Virtual Library

Start Your Search

Fugen Li

Author of

  • +

    P2.11 - Screening and Early Detection (ID 178)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Screening and Early Detection
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.11-11 - Exosomal miRNAs as Diagnosis Biomarkers for Distinguishing Benign and Malignant Nodules in Non-Small Cell Lung Cancer (ID 2075)

      10:15 - 18:15  |  Author(s): Fugen Li

      • Abstract


      Lung cancer is the most common malignancy worldwide with the highest morbidity and mortality. The milestone National Lung Screening Trial (NLST) demonstrated a 20% reduction in lung cancer-specific mortality in high-risk individuals with screening low-dose computed tomography (LDCT). However, the high false positive rate of LDCT demands more accurate diagnostic technique. Liquid biopsy represent a valuable non-invasive approach when tumors are not suitable for biopsy or resection. Till now, it is rare to see the studies on plasma exosome-derived miRNAs as early diagnosis biomarkers to distinguish benign and malignant nodules based on miRNA sequencing.


      Forty-eight patients including twenty-eight lung adenocarcinoma and twenty benign nodules with various pathological characteristics (granuloma, atypical neoplasia, fibrosis, tuberculosis, and other benign types) were enrolled as a training cohort. A testing cohort consisted of 46 patients with benign and malignant nodules. Exosomes were precipitated from the plasma of patients, and RNA sequencing were performed to identify the differential expressed miRNAs. The statistical model was trained and tested to discriminate benign nodules from malignant ones.


      The typical exosome markers CD9 and CD63 were successfully confirmed by western blotting of extracted exosomes. Seven differential expressed miRNAs (let-7b-3p, let-7d-3p, miR-150-5p, miR-30e-5p, miR-125b-5p, miR-361-5p, and miR-378c) with median expression > 50 were selected by LASSO-penalized regression to classify the samples into correct groups. LOO-CV (leave one out cross validation) was used to stabilize the model parameter with minimal error rate (16.7%). ROC curves with area under curve (AUC) of the model are achieved to 0.952, while sensitivity and specificity of distinguishing between benign and malignant nodules were 93% and 85%, respectively. In addition, the AUC in a part of testing cohort (1 benign and 11 tumor) was up to 1.000. The rest samples of testing cohort were being detected using RNA sequencing.


      This study using plasma exosome-derived miRNA sequencing identified seven miRNAs to train and test a model to distinguish benign and malignant nodules, which provides insights into the feasibility of exosomal miRNA as a novel early diagnosis approach for lung cancer.