 2016⁸⁴	Case–control	I/II/III/IV	miR21	Plasma	RT-PCR	Asian	48	9	15	21	2	LC
Leidinger P 2016⁸⁵	Case–control	I/II/III	panel miRNA	Whole blood	RT-PCR	Caucasian	67	0	7	20	2	LC
Wang W 2016⁵³	Case–control	I/II/III/IV	panel miRNA	Serum	RT-PCR	Asian	44	3	10	12	2	LC
Nadal E 2015⁸⁶	Case–control	I/II/III	panel miR193b, 301, 141, 200b	Serum	RT-PCR	Caucasian	81	1	3	22	2	NSCLC
Ma J 2015⁸⁷	Case–control	I/II/III/IV	PANEL miR19b,29b	Peripheral blood mononulceated cells	RT-PCR	Caucasian	61	12	23	57	2	NSCLC
Yang JS 2015⁵⁶	Case–control	I/II/III/IV	Paneb, miR152, 148a,148b,21	Serum	RT-PCR	Asian	146	27	6	273	2	NSCLC
Dou H 2015⁸⁸	Case–control	I/II/III/IV	panel miR152, let-7c	Plasma	dPCR	Asian	103	67	17	293	2	NSCLC
Yang YL 2015²⁸	Case–control	I/II/III/IV	miR-10b	Peripheral blood mononulceated cells	RT-PCR	Asian	64	12	10	40	2	NSCLC

Fifty-four studies (32.7%) dealt with early stages (I/II) and 85 studies (51.5%) were about late and metastatic stages (III/IV) and 26 studies focused on all stages. The controls included consisted of healthy participants, participants with inflammatory pulmonary diseases or with benign pulmonary nodules.

Eighty-seven studies (52.7%) assessed a specific biomarker and 78 studies (47, 3%) assessed panels or classifiers. Biomarkers consisted of microRNA in 112 studies (67.9%), long non-coding RNA in 2 studies (1.2%), circular RNA in 4 studies (2.4%), exosomal micro-RNA in 31 studies (18.8%), metabolites in 9 studies (5.5%) and extracellular vesicle micro-RNA in 7 studies (4.2%). The techniques of diagnosis used consisted mainly of RT-PCR in 144 studies (87.3%). The other techniques used consisted of digital PCR in 11 studies (6.7%), mass spectrometry for proteomic assessment in 8 studies (4.8%) and tethered cationic lipoplex nanoparticles (tCLN) in 2 studies (1.2%).

The samples used consisted of plasma in 60 studies (36.4%), serum in 73 studies (44.2%), whole blood in 2 studies (1.2%), sputum in 21 studies (12.7%), BAL in 1 study (0.6%), exhaled breath in 2 studies (1.2%), peripheral blood mononucleated cells in 2 studies (1.2%) and pleural effusion in 4 studies (2.4%).

Quality assessment

Meta-disc 5.4 software was used to analyze the data by two writers. Different items related to four domains (patient selection, reference test, index test, flow and time) were picked up from the QUADAS-2 and were used to perform the quality assessment of the eligible studies. The maximum score accounted for 11. The mean score of the different studies included accounted for 3.45 (DS = 1.87) with extremes ranging from 2 to 11. The mean scores attributed, respectively, to the patient selection domain, the flow and timing domain, the reference standard domain and the index test domain accounted for 0/2 (SD = 0.0), 0.18/1 (SD = 0.384), 1.28/4 (SD = 0.451) and 1.05/4 (SD = 0.221).

Diagnostic performance

The Spearman correlation coefficient was −0.129 (p = 0.099), suggesting that the threshold effects were not statistically significant. The heterogeneity in the pooled SEN (I² = 85.9%, p = 0.000) and SPE (I² = 80.1%, p = 0.000) was statistically significant. Therefore, the pooled effect was analyzed by random effects model. The indexes were shown as follows: pooled SEN and SPE were 0.76 (95% confidence interval [CI]: 0.752–0.767) and 0.79 (95% CI: 0.782–0.798), pooled PLR and NLR were 3.672 (95% CI: 3.333–4.046) and 0.299 (95% CI: 0.273–0.327) and pooled DOR accounted for 13.927 (95%CI: 11.836–16.387). In addition, the sROC curve was shown to appraise diagnostic accuracy and the AUC was 0.859 suggesting that the different biomarkers assessed have a good diagnostic accuracy for diagnosing lung cancers (Figure 2). Table 3 illustrates the details of the different diagnostic parameters. To assess the different sources of heterogeneity, we conducted a series of analyses including publication bias, meta-regression and sub-group analysis.

Figure 2.

(a) The forest plot illustrating the SEN, (b) the SPE, (c) DOR, (d) the sROC curve of the different studies included.

Table 3.

