Editorial 

March 29, 2025

Is It Time to Replace Plasma With Prothrombin Complex Concentrate in Cardiac Surgery?

Ryan Wang, Elliott Bennett-Guerrero

JAMA. Published online March 29, 2025. doi:10.1001/jama.2025.3644

Bleeding after cardiac surgery with cardiopulmonary bypass is associated with adverse clinical outcomes and increased resource utilization. Broadly speaking, this bleeding can be due to inadequate surgical hemostasis or microvascular bleeding caused by bypass-induced platelet dysfunction and/or depletion of coagulation factors. The most common method of restoring coagulation factors in this setting is through the transfusion of thawed frozen plasma, which occurs in approximately 25% of cardiac surgeries.¹ Plasma is generally regarded as a safe treatment, with very low rates of transfusion-related acute lung injury (<0.01%) and severe allergic reactions (<0.01%).² The most common adverse event in this setting is transfusion-associated circulatory overload (<0.1%).² The rapid administration of the amount of plasma needed to treat coagulopathy (750-1000 mL of thawed frozen plasma) can be particularly detrimental in some patients, including those with significant right ventricular dysfunction or those who are euvolemic.

Prothrombin complex concentrates (PCCs) are a class of blood products that attempt to address this limitation of plasma by providing concentrated doses of factors II, VII, IX, and X in a small volume, typically 80 mL for a standard dose (2000 IU expressed as units of factor IX). Commercially available 4-factor PCCs (eg, Kcentra, CSL Behring; and Octaplex, Octapharma AG) contain the 4 coagulation factors plus small amounts of proteins C and S and heparin, whereas 3-factor products (eg, Bebulin, Baxter) omit factor VII. PCCs are purified from pooled human plasma, with solvent or detergent viral inactivation. Similar to thawed frozen plasma, PCCs carry a low risk of adverse reactions (<0.03%).³ The most significant clinical benefit of PCCs is their ability to more rapidly restore factor levels because their small dosing volume allows for more rapid administration. However, the main concerns regarding widespread use of PCCs are their cost and potential to cause thromboembolic complications.³ In addition, at least one commercially available PCC states that it “should not be used in patients with recent myocardial infarction, with a high risk of thrombosis or with angina pectoris…” Such an admonition could discourage its use in cardiac surgery such as coronary artery bypass graft surgery.⁴ Finally, PCCs are only approved in some countries for rapid reversal of vitamin K antagonists, making their use in most surgeries off-label.

In this issue of JAMA, Karkouti and colleagues report the findings of the Factor Replacement in Surgery II (FARES-II) trial that compared the effectiveness and safety of PCC with frozen plasma in the management of postcardiopulmonary bypass bleeding due to coagulation factor deficiency.⁵ The unblinded study, which took place across 10 Canadian and 2 US hospitals, randomized 420 patients undergoing cardiac surgery to receive PCC or thawed frozen plasma if they were bleeding and had a documented point-of-care international normalized ratio (INR) of 1.5 or higher. The primary outcome was hemostatic response, a composite end point defined as no need for further hemostatic interventions (ie, surgical reexploration for bleeding, transfusion of coagulation factor–containing products or platelets) in the 24 hours after drug administration. Consistent with its more concentrated formulation, completion of the first dose of PCC occurred 19 minutes faster than thawed frozen plasma (median, 7 minutes vs 26 minutes). The PCC group reached the primary end point of hemostatic response more frequently than the frozen plasma group (77.9% vs 60.4%, P < .001) and also achieved reductions in related secondary outcomes, such as chest tube output and overall transfusion requirements. The differences in blood products administered were modest, and no differences were observed in mortality or in intensive care unit (ICU) or hospital length of stay. Acute kidney injury was less common in patients treated with PCC patients, although this was an exploratory end point and was not associated with improved clinical outcomes.

The positive findings of the FARES-II trial contrast those of 2 randomized trials evaluating the efficacy of PCC compared with thawed frozen plasma in cardiac surgery, which did not find a difference in 24-hour transfusion requirements or chest tube output.^6,7 The superior hemostatic effect observed in the FARES-II trial is potentially explained by the larger sample size, and most likely, the use of higher doses of PCC (24 IU/kg) vs dosing used in previous trials (15 IU/kg⁶ and 13 IU/kg⁷).

