The concordance and correlation of measurements by multiple electrode and light transmittance aggregometries based on the pre-defined cutoffs of high and low on-treatment platelet reactivity

Abstract

The consensus document suggested the definition of high on-treatment platelet reactivity (HPR) and future directions. Although multiple platelet function assays have developed based on different mechanisms, inter-assay concordance of HPR identification may be an important pressing need. This study was performed to correlate between the cutoffs of HPR suggested by multiple electrode (MEA) and light transmittance aggregometries (LTA). We enrolled 246 consecutive patients undergoing non-emergent percutaneous coronary intervention after dual antiplatelet therapy. On the basis of consensus document, the cutoffs of HPR to adenosine diphosphate (ADP) were defined as ADPtest >= 47 U, and 5 and 20 mu M ADP-induced maximal platelet aggregation (MPA) >= 46% and 59%, respectively. In addition, the cutoff of low PR (LPR) for major bleeding was selected as ADPtest <= 19 U. ADPtest showed moderate correlations with ADP-based LTA data (0.663 <= r <= 0.710). In the receiver-operating characteristics (ROC) curve analysis, ADPtest >= 47U was corresponded to 5 and 20 mu M ADP-induced MPAs >= 46.4% and >= 56.8%, respectively. Good agreements were observed between ADPtest >= 47 U, and 5 mu M ADP-induced MPA >= 46% (kappa = 0.537, 80.5% of concordance rate) and 20 mu M ADP-induced MPA >= 59% (kappa = 0.564, 81.7% of concordance rate). In the ROC curve analysis for the cutoff of LPR (ADPtest <= 19 U), 5 and 20 mu M ADP-induced MPAs <= 26.6% and <= 35.3%, respectively, were suggested as the hypothetical threshold for major bleeding. On the basis of consensus document, the cutoffs of MEA- and LTA-based HPR are well matched. However, the agreement of HPR between assays is moderate, which may implicate the limitation of risk stratification by platelet function testing.