The detailed pooled SEN, SPE, PLR, NLR and DOR of the different included studies.

	Pooled SEN	Pooled SPE	PLR	NLR	DOR
165 STUDIES INCLUDED	- 0.760 - 95%CI = [0.752–0.767] - Heterogeneity chi-squared = 1161.96 (df = 164) – - p = 0.000 - Inconsistency (I-square) = 85.9%	- 0.790 - 95%CI = [0.782–0.798] - Heterogeneity chi-squared = 822.36 (df = 164) - p = 0.000 - Inconsistency (I-square) = 80.1%	- 3.672 - 95%CI = [3.333–4.046] - Heterogeneity chi-squared = 801.25 (df = 164) - p = 0.000 - Inconsistency (I-square) = 79.5%	- 0.299 - 95%CI = [0.273–0.327] - Heterogeneity chi-squared = 1080.59 (df = 164) - p = 0.000 - Inconsistency (I-square) = 84.8%	- 13.927 - 95%CI = [11.836–16.387] - Heterogeneity chi-squared = 706.88 (df = 164) - p = 0.000 - Inconsistency (I-square) = 76.8%

Meta-regression analysis and publication bias

We conducted a meta-regression analysis to assess the potential heterogeneity due to the detection technique, the ethnicity of the participants, the number of biomarkers assessed (one biomarker or a panel), the biomarker subtype, the specimen used and the tumor stage. We found that both SEN and SPE were influenced by the biomarker number (p = 0.009) and the tumor stage (p = 0.0241). Table 4 illustrates these different findings.

Table 4.

The different variables assessed for the meta-regression analysis.

Variable	Standard error	p value
Sample	0.06	0.49
Ethnicity	0.2	0.62
Biomarkers number	0.18	0.01
Biomarkers sub-type	0.08	0.22
Tumor stage	0.13	0.02
Technique analysis used	0.19	0.42
Microscopic subtype of lung cancer	0.14	0.2

The funnel plot asymmetry test was used to perform the publication bias analysis which showed a symmetric funnel plot in Figure 3, meaning that the potential publication did not exist (Egger's test p = 0.341).

Figure 3.

The funnel plot illustrating the absence of publication bias.

Sub-group analysis

A sub-group analysis was performed in order to assess the diagnostic performance according to the technique used, biomarkers number (one marker or a panel), the tumor stage, the biomarker subtype, the sample used and the biomarker subtype.

According to the technique used, pooled SEN, SPE and DOR accounted, respectively, for 0.758 (95%CI: 0.750–0.766), 0.800 (95% CI: 0.791–0.809) and 14.242 (95% CI: 11.996–16.910) in the RT-PCR group, 0.748 (95%CI: 0.713–0.781), 0.815 (95% CI: 0.784–0.843) and 14,522 (95% CI: 8.402–25.101) in the digital PCR group (Figure 4(a)–(c)), 0.801 (95%CI: 0.768–0.832), 0.664 (95% CI: 0.629–0.697) and 12.173 (95% CI: 4.805–30.837) in the mass spectrometry group and 0.708 (95%CI: 0.607–0.797), 0.667 (95% CI: 0.472–0.827) and 4.772 (95% CI: 1.968–11.572) in tCLN group.

Figure 4.

(a) The forest plot of the sensitivities of the studies using digital PCR, (b) the forest plot of the specificities of the studies using digital PCR, (c) the sROC curve of the studies using digital PCR, (d) the sROC curve of the studies using multiple biomarkers or classifiers, (e) the sROC curve of the studies using 1 biomarker, (f) the sROC curve of the studies focusing on early-stage lung cancers, (g) the sROC curve of the studies focusing on late-stage lung cancers.

In the group of studies using one biomarker, the pooled SEN, SPE and dOR accounted, respectively, for 0.737 (95%CI: 0.726–0.748), 0.782 (95% CI: 0.770–0.793) and 11.508 (95% CI: 9.425–14.051). In the studies reporting a panel of biomarkers or classifiers, the pooled SEN, SPE and dOR accounted for 0.788 (95%CI: 0.777–0.799), 0.799 (95% CI: 0.787–0.811) and 17.914 (95% CI: 13.693–23.437) (Figure 4(d) and (e)).

The diagnostic performances (SEN, SPE, DOR) according to the biomarker subtype were estimated to 0.759 (95%CI: 0.749–0.768), 0.802 (95% CI: 0.793–0.812) and 14.454 (95% CI: 11.866–17.607) in the micro-RNA group, 0.689 (95%CI: 0.603–0.767), 0.864 (95% CI: 0.757–0.936) and 14.952 (95% CI: 6.633–33.708) in the long non-coding RNA, 0.724 (95%CI: 0.679–0.766), 0.738 (95% CI: 0.677–0.793) and 7.163 (95% CI: 3.860–13.292) in the circular RNA group, 0.756 (95%CI: 0.736–0.755), 0.806 (95% CI: 0.782–0.829) and 14.006 (95% CI: 10.167–19.294) in the exosomal RNA group, 0.834 (95%CI: 0.805–0.859), 0.683 (95% CI: 0.651–0.715) and 19.559 (95% CI: 6.961–54.957) in the metabolites group and 0.701 (95%CI: 0.649–0.750), 0.760 (95% CI: 0.705–0.809) and 9.875 (95% CI: 4.749–20.534) in the group of extra-cellular vesicle micro-RNA.

