Diagnosis of pulmonary nodules by DNA methylation analysis in bronchoalveolar lavage fluids

Lei Li,Weihe Liang,Hao Yang,Qiaoyun Huang,Jian-Bing Fan,Bo Wang,Zhujia Ye,Yanying Liu,Mengzhu Yang,Yingying Zhu,Jinsheng Tao,Weimin Li,Jing Jin,Dan Liu,Zhiwei Chen,Sai Yang,Siyu Chen,Jiehan Xu

doi:10.1186/s13148-021-01163-w

Abstract

BackgroundLung cancer is the leading cause of cancer-related mortality. The alteration of DNA methylation plays a major role in the development of lung cancer. Methylation biomarkers become a possible method for lung cancer diagnosis.ResultsWe identified eleven lung cancer-specific methylation markers (CDO1, GSHR, HOXA11, HOXB4-1, HOXB4-2, HOXB4-3, HOXB4-4, LHX9, MIR196A1, PTGER4-1, and PTGER4-2), which could differentiate benign and malignant pulmonary nodules. The methylation levels of these markers are significantly higher in malignant tissues. In bronchoalveolar lavage fluid (BALF) samples, the methylation signals maintain the same differential trend as in tissues. An optimal 5-marker model for pulmonary nodule diagnosis (malignant vs. benign) was developed from all possible combinations of the eleven markers. In the test set (57 tissue and 71 BALF samples), the area under curve (AUC) value achieves 0.93, and the overall sensitivity is 82% at the specificity of 91%. In an independent validation set (111 BALF samples), the AUC is 0.82 with a specificity of 82% and a sensitivity of 70%.ConclusionsThis model can differentiate pulmonary adenocarcinoma and squamous carcinoma from benign diseases, especially for infection, inflammation, and tuberculosis. The model’s performance is not affected by gender, age, smoking history, or the solid components of nodules.

Highlights

Based on the published data in 2020 [1], lung cancer is one of the most dangerous malignant tumors for human health and life, with the highest mortality rates, and the 5-year relative survival rate for lung cancer is only 19%
Clustering of lung adenocarcinoma (LUAD) and lung squamous carcinoma (LUSC) together suggested that these CpG sites have consistent methylation status between these two groups
Validation of target DNA methylation markers Besides the performance on The Cancer Genome Atlas (TCGA) datasets, we further investigated the potentials of these CpG sites on the classification of malignant and benign pulmonary nodules using clinical pulmonary tissue and bronchoalveolar lavage fluid (BALF) samples, respectively

Summary

Introduction

Based on the published data in 2020 [1], lung cancer is one of the most dangerous malignant tumors for human health and life, with the highest mortality rates, and the 5-year relative survival rate for lung cancer is only 19%. If lung cancer can be diagnosed at the localized stage, especially for non-small cell lung cancer (NSCLC) on stage IA, the 5-year relative survival rate can achieve 92%. It is an effective and essential way to prolong lung cancer patients’ lives by early diagnosis with appropriate treatments. The application of LDCT increases the detection rate of pulmonary nodules and reduces the mortality of lung cancer. Methylation biomarkers become a possible method for lung cancer diagnosis

Methods

Results

Discussion

Conclusion