Haemostatic Agent Nature and Administration

Tranexamic acid (TXA; Cyklokapron) is a synthetic lysine analogue available as a 100 mg/mL solution. It now resides on the World Health Organization’s List of Essential Medicines [2]. Recognised side effects are reported to include gastrointestinal upset, nausea, allergic reaction, visual disturbance, hypotension with rapid i.v. administration and seizures at high doses. Relative contraindications include a history of previous convulsions, renal failure (dose adjustment required), pregnancy, patients at high risk of thrombosis and massive haematuria. There is no evidence of teratogenicity in animal studies with the manufacturer recommendation that use in pregnancy should only occur if potential benefit outweighs risk. TXA crosses the placenta [3].

Pharmacology and Action

The antifibrinolytic action of TXA is mediated by reversible binding to lysine binding sites on plasmin and plasminogen and thereby competitively inhibiting binding to fibrin. This impairs the ability of plasmin to break down fibrin, inhibiting fibrinolysis and thereby maintaining the fibrin meshwork and promoting clot stability. The small size of TXA means that distribution is to all body compartments. Clearance is predominately through renal excretion of the unchanged drug with a plasma half-life of two hours [4].

TXA has approximately eight times the antifibrinolytic efficacy of the older synthetic lysine analogue ε-aminocaproic acid that requires a loading dose followed by continuous infusion to maintain therapeutic plasma concentration. As such, the use of ε-aminocaproic acid is now reserved for situations where TXA is not immediately available. The older antifibrinolytic agent aprotinin is no longer recommended in the trauma scenario due to a lack of benefit [5] and a consistently negative impact on mortality [6].

Indications, Efficacy and Safety

Hyperfibrinolysis, driven by protein C activation secondary to tissue hypoperfusion and injury, is a key component of trauma-induced coagulopathy (TIC) and is associated with a high mortality [7–10]. The CRASH-2 trial [5] (Clinical Randomisation of Antifibrinolytic therapy in Significant Haemorrhage) provides the main evidence base for antifibrinolytic intervention in adult blunt and penetrating trauma. This randomised controlled trial of over 20,000 trauma patients with, or at risk of, significant bleeding allocated patients to clinician-blinded doses of either placebo or 1 g TXA over 10 minutes followed by a further placebo or 1 g dose over 8 hours. TXA significantly reduced all-cause mortality from 16.0% to 14.5% (relative risk 0.91) and haemorrhagic deaths by 0.8% (relative risk 0.85), predominately through an impact on exsanguination on the day of injury in those receiving TXA within 3 hours of injury [11, 12]. Importantly, administration after 3 hours significantly increased the risk of mortality secondary to bleeding (4.4% vs. 3.1% placebo mortality; relative risk 1.44).

While evidence in elective sugery shows that TXA reduces the need for blood transfusion [13], this is not borne out in Cochrane analysis of randomised trials in acute trauma [12]. However, the survival benefit of TXA in CRASH-2 may have led to surviving patients receiving blood where equivalent patients in the placebo arm would have died and hence not been transfused – eliminating any benefit of TXA on overall transfusion. Meta-analysis of 12 studies investigating TXA in orthopaedic trauma surgery showed a reduction in packed red blood cell (PRBC) for transfusion requirement (odds ratio 0.41) and bleeding mass without any increase in symptomatic thromboembolism (TE) [14]. Additional data supports the use of TXA in mature civilian trauma systems with a reduction in all-cause mortality and multiorgan failure in severely shocked patients, but not those without shock [15]. However, a post-hoc analysis of CRASH-2 suggests that TXA is of benefit regardless of severity of injury and should not be restricted to high risk or major haemorrhage patients [16].

The beneficial effects of TXA appear to be more pronounced the sooner it is given after injury, with delivery within an hour providing the greatest mortality benefit [11]. Meta-analysis of the CRASH-2 and WOMAN trials of TXA in trauma and major obstetric haemorrhage (MOH), respectively, showed that immediate TXA administration improved survival by more than 70% with efficacy decreasing by 10% for every 15-minute treatment delay until 3 hours, after which there was no benefit [17]. Administration of TXA beyond 3 hours after injury appears to be associated with an increased risk of harm [11] through fatal haemorrhage, although this was a post-hoc subgroup analysis with reduced precision. This has led the UK National Clinical Guideline Centre to recommend that empiric administration of TXA should be avoided more than 3 hours after injury without evidence of ongoing hyperfibrinolysis [18, 19].

