Multi-site DNA methylation alterations of peripheral blood mononuclear cells serve as novel biomarkers for the diagnosis of AIS/stage I lung adenocarcinoma: A multi-center cohort study
Peilong Li,Shibiao Liu,Tiantian Wang,Fang Wang,Juan Li,Qiuchen Qi,Shujun Zhang,Yan Xie,Jianping Li,Yongcai Zhu,Suli Yang,Guotao Yin,Xiaoyi He,Shijun Li,Huiting Xu,Mengqiu Xiong,Guanghua Li,Yi Zhang,Lutao Du,Chuanxin Wang
DOI: https://doi.org/10.1097/JS9.0000000000002101
2024-10-01
Abstract:Background: Early diagnosis remains an obstacle for improving the outcome of lung adenocarcinoma (LUAD). DNA methylation changes in peripheral blood mononuclear cells (PBMCs) could reflect immune response to tumorigenesis, providing the theoretical basis for early cancer diagnosis based on immune cell profiling. Methods: This multi-center study evaluated the DNA methylation patterns based on PBMCs samples from 1115 individuals at nine medical centers. Genome-wide DNA methylation profiling of PBMCs in a discovery cohort (35 LUAD patients and 50 healthy controls) was performed using Illumina 850K microarray. Candidate differentially methylated CpG positions (DMPs) were selected and validated in a two-step DMPs screening cohort (65 LUAD patients and 80 healthy controls) by pyrosequencing and multiple target region methylation enrichment sequencing (MTRMES). Then, an early LUAD Diagnostic Panel (LDP score) based on multi-site methylation-specific chip-based digital PCR was constructed in a training set and then confirmed in a validation set from the LDP score development cohort (389 AIS/stage I LUAD patients and 293 healthy controls). Besides, we included 157 other cancer patients, including 52 gastric cancer (GC) patients, 50 breast cancer (BC) patients, and 55 colorectal cancer (CRC) patients to assess the specificity of LDP score. In addition, we also evaluated the early warning ability of LDP score for LUAD in a prospective cohort (46 people who were at high-risk of developing LC). Results: A total of 1415 LUAD-specific DMPs were identified. Then, six DMPs were selected for validation and three DMPs were finally verified. The LDP score was constructed by combining the three DMPs, age, and gender, and showed an AUC of 0.916, sensitivity of 88.17%, and specificity of 80.20% in combined set, outperforming traditional methods, such as CEA and CT (detection rate: 87.79% vs. 4.69%; 87.79% vs. 35.21%). This diagnostic performance was confirmed in sub-types of LUAD with clinical challenges, such as 6-20 mm LUAD (AUC: 0.914, 95%CI: 0.889-0.934) and ground-glass nodules (AUC: 0.916, 95%CI: 0.889-0.938). Importantly, our LDP score had significant improvement in terms of selecting high-risk individuals who should receive low-dose computed tomography (87.80% vs. 9.28%). Remarkably, LDP score could predict LUAD around two years before clinical diagnosis in our prospective cohort. Conclusions: The novel developed LDP score represented a convenient and effective assay for the detection of AIS/stage I LUAD with high sensitivity and specificity, and had demonstrated unique advantages over traditional detection methods.