The FARES-II trial has numerous strengths and high internal validity. The 2 study arms were generally well balanced with regard to potential confounders. Although not powered for the individual components of the primary end point, it is reassuring that each component of the composite outcome showed benefit. Multiple secondary measures of hemostatic effectiveness were reported (eg, red blood cell transfusions, total blood products given, chest tube output) and were consistent with the primary end point showing benefit in patients treated with PCC. Another strength of the study was the consistent finding of benefit for PCC-treatment across most of the study sites. Additionally, the authors prespecified the statistical analysis to first test whether PCC was noninferior, then whether it was superior, compared with frozen plasma, prioritizing safety in the evaluation of PCC treatment against current practice.

Some readers may find the 20% dropout rate after randomization concerning. However, the majority of these were patients who did not end up needing factor replacement or declined consent after surgery. Delayed consent was approved by Canadian institutional review boards because PCC use in the study was consistent with approved indications and practice standards in Canada. Although consent obtained this way improved the efficiency of enrollment of bleeding patients into the trial, it prevented analysis of a true intent-to-treat population, which could introduce bias.

A potential strength, but also possible limitation, of the study is the requirement for point-of-care INR testing, which may not be widely available. Transfusion protocols requiring specialized testing and targeting treatment for the various causes of coagulopathy have been shown to reduce blood loss and transfusion requirements in cardiac surgery.^8,9 Therefore, testing the safety and efficacy of PCC in the FARES-II trial using a protocolized approach likely ensured that only patients most likely to benefit from PCC were randomized and included in the analysis. However, requiring point-of-care INR testing to determine the need for factor replacement may limit the generalizability of this study. An international survey of cardiovascular anesthesiologists, with 83% of the 536 respondents being from North America, reported that a majority of cardiac surgery centers had some type of point-of-care coagulation testing available in the operating room; however, they did not specify if this included point-of-care INR as required by the FARES-II trial.¹⁰ Additionally, only 19% to 35% of the 536 respondents reported using laboratory-driven transfusion protocols highlighting the potential challenges of implementing the required algorithm used in the FARES-II trial; lower-resource institutions may encounter challenges to using PCCs as was protocolized in this study. Of note, the FARES-II trial allowed PCC or thawed frozen plasma to be administered based on clinical judgment alone in cases of severe bleeding; however, the authors did not report what percentage of patients were treated in this manner.

Clinicians seeking to use PCCs may be limited by institutional constraints, potentially related to high acquisition costs for PCC relative to plasma. The direct costs for these products are approximately $2500 for 2000 IU of PCC (eg, $1.32 per IU),¹¹ roughly 10 times as much as the cost of frozen plasma ($240 for 1000 mL).¹² However, a rigorous cost analysis would need to consider potential savings associated with the use of PCC, such as decreased need for additional blood transfusions, as shown in the FARES-II trial.

One of the concerns with using PCCs is the inherent thromboembolic risk associated with the administration of coagulation factors. Thromboembolic events appear to be very low (~0.01%),³ and in many cases patients treated with PCC have other risk factors for thrombosis. Clinicians should note that the FARES-II trial excluded high-risk patients with a recent history of thromboembolism. Importantly, the standardized use of laboratory-driven transfusion thresholds in this study may have reduced the thromboembolic risk of administering PCC to patients without factor deficiency. In another study (PROCOAG),¹³ PCC was given to bleeding trauma patients without first confirming factor deficiency. Prothrombin times were significantly elevated in only 25% of their participants. Patients randomized to PCC (vs placebo) demonstrated significantly higher rates of thromboembolic events, suggesting that administering PCC to patients without factor deficiency might increase thrombotic risk.

Is it time to replace plasma with prothrombin complex concentrate in cardiac surgery? The FARES-II trial provides substantial evidence that PCC, when used with a structured algorithm and point-of-care INR testing, is more effective than thawed frozen plasma at treating bleeding after cardiac surgery due to factor deficiency. Administration of PCC may be beneficial in coagulopathic patients who cannot receive a large volume of thawed frozen plasma or for whom rapid reversal is important. However, the differences in blood products administered were modest in the FARES-II trial, and no differences were observed in mortality or in ICU or hospital length-of-stay, which may argue against a major clinical benefit for most patients.