Raising the bar on fibrinogen: a retrospective assessment of critical hypofibrinogenemia in severely injured trauma patients

Introduction

Approximately one-quarter of severely injured trauma patients demonstrate an acute traumatic coagulopathy (ATC) at hospital admission.1 This coagulopathy is associated with significantly greater resource utilization, including blood product administration, and an increased risk of mortality. Historically, ATC was identified by specific conventional coagulation tests (CCTs), such as partial thromboplastin time, and international normalized ratio (INR), with specific focus on coagulation factor depletion as the underlying etiology of post-traumatic coagulopathy.1 2 As the mechanisms and complexity involved in ATC further evolved, a cell-based model of coagulation was developed and incorporated multiple pathways that control post-traumatic hemostasis and that may ultimately contribute to early disorders in coagulation, serve as markers of clinical outcome and provide targets for potential areas of therapeutic intervention.3 4

Fibrinogen is a serum protein that on activation by thrombin is converted to fibrin to form a mesh-work, which binds platelets and contributes to a hemostatic clot.5 Fibrinogen levels were traditionally not included as part of the CCTs that defined ATC. Interestingly, fibrinogen is depleted earliest and most rapidly among the coagulation factors after hemorrhage.6 7 Within the last decade, multiple investigations describe that admission fibrinogen levels are associated with increased mortality in severely injured trauma patients.7–11 Inaba et al9 reported that critically low fibrinogen (ie, <100 mg/dL) was the most significant predictor of in-hospital death. Similarly, Rourke et al11 noted that decreased fibrinogen levels were an independent predictor of 24-hour and 28-day mortality. Low admission fibrinogen is also associated with worse injury severity, increasing depth of clinical shock and volume of prehospital crystalloid resuscitation.8–11 Lastly, there are reports linking the need for massive transfusion to admission fibrinogen levels.12

Perhaps most striking is the level at which admission serum fibrinogen is considered clinically relevant and associated with adverse outcomes in severely injured patients. Historically, fibrinogen was not considered critically low until levels were <100 mg/dL.6 13 Although hypofibrinogenemia is well recognized to correlate with mortality, less is known about the critical level at which it does so. More recent investigations suggest that critical fibrinogen levels occur at greater thresholds than previously thought, perhaps even as high as 229 mg/dL.14 Additionally, the association with hypofibrinogenemia and multiple organ failure (MOF) has not been well established. Therefore, the purpose of the present investigation was to evaluate the association of admission fibrinogen levels on MOF and mortality in a severely injured cohort.

Methods

This is a retrospective investigation from a single academic trauma center. Following approval by the Institutional Review Board (University of Maryland, Baltimore: HP-00078011), we searched the institution’s trauma database over a 2-year period from October 1, 2015 to October 31, 2017. Due to the retrospective nature of the study, a waiver of informed consent was granted and therefore individual patient consent was not obtained. This study fulfilled Strengthening the Reporting of Observational studies in Epidemiology guidelines. Inclusion criteria were patients 18–89 years of age, Injury Severity Score (ISS) >15 and transferred from the scene of injury. In addition, included patients were also required to have a plasma fibrinogen level obtained within 30 minutes of admission. Patients who were transferred from an outside facility, sustained a cardiac arrest prior to arrival or who did not have a plasma fibrinogen level obtained within 30 minutes of presentation to the trauma center were excluded.

We collected patient demographics, injury characteristics, admission physiology data, admission Glasgow Coma Scale (GCS) score and CCTs. The admission Shock Index (SI) was calculated as the quotient of the admission heart rate over the admission systolic blood pressure (SBP) for each patient prior to blood transfusion. This value is a well-described marker for injury severity that is associated with increased transfusion requirements and mortality.15 Volume of transfused packed red blood cells (PRBCs) was collected, and the total amount of each product was calculated at 1 hour, 3 hours and 24 hours after admission. The Critical Administration Threshold (CAT) was also calculated, indicating transfusion of ≥3 units PRBCs in 1 hour, which represents an early marker for massive transfusion.16 At the time of data collection, our statewide trauma system did not use prehospital blood products. In addition, the time of each CCT was determined, and all CCT were obtained within 5 minutes of each other. These CCT laboratory values included platelet count, INR and plasma fibrinogen levels. Plasma fibrinogen was measured using the STA-Fibrinogen 5 assay (Diagnostica Stago, Parsippany, NJ, USA) and an STA Evolution device (Diagnostica Stago). Measurement was based on the Clauss method, where diluted plasma is treated with concentrated thrombin, and the clotting time is referenced against a calibrated curve to determine fibrinogen level.