According to the tumor stage, the pooled SEN, SPE and DOR accounted, respectively, for 0.777 (95%CI: 0.767–0.788), 0.806 (95% CI: 0.794–0.817) and 15.587 (95% CI: 12.367–19.644) in the group of stages III and IV. In early stages (I/II), these diagnostic performances were estimated, respectively, to 0.750 (95%CI: 0.735–0.764), 0.790 (95% CI: 0.776–0.803) and 15.385 (95% CI: 11.695–20.239) (Figure 4(f) and (g)). In the studies, assessing all the stages, pooled SEN, SPE and dOR accounted for 0.717 (95%CI: 0.695–0.738), 0.729 (95% CI: 0.703–0.753) and 7.675 (95% CI: 5.406–10.895).

According to the sample used, the pooled SEN, SPE and dOR were estimated, respectively, to 0.772 (95%CI: 0.759–0.785), 0.776 (95% CI: 0.761–0.791) and 15.829 (95% CI: 11.869–21.111) in the studies assessing biomarkers in plasma, 0.756 (95%CI: 0.745–0.767), 0.794 (95% CI: 0.782–0.805) and 12.912 (95% CI: 10.028–16.625) in the studies using serum, 0.794 (95%CI: 0.723–0.854), 0.909 (95% CI: 0.783–0.975) and 50.911 (95% CI: 1.572–(−)1649.1) in the studies using whole blood samples, 0.743 (95%CI: 0.717–0.768), 0.814 (95% CI: 0.792–0.833) and 14.045 (95% CI: 10.531–18.730) in the group using sputum, 0.667 (95%CI: 0.569–0.754), 0.630 (95% CI: 0.523–0.729) and 3.413 (95% CI: 1.904–6.117) in the group using exhaled breath and 0.671 (95%CI: 0.593–0.742), 0.767 (95% CI: 0.640–0.866) and 10.774 (95% CI: 3.466–33.491) in the group using pleural effusion. Only one study used BAL and one study used peripheral blood mononucleated cells inducing their exclusion from the subgroup analysis.

All the details of the subgroup analysis are represented in Table 4.

Discussion

In this study, we aimed to assess the diagnostic performances of the micro-RNA and the metabolites in the diagnosis of lung cancers. For that issue, we performed a meta-analysis during a nine-year-period (2013–2022). According to our inclusion criteria, 165 studies from 79 articles were included. The pooled SEN, SPE and DOR accounted, respectively, for 0.76, 0.79 and 13.927. The AUC was estimated to 0.859 suggesting a good diagnostic accuracy. The heterogeneity in the pooled SEN and SPE was statistically significant. The meta-regression analysis focusing on the technique used, the sample, the number of biomarkers, the biomarker subtype, the tumor stage and the ethnicity revealed the biomarker number (p = 0.009) and the tumor stage (p = 0.0241) as potential sources of heterogeneity. A sub-group analysis was performed in order to assess the diagnostic performances according to the technique used, the biomarkers number, the tumor stage, the biomarker subtype and the sample used. The best diagnostic performances were highlighted in the studies using PCR digital techniques (SEN = 0.748, SPE = 0.815, DOR = 14.522), studies using panel of biomarkers or classifiers (SEN = 0.788, SPE-0.799, DOR = 17.914), studies analyzing long non-coding RNA (SEN = 0.689, SPE = 0.864, DOR = 14.952), studies focusing on metastatic and late stages (SEN = 0.777, SPE = 0.806, DOR = 15.587) and studies including whole blood samples (SEN = 0.794, SPE = 0.909, DOR = 50.911). We also assessed the quality of the different studies included using the most relevant items from the QUADAS-2 criteria. The included studies were rated taking into account the four key domains: patient selection, index test, reference standard and flow and timing.¹² The mean score of the different studies included accounted for 3.45/11 highlighting studies with the lowest scores attributed to the lack of information about the patient selection domain. This fact couldn’t be attributed to the low quality of the studies but should encourage authors to mention more details about the inclusion of the patients, the reference test and the index test in the methods section. In a meta-analysis focusing on the diagnostic and prognostic value of miR-155, Sha et al. reported that the quality assessment according to the QUADAS revealed that most studies were consistent with the quality criteria. The discordance with our results may be explained by the few number of studies included (12 studies). Concerning the different techniques of assessment used, the review of the literature revealed RT-PCR for micro-RNA and mass spectrometry for metabolites as the most used techniques. In our sub-group analysis, the digital PCR group seemed to have better diagnostic performances than the others. The plasma and the whole blood were also the most reported samples with high diagnostic performances in some studies.⁸⁹ In our meta-analysis, whole blood-based studies seemed to have higher diagnostic performances. The diagnostic performance of panels has been compared to the use of single biomarkers and revealed in many studies the superiority of using multiple biomarkers.⁹⁰ This fact was also highlighted in our meta-regression analysis and in the sub-group analysis. In this study, the sub-group analysis revealed better diagnostic performances of biomarkers in late lung cancers. Zhang et al. reported in a study assessing the diagnostic performances of exosomal miR-20b-5p and miR-3187, an AUC of 0.838 in early-stage cancers but they didn’t include late-stage cancers to compare them.⁹¹ In a meta-analysis about the diagnostic value of microRNA-486, Shen et al. reported a better diagnostic value of mi-R in early-stage cancers but they included only two studies assessing late-stage cancers.⁹² In another meta-analysis about the diagnostic value of micro-RNA 25 in patients with NSCLC, Li et al. reported that these biomarkers could be potential biomarkers in early lung cancers but, in their meta-analysis including six studies, only one study focused on early-stage lung cancer.⁹³

