Introduction
Factor I (fibrinogen) is said to be the first coagulation factor to reach critically
low levels in acute bleeding[1] and current guidelines recommend maintaining the plasma fibrinogen level above 1.5 g/L.[2] However, other coagulation components such as platelets, factor V (FV), and factor
XIII (FXIII) are influenced by thrombin generation earlier during the cascade of events.[3]
Among others, fibrinogen has a critical role in the maintenance of hemostasis as it
is converted to soluble fibrin to form a clot in concert with activated platelets;
it is further stabilized by FXIII-mediated fibrin cross-linking.[4] Fibrinogen is an important biomarker of acute inflammation.[5] In acutely ill patients, low fibrinogen levels have been associated with poorer
outcomes and increased mortality. In fact, hypofibrinogenemia is an independent risk
factor for increased mortality in trauma patients, suggesting the requirement of massive
transfusion[6]
[7]; and it is also strongly associated with shock severity.[8]
In the trauma setting, empirically based transfusion protocols used worldwide are
suggesting to use early administration of plasma, platelets, and red blood cells in
a 1:1:1 ratio,[9]
[10]
[11] while in viscoelastic assays (VEAs)-guided transfusion protocols, it seems that
plasma and platelets used can be reduced, and with better results.[12]
[13]
[14] In addition, it seems that overtransfusion occurs more frequently when empirical
massive transfusions are used as compared with VEA-driven protocols.[15]
[16] When analyzing these differences in detail, it becomes evident that none of the
large trials evaluating massive transfusion in severe trauma[9]
[10] mention the type of fibrinogen replenishment in detail, fibrinogen concentrates
or cryoprecipitate. Thus, the question arises whether the main difference between
the two earlier-mentioned approaches (fixed ratio protocols vs. VEA-guided protocols)
is, in fact, the use of specific fibrinogen supplementation.
In ongoing obstetric hemorrhage, low fibrinogen levels are associated with severe
postpartum hemorrhage[17]
[18]; prepartum fibrinogen levels and VEAs, however, are not predictive of postpartum
hemorrhage.[19]
[20] While low fibrinogen levels are also associated with increased bleeding in cardiac
surgeries,[21]
[22] VEAs do not improve the ability to predict bleeding,[23]
[24] and preoperative fibrinogen level does not predict future transfusion needs.[25] In hip fractures,[26] only postoperative VEAs but not preoperative ones may be a predictor of transfusion
requirements.
On the other hand, VEAs have been demonstrated to effectively detect hypofibrinogenemia
and serve as a guide for fibrinogen replacement.[27]
[28] In thromboelastometry (ROTEM), the FIBTEM assay includes cytochalasin in the TF-induced
assay, a cell-permeable alkaloid interacting with the actin filament component of
cytoskeletal networks to inhibit the platelet effect on the clot. Therefore, the result
reflects the nonplatelet, TF-induced clot formation supported by fibrinogen as well
as FXIII activity. F XIII activity is responsible for roughly 30% of the FIBTEM amplitude.[29] In addition, FIBTEM predicts the final strength of that clot.[30]
In healthy individuals, the FIBTEM assay has an excellent correlation with fibrinogen
levels as assessed by the Clauss method[31] and, accordingly, fibrinogen plasma levels are statistically very well correlated
with thromboelastographic assays.[32] The Clauss fibrinogen assay is the most often used laboratory method to measure
plasma fibrinogen levels. The main advantage of the FIBTEM assay over the Clauss method
is that it provides useful combined information on FVIII activity, fibrinogen concentration,
and, to a lesser extent, FXIII activity within 5 to 10 minutes, while the standard
laboratory assays require 30 to 60 minutes before the results are available.
