A Novel Dried Blood Spot Detection Strategy for Characterizing Cardiovascular Diseases

Linsheng Liu; Xurui Jin; Yangfeng Wu; Mei Yang; Tao Xu; Xianglian Li; Jianhong Ren; Lijing L Yan

doi:10.3389/fcvm.2020.542519

A Novel Dried Blood Spot Detection Strategy for Characterizing Cardiovascular Diseases

Front Cardiovasc Med. 2020 Oct 9:7:542519. doi: 10.3389/fcvm.2020.542519. eCollection 2020.

Authors

Linsheng Liu¹, Xurui Jin², Yangfeng Wu³, Mei Yang⁴, Tao Xu^{5

6}, Xianglian Li¹, Jianhong Ren⁴, Lijing L Yan²

Affiliations

¹ Clinical Pharmacology Research Laboratory, The First Affiliated Hospital of Soochow University, Suzhou, China.
² Global Health Research Center, Duke Kunshan University, Kunshan, China.
³ Peking University Clinical Research Institute, Beijing, China.
⁴ Suzhou BioNovoGene Metabolomics Platform, Suzhou, China.
⁵ The Key Laboratory of Developmental Genes and Human Disease, Institute of Life Sciences, Southeast University, Nanjing, China.
⁶ The Therapeutic Antibody Research Center of SEU-Alphamab, Southeast University, Nanjing, China.

Abstract

Cardiovascular diseases (CVDs) are the leading cause of death in China. Conventional diagnostic methods are dependent on advanced instruments, which are expensive, inaccessible, and inconvenient in underdeveloped areas. To build a novel dried blood spot (DBS) detection strategy for imaging CVDs, in this study, a total of 12 compounds, including seven amino acids [homocysteine (Hcy), isoleucine (Ile), leucine (Leu), valine (Val), phenylalanine (Phe), tyrosine (Tyr), and tryptophan (Trp)], three amino acid derivatives [choline, betaine, and trimethylamine N-oxide (TMAO)], L-carnitine, and creatinine, were screened for their ability to identify CVD. A rapid and reliable method was established for the quantitative analysis of the 12 compounds in DBS. A total of 526 CVD patients and 200 healthy volunteers in five provinces of China were recruited and divided into the following groups: stroke, coronary heart disease, diabetes, and high blood pressure. The orthogonal projection to latent structures-discriminant analysis (OPLSDA) model was used to characterize the difference between each CVD group. Marked differences between the groups based on the OPLSDA model were observed. Based on the model, the patients in the three training sets were mostly accurately categorized into the appropriate group. In addition, the receiver operating characteristic (ROC) curves and logistic regression of each metabolite chosen by the OPLSDA model had an excellent predictive value in both the test and validation groups. DBS detection of 12 biomarkers was sensitive and powerful for characterizing different types of CVD. Such differentiation may reduce unnecessary invasive coronary angiography, enhance predictive value, and complement current diagnostic methods.

Keywords: DBS; biomarker; cardiovascular diseases; metabolomics; risk prediction.