There is a clear consensus that treatment with FFP is inappropriate in cases of severe fibrinogen deficiency as it contains insufficient concentrations of fibrinogen (Chowdhury et al, 2004; Danes et al, 2008). Although scientific evidence regarding the clinical efficacy of fibrinogen concentrate is limited, this is even more the case for cryoprecipitate (Danes et al, 2008), with no published studies specifically addressing the efficacy of cryoprecipitate in the management of perioperative bleeding.
There is evidence for the efficacy and safety of fibrinogen concentrate in congenital fibrinogen deficiency (Rodriguez et al, 1988; Kreuz et al, 2005; Aygoren-Pursun et al, 2007), and data on its use in patients with acquired deficiencies in a variety of surgical settings is beginning to emerge (Heindl et al, 2005; Danes et al, 2008; Fenger-Eriksen et al, 2008, 2009b; Haas et al, 2008; Weinkove & Rangarajan, 2008; Rahe-Meyer et al, 2009a,b). In an observational study of 69 patients suffering from various forms of acquired severe hypofibrinogenaemia, most (62%) had consumptive hypofibrinogenaemia (Danes et al, 2008). After a median dose of 4 g of fibrinogen concentrate, a mean absolute increase of 1·09 g/l of plasma fibrinogen was measured. Coagulation parameters, prothrombin and activated partial thromboplastin time, were significantly improved (P < 0·001) at 24 and 72 h after fibrinogen concentrate administration (Danes et al, 2008). Furthermore, there was an association between plasma fibrinogen concentrations after treatment and 7-d patient survival. In another retrospective study of 43 patients, a similar increase in fibrinogen levels (1·01 g/l) was achieved with half the average dose of fibrinogen (Fenger-Eriksen et al, 2008). This may reflect differences between the two study populations, including underlying clinical conditions and proportion of paediatric patients.
There are no published studies comparing the efficacy of fibrinogen concentrate with cryoprecipitate. The specification for cryoprecipitate in the UK requires that 75% of units contain at least 140 mg of fibrinogen, in other words, there can be wide variation in fibrinogen content between units (O’Shaughnessy et al, 2004). This confounds dose calculations in comparative studies.
In the past, nearly all patients with haemophilia who received factor VIII concentrates were exposed to transfusion-transmitted viruses, due to factor VIII being unable to withstand heating at 60°C during the standard pasteurization process. The same is true for fibrinogen. In 1985, CSL Behring (then called Behringwerke), invented a pasteurization process using stabilizers to protect fragile proteins (Heimburger, 2002). A sterile filtration step also removed pathogens (Fig 2). As a result, the fibrinogen concentrate available today has a superior viral safety over cryoprecipitate, despite being plasma-derived. The overall reduction of titres during the production of fibrinogen concentrate is shown for some representative viruses in Table I (Groner, 2008).
Table I. Mean reduction of virus titres during the production of Haemocomplettan® P/Riastap (CSL Behring, Marburg, Germany) (Groner, 2008).
|Virus||Mean reduction of titre (log10)|
|Human immunodeficiency virus||≥9·6|
|Bovine viral diarrhoea virus*||≥11·2|
|Herpes simplex virus 1||≥9·1|
|Hepatitis A virus||≥7·6|
The introduction of viral testing of plasma has greatly reduced the risk of viral transmission with cryoprecipitate. It has been estimated that in the UK, where FFP is not manufactured from first-time or lapsed donors, the residual risks from a single unit of FFP are 1 in 10 million for HIV, 1 in 50 million for hepatitis C virus and 1 in 1·2 million for hepatitis B virus (Williamson et al, 2003). However, in the absence of pathogen inactivation processes, the potential remains for contamination of plasma with an emerging, potentially lethal agent, for example variant Creutzfeld-Jakob disease (vCJD) in the UK, and West Nile virus (WNV) in the USA. The first case of possible transmission of vCJD as a result of transfusion was reported in 2004 (Pincock, 2004). Countermeasures taken to minimize the risk of vCJD transmission by transfusion, such as the use of leucocyte-depleted blood components and sourcing of FFP for neonates and children born after 1 January 1996 from outside the UK (O’Shaughnessy et al, 2004), increase the cost of blood products. By 2002, four transplant-associated cases and 23 transfusion-transmitted symptomatic cases of WNV had been identified. A test to detect genomic material of WNV was developed very rapidly, and transfusion-related transmission of WNV has been very rare since the implementation of testing in 2003 (Alter & Klein, 2008).
