Comparison of platelet function tests for the in vitro quality assessment of platelet concentrates produced under real-life conditions.

Platelet quality in different platelet concentrates (PCs) has been the subject of several studies. Nonetheless, there is a lack of robust data on the correlation and agreement among platelet function tests as a prerequisite for the association of PC functionality in vitro with platelet function in vivo post PC transfusion. The purpose of our study was to correlate a larger panel of platelet function assays in PCs and to assess whether the methods agree sufficiently and can be used interchangeably. Twelve apheresis platelet concentrates in plasma (APC), 16 pooled platelet concentrates in plasma (PPC), and 12 PPC in T-sol (PPCA) were examined on days 1 and 4 after production. PCs were tested for platelet count, light transmission aggregation (LTA) induced by ADP, collagen, or TRAP; platelet ATP release induced by collagen; and spontaneous and ADP and TRAP-induced increase in CD62P and PAC1 expression measured by flow cytometry. All tests were performed in undiluted platelet-rich plasma, recalcified and mixed with an inhibitor of factor Xa and thrombin. Most platelet function parameters correlated significantly with each other, but agreement among methods was insufficient. A proper inverse correlation was observed between ADP-induced LTA and spontaneous platelet activation assessed by CD62P expression ( = -0.61,  < 0.0001). Spontaneous CD62P correlated also significantly with spontaneous PAC1 ( = 0.69,  < 0.0001) and inversely with TRAP-induced CD62P expression ( = -0.86,  < 0.0001). We found significant correlations among all flow cytometric assays measuring platelet CD62P and PAC1 expression induced by ADP or TRAP. Subsequent Bland Altman analysis revealed insufficient agreement between methods. With one exception (collagen-induced LTA compared with TRAP-induced LTA, percentage error = 16%) the limits of agreement expressed as percentage error exceeded the chosen acceptable difference of 30%. In APC, platelet count was 41% and 44% higher, respectively, than in PPC and PPCA ( < 0.0001). Spontaneous CD62P and PAC1 expression were significantly greater, and ADP-induced aggregation and agonist-induced increase in CD62P and PAC1 were significantly lower in PPCA compared to APC and PPC on day 4 of storage. ADP and TRAP-induced CD62P and PAC1 activatability fell significantly during storage between day 1 and day 4 in APC and PPCA, but not in PPC. In conclusion, different platelet function tests capture different aspects of platelet function and do not correlate and agree sufficiently to be used interchangeably.