The primary outcome was 28-day mortality. Secondary outcomes were 28-day MOF, CAT+ within 1 hour of admission, total PRBCs transfusion during the first 3 hours of admission and total 24-hour PRBCs transfused. MOF was defined by the Denver organ failure score.17 This score is well validated in the trauma literature and consists of an evaluation of four organ systems (cardiac, pulmonary, hepatic and renal) on each day of intensive care unit (ICU) stay. The organ failure score for each system is calculated daily on a scale of 0 to 3, with the summation of scores for the four organ systems comprising the total organ failure score. MOF was considered present if the total daily sum of the worst scores from each organ system was >3, with the first event of MOF occurring at least 48 hours after admission to the ICU. Time from hospital admission until the first day on which MOF occurred was documented.

Results

Four hundred and seventeen patients fulfilled inclusion and exclusion criteria. The median age was 37 years (IQR: 26–54), median ISS was 26 (IQR: 21–33), median GCS score on arrival to the trauma center was 14 (IQR: 6–15) and the mean admission SI was 0.79 (SD: 0.41). The majority of patients were male (354 of 417, 84.9%), sustained a blunt mechanism of injury (298 of 417, 71.5%) and were transported to the trauma center by ground (243 of 417, 58.3%). The mean admission fibrinogen was 262.3 mg/dL (SD: 87.6). The time from admission until the first fibrinogen value was obtained occurred at a mean of 12.0 minutes (SD: 6.0). Admission fibrinogen values were distributed among <150 mg/dL: 20 of 417 (4.8%), 150 mg/dL to 200 mg/dL: 76 of 417 (18.2%), and >200 mg/dL: 321 of 417 (77.0%). Patients in the lower admission fibrinogen strata (ie, <150 mg/dL and 150 mg/dL to 200 mg/dL) were significantly younger, had a greater admission SI, lower GCS, higher ISS, greater lactate and worse coagulation status, compared with patients with fibrinogen >200 mg/dL (table 1).

Discussion

In the current study, we described the phenotype of hypofibrinogenemia as younger, more severely injured patients, presenting with a higher lactate and who were more coagulopathic. Admission fibrinogen <150 mg/dL was associated with a greater risk of 28-day mortality, compared with 150 mg/dL to 200 mg/dL or >200 mg/dL. Patients with an admission fibrinogen <150 mg/dL and 150 mg/dL to 200 mg/dL, compared with >200 mg/dL, were more likely to receive greater amounts of PRBCs in the first few hours after hospital admission and achieve CAT+ status within 1 hour, which reflect a valuable marker to identify patients that receive a massive resuscitation, are at greater risk of death and require intense resource utilization. These patients represent an important population that may benefit from early, and potentially prehospital, blood product transfusion.20–22 An interesting observation was that admission fibrinogen was associated with an increased incidence of MOF. This is particularly relevant given recent discussion regarding hypofibrinogenemia and endothelial glycocalyx damage after severe injury.23 There is documented evidence that endothelial glycocalyx degradation is associated with increased MOF.23 Although our observation is strictly hypothesis generating, this should encourage future investigations to study the implications of admission fibrinogen on subsequent post-traumatic MOF.