In almost all meta-analyses focusing on the biomarkers of lung cancer, in diagnostic, prognostic or therapeutic fields, the authors reported a significant heterogeneity. Huang et al. reported in a meta-analysis, about the potential diagnostic value of extracellular vesicle miRNA in NSCLC, that the meta-regression analysis revealed the origin countries, the publication year, the sample size, the source of control, the detection method and the specimen used as potential sources of heterogeneity.⁹⁴ In another systematic review about the circulating micro-RNA-145 as a diagnostic biomarker, Tao et al. reported the publication date and the sample as potential sources of heterogeneity.⁹⁵ Jiang et al. reported the smoking status, the number of biomarkers and the ethnicity as potential sources of heterogeneity.⁹⁶ In a meta-analysis focusing on micro-RNA-486, Shen et al. reported that the microscopic subtype was also a potential source of heterogeneity.⁹²

The diagnostic performances in this study are almost the same than those published in other meta-analyses focusing of micro-RNA, exosomal micro-RNA or extracellular vescile micro-RNA even if the inclusion criteria or the period of study may be different.^94,96–100

Conclusion

Even if this meta-analysis highlights the diagnostic potential of microRNA and metabolites in lung cancer, it also reinforces the need for consistent prospective studies in order to assess that finding especially in early-stage lung cancers in order to increase the rate of lung cancers resected totally surgically and to improve their prognoses.

Footnotes

Authors’ contributions

MM performed the review of the literature, the statistical analysis and wrote the manuscript, MMZ performed the review of the literature and red the final version of the manuscript, MA and IB contributed to the writing of the manuscript and red the final version, FM participated to the quality analysis of the articles included and corrected the final version of the manuscript, AM validated the statistical study and corrected the final version of the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Not applicable.

Informed consent

Not applicable.

References

Hofman

Popper

. Pathologists and liquid biopsies: to be or not to be? Virchows Arch 2016; 469: 601–609.

Wadowska

Bil-Lula

Trembecki

, et al. Genetic markers in lung cancer diagnosis: a review. Int J Mol Sci 2020; 21: 1–24.

Pérol

Arpin

Soria

. Biomarqueurs du cancer bronchique non à petites cellules : du concept à la pratique clinique – Biomarkers for non-small-cell lung cancer: from concept to molecular analysis-based treatment strategies. 2008.

Seijo

Peled

Ajona

, et al. Biomarkers in lung cancer screening: achievements, promises, and challenges. J Thorac Oncol 2019; 14: 343–357.

Ostrin

Sidransky

Spira

, et al. Biomarkers for lung cancer screening and detection. Cancer Epidemiol Biomarkers Prev 2020; 29: 2411–2415.

Maly

Kolostova

, et al. Circulating tumor cells in diagnosis and treatment of lung cancer. In Vivo 2019; 33: 1027–1037.

Hofman

Heeke

Marquette

, et al.

Circulating tumor cell detection in lung cancer: But to what end?

Cancers (Basel) 2019; 11: 1–10. DOI:10.3390/cancers11020262

Heeke

Mograbi

Alix-Panabières

, et al. Never travel alone: The crosstalk of circulating tumor cells and the blood microenvironment. Cells 2019; 8: 24–35. DOI:10.3390/cells8070714

Marquette

Boutros

Benzaquen

, et al. Circulating tumour cells as a potential biomarker for lung cancer screening: a prospective cohort study. Lancet Respir Med 2020; 8: 709–716.

