Fresh frozen plasma versus prothrombin complex concentrate in patients with intracranial haemorrhage related to vitamin K antagonists (INCH): a randomised trial

ICH is a significant cause of mortality in patients on VKAs, causing up to 36% of bleeding-related deaths, with an annual incidence of between 0.33% and 1.9% in that group. VKAs generate their therapeutic effect by depleting clotting factors; repletion of those factors can normalise coagulation, preventing haematoma expansion. Previous trials have demonstrated the superiority of PCC compared with FFP for VKA reversal, although a retrospective study from 2015 comparing the treatments in ICH showed no mortality difference. In the absence of evidence, treatment guidelines have suggested either could be used in ICH.

This trial involved any adult patient who presented with a CT diagnosed VKA-related ICH and an INR greater than 2.0 within 12 hours of symptom onset. There were numerous exclusion criteria, including concurrent ischaemic disease, trauma, heart failure or premorbid disability. Each patient was randomly assigned either 20 mL/kg FFP or 30 IU/kg PCC (Octaplex). The treatment protocol dictated that the infusion should start within one hour of diagnosis, and be administered as rapidly as the clinical situation permitted. INR was measured prior to the infusion and at 3 hours post-treatment, while serial CT scans were analysed by the same radiologist in all cases to estimate haematoma volume.

The primary outcome was INR at 3 hours. Any patients who had an INR ≥1.2 at this point were administered PCC regardless of randomisation. Secondary outcomes included mortality, absolute change in haematoma volume and relative haematoma growth. Adverse events were also monitored and recorded. As the interim safety analysis resulted in premature termination of the study, only 54 patients were recruited – 4 of whom were withdrawn prior to infusion.

• Results showed that 18 of 27 patients treated with PCC, and 2 of 23 patients treated with FFP had an INR of equal to, or lower than 1.2 at 3 hours (adjusted odds ratio [OR] 30.6; 95% confidence interval [CI] 4.7–197.9; p=0.003). 
• Adjusted difference of haematoma expansion at 3 hours was greater with FFP (adjusted difference 16.9 mL; 95% CI 2.5–31.3; p=0.023).
• Haematoma expansion of at least 15% at 3 hours occurred in 16/22 FFP patients (some data were not available for all patients) and 15/26 PCC patients, this was not statistically significant (OR 2.0; 95% CI 0.6–7.3; p=0.28).
• Adjusted difference of haematoma expansion at 24 hours was greater with FFP (adjusted difference 16.4 ml; 95% CI 2.9–29.9; p=0.018).
• At day 90, 8/23 patients in the FFP group, and 5/27 in the PCC group had died, although this was not statistically significant (p=0.14).

Of the deaths, 5 of the FFP group were deemed to be due to haematoma expansion, while none from the PCC group were. There were 6 serious adverse events attributed to FFP, and 2 to PCC. Thromboembolic events were also monitored; there was one ischaemic stroke in each group, and one pulmonary embolism in the PCC group.

The findings of this study suggest PCC is better than FFP at normalising INR, and leads to less haematoma expansion. The authors also emphasise how important the rapid reversal of anticoagulation is in this condition. There were some limitations: there was a relatively small sample size, the trial was terminated prematurely, and the assessors were not blinded. The results are clear enough to suggest a shift in departmental practice towards using PCC for VKA reversal in this group of patients – pending further evidence.