Consideration should be given to the provision of TXA in a pre-hospital setting to minimise the time to delivery, although at present there is no robust evidence to support this approach [20–22]. Prospective data demonstrates that pre-hospital administration of TXA has beneficial effects on ROTEM indices of TIC with enhanced clot stability and reduced fibrinolysis [23]. TXA may also have a role in isolated traumatic brain injury. Prospective data from two randomised controlled trials (RCTs) of TXA vs. placebo in isolated traumatic brain injury demonstrates a reduction in haemorrhagic progression [24, 25] with a further RCT showing a non-significant trend towards improved mortality and outcome [26]. Retrospective observational data from Japan shows a significant reduction in 28-day mortality (10.0% vs. 18.4%) including in a subgroup with primary brain injury (6.0% vs. 13.2%) where TXA was given within 3 hours of injury [27].

The European guidelines on management of major bleeding and coagulopathy following trauma [28, 29] recommend TXA administration as early as possible to the trauma patient who is bleeding or at risk of significant haemorrhage (GRADE 1A [30]). These guidelines also highlight administration within 3 hours of injury in the bleeding trauma patient (Grade 1B) and that consideration should be given to administration of the first dose of TXA en route to hospital (Grade 2C). The UK National Clinical Guideline Centre also recommends that TXA may be used in paediatric trauma by extrapolation from the CRASH-2 study and the good safety profile of TXA in paediatric patients [18].

TXA is a cheap and readily available intervention with a cost of around £45 [18, 31] per life year gained. However, Europe-wide audit [32] of trauma practice in 2015 showed that only 66% of centres are using tranexamic acid frequently; future audits will hopefully demonstrate an increase in uptake.

Future Studies

The prospective Cal-PAT study assessing the feasibility of pre-hospital TXA administration to trauma patients with evidence of haemorrhagic shock has released interim data. This shows a significant reduction in blood product usage, a trend towards lower 24-hour mortality (3.9% vs. 7.2%) without any increase in adverse events [43]. The ongoing PATCH study of pre-hospital TXA will also give further insight into this area [44].

The CRASH-3 study, [45] an international, randomised, placebo-controlled trial that has now completed recruitment, will assess the impact of early TXA administration on death and disability in isolated traumatic brain injury (TBI) patients. This will complement the ongoing study into pre-hospital TXA in moderate-to-severe TBI [46].

Further insight in this area may be derived from the ongoing ULTRA study [47] of the efficacy of TXA in preventing aneurysm rebleeding in non-traumatic subarachnoid haemorrhage and the recently analysed TICH-2 trial [48] of TXA in spontaneous intracerebral haemorrhage.

Haemostatic Agent Nature and Administration

Fibrinogen concentrate (FgC ; Fibryga, RiaSTAP) is provided in variable pathogen-inactivated doses around 1 g per vial in powdered form for i.v. use once reconstituted. Fibryga should be administered within 4 hours of reconstitution, while RiaSTAP contains human albumin and should be administered within 8 hours of reconstitution. FgC is derived from pooled human plasma and demonstrates comparable pharmacokinetics to native fibrinogen with a half-life of 80 hours (70 hours in paediatric patients).

Cryoprecipitated AHF (antihaemophilic factor), commonly known as cryoprecipitate , is derived from separating the cold-insoluble protein fraction from fresh frozen plasma (FFP) by centrifugation and constitutes a subset of clotting proteins present in a single unit of whole blood. This 5–20 mL volume can be frozen and stored for up to 1 year prior to use. A single unit of cryoprecipitate contains 150–250 mg fibrinogen, 100–150 units of vWF, fibronectin, 80 units of factor VIII and 50–75 units of factor XIII. Pooled cryoprecipitate, as used clinically, contains the cryoprecipitate from 5 to 10 units of FFP. Cryoprecipitate requires 10–30 minutes to thaw and should be administered through a standard transfusion filter within 4 hours. Blood group compatibility (but not Rhesus type) is preferred.

Indications, Efficacy and Safety

Fibrinogen, clotting factor 1, is the final common substrate in the clotting pathway and is enzymatically converted by thrombin to fibrin. Fibrin promotes clot stability and platelet activation through GPIIb/IIIa activation and also functions as antithrombin I, thereby regulating clot proliferation. Fibrinogen is distributed solely to the vascular compartment without any systemic reserve; hence, the 8–10 g of fibrinogen in the pre-morbid patient is very sensitive to haemorrhagic losses, consumption and dilution. Protein C activation is a key component of TIC and leads to significant hypofibrinogenaemia [49, 50] such that fibrinogen deficiency is the commonest and earliest-detected clotting factor abnormality in TIC. Additionally, the physiological environment of trauma with haemodilution, acidosis, hyperfibrinolysis and hypothermia compound the situation through impaired fibrinogen function [51, 52].