10.

Devriese

Bosma

van de Heuvel

, et al. Circulating tumor cell detection in advanced non-small cell lung cancer patients by multi-marker QPCR analysis. Lung Cancer 2012; 75: 242–247.

11.

Shen

. Exosomal miRNAs as biomarkers for diagnostic and prognosticin lung cancer. Cancer Med 2020; 9: 6909–6922.

12.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev 2021; 10: 21–32. DOI:10.1186/s13643-021-01626-4

13.

Whiting

Rutjes

AWS

Westwood

, et al. QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies. 2011. www.annals.org.

14.

Tang

Shen

Wang

, et al. Identification of plasma microRNAs as novel noninvasive biomarkers for early detection of lung cancer. Eur J Cancer Prev 2013; 22: 540–548.

15.

Geng

Fan

Zhang

, et al. Five microRNAs in plasma as novel biomarkers for screening of early-stage non-small cell lung cancer. Respir Res 2014; 15: 1–9.

16.

Wang

Yang

Zhang

, et al. Early detection of lung cancer in serum by a panel of microRNA biomarkers. Clin Lung Cancer 2015; 16: 313–319.e1.

17.

Zhu

Zhou

Zha

, et al. Diagnostic value of serum miR-182, miR-183, miR-210, and miR-126 levels in patients with early-stage non-small cell lung cancer. PLoS One 2016; 11(1): 1–40. DOI:10.1371/journal.pone.0153046

18.

Shang

Xie

Wang

, et al. Predicative values of serum microRNA-22 and microRNA-126 levels for non-small cell lung cancer development and metastasis: a case-control study. Neoplasma 2017; 64: 453–459.

19.

Bagheri

Khorshid

HRK

Tavallaie

, et al. A panel of noncoding RNAs in non–small-cell lung cancer. J Cell Biochem 2019; 120: 8280–8290.

20.

Leng

Lin

Zhan

, et al. An integromic signature for lung cancer early detection. Oncotarget 2018; 9: 24684–24692.

21.

Abdollahi

Rahmati

Ghaderi

, et al. A combined panel of circulating microRNA as a diagnostic tool for detection of the non-small cell lung cancer. QJM 2019; 112: 779–785.

22.

Wang

, et al. Clinical significance of the expression of miRNA-21, miRNA-31 and miRNA-let7 in patients with lung cancer. Saudi J Biol Sci 2019; 26: 777–781.

23.

Wang

Jia

Shi

, et al. Identification and evaluation of 2 circulating microRNAs for non-small cell lung cancer diagnosis. Clin Exp Pharmacol Physiol 2018; 45: 1083–1086.

24.

Zhang

Mao

Shen

, et al. Plasma miR-145, miR-20a, miR-21 and miR-223 as novel biomarkers for screening early-stage non-small cell lung cancer. Oncol Lett 2017; 13: 669–676.

25.

Sun

Mei

, et al. Circulating microRNA-339-5p and −21 in plasma as an early detection predictors of lung adenocarcinoma. Pathol Res Pract 2018; 214: 119–125.

26.

Abd-El-Fattah

Sadik

NAH

Shaker

, et al. Differential MicroRNAs expression in serum of patients with lung cancer, pulmonary tuberculosis, and pneumonia. Cell Biochem Biophys 2013; 67: 875–884.

27.

Chen

, et al. Expression of mir-148/152 family as potential biomarkers in non-small-cell lung cancer. Med Sci Monit 2015; 21: 1155–1161.

28.

Yang

Zhuo

, et al. Prognostic value of microRNA-10b overexpression in peripheral blood mononuclear cells of nonsmall-cell lung cancer patients. Tumor Biol 2015; 36: 7069–7075.

29.

Anjuman

Guarnera

, et al. Analysis of lung flute-collected sputum for lung cancer diagnosis. Biomark Insights 2015; 10: 56–61.

30.

Fang

Jiang

. Integrating DNA methylation and microRNA biomarkers in sputum for lung cancer detection. Clin Epigenetics 2016; 8: 12–40. DOI:10.1186/s13148-016-0275-5

31.

Sui

Zhang

Zhu

. Diagnostic value of serum miR197 and miR145 in non-small cell lung cancer. Oncol Lett 2019; 17: 3247–3252.

32.

Zhou

Wen

Shan

, et al. A six-microRNA panel in plasma was identified as a potential biomarker for lung adenocarcinoma diagnosis. Oncotarget 2017; 8: 6513–6525.

33.

Wang

Guo

Song

, et al. Tumor-associated exosomal miRNA biomarkers to differentiate metastatic vs. Nonmetastatic non-small cell lung cancer. Clin Chem Lab Med 2020; 58: 1535–1545.