Replacement therapy with coagulation factor concentrates has several advantages, such
as allowing standardized doses to be rapidly administered in a small volume compared
with fresh frozen plasma (FFP), which requires a larger volume to replace clotting
factors that potentially lead to fluid overload. Coagulation factor replacement has
a very good safety profile, as all products are being virally inactivated during the
manufacturing process. The RETIC trial,[33] a single-center randomized trial, randomly allocated 48 patients to receive FFP
and 52 to receive fibrinogen concentrate (FC) plus FXIII concentrate (with every second
fibrinogen dose), looking for differences in transfusion requirements and development
of multiorgan failure (MOF). This trial was prematurely stopped due to safety reasons
in the FFP arm, as half of these patients required rescue therapy, 30% needed massive
transfusion, and 66% developed MOF compared with 4, 12, and 50%, respectively, in
the factor concentrates arm. The study concludes that early and effective coagulation
factor supplementation, as per the study protocol, is essential when managing severe
bleeding in multiple trauma.
With this whole body of evidence in the literature, there has been a change in the
treatment of severe bleeding in the acute care setting in the last decade, prioritizing
early and aggressive factor concentrate replacement over FFP. However, before continuing
down this path, clinicians might want to answer these two questions:
-
Has fibrinogen replacement been linked to better outcomes?
-
Are VEAs fibrinogen-specific?
Regarding the first question, the CRYOSTAT-2 randomized clinical trial,[34] performed in 26 major trauma centers in the United Kingdom and the United States,
has recently allocated 1,604 patients with major trauma hemorrhage to receive as early
as possible cryoprecipitate (n = 799) versus standard care (n = 805). There has not been an improvement in clinical outcomes, the main target being
all-cause mortality at day 28 after trauma (OR = 0.96 (95% CI = 0.75–1.23), p = 0.74). These patients had no fibrinogen levels measured before supplementation,
but all patients in both arms received early hemostatic resuscitation and damage control
surgery. This result is likely pointing out that treatment with fibrinogen as a sole
factor is not as important as the whole early and coordinated damage control approach.
Also, fibrinogen replacement by itself has not been shown to improve clinical outcomes
in cardiac surgery, obstetrics, or liver transplant.
In relation to the second question, beyond healthy individuals, the correlation between
VEAs and fibrinogen plasma levels is not uniform across the literature, suggesting
that VEA results may be less specific to fibrinogen levels than frequently assumed.
For example, in surgical patients, FVIII activity, and not fibrinogen, shows the highest
correlation with thromboelastographic assays.[35] In the trauma setting, large interindividual variabilities have been reported, where
a Fibtem A10 of 5 mm corresponds to a Clauss assay result anywhere between 0.6 and
2.0 g/L, and a Clauss result of 2 g/L to a Fibtem A10 between 5 and 15 mm.[36] Consequently, VEAs have shown a moderate correlation with fibrinogen levels in severely
injured patients,[37] as well as in obstetric hemorrhage[38]
[39] and cardiac surgery.[40] In liver transplant, the FIBTEM assay correlates well with fibrinogen levels in
the pre-reperfusion period, but not after graft reperfusion.[41] In addition, a FIBTEM assay may overestimate fibrinogen contribution to clot firmness
in the presence of a high platelet count,[42] despite platelet inhibition. However, this is an unlikely scenario in a trauma setting
or liver transplant, but one that may occur in cardiac surgery. Also, in this particular
setting, VEAs have shown a very high negative predictive value for hypofibrinogenemia,
as a FIBTEM amplitude above a defined cut-off level (usually 8 mm at 10 minutes) virtually
excludes the likelihood of low plasma levels of fibrinogen. However, the positive
predictive value is rather low, below 0.7.[43] To sum up, there is a varying degree of correlation between VEAs and fibrinogen
levels in different clinical contexts, suggesting that VEAS believed to be relatively
specific for fibrinogen indeed are probably not.
