Abstract:e20014 Background: Lung cancer is a prevalent and fatal disease, ranking second among common cancers globally. Early detection and prompt treatment significantly improve survival rates for most cancers, including lung cancer. Presently, low-dose computed tomography (LDCT) is a widely accepted tool for early lung cancer screening. However, it exhibits a high false-positive rate of up to 95%, resulting in overdiagnosis and causing anxiety among patients awaiting diagnosis results. Therefore, investigating alternative effective interventions would provide physicians with additional tools for evaluating lung cancer and making informed decisions. This study employs the concept of liquid biopsy, utilizing next-generation sequencing (NGS) to gather miRNA profiles, aiming to construct a machine learning-driven model for the detection of lung cancer. Methods: We assembled case-control cohorts, comprising 74 lung cancer cases (stage 0-II: 66; III-IV: 8) and 74 age- and gender-matched healthy control subjects, recruited from Chang Gung Memorial Hospital in Linkuo, Taiwan. Plasma derived from whole blood was collected from each subject before any cancer surgery or treatment. Subsequently, we extracted miRNA from the plasma and constructed libraries using the QIAseq miRNA Library kit for miRNA sequencing conducted on the Illumina NextSeq550 platform. RNA-seq data analysis was performed on miRNA sequencing data using the QIAGEN RNA-seq Analysis Portal 5.0. The count per million base (CPM) derived from the quantification results was utilized for differential analysis and constructing the machine learning model. To assess the model's performance, we reserved 10% of the subjects for isolated testing, while the remaining 90% underwent 10-fold cross-validation (CV). Results: The final model, built using logistic regression with six significant differentially expressed miRNAs (fold change > = 2 and P-value < = 0.01), achieved optimal performance. The 10-fold CV demonstrated accuracy: 97.73%; sensitivity: 98.48%; specificity: 96.97%; positive predictive value: 97.01%; negative predictive value: 98.46%. In isolated testing, all 16 subjects, including 8 cases and 8 controls, were correctly identified. Conclusions: This study with 148 subjects showcases the potential of NGS and machine learning for early-stage lung cancer detection via liquid biopsy. However, the sample size is relatively small, lacking cross-site subjects. Therefore, an additional multi-center clinical trial is underway to gather more subjects, aiming for a deeper investigation and more concrete conclusions in the near future.

Detection of miR-155-5p and imaging lung cancer for early diagnosis: in vitro and in vivo study

Evaluation of miR-155 silencing using a molecular beacon in human lung adenocarcinoma cell line

Hairpin allosteric molecular beacons-based cascaded amplification for effective detection of lung cancer-associated microRNA

Detecting mir-155-3p through a Molecular Beacon Bead-Based Assay

Screening and Validation of Mirnas As Biomarkers for Diagnosis of Lung Cancers

Detection of MicroRNA-155 based on lambda exonuclease selective digestion and CRISPR/cas12a-assisted amplification

Highly sensitive fluorescence multiplexed miRNAs biosensors for accurate clinically diagnosis lung cancer disease using LNA-modified DNA probe and DSN enzyme

Visualizing miR-155 To Monitor Breast Tumorigenesis and Response to Chemotherapeutic Drugs by a Self-Assembled Photoacoustic Nanoprobe

Ultrasensitive miRNA detection: A novel DNA nanomachine using split-type molecular beacons-mediated cascade amplification for cancer diagnostics

MiR ‐205‐5p promotes lung cancer progression and is valuable for the diagnosis of lung cancer

In Vitro Research on MicroRNA-155 Targeted Radiolabeled Oligonucleotide in Breast Cancer Cells

Combined SERS Microfluidic Chip with Gold Nanocone Array for Effective Early Lung Cancer Prognosis in Mice Model

Altered Mir-143 and Mir-150 Expressions in Peripheral Blood Mononuclear Cells for Diagnosis of Non-Small Cell Lung Cancer.

Early Detection of Lung Cancer in Serum by a Panel of MicroRNA Biomarkers.

A novel microRNA miR-4433a-3p as a potential diagnostic biomarker for lung adenocarcinoma

The clinical use of circulating microRNAs as non-invasive diagnostic biomarkers for lung cancers.

Application of CRISPR/Cas12a in miRNA-155 Detection: A Novel Homogeneous Electrochemiluminescence Biosensor

MicroRNA biomarkers and their use in evaluating the prognosis of lung cancer

Pump-free microfluidic chip based laryngeal squamous cell carcinoma-related microRNAs detection through the combination of surface-enhanced Raman scattering techniques and catalytic hairpin assembly amplification

[In Vivo Imaging of Breast Tumors by a 99mtc Radiolabeled Probe Targeting Microrna-155 in Mice Models].

Constructing a machine learning-based non-invasive liquid biopsy model for lung cancer detection via miRNA sequencing.