Introduction: 1% of patients of warfarin experience life-threatening bleeding per year and require emergency reversal of their anticoagulation. This is usually done by using a combination of vitamin K and prothrombin complex concentrate (PCC), FVIIa or FFP. Algorithms for warfarin reversal commonly use INR thresholds to guide the PCC dose. However, the INR system has only been validated for values up to 4.5.

Aim: To compare the ability of PCC, FVIIa and FFP to reverse the warfarin effects on the INR and thrombin generation (TG) as measured by the calibrated automated thrombography (CAT). We hypothesised that INRs greater than 4.5 would not correlate with TG and in which case one concentration of reversal agents would be suitable to reverse warfarin in all these patients.

Methods and Results: A pool of plasma from patients on warfarin (final INR 4.8) was spiked with Beriplex P/N (25Units/kg), FVIIa (~140μg/kg) and FFP (~20ml/kg). The INR normalised with Beriplex and FVIIa and was reduced to 1.4 with FFP. However, only Beriplex completely normalised TG parameters. In view of this Beriplex was used for subsequent reversal experiments. We analysed 48 plasma samples with INR >4.5 (median 5.5; range 4.6–8.0) and measured TG at 5pM tissue factor. We found a significant moderate correlation between the INR and the endogenous thrombin potential (ETP) which is the maximum amount of thrombin generated (r=−0.59; P= <0.0001). The mean ETP was 279.4nM.min (95%CI 240 – 318.9; SD 135.9). 2 plasma samples with INR >8.0 were spiked with 5 concentrations of Beriplex P/N (7.5, 15, 25, 35 & 50Units/kg) in order to decide which is the ideal concentration of this PCC to reverse warfarin effects on the CAT. Both samples showed that a concentration between 25 and 35Units/kg could normalise TG without producing hypercoagulability. Hence we chose 30Units/kg to spike another 10 samples with INR >4.5 (mean 6.4; range 4.7–8.0). The INR, TG and FII, VII, IX and X were tested after the addition of Beriplex P/N. The INR normalised in all samples (median 1.2; range 1.08–1.3) as did the ETP (median 1942nM.min; range 1546–2753nM.min), FII (mean 114U/dl; range 76–126U/dl), FIX (mean 127U/dl; range 106–154U/dl) and FX (mean 131U/dl; range 120–140U/dl). FVII increased to a mean 52U/dl (range 20–68U/dl). A lower concentration of Beriplex (25U/ kg) did not satisfactorily normalise the INR and TG in all samples tested.

Conclusion: PCCs fully reverse the effects of warfarin when compared to standard doses of FFP and FVIIa. INRs >4.5 still correlate with TG but the amount of thrombin generated was comparatively low making the use of one dose of PCC (30U/kg) suitable for reversing the effect of all INRs >4.5. This not only simplifies the reversal algorithms but reduces cost since doses up to 50U/kg are sometimes advocated for INRs >6.0.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author