Fibrinogen exists in plasma as a serum protein that is cleaved by thrombin and subsequently polymerized to form strands of fibrin.5 6 24 These fibrin strands serve as the framework on which platelets attach and are activated via glycoprotein IIb/IIIa receptors, further contributing to platelet activity, thrombin generation and clot amplification.3 6 Therefore, fibrinogen is crucial to functional hemostasis after injury. The reported physiologic range of fibrinogen is between 1.5 and 4.0 g/L (eg, 150 mg/dL to 400 mg/dL)3–5; however, the serum concentration decreases rapidly during active bleeding and reaches critically low levels more quickly than other clotting factors, despite the fact it is an acute phase reactant protein. Additional processes involved with traumatic coagulopathy, such as endothelial glycocalyx degradation, autoheparinization, protein C activation and hyperfibrinolysis contribute to impaired and pathologic fibrinogen utilization after severe injury.3 5 Iatrogenic causes are also responsible for rapid depletion of fibrinogen during hemorrhage such as excessive crystalloid resuscitation that dilutes serum coagulation factor concentrations and further exacerbates trauma- induced coagulopathy.5 6 11 Similarly, blood product resuscitation may also contribute to further decreases in fibrinogen levels, as red cells, plasma and platelets are typically not sufficient to replace hypofibrinogenemia.6 25

Hypofibrinogenemia is an established marker of adverse outcomes; however, the precise threshold and quantitative fibrinogen level associated with clinical outcomes has not been clearly defined. Rourke et al11 described that for every 1 mg/dL increase in fibrinogen, the risk of 28-day mortality decreased by 22%. The American Society of Anesthesiologists (ASA) Practice Guidelines for Perioperative Blood Management suggest that clinically relevant thresholds of hypofibrinogenemia exist at 80 mg/dL to 100 mg/dL and recommend that fibrinogen replacement rarely is indicated at levels <150 mg/dL.13 Recent European guidelines suggest critical levels of fibrinogen are less than 150 mg/dL in bleeding patients; however, data to support this threshold are based on limited evidence.26 The results of the current study demonstrate that admission fibrinogen <150 mg/dL is associated with a significantly increased risk of 28-day mortality, after adjusting for age, injury severity and admission GCS. It is important to acknowledge that our study population size may not have been large enough to detect differences in mortality at other fibrinogen thresholds, such as 150 mg/dL to 200 mg/dL. Hagemo et al observed in a multicenter study of 1133 trauma patients that a threshold for admission fibrinogen of 229 mg/dL was associated with a significant increase in mortality, suggesting that the critical value for low fibrinogen be reassessed.14 Future studies would benefit from larger, multicenter populations to potentially identify an optimal cut-off at which fibrinogen is associated with worse clinical outcomes in trauma patients. Nonetheless, the results of our investigation provide further evidence that warrants reassessment of the presently suggested thresholds for critical levels of hypofibrinogenemia that are recommended by the ASA13 and American College of Surgeons Committee on Trauma in severely injured trauma patients.27

Perhaps more interesting from the present study is that fibrinogen levels were associated with MOF. A potential explanation for this finding may in part be associated with the endothelial dysfunction that occurs after traumatic injury and coagulopathy.23 Endothelial glycocalyx damage is associated with the release of tissue plasminogen activator, dissolution of fibrin strands and subsequent hypofibrinogenemia.28 In addition, destruction of the glycocalyx is also associated with increased levels of systemic syndecan-1, which correlates with organ dysfunction.29 It has been observed in vitro that fibrinogen stabilizes syndecan-1 on the endothelial cell membrane to restore endothelial cell barrier integrity.30 Additionally, in a mouse model of hemorrhagic shock, fibrinogen decreased microRNA-19b, a pathologic microRNA that contributes to endothelial cell dysfunction.31 Barry et al32 demonstrated in a similar mouse model of hemorrhagic shock that resuscitation with cryoprecipitate preserved endothelial cell function and reduced vascular permeability. There is also human data suggesting that fibrinogen may mitigate the development of MOF. The Reversal of Trauma-Induced Coagulopathy using First Line Coagulation Factor Concentrates or Fresh Frozen Plasma trial randomized severely injured patients to early fibrinogen or factor concentrate compared with fresh frozen plasma resuscitation and found that patients receiving fibrinogen had a significantly lower odds of developing MOF.20 Due to the smaller sample size of our population, we are unable to comment on the effect of fibrinogen replacement via cryoprecipitate or to perform an adjusted analysis with MOF as an outcome. We selected the Denver MOF score because this is a validated scoring system in trauma patients with greater specificity for identifying prolonged mechanical ventilation, ICU length of stay and mortality.17 33 34 However, other scoring systems, such as the Sequential Organ Failure Assessment and Multiple Organ Dysfunction Score, are also described in the tr

