BACKGROUND: There is a paucity of data on the underlying procoagulant–anticoagulant balance during extracorporeal membrane oxygenation (ECMO). We hypothesized that adult ECMO patients would have an imbalance between procoagulant and anticoagulant factors, leading to an abnormal underlying thrombin generation (TG) pattern. METHODS: Twenty adult venoarterial (VA) ECMO patients had procoagulant and anticoagulant factor levels measured temporally on ECMO day 1 or 2, day 3, and day 5. In heparin-neutralized plasma, underlying TG patterns, and sensitivity to activated protein C were assessed using calibrated automated thrombogram. TG parameters including lag time, peak TG, and endogenous thrombin potential (ETP) were compared against 5 normal plasma controls (3 males and 2 females) obtained from a commercial supplier. Thrombomodulin (TM) was added to some samples to evaluate for activated protein C resistance. RESULTS: Procoagulant factors (factor [F] II, FV, and FX) were mostly in normal reference ranges and gradually increased during the first 5 ECMO days (P = .022, <.001, <.001). FVIII levels were elevated at all time points and did not change (P = .766). In contrast, FXI was in the low-normal range but did not increase during ECMO (P = .093). Antithrombin (AT) and protein C levels were below normal but increased during the first 5 ECMO days (P = .002 and P = .014). Heparinase-treated samples showed prolonged lag time, increased peak TG, and increased ETP compared to controls; mean difference in lag time on ECMO day 1 or 2 = 6.0 minutes (99% confidence interval [CI], 2.8–9.2), peak TG = 193.4 (99% CI, 122.5–264.3), and ETP = 1170.4 (99% CI, 723.2–1617.6). After in vitro TM treatment, differences in TG parameters were accentuated and ECMO samples appeared insensitive to TM treatment; mean difference in lag time on ECMO day 1 or 2 = 9.3 minutes (99% CI, 6.2–12.4), peak TG = 233.0 (99% CI, 140.9–325.1), and ETP = 1322.5 (99% CI, 764.8–1880.2). Similar differences in TG parameters were observed on ECMO days 3 and 5. CONCLUSIONS: Contact activation occurs during ECMO, but procoagulant factor levels are generally preserved. Although heparin-neutralized TG is delayed, peak TG and ETP are supranormal in the setting of high FVIII and low AT and protein C levels. Resistance to TM is also apparent. These changes demonstrate a possible mechanism for hypercoagulability during adult VA ECMO.

Reduced TG in LT can be more effectively restored by using PCC rather than plasma. The required doses of PCC for LT patients seem to be lower than warfarin reversal due to slow thrombin inhibition.

No abstract

Background: Gestational changes in coagulation factor concentrations include elevations in fibrinogen, Factor VIII, and von Willebrand factor (vWF). We hypothesised that blood samples from term pregnant (TP) subjects are less prone to coagulation disturbances from haemodilution compared with those from non-pregnant (NP) females. Methods: Blood samples were collected from 15 NP and 15 TP subjects. In vitro haemodilution with normal saline was assessed by modified Clauss fibrinogen assay, factor activity, flow-chamber assay, and thromboelastometry. The impact of human fibrinogen concentrate (hFC), cryoprecipitate, and vWF/Factor VIII (FVIII) concentrate replacement in diluted TP and NP blood was compared. Thrombin generation and activated protein C sensitivity were assessed. Results: TP blood contained twice the concentrations of fibrinogen, FVIII, and vWF relative to NP blood (P<0.0001). Platelet thrombus formation (PTF) under flow was reduced by 99.2% and 69.2% in diluted NP and TP blood, respectively. Platelet thrombus formation was partially restored by adding vWF/FVIII, but not hFC or cryoprecipitate. Fibrin clot firmness approached the threshold of 10 mm in diluted NP blood, and clot firmness was effectively restored by hFC, but not by vWF/FVIII. In the presence of thrombomodulin, peak thrombin generation was decreased by 86.7% in NP plasma, but by 31.8% in TP plasma (P<0.0001 vs NP plasma), indicating reduced activated protein C sensitivity in TP plasma. Both elevated FVIII and haemodilution contributed to activated protein C insensitivity. Conclusions: Our in vitro model showed relative resistance of TP blood to dilutional coagulation changes with respect to platelet adhesion, fibrin polymerisation, and thrombin generation. Careful therapeutic monitoring for different prohaemostatic agents in pregnant women is warranted.

FIX derived from PCC strongly enhances tissue factor-triggered thrombin generation in the presence of elevated FVIII activity. Haemodilution further enhances procoagulant effects of FIX and FVIII by slowing down inhibition of procoagulant enzymes. Dosing of PCC per prothrombin time may underestimate PCC's procoagulant potential because it does not account for intrinsic tenase or antithrombin activity.