Assessment of hemostatic changes after initiation of continuous venovenous hemodialysis.

Background: Continuous renal replacement therapy (CRRT) is a standard therapy in critically ill patients suffering from acute kidney injury (AKI). Extracorporeal circulation and exposure to foreign surfaces during CRRT may induce disturbances in hemostasis, particularly in platelet function. The present study described the hemostatic changes associated with CRRT and aimed to identify the independent predictors of premature clotting of the circuit.

Methods: In a prospective cohort mono-center study, patients were assessed for eligibility if they were i) diagnosed with AKI and ii) assigned to receive CRRT for the first time. Patients were included in the study if their platelet count was greater than 100/nL prior to inclusion in the study. After initiation of CRRT, aggregometric [Multiplate, Roche, Grenzach, Germany: Arachidonic acid (ASPItest)-, ADP (ADPtest)- and Thrombin (TRAPtest)-induced platelet aggregation] and viscoelastic (ROTEM; TEM International, Munich, Germany) analyses were performed immediately before (Baseline, T1) and 6 hours (T2), 12 hours (T3), 24 hours (T4), and 48 hours (T5) after initiation of CRRT. Conventional laboratory coagulation analyses were routinely performed twice a day. Arachidonic acid- and ADP-induced platelet aggregations were defined as primary endpoints.

Results: A total of 127 patients were screened for eligibility, and 50 patients were enrolled in this study. Aggregometric analyses showed that arachidonic acid-induced platelet aggregation was significantly reduced at T2 [532 (210/1105) median (25th/75th percentile) AU*min] compared to the Baseline at T1 [780 (297/1156), p = 0.007] and remained unchanged from T2 onward. Platelet aggregation in the ADPtest and TRAPtest remained unchanged during the study period. Viscoelastic and conventional coagulation analyses indicated a progredient increase of clot firmness. In total, 76 filter sets (an average of 1.5 per patient) were used, and 26 filter sets occluded prematurely after an average treatment time of 17 ± 12 hours. No predictors for premature clotting of the circuit were identified.

Conclusions: The results of the present study indicate that CRRT may lead to impaired primary hemostasis as shown by a decrease in ex vivo arachidonic acid-induced platelet aggregation. Moreover, viscoelastic measure indicate a fibrinogen-associated trend of increasing clot firmness during the study period. Further studies are needed to analyze whether these findings are of hemostatic relevance.