In addition, functional fibrinogen polymerization assays are not equally efficient
in eliminating platelet contribution to clot strength. Therefore, they seem not uniformly
accurate to evaluate the fibrinogen part of the result, with cytochalasin-D-based
assays such as FIBTEM seemingly being more accurate than glycoprotein-IIb/IIIa inhibition-based
assays, such as functional fibrinogen-TEG.[44]
The most relevant point, however, is assuming the (misleading) concept that clot strength
is determined only by fibrinogen and platelets; misleading because activated factor
XIII (FXIII), hematocrit, and FVIII all have a relevant impact on the amplitude and
clot strength.[32]
[45]
[46] As said, FXIII activity is responsible for around 30% of the FIBTEM amplitude.[29] In fact, in the above-mentioned RETIC trial,[33] FXIII concentration measurements were obtained on admission, and FXIII concentrate
was administered with each second fibrinogen dose and in patients with bleeding scores
2 to 3, exhibiting FXIII concentrations below 60%. By the end of the study, more than
40% of the recruited patients received FXIII concentrate (27 in the CFC group and
11 in the FFP group). In obstetric hemorrhage, the largest prospective observational
study (n = 1,309) in which postpartum blood loss was objectively measured, FXIII was found
to be the only prepartum coagulation factor associated with postpartum blood loss,
as compared with fibrinogen and prothrombin.[20] FXIII stabilizes the fibrin mesh and increases clot resistance,[47] which is why it is important that acquired FXIII deficit is common in the acute
care setting and might explain the associated poorer outcomes.[48]
What is more, in the RETIC trial,[33] fibrinogen plasma levels at 48 hours after trauma increase while FXIII levels decrease
at the same time point, indicating continued consumption. In the authors' clinical
experience, in patients with an ongoing inflammatory response or in the prepartum
setting, VEAs and fibrinogen measured by Clauss assay are usually high, and fibrinogen
replacement is not required. On the other hand, FXIII has the longest half-life of
clotting factors, and the transcription of the gene that regulates its synthesis is
slow, inducing only a slower increase of plasma levels as compared with other coagulation
factors. This seems to explain the continued evolution of FXIII deficiency over a
longer period; besides the RETIC trial, more data on a correlation between time after
trauma and the reduction of FXIII activity beyond 7 days after trauma have been reported.[49]
With all the available data, including some solid evidence that reduction of F XIII
is associated with a poorer outcome, we want to stress that FXIII levels need to be
assessed early in potential deficiency settings to avoid delays in diagnosing the
deficiency and initiating replacement if necessary.[50] Indeed, in surgical patients at high risk for bleeding, the early intraoperative
use of FXIII has been beneficial in reducing blood loss.[51]
To sum up, VEAs are important to detect coagulation factor deficiencies early on;
unfortunately, they do not provide factor-specific measurements yet. If low fibrinogen
levels (as evidenced by factor-specific assays) occur, they are associated with impaired
clinical outcomes and linked to increased morbidity and mortality. However, there
is so far no proof of causality between these observations, as fibrinogen levels are
not predictors of bleeding or transfusion requirements, and replacing fibrinogen does
not improve outcome. Thus, data available to date do not suggest that fibrinogen replacement
by itself will improve outcomes. Over the last years, evidence has been accumulating
that strongly suggests that acquired FXIII deficiency is very prevalent in the perioperative,
perinatal, and trauma setting and that FXIII replacement can improve clinical outcome.
However, early FXIII administration and FXIII threshold levels for administration
are still under investigation in different clinical contexts. In the acute care setting,
it is suggested to be supplemented for a trigger of 60 to 70% as levels below 40%
are associated with increased morbidity and major bleeding.[50] In postpartum hemorrhage (PPH), the SWIss Factor XIII Trial in PPH (SWIFT) trial
is an ongoing phase 4 study; its goal is to determine if postpartum blood loss can
be reduced by replenishing coagulation factor XIII (FXIII) at an early stage of PPH.
Considering that VEA use has been associated with both reduction in blood product
use and increased fibrinogen use, clinicians should be aware of the potential bias
when VEAs are used to guide coagulation factor replacement therapy. A general hemostatic
resuscitation approach should include early control damage and escalating hemostatic
replacement, primarily based on using factor concentrates (FXIII, fibrinogen) along
with optimization of predisposing factors (anemia, hypocalcemia, and hypothermia)
as the first step.