Hypofibrinogenaemia , variably defined but typically as less than 1.5 or 2 g/L [53], is present in 25–70% of patients with haemorrhagic shock [49], more severe injuries and anaemia [54] and is associated with greater transfusion requirement [55] and worse outcomes [49, 53, 56, 57]. While it is clear that fibrinogen replacement in hypofibrinogenaemic coagulopathy improves outcomes [49, 56, 58, 59], prior to the availability of clotting results the initial approach to management of TIC and bleeding is contentious [60, 61]. The source of fibrinogen in initial resuscitation may be provided through either fibrinogen concentrate, FFP (which contains approximately 2 g/L fibrinogen), or cryoprecipitate.

While FFP administration maintains fibrinogen concentrations, it will not correct hypofibrinogenaemia [62] without infusion of around 30 mL/kg [63] and may worsen coagulopathy [50, 64, 65] and outcomes [61] with associated multiorgan failure [66, 67]. Because of these issues, FgC and cryoprecipitate have been assessed as alternative sources of fibrinogen. The choice between FgC or cryoprecipitate for fibrinogen supplementation is not clear with systematic reviews demonstrating no significant difference in outcomes [68, 69]. There is broad heterogeneity in institutional practice with cryoprecipitate [70], which is currently the leading source of fibrinogen in the UK, in contrast to the rest of Europe where FgC is more prevalent [71]. A key limitation of cryoprecipitate is delay in administration of typically 2 or more hours, by which time mortality may be 50% [72]. However, the CRYOSTAT study demonstrated that with appropriate infrastructure it is possible to administer cryoprecipitate within 90 minutes and generate fibrinogen levels above 1.8 g/L. [73]

Fibrinogen concentrate has the advantage of more rapid availability due to easy reconstitution and the lack of need for blood group matching as well as a smaller administration volume and hence lower risk of TRALI and TACO than FFP. FiiRST, a prospective study of FgC in haemorrhagic shock secondary to trauma, has shown the feasibility of administration within one hour of admission with no evidence of an increase in complications [74]. The early-fibrinogen in trauma (E-FIT) study will shortly report on the feasibility of providing FgC within 45 minutes of admission and maintenance of fibrinogen levels above 2 g/L in the face of ongoing bleeding across multiple UK trauma centres. While retrospective data of FgC administration in trauma shows a likely impact on survival, especially in patients with a higher injury severity score [75, 76], systematic reviews of FgC in surgery [77], trauma [78, 79] or haemorrhage [80] have not found any mortality benefit so firm conclusions cannot be drawn as yet. Similarly, while FgC in comparison to FFP [61] or cryoprecipitate [69] in broader haemorrhage settings demonstrates clinical efficacy, direct prospective comparisons are required.

Two grams of fibrinogen rather than high-dose FFP is appropriate in significant haemorrhage while clotting results are pending [81]. This is sufficient to maintain fibrinogen levels in the face of dilution by an initial transfusion of four units of PRBC. Clotting factor-only approaches to the management of TIC with FgC and/or prothrombin complex concentrate (PCC) show clinical and viscoelastic testing (VET) efficacy with reduced transfusion requirement and mortality [82, 83]. Consequently, some national guidelines [84, 85] now support FgC as the primary intervention for fibrinogen replacement although the European guidelines [28] for the initial resuscitation of coagulopathy still support an FFP:RBC ratio of 1:2 or greater (Grade 1B evidence). Initial resuscitation based on admission Hb level is also likely to be of value [28] (Grade 1C) with consideration of a threshold of 10 g/L for intervention [54]. At present, robust, prospective data supporting either FFP-based strategies or direct fibrinogen supplementation is lacking [86]. The choice of the form of fibrinogen administered to patients will be dependent on local guidelines, institutional practice, product availability, clinician preference and expertise. Health economics are also likely to play a role as per gram of fibrinogen FgC is four times the cost of cryoprecipitate ($1140 vs. approximately $414) [87].

Where FgC or cryoprecipitate is administered, a point of care (POC)-based targeted approach to transfusion should be implemented. VET-guided fibrinogen substitution permits a reduction in blood product exposure and treatment cost without impairment of clinical outcomes [81, 88, 89]. Furthermore, POC- or VET-guided administration of FgC and PCC, without the use of FFP, may lead to a reduced requirement for PRBC and platelet transfusion with good clinical outcomes [59, 89]. Similar findings have been demonstrated in a range of perioperative settings with a 90% reduction in FFP usage and a reduction in the incidence of massive transfusion [90].