34.

Mozzoni

Banda

Goldoni

, et al. Plasma and EBC microRNAs as early biomarkers of non-small-cell lung cancer. Biomarkers 2013; 18: 679–686.

35.

Razzak

Bédard

ELR

Kim

, et al. MicroRNA expression profiling of sputum for the detection of early and locally advanced non-small-cell lung cancer: a prospective case-control study. Current Oncol 2016; 23: e86–e94.

36.

Tamiya

Mitani

Saito

, et al. Exosomal microRNA expression profiling in patients with lung adenocarcinoma-associated malignant pleural effusion. Anticancer Res 2018; 38: 6707–6714.

37.

Fang

Liu

, et al. Extracellular vesicle-delivered miR-505-5p, as a diagnostic biomarker of early lung adenocarcinoma, inhibits cell apoptosis by targeting TP53AIP1. Int J Oncol 2019; 54: 1821–1832.

38.

Yang

Wang

, et al. Exosomal miR-21/Let-7a ratio distinguishes non-small cell lung cancer from benign pulmonary diseases. Asia Pac J Clin Oncol 2020; 16: 280–286.

39.

Halvorsen

Bjaanaes

Leblanc

, et al. A unique set of 6 circulating microRNAs for early detection of non-small cell lung cancer. Oncotarget 2016; 7: 37250–37259.

40.

Kim

Gazala

Razzak

, et al. Non-small cell lung cancer detection using microRNA expression profiling of bronchoalveolar lavage fluid and sputum. Anticancer Res 2015; 35: 1873–1880.

41.

Jin

Chen

, et al. Evaluation of tumor-derived exosomal miRNA as potential diagnostic biomarkers for early-stage non-small cell lung cancer using next-generation sequencing. Clin Cancer Res 2017; 23: 5311–5319.

42.

Liu

Kannisto

, et al. Non-invasive detection of exosomal MicroRNAs via tethered cationic lipoplex nanoparticles (tCLN) biochip for lung cancer early detection. Front Genet 2020; 11: 12–22. DOI:10.3389/fgene.2020.00258

43.

Sanfiorenzo

Ilie

Belaid

, et al. Two panels of plasma MicroRNAs as non-invasive biomarkers for prediction of recurrence in resectable NSCLC. PLoS One 2013; 8: 12–24. DOI:10.1371/journal.pone.0054596

44.

Ulivi

Foschi

Mengozzi

, et al. Peripheral blood miR-328 expression as a potential biomarker for the early diagnosis of NSCLC. Int J Mol Sci 2013; 14: 10332–10342.

45.

Xing

Guarnera

, et al. Sputum microRNA biomarkers for identifying lung cancer in indeterminate solitary pulmonary nodules. Clin Cancer Res 2015; 21: 484–489.

46.

Shen

Liao

Guarnera

, et al. Analysis of microRNAs in sputum to improve computed tomography for lung cancer diagnosis. J Thorac Oncol 2014; 9: 33–40.

47.

Guarnera

, et al. Digital PCR quantification of miRNAs in sputum for diagnosis of lung cancer. J Cancer Res Clin Oncol 2014; 140: 145–150.

48.

Zhang

Yang

, et al. Prognostic significance of serum microRNA-210 levels in nonsmall-cell lung cancer. J Int Med Res 2013; 41: 1437–1444.

49.

Anjuman

Guarnera

, et al. Evaluation of lung flute in sputum samples for molecular analysis of lung cancer. Clin Transl Med 2013; 2: 1–6.

50.

Gao

Chang

Wang

, et al. Potential diagnostic value of miR-155 in serum from lung adenocarcinoma patients. Oncol Rep 2014; 31: 351–357.

51.

Dong

Zhang

, et al. The role of microRNA-2I in predicting brain metastases from non-small cell lung cancer. Onco Targets Ther 2017; 10: 185–194.

52.

Abu-Duhier

Javid

Sughayer

, et al. Clinical significance of circulatory miRNA-21 as an efficient non-invasive biomarker for the screening of lung cancer patients. Asian Pac J Cancer Prev 2018; 19: 2607–2611.

53.

Wang

Ding

, et al. Identification of miRNAs as non-invasive biomarkers for early diagnosis of lung cancers. Tumor Biol 2016; 37: 16287–16293.

54.

Zhang

Yin

, et al. Detection of circulating exosomal miR-17-5p serves as a novel non-invasive diagnostic marker for non-small cell lung cancer patients. Pathol Res Pract 2019; 215: 12–20. DOI:10.1016/j.prp.2019.152466

55.

Liu

, et al. Clinical significance of serum miR-25 in non-small-cell lung cancer. Br J Biomed Sci 2019; 76: 111–116.

56.

