Secondary Logo

Journal Logo

2019 WTA PODIUM PAPER

Dense and dangerous

The tissue plasminogen activator-resistant fibrinolysis shutdown phenotype is due to abnormal fibrin polymerization

Dow, Nathan MS; Coleman, Julia R. MD, MPH; Moore, Hunter MD, PhD; Osborn, Zachary T. BA; Sackheim, Adrian M. BS; Hennig, Grant PhD; Butenas, Saulius PhD; Nelson, Mark T. PhD; Moore, Ernest E. MD; Freeman, Kalev MD, PhD

Author Information
Journal of Trauma and Acute Care Surgery: February 2020 - Volume 88 - Issue 2 - p 258-265
doi: 10.1097/TA.0000000000002554

Abstract

Trauma is a leading cause of death worldwide and is the leading cause of death in the United States for persons between 1 year and 44 years of age.1 Additionally, most early preventable deaths are due to uncontrolled hemorrhage.2 Coagulopathy related to fibrinolysis following major traumatic injury complicates postinjury care and