Future Studies

The Pilot Randomized trial of Fibrinogen in Trauma Haemorrhage (PRooF-iTH) study is currently enrolling and will report on the efficacy and safety of first-line treatment with fibrinogen concentrate in trauma haemorrhage [100].

The Fibrinogen Concentrate in Trauma Patients, Presumed to Bleed (FI in TIC) study has completed recruitment and will provide RCT evidence comparing pre-hospital FgC to placebo in trauma [101].

The ongoing FEISTY trial is assessing the feasibility and efficacy of fibrinogen supplementation using VET-guided administration of FgC or cryoprecipitate [102].

CRYOSTAT-2 is currently recruiting to assess the impact of high-dose cryoprecipitate in the initial resuscitation of major traumatic haemorrhage [103].

Haemostatic Agent Nature and Administration

Four-factor prothrombin complex concentrate (PCC; Beriplex, Octaplex, Kcentra) is derived from donor-pooled human plasma and contains human albumin, human antithrombin III and heparin. All formulations are reconstituted from lyophilised powder and infused slowly at room temperature.

Of note, mass spectrometric analysis of Kcentra demonstrated 92 plasma proteins not included on the product insert, any number of which may contribute to the action of PCC in TIC and other settings [104]. Interestingly, in this analysis, Factor VII concentrations were very low (less than 1% that of prothrombin) highlighting the batch variability of this complex.

Indications, Efficacy and Safety

The role of PCC may prove to be limited outside of vitamin K antagonist (VKA)-related coagulopathy as thrombin generation is generally well maintained during trauma [105]. Furthermore, as clot instability is the major issue in TIC, platelets and fibrinogen are likely to be paramount in importance [106, 107]. However, PCC is likely to be an important part of concentrate-based approaches to TIC management, especially where there is evidence of delayed onset of thrombin generation.

PCC provides rapid factor replacement within a small administration volume and is indicated in patients with VKA-acquired coagulopathy, VKA-associated bleeding and urgent reversal of VKAs. There is very little evidence regarding the use of PCC outside the context of emergency reversal of VKA and potentially novel oral anticoagulants (NOACs), although it is licensed in most European countries for the management of acquired coagulopathy. Reversal of VKA-associated coagulopathy can be achieved within minutes [108, 109] compared to a number of hours with FFP. PCC can be administered rapidly and is a negligible volume in comparison to other transfusion products.

The rapid action of PCC and factor concentrates in enhancing thrombin generation and clot strength coupled with a minimisation of volume expansion and factor dilution might be expected to reverse coagulopathy more quickly than FFP.

Support for this comes from a limited number of retrospective studies assessing the empirical management of coagulopathic patients (defined as an INR ≥ 1.5). The addition of PCC to FFP-based management of TIC led to a reduction in PRBC and FFP use, more rapid correction of coagulopathy and a 5% absolute reduction in mortality [110]. Similar benefits were seen with the addition of PCC to FFP-based management of coagulopathic traumatic brain injury [111] with the additional advantage of shorter time to surgical intervention, an effect also seen in emergency general surgery [112]. A further retrospective study showed that the use of PCC in high-energy pelvic and limb fractures led to more rapid correction of coagulopathy and consequent surgical intervention as well as a reduction in PRBC and FFP requirement [113].

PCC is most commonly used as part of a factor concentrate-based management approach with fibrinogen and other concentrates, with additive effects on VET markers of coagulopathy [114]. A large retrospective database review [59] of clotting factor concentrates (FgC and PCC) compared to FFP in TIC showed a comparable mortality, but the concentrate-only approach required significantly less PRBC and had an 80% absolute risk reduction for the incidence of multiorgan failure and shorter ventilation requirement. Similarly, a retrospective analysis [82] of data from the DIA-TRE-TIC study [115] showed that a clotting factor-based approach reduced PRBC and platelet exposure with an attendant reduction in multiorgan failure (18% vs. 37%) and sepsis compared to FFP-based management. A further retrospective study [88] showed TEG-guided concentrate-based management led to an avoidance of RBC transfusion in 29% of patients in the FgC-PCC group vs. 3% in the FFP group with comparable data for platelet transfusion.

The only prospective data regarding PCC use comes from the RETIC trial [116], a single-centre open-label RCT in Austria that assessed the impact of VET-guided management using clotting factor concentrates vs. FFP in trauma. While the limitations of this study make it difficult to draw conclusions regarding the efficacy of the two approaches, PCC was required in 16% of patients in the clotting factor concentrate arm compared to 100% of those receiving FFP, demonstrating it may play a key role as a component of algorithm- and VET-based strategies.

