This site is intended for healthcare professionals
Oral Anticoagulation Reversal Learning Zone

A retrospective propensity score-matched early thromboembolic event analysis of prothrombin complex concentrate vs. fresh frozen plasma for warfarin reversal prior to emergency neurosurgical procedure

Read time: 3 mins
Last updated:12th Mar 2020
Published:12th Feb 2020
This retrospective study of 63 patients taking warfarin who presented at a level I trauma centre and required emergency neurological procedures attempts to determine the difference in thrombotic complication rate following anticoagulation reversal with prothrombin complex concentrate (PCC) or fresh frozen plasma (FFP).

The oral vitamin K antagonist, warfarin, is often used to reduce the risk of stroke and systemic embolism in patients with a history of atrial fibrillation, and in the prevention of venous thromboembolism. However, the effects of this common treatment can be problematic as patients experiencing a spontaneous or traumatic intracranial haemorrhage have an increased risk of clinical and radiographic deterioration.

In order to reduce the associated risk of increased bleeding and haematoma expansion, patients are typically given one of two anticoagulation reversal treatments to allow the necessary neurosurgical interventions. The traditional treatment is the administration of vitamin K and the infusion of thawed plasma or FFP. When administered with vitamin K, FFP contains all the soluble coagulation factors and proteins to promote vitamin K-dependent clotting. However, the treatment has limitations with a gradual International Normalised Ratio (INR) reversal resulting in possible delays when urgent neurosurgical intervention is needed. This is in addition to an associated risk of transfusion-related acute lung injury, circulatory overload, and allergic or anaphylactic reactions.

An alternative to FFP is the administration of PCC. There are various combinations of PCC containing different amounts of clotting factors with or without the addition of proteins C and S. Some studies have shown in neurosurgical populations that PCC is both effective and superior to FFP for achieving INR reversal and time to reversal. Despite this, PCC and FFP had not been directly compared previously in patients requiring emergency neurosurgical procedures where urgent reversal is critical.

This retrospective cohort analysis compared acute thrombotic complications following administration of FFP or PCC in anticoagulated patients needing urgent reversal to undergo emergency neurosurgery to treat intracranial haemorrhage. Between the years 2007–2016, 63 patients from a level I trauma centre were identified as being on warfarin therapy, presenting with an intracranial haemorrhage and subsequently underwent anticoagulation reversal via PCC or FFP treatment prior to a neurosurgical procedure. The primary endpoint in this study was the incidence of thrombotic complications within 72 hours of anticoagulation reversal with thrombotic complications defined as deep vein thrombosis, pulmonary embolism, ischaemic stroke, or myocardial infarction. Secondary endpoints were INR reversal, mortality rate during hospitalisation, and rate of pulmonary complications within 1 week of treatment (pneumonia, pneumothorax, respiratory distress requiring mechanical ventilation, and pulmonary edema and effusion).

A multivariate propensity score matching (PSM) analysis was used to correct for any bias between the two groups (PCC n=28, FFP n=35), closely matching patients in each group based on age, gender, INR pre- and post-treatment, trauma, altered mental status, and existence of heart failure. Although patients were well matched for the majority of factors that are known to predispose neurosurgical patients to thrombosis, there was a significant difference in the rate of trauma between the two groups (PCC n=6/26, FFP n=20/35; p=0.013) and this should be considered when looking at the results.

Thrombotic complications

Overall thrombotic complications within 72 hours were seen in 3.57% (n=1/28) of the PCC group and 0% (n=0/35) in the FFP group – there was no significant difference between the two groups. The results found in this study associated with PCC reversal are consistent with the few reports investigating warfarin therapy and neurosurgical procedures; in previous studies, thrombotic complication rates ranged from 0–13% (Agarwal et al., 2017). Similar results are seen in the few neurological studies comparing PCC and FFP (table 1). 

Table 1: Rate of thrombotic complication following use of PCC and FFP for anticoagulation reversal (Agarwal et al., 2017).


Thrombotic complications following anticoagulation reversal appear uncommon and similar between the two treatment methods, regardless of whether the neurosurgical case is operative or nonoperative.  Even in general surgery similar rates have been seen (PCC 7% n=6/88, FFP 8% n=7/88) (Goldstein et al., 2015).

Pulmonary complications

The most notable of the secondary endpoints was the rate of pulmonary complications. Patients receiving PCC had a significantly higher rate of pulmonary complications within 1 week of treatment (39.3 %, n=11) in comparison to the FFP group (11.4 %, n=4) (figure 1). The authors found this result surprising given previous studies suggested that a benefit of not using FFP is the reduced incidence of volume overload and secondary pleural effusions or pulmonary edema, something not seen in the FFP group in this study.


The remaining secondary endpoints, INR reversal and mortality rates, were seen to be similar between both treatment groups with no significant difference found.

In summary, Agarwal et al., have demonstrated that thrombotic complication risk is uncommon but not rare in patients requiring anticoagulation reversal with FFP or PCC for emergency neurosurgery. A significantly higher rate of pulmonary complications was observed in the PCC group, a surprising result that needs further investigation.

Given the nature of the retrospective study analysing data from a single level 1 trauma center, the study was unable to account for all possible confounding factors such as intracranial haematoma volume. Despite the PSM analysis matching the two groups a significant difference in the rate of trauma was observed between the groups and this may have impacted the results. For these reasons, further research is required in a larger neurosurgical population.