Platelet transfusion in severely ill patients

Severely ill patients, such as those with hematologic malignancy or the critically ill, often suffer from thrombocytopenia and therefore require platelet transfusions either prophylactically or therapeutically. Platelet transfusions are associated with scarcity, high costs and a risk of transfusion-related adverse events. In this thesis we studied the sense and side-effects of platelet transfusion in severely ill patients.
In the first part of this thesis we determined the value of prophylactic platelet transfusion before central venous catheter (CVC) placement. We compared different methods of post-procedural bleeding assessment and found that retrospective assessment leads to underreporting of bleeding complications. We also found that differences in definitions and methodology likely influence reported bleeding incidences to a greater extent than differences in patient characteristics. We then compared, in a single-center retrospective cohort, the post-CVC placement bleeding risk of a combined severe thrombocytopenia and coagulopathy versus severe thrombocytopenia alone, and found no difference. Lastly we found, in a multicenter randomized controlled trial (PACER), that prophylactic platelet transfusion before CVC placement in severely ill patients admitted to the hematology ward or intensive care unit, prevents bleeding complications but also comes with higher overall costs.
In the second part of this thesis we investigated the relationship between the platelet storage lesion and transfusion efficacy and transfusion-related acute lung injury. During collection and storage of platelet concentrate, platelets undergo biochemical changes collectively referred to as the platelet storage lesion. In a secondary analysis of the PACER trial we found that bone marrow depression, sepsis, disseminated intravascular coagulation and platelet concentrate irradiation, but not storage duration, were negatively with the corrected platelet count increment. In a small randomized trial comparing autologous transfusion of fresh (2 days) and aged (7 days) platelets in healthy male volunteers pre-medicated with lipopolysaccharide to mimic a mild sepsis (DIVA), we found that storage duration did not influence hemostatic measurements, nor the peak concentration of transfused platelets, but the concentration of transfused aged platelets did show a steeper decline reaching peak levels. Also in the DIVA trial, we found no difference in pulmonary inflammation or edema, as determined using bronchoalveolar lavage. Finally, we studied the relationship between platelet concentrate acidity and pulmonary microvascular endothelial inflammation and barrier function in an in vitro model, and found that lower pH is associated with more endothelial inflammation, but not with differences in barrier function.