Mathematical modeling of flow-mediated thrombin generation after treatment of postoperative bleeding in cardiac surgery with cardiopulmonary bypass.

BACKGROUND

Postcardiopulmonary bypass (CPB) bleeding is partly driven by impaired thrombin generation. Recent trials suggest differing clinical effects of fresh frozen plasma (FFP) and prothrombin complex concentrate (PCC), but conventional and static assays do not resolve their hemostatic mechanisms.

OBJECTIVES

This study aimed to compare FFP and PCC effects on simulated post-CPB thrombin generation and identify mechanisms underlying varied treatment response.

METHODS

We used a mathematical model of flow-mediated coagulation to simulate thrombin generation in virtual patients with physiologic variation in coagulation proteins and platelets. A virtual population of 4891 virtual patients was selected for preserved pre-CPB but delayed post-CPB thrombin generation. FFP and PCC at therapeutic doses were simulated in the virtual population.

RESULTS

Both FFP and PCC improved simulated thrombin generation dose-dependently. Prothrombin complex concentrate produced faster and greater thrombin generation overall, with responses to low PCC comparable with those achieved with high FFP. Poor response to high FFP was associated with lower pre-CPB platelet and factor (F)IX levels, whereas poor response to high PCC was associated with lower FVIII and higher tissue factor pathway inhibitor. Most FFP poor responders were rescued by PCC; PCC poor responders were only partially corrected by FFP.

CONCLUSION

A mathematical model captured the clinically observed differences between FFP and PCC efficacy after CPB and further showed that they were dependent on innate coagulation protein and platelet levels. Poor response to FFP was associated with low platelet and FIX levels, whereas poor response to PCC was associated with low FVIII and elevated tissue factor pathway inhibitor.