There are a number of additional limitations to this investigation. Due to the retrospective nature, we are unable to determine if low admission fibrinogen is the cause of the observed increased mortality. We also evaluate quantitative fibrinogen levels and were unable to investigate qualitative studies, such as functional fibrinogen. Furthermore, although we adjusted for clinically relevant confounding variables, there may be factors associated with death that we may have not captured in our analysis. For example, our trauma registry was not able to provide data on prehospital crystalloid content or volume. Clinical practice in our statewide trauma system is limited to prehospital crystalloid, and there is no administration of colloid; however, the precise effect of prehospital resuscitation volume in our cohort is unknown. We were also unable to obtain consistent temperature and serum pH values on admission for each patient. Hypothermia and acidemia are variables that may affect fibrin polymerization and function in bleeding patients.6 Serum ionized calcium levels were not immediately obtained in every patient. Hypocalcemia affects hemostatic clot function and is associated with increased mortality.35 The cause of death in a majority of patients in our cohort was due to TBI. Although TBI is associated with an inherent coagulopathy,36 the results from this investigation may not be applicable to all trauma patients. It is possible that low fibrinogen is simply a specific marker of injury severity in patients with severe TBI that is not captured by other traditional measures of coagulopathy, such as INR.37

We were also unable to determine from the present study if fibrinogen supplementation in patients with hypofibrinogenemia results in decreased 28-day mortality. In the USA and UK, cryoprecipitate is used to treat acquired hypofibrinogenemia, with pooled units providing approximately 2g to 3 g of fibrinogen, whereas fibrinogen concentrate is primarily used and approved for trauma patients in Europe.38 A previous retrospective study observed that fibrinogen supplementation with higher ratios of cryoprecipitate: PRBCs in patients receiving a massive resuscitation was associated with decreased mortality in a combat support hospital21; however, there is limited prospective data on fibrinogen replacement and mortality in trauma patients. Results of the Early Cryoprecipitate for Major Hemorrhage in Trauma (CRYOSTAT-2) trial should be available in the near future, which compares early (within 90 minutes of arrival) high-dose cryoprecipitate to standard of care massive transfusion and the effect on mortality.22 There is also current development of cryoprecipitate with a longer shelf-life that would allow for more readily available product in emergency settings. Our data collection also occurred prior to implementation of whole blood administration at our institution, and we are unable to draw conclusions on the association of whole blood transfusion in patients with low fibrinogen. We did not administer prehospital blood products in our statewide trauma system, and it is conceivable that earlier transfusion of fibrinogen containing blood products may impact the severity of hypofibrinogenemia at hospital arrival.39–41 Although the current study does not address fibrinogen repletion, it does highlight the gap in current knowledge as to whether low fibrinogen is a biomarker of severe injury or a target for intervention.

In conclusion, we observed that traditional thresholds of hypofibrinogenemia are inadequate to define critical levels of fibrinogen in severely injured patients. The current data suggest that the association with hypofibrinogenemia and mortality may occur at <150 mg/dL. We have also shown an important association between hypofibrinogenemia and MOF. This may suggest that the effect of low fibrinogen levels extends beyond its effect on bleeding. Early identification of patients with hypofibrinogenemia may represent an important clinical target for specific early and targeted blood product transfusion, such as cryoprecipitate, fibrinogen concentrate or whole blood. Larger clinical studies that evaluate the full spectrum of trauma care and patient outcomes are necessary to identify the true impact of hypofibrinogenemia in the severely injured trauma population, which is necessary to guide replacement strategies and improve outcomes.