Integration of miRNA expression analysis of purified leukocytes and whole blood reveals blood-borne candidate biomarkers for lung cancer

Guini Hong Yue Huo Yaru Gao Liyuan Ma Shuang Li Tian Tian Haijian Zhong Hongdong Li a School of Medical Information Engineering,Gannan Medical University,Ganzhou,Chinab School of Public Health and Health Management,Gannan Medical University,Ganzhou,China
DOI: https://doi.org/10.1080/15592294.2024.2393948
2024-08-22
Epigenetics
Abstract:Changes in leukocyte populations may confound the disease-associated miRNA signals in the blood of cancer patients. We aimed to develop a method to detect differentially expressed miRNAs from lung cancer whole blood samples that are not influenced by variations in leukocyte proportions. The Ref-miREO method identifies differential miRNAs unaffected by changes in leukocyte populations by comparing the within-sample relative expression orderings (REOs) of miRNAs from healthy leukocyte subtypes and those from lung cancer blood samples. Over 77% of the differential miRNAs observed between lung cancer and healthy blood samples overlapped with those between myeloid-derived and lymphoid-derived leukocytes, suggesting the potential impact of changes in leukocyte populations on miRNA profile. Ref-miREO identified 16 differential miRNAs that target 19 lung adenocarcinoma-related genes previously linked to leukocytes. These miRNAs showed enrichment in cancer-related pathways and demonstrated high potential as diagnostic biomarkers, with the LASSO regression models effectively distinguishing between healthy and lung cancer blood or serum samples (all AUC > 0.85). Additionally, 12 of these miRNAs exhibited significant prognostic correlations. The Ref-miREO method offers valuable candidates for circulating biomarker detection in cancer that are not affected by changes in leukocyte populations.
genetics & heredity,biochemistry & molecular biology
What problem does this paper attempt to address?