There are two main methods available for viral inactivation of plasma: treatment with solvent/detergent or methylene blue. However, at present, cryoprecipitate available in the UK is all derived from untreated FFP (O’Shaughnessy et al, 2004). The use of methylene blue for viral inactivation of plasma was first described by Lambrecht et al (1991). Reactive oxygen species that are generated when methylene blue is exposed to visible light inactivate viruses. However, it is well documented that methylene blue treatment also reduces coagulation factor levels, with fibrinogen one of the factors most sensitive to depletion. Compared with cryoprecipitate derived from untreated plasma, the loss of fibrinogen in methylene blue-inactivated cryoprecipitate ranged from 18% to 41% (Aznar et al, 2000; Hornsey et al, 2000; Seghatchian & Krailadsiri, 2001). A retrospective analysis of blood product use before and after the introduction of methylene blue-inactivated plasma in a University hospital in Spain found that the demand for plasma and cryoprecipitate increased across all diagnostic categories (Atance et al, 2001). The authors proposed compensation for low fibrinogen content as the most plausible explanation for this increase. No loss of coagulation factors is associated with solvent/detergent treatment of plasma, but there have been batch withdrawals due to possible contamination with parvovirus B19 (O’Shaughnessy et al, 2004). Both the methylene-blue and solvent/detergent methods of inactivation are primarily effective against enveloped viruses. Therefore, agents that are inactivated by pasteurization or removed by filtration during the processing of fibrinogen concentrates may still be transmitted by cryoprecipitate prepared from virally-inactivated plasma.
Aside from the risk of viral transmission, the same immune-mediated risks associated with transfusion of plasma exist for cryoprecipitate (MacLennan & Barbara, 2006). These risks include anaphylactic or anaphyloid reactions, allergic reactions, haemolysis, and TRALI. However, the non-specific allergic reactions associated with FFP transfusion are rarely seen with cryoprecipitate as only a small volume of plasma is used to resuspend the cryoprecipitated proteins (Ahrons et al, 1970; Reese et al, 1975). TRALI is believed to be caused by donor leucocyte antibodies, which are produced mainly as a result of pregnancy. Although it is under-diagnosed (Wallis, 2003), not least because the same clinical features are seen in acute lung injury resulting from other causes such as sepsis, trauma and shock, TRALI is a common cause of transfusion-related death. The UK National Blood Service announced its decision not to supply FFP from female blood donors in an effort to reduce the risk of TRALI in 2003. A recent summary of 10 years of haemovigilance reports of TRALI in the UK suggested that this policy has been effective in reducing the incidence of TRALI due to FFP (Chapman et al, 2008). However, the report also found that, although low, the risk from cryoprecipitate seems to have increased (from 4·3 cases per 106 cryoprecipitate components issued from 1999 to 2004 to 9·6 cases per 106 from 2004 to 2006). This could be because there is now a greater proportion of female donor plasma being used to prepare cryoprecipitate (Chapman et al, 2008).
A further safety concern in the management of perioperative bleeding is the risk of a thrombogenic event occurring. Solvent/detergent-treated plasma has reduced activity of the anticoagulant protein S, and has been associated with deep vein thromboses when used to treat patients with thrombotic thrombocytopenic purpura (PLAS+SD®; VI Technologies, Melvin, NY, USA) (Flamholz et al, 2000). Increased fibrinogen levels are associated with increased risk of coronary heart disease and myocardial infarction (MI). However, there is evidence to suggest that, in this instance, fibrinogen is acting as a marker rather than a mediator (Reinhart, 2003). For example, there is no increased risk of MI in people with fibrinogen gene polymorphisms that result in high fibrinogen levels (Reinhart, 2003). There have been nine reports of thromboembolic events in patients with congenital or acquired afibrinogenaemia during postmarketing surveillance of a fibrinogen concentrate (Dickneite et al, 2008). Although the patients had additional risk factors in most of the cases, a causal relationship could not be definitely excluded. Nonetheless, fibrinogen concentrate appears to have a very low risk of thrombogenicity, corresponding to one case report for every 13 655 treatments of 8 g (Dickneite et al, 2008). Thrombin generation in plasma is inhibited by formation of fibrin (hence fibrin has been referred to as antithrombin I), and increased thrombin generation in afibrinogenaemic patients can be normalized by fibrinogen supplementation (Korte & Feldges, 1994; Mosesson, 2003). Cryoprecipitate is not subject to the same postmarketing surveillance as fibrinogen concentrate, but has been associated with thrombotic events (Nizzi et al, 2002). Administration of cryoprecipitate to substitute fibrinogen could cause thrombosis as a result of supraphysiological levels of other proteins present in the precipitate (e.g. VWF).
