Agonist-induced platelet reactivity correlates with bleeding in haemato-oncological patients.

Prophylactic platelet transfusions are administered to prevent bleeding in haemato-oncological patients. However, bleeding still occurs, despite these transfusions. This practice is costly and not without risk. Better predictors of bleeding are needed, and flow cytometric evaluation of platelet function might aid the clinician in identifying patients at risk of bleeding. This evaluation can be performed within the hour and is not hampered by low platelet count. Our objective was to assess a possible correlation between bleeding and platelet function in thrombocytopenic haemato-oncological patients. Inclusion was possible for admitted haemato-oncology patients aged 18 years and above. Furthermore, an expected need for platelet transfusions was necessary. Bleeding was graded according to the WHO bleeding scale. Platelet reactivity to stimulation by either adenosine diphosphate (ADP), cross-linked collagen-related peptide (CRP-xL), PAR1- or PAR4-activating peptide (AP) was measured using flow cytometry. A total of 114 evaluations were available from 21 consecutive patients. Platelet reactivity in response to stimulation by all four studied agonists was inversely correlated with significant bleeding. Odds ratios (OR) for bleeding were 0·28 for every unit increase in median fluorescence intensity (MFI) [95% confidence interval (CI) 0·11-0·73] for ADP; 0·59 [0·40-0·87] for CRP-xL; 0·59 [0·37-0·94] for PAR1-AP; and 0·43 [0·23-0·79] for PAR4-AP. The platelet count was not correlated with bleeding (OR 0·99 [0·96-1·02]). Agonist-induced platelet reactivity was significantly correlated to bleeding. Platelet function testing could provide a basis for a personalized transfusion regimen, in which platelet transfusions are limited to those at risk of bleeding.