Background: Hemorrhagic and thromboembolic complications are common during treatment with extracorporeal membrane oxygenation (ECMO), resulting in considerable morbidity and mortality. This emphasizes the clinical relevance of understanding hemostatic changes occurring during ECMO treatment. As platelets are key players in hemostasis, detailed knowledge on how ECMO treatment affects platelet function is of great importance. We therefore aimed to systematically summarize and discuss existing knowledge on platelet function during ECMO treatment in adult patients.Methods: Systematic review complying with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Objectives and methods were specified in a PROSPERO protocol (ID no CRD42018084059). The MEDLINE/PubMed, EMBASE, and Web of Science databases were systematically searched on September 10, 2018. A standardized quality assessment tool was used to assess the risk of bias in included studies. Primary outcome was platelet function during ECMO treatment, measured as platelet adhesion, activation or aggregation. Secondary outcomes were thrombosis, bleeding, and mortality during ECMO treatment.Results: A total of 591 studies were identified, of which seven were eligible for inclusion in the qualitative synthesis. Of these, one study investigated expression of platelet adhesion receptors and found them to be reduced during ECMO treatment; two studies reported a decrease in platelet activation markers during ECMO treatment; and five studies demonstrated reduced platelet aggregation during ECMO treatment. Three studies reported on thrombosis, mortality and/or bleeding during ECMO treatment; no thromboembolic events were reported; all three studies reported frequent bleeding episodes defined on basis of transfusion requirements. An in-hospital mortality of 35–40% and a 30-day mortality of roughly 30% were reported in three different studies.Conclusions: The present systematic review reveals a substantial knowledge gap regarding platelet function during ECMO treatment in adult patients and underscores the demand for more and well-designed studies on this topic. There is suggested evidence of reduced platelet adhesion, decreased platelet activation, and reduced platelet aggregation in adult patients during ECMO treatment. Importantly, platelet aggregation results need to be interpreted in the light of low platelet counts. The associations of platelet function and bleeding and/or thromboembolic complications during ECMO treatment remain to be fully elucidated.

Objective: Hemorrhagic and thromboembolic complications are common during support with extracorporeal membrane oxygenation (ECMO). As platelets play a pivotal role in hemostasis, we aimed to clarify how ECMO support affects platelet function. Methods: We included 33 adult patients undergoing ECMO support at a tertiary ECMO referral center at Aarhus University Hospital, Denmark. Blood samples were collected on the first morning following ECMO initiation, and subsequently every morning until the 7th (±1) day. Platelet aggregation was evaluated by whole blood impedance aggregometry (Multiplate ® Analyzer) using adenosine diphosphate (ADPtest), arachidonic acid (ASPItest), and thrombin-receptor-agonist-peptide-6 (TRAPtest) as agonists. A new model was applied, taking platelet count into consideration in interpretation of impedance aggregometry analyses. On the 1st and 3rd day, platelet activation was assessed by flow cytometry (Navios) using collagen-related peptide, ADP, TRAP, and arachidonic acid as agonists. Results: Blood samples from all 33 patients were analyzed on day 1 of ECMO support; 24 patients were still receiving ECMO and analyzed on day 3; 12 patients were analyzed on day 7 (±1). After ECMO initiation, platelet counts decreased significantly ( p < 0.002) and remained low during ECMO support. ECMO patients demonstrated significantly reduced platelet aggregation on day 1 compared with healthy controls (all p < 0.001). However, when taking platelet count into consideration, platelet aggregation relative to platelet count did not differ from healthy controls. Flow cytometry analyses demonstrated impaired platelet activation in ECMO patients on day 1 compared with healthy controls (all p < 0.03). No substantial difference was found in platelet activation from day 1 to day 3 on ECMO support. Conclusions: Employing impedance aggregometry and flow cytometry, we found both impaired platelet aggregation and decreased platelet activation on day 1 of ECMO support compared with healthy controls. However, platelet aggregation was not impaired, when interpreted relative to the low platelet counts. Furthermore, levels of bound fibrinogen, on the surface of activated platelets in ECMO patients, were higher than in healthy controls. Together, these findings suggestively oppose that platelets are universally impaired during ECMO support. No marked difference in activation from day 1 to day 3 was seen during ECMO support.