Yang

J-S

B-J

H-W

, et al. Serum miR-152, miR-148a, miR-148b, and miR-21 as novel biomarkers in non-small cell lung cancer screening. Tumor Biol 2015; 36: 3035–3042.

57.

Zhu

Wang

Wei

, et al. Hsa_circ_0013958: a circular RNA and potential novel biomarker for lung adenocarcinoma. FEBS J 2017; 284: 2170–2182.

58.

Wang

Chen

, et al. Hsa_circ_0079530 promotes cell proliferation and invasion in non-small cell lung cancer. Gene 2018; 665: 1–5.

59.

Hang

Zhou

Qin

, et al. A novel plasma circular RNA circFARSA is a potential biomarker for non-small cell lung cancer. Cancer Med 2018; 7: 2783–2791.

60.

Liu

Zhang

, et al. Downregulation of serum exosomal miR-216b predicts unfavorable prognosis in patients with non-small cell lung cancer. Cancer Biomark 2019; 27: 113–120.

61.

Zhao

J-J

Zhang

, et al. Serum miR-21 level: a potential diagnostic and prognostic biomarker for non-small cell lung cancer. 2015.

62.

Cazzoli

Buttitta

Di Nicola

, et al. microRNAs derived from circulating exosomes as noninvasive biomarkers for screening and diagnosing lung cancer. J Thorac Oncol 2013; 8: 1156–1162.

63.

Xie

Meng

, et al. Early lung cancer diagnostic biomarker discovery by machine learning methods. Transl Oncol 2020; 14: 12–24. DOI:10.1016/j.tranon.2020.100907

64.

X-J

Hayward

Fong

P-Y

, et al. A blood-based proteomic classifier for the molecular characterization of pulmonary nodules. Sci Transl Med 2013; 5: 1–10.

65.

Klupczynska

Dereziński

Garrett

, et al. Study of early stage non-small-cell lung cancer using orbitrap-based global serum metabolomics. J Cancer Res Clin Oncol 2017; 143: 649–659.

66.

Zhang

Zheng

Ahmed

, et al. A high-performing plasma metabolite panel for early-stage lung cancer detection. Cancers (Basel) 2020; 12: 34–40. DOI:10.3390/cancers12030622

67.

Terlizzi

Colarusso

De Rosa

, et al. Circulating and tumor-associated caspase-4: a novel diagnostic and prognostic biomarker for non-small cell lung cancer. 2018. www.oncotarget.com.

68.

Chen

, et al. Global lipidomics identified plasma lipids as novel biomarkers for early detection of lung cancer. Oncotarget 2017; 8: 107899–107906.

69.

Sun

Zhao

Wang

, et al. Integration of metabolomic and transcriptomic profiles to identify biomarkers in serum of lung cancer. J Cell Biochem 2019; 120: 11981–11989.

70.

Fahrmann

Kim

DeFelice

, et al. Investigation of metabolomic blood biomarkers for detection of adenocarcinoma lung cancer. Cancer Epidemiol Biomarkers Prev 2015; 24: 1716–1723.

71.

Ros-Mazurczyk

Jelonek

Marczyk

, et al. Serum lipid profile discriminates patients with early lung cancer from healthy controls. Lung Cancer 2017; 112: 69–74.

72.

Guarnera

, et al. Quantification of plasma miRNAs by digital PCR for cancer diagnosis. Biomark Insights 2013; 8: 1–7. DOI:10.4137/BMI.S13154

73.

Zhang

Mao

Shen

, et al. Plasma miR-145, miR-20a, miR-21 and miR-223 as novel biomarkers for screening early-stage non-small cell lung cancer. Oncol Lett 2017; 13: 669–676.

74.

Zhang

, et al. The role of plasma miRNAs in the diagnosis of pulmonary nodules. J Thorac Dis 2018; 10: 4032–4041.

75.

Aiso

Ohtsuka

Ueda

, et al. Serum levels of candidate microRNA diagnostic markers differ among the stages of non-small-cell lung cancer. Oncol Lett 2018; 16: 6643–6651.

76.

Zhou

Huang

Zhang

, et al. MicroRNAs: A novel potential biomarker for diagnosis and therapy in patients with non-small cell lung cancer. Cell Prolif 2017; 50: 23–35. DOI:10.1111/cpr.12394

77.

Czarnecka

Szmyd

Barańska

, et al. A strong decrease in TIMP3 expression mediated by the presence of miR-17 and 20a enables extracellular matrix remodeling in the NSCLC lesion surroundings. Front Oncol 2019; 9: 23–34. DOI:10.3389/fonc.2019.01372

78.

Yao

, et al. A panel of miRNAs derived from plasma extracellular vesicles as novel diagnostic biomarkers of lung adenocarcinoma. FEBS Open Bio 2019; 9: 2149–2158.

79.