Future improvements in management of perioperative bleeding
General uptake of fibrinogen concentrate in place of cryoprecipitate for the management of perioperative bleeding would bring about the benefits described above and summarized in Table II. There is also scope for further improvement through clarification of critical fibrinogen levels and more rapid and reliable diagnostic tests.
Table II. Attributes to consider when choosing between cryoprecipitate and fibrinogen concentrate for the management of perioperative bleeding.
|Immunological reactions||Risk of ABO incompatibility, TRALI and severe anaphylaxis, albeit low because resuspended in small volume of plasma.||Negligible risk of immunological reactions because purification steps during preparation remove donor antibodies|
|Risk of viral/pathogen transmission||Usually not virally inactivated. Viral inactivation using MB or SD not as effective as pasteurization||Manufacturing process includes pasteurization* and filtration steps that minimize the risk|
|Accuracy of dosing||Inconsistent fibrinogen content between units. Reduced fibrinogen content in cryoprecipitate from MB-treated plasma||Standardized fibrinogen content|
|Number of units/vials required to provide 4 g dose||29 units (based on 140 mg/unit)||4 vials|
|Reconstituted volume of 4 g dose||Approximately 375 ml (based on 15 g/l)||200 ml|
|Time to administration||Must first be thawed in a water bath||Can be reconstituted rapidly|
|Cost||Perceived to be cheaper than fibrinogen concentrate||Cost-effective and overall cost may be similar to cryoprecipitate|
|Availability||Not available in many European countries. Mainly used in UK and USA||More widely available than cryoprecipitate in most European countries|
Although fibrinogen levels of around 1 g/l are widely cited as a threshold for triggering supplementation, dating back to the late 1980s (Ciavarella et al, 1987), the optimal value for this is not clear (O’Shaughnessy et al, 2004; Stainsby et al, 2006). The critical threshold may vary with the clinical situation and patient characteristics, for example, normal fibrinogen levels increase with age (Oswald et al, 1983; Balleisen et al, 1985; Laharrague et al, 1993; Ishikawa et al, 1997; Coppola et al, 2000). Several authors consider a threshold of 1 g/l to be too low when blood loss is continuing (Haas et al, 2008). In an in vitro study, there was no clot formation at <0·5 g/l fibrinogen and at 0·75 g/l only weak clots were formed (Nielsen et al, 2005). In a study of postpartum haemorrhage (PPH), fibrinogen concentrations less than 2 g/l had a 100% positive predictive value for severe PPH (Charbit et al, 2007).
Point-of-care monitoring may further improve the management of patients with peri-operative bleeding. Certainly, evidence is growing that haemostasis management guided by thromboelastographic techniques is associated with a lower requirement for allogeneic transfusions (Shore-Lesserson et al, 1999; Royston & von Kier, 2001; Kozek-Langenecker, 2007; Spalding et al, 2007). Recently, assays that specifically measure the contribution of fibrinogen to clot formation by inhibiting platelets have been developed for both available point-of-care systems – the TEG® Coagulation Analyzer, Haemoscope Corporation, Niles, IL, USA and the ROTEM® Whole Blood Haemostasis Analyser, Pentapharm GmbH, Munich, Germany. These methods provide faster results than the standard laboratory tests and also offer greater accuracy. The Clauss method for measuring fibrinogen may give falsely high fibrinogen levels if colloids have been used for volume replacement because they impair fibrin polymerization (Hiippala, 1995; Mardel et al, 1998; Fries et al, 2002; Innerhofer et al, 2002; Mittermayr et al, 2007).