There is insufficient data available to make firm recommendations on the best approach to the management of either initial resuscitation or that following availability of clotting results when considering use of PCC. The current European guidelines [28] state that management could include FFP or clotting factors or a combination of both and supports the use of PCC in the setting of delayed coagulation initiation as represented by prolonged clot initiation (ROTEM-CT or TEG-R) times where fibrinogen levels are normal (Grade 2C).

Haemostatic Agent Nature and Administration

Recombinant activated human factor VII (rFVIIa; eptacog alfa, Novo-Seven) is produced through recombinant cell lines. It is comparable to the endogenous activated clotting factor VII although with a shorter plasma half-life (2 vs. 5 hours).

Physiological levels of the activated serine protease FVIIa, complexed with tissue factor (TF), act through catalysis of the activation of clotting factors IX and X in the extrinsic clotting pathway. At supraphysiological levels, it appears that rFVIIa functions on the activated platelet surface to enhance factor X activation and thrombin generation, independently of TF [127]. As such, deranged levels of substrates or cofactors like calcium, factor X, fibrinogen, prothrombin and platelets are likely to impair rFVIIa efficacy [128]. The activity of rFVIIa in vitro is very dependent on the physiological environment which should be borne in mind regarding its likely clinical efficacy. rFVIIa activity is reduced by 90% and the rFVIIa:TF complex by 60% in a pH of 7.0 [129], although rFVIIa appears to be resistant to temperature changes with full activity in vitro at 33 °C.

Indications, Efficacy and Safety

Prospective controlled interventional studies in trauma are challenging to perform [130] and as such the majority of evidence for rFVIIa is of low quality. Two parallel RCTs [131] assessing rFVIIa efficacy in blunt and penetrating trauma requiring six or more units of PRBC within 4 hours have been performed using high rFVIIa doses (200, 100 and a further 100 mcg/kg). Both cohorts receiving rFVIIa required less PRBC transfusions and were more likely to avoid massive transfusion (above 20 PRBC units; 14% vs. 33%), although only the blunt trauma cohort reached significance. A subgroup analysis of those patients requiring higher FFP doses showed that rFVIIa reduced the incidence of MOH and/or ARDS (3% vs. 20%). However, no mortality benefit from rFVIIa was found in these studies or the larger follow-up RCT [130], although the reduction in PRBC requirement was reproduced in keeping with retrospective studies in traumatic [132–134] and other forms [135] of haemorrhage. Additionally, some retrospective studies have identified that, while rFVIIa significantly reduces initial haemorrhagic mortality, longer-term mortality is unchanged due to multiorgan failure [136, 137]. Additionally, while early studies suggested efficacy in isolated traumatic intra-cranial haemorrhage, the literature is now inconsistent [138–141] with some evidence of harm and a lack of prospective data to support its use [142].

Two meta-analyses [143, 144] and a systematic review [145] of rFVIIa across all off-license uses confirmed a lack of mortality benefit despite a reduction in blood loss and PRBC transfusion requirement. A trend towards better outcomes was noted where rFVIIa was used therapeutically rather than prophylactically and at doses no greater than 90 μg/kg. Interestingly, this systematic review also identified a reduction in the incidence of ARDS in trauma patients treated with rFVIIa (risk difference − 0.05) which may be attributable to the reduction in allogenic blood product transfusion required.

In keeping with the physiological sensitivity of rFVIIa, some experts feel that only a subset of patients respond to rFVIIa and this may explain the mixed results to date and the failure to demonstrate a mortality benefit in heterogeneous trauma cohorts. This has been noted in retrospective case reviews [146–149] demonstrating that patients with a pH under 7.2, platelet count under 100 × 10⁹/L and systolic blood pressure under 90 mmHg at the time of rFVIIa administration have a poor response. For maximum efficacy, rFVIIa should therefore be given concurrent to efforts to optimise these conditions.

The high cost of rFVIIa at £3,700 in 2007 [150] for a single dose and limited evidence base for efficacy mean that guidelines for off-license use in trauma should be instituted and administration reserved for where life-threatening haemorrhage and coagulopathy have persisted despite all other routine measures. The European guidelines support the use of rFVIIa where resuscitation has achieved a platelet count above 50 × 10⁹/L, fibrinogen above 1.5–2.0 g/L, haematocrit above 0.24 with concomitant administration of antifibrinolytics, correction of acidosis, core body temperature and ionised calcium levels along with surgical haemostasis have failed [28, 29]. This is supported by prospective multicentre data showing that an rFVIIa dose of 100 μg/kg was increasingly effective and led to better survival when a larger number of these criteria were met prior to rFVIIa administration [151]. Additional indications for use may include life-threatening bleeding where conventional therapy cannot be tolerated, is inappropriate or refused or where no other therapy is available.