Xia

Wei

, et al. Exosome-mediated transfer of miR-1260b promotes cell invasion through Wnt/β–catenin signaling pathway in lung adenocarcinoma. J Cell Physiol 2020; 235: 6843–6853.

80.

Lin

, et al. Combination of serum miRNAs with serum exosomal miRNAs in early diagnosis for non-small-cell lung cancer. Cancer Manag Res 2020; 12: 485–495.

81.

Han

Zhang

, et al. Tumor-derived circulating exosomal miR-342-5p and miR-574-5p as promising diagnostic biomarkers for early-stage lung adenocarcinoma. Int J Med Sci 2020; 17: 1428–1438.

82.

Sun

Chen

, et al. Increased plasma miRNA-30a as a biomarker for non-small cell lung cancer. Med Sci Monit 2016; 22: 647–655.

83.

Gao

Wang

Zhao

, et al. Plasma miR-324-3p and miR-1285 as diagnostic and prognostic biomarkers for early stage lung squamous cell carcinoma. Oncotarget 2016; 7: 59664–59675.

84.

Wei

Gao

Zhu

C-J

, et al. Identification of plasma microRNA-21 as a biomarker for early detection and chemosensitivity of non-small cell lung cancer. Chin J Cancer 2011; 30: 407–414.

85.

Leidinger

Brefort

Backes

, et al. High-throughput qRT-PCR validation of blood microRNAs in non-small cell lung cancer. Oncotarget 2015; 7: 4611–4623.

86.

Nadal

Truini

Nakata

, et al. A novel Serum 4-microRNA signature for lung cancer detection. Sci Rep 2015; 5:12-23: 12–23. DOI:10.1038/srep12464

87.

Lin

Zhan

, et al. Differential miRNA expressions in peripheral blood mononuclear cells for diagnosis of lung cancer. Lab Invest 2015; 95: 1197–1206.

88.

Dou

Wang

, et al. Decreased plasma let-7c and miR-152 as noninvasive biomarker for non-small-cell lung cancer. Int J Clin Exp Med 2015; 8: 9291–9298.

89.

Shen

Wang

Yang

, et al. Diagnostic value of circulating microRNAs for lung cancer: a meta-analysis. Genet Test Mol Biomarkers 2013; 17: 359–366.

90.

Gao

Shi

, et al. Relationship between diagnostic accuracy of microRNAs for NSCLC and number of combined microRNAs: an overview with meta-analysis. Expert Rev Mol Diagn 2021; 21: 983–993.

91.

Zhang

Song

Xie

, et al. Circulating serum exosomal miR-20b-5p and miR-3187-5p as efficient diagnostic biomarkers for early-stage non-small cell lung cancer. Exp Biol Med 2020; 245: 1428–1436.

92.

Shen

Zhang

, et al. Diagnostic and prognostic value of microRNA-486 in patients with lung cancer: a systematic review and meta-analysis. Int J Biol Markers 2022; 37: 377–385.

93.

Sun

Zhou

, et al. Diagnostic value of microRNA-25 in patients with non-small cell lung cancer in Chinese population: A systematic review and meta-analysis. Medicine (Baltimore) 2020; 99: 23–30. DOI:10.1097/MD.0000000000023425

94.

Huang

Zhu

Lin

, et al. The potential diagnostic value of extracellular vesicle miRNA for human non-small cell lung cancer: a systematic review and meta-analysis. Expert Rev Mol Diagn 2021; 21: 823–836.

95.

Tao

Zhou

, et al. Circulating microRNA-145 as a diagnostic biomarker for non-small-cell lung cancer: a systemic review and meta-analysis. Int J Biol Markers 2020; 35: 51–60.

96.

Jiang

Quan

, et al. Clinically correlated MicroRNAs in the diagnosis of non-small cell lung cancer: a systematic review and meta-analysis. Biomed Res Int 2018; 12: 12–34. DOI:10.1155/2018/5930951

97.

Wang

Zhao

, et al. Clinical use of microRNAs as potential non-invasive biomarkers for detecting non-small cell lung cancer: a meta-analysis. Respirology 2015; 20: 56–65.

98.

Yang

Wang

Zhang

, et al. The accuracy of microRNA-210 in diagnosing lung cancer: a systematic review and meta-analysis. Oncotarget 2016; 7: 63283–63293.

99.

Huang

Zhu

Lin

, et al. The potential diagnostic value of extracellular vesicle miRNA for human non-small cell lung cancer: a systematic review and meta-analysis. Expert Rev Mol Diagn 2021; 21: 823–836.

100.

Wang

Liu

, et al. MicroRNA-21 as a diagnostic and prognostic biomarker of lung cancer: a systematic review and meta-analysis. Biosci Rep 2022; 42: 56–63. DOI:10.1042/BSR20211653