Proteomic Discovery and Validation of the Confounding Effect of Heparin Administration on the Analysis of Candidate Cardiovascular Biomarkers

Abstract

Abstract BACKGROUND Several plasma proteins have been suggested as markers for a variety of cardiovascular conditions but fail to qualify in independent patient cohorts. This may relate to interference of medication on plasma protein concentrations. We used proteomics to identify plasma proteins that changed in concentration with heparin administration and therefore potentially may confound their evaluation as biomarkers in situations in which heparin is used. METHODS We used a proteomic approach based on isobaric tagging and nano-LC-MS/MS analysis to quantify several hundred proteins in a discovery study in which individual plasma samples from 9 patients at intravascular ultrasound follow-up 12 months after an acute myocardial infarction before heparin administration and 2, 15, and 60 min after heparin administration; we validated our findings in 500 individual plasma samples obtained at admission from patients with suspected ST segment elevation myocardial infarction (STEMI), of whom 363 were treated with heparin before admission. RESULTS In the discovery study, 25 of 653 identified plasma proteins displayed a changed concentration after heparin administration (Bonferroni-corrected P value at P < 7.66 × 10−5). Fourteen of the proteins changed significantly among heparin-treated patients in the validation study (nominal significance level of P < 6.92 × 10−5). Among heparin-affected proteins in both the discovery study and the validation study were midkine, spondin 1, secreted frizzled-like protein 1, lipoprotein lipase, and follistatin, all previously associated with STEMI. CONCLUSIONS Medications such as heparin administration given before blood sampling may confound biomarker discovery and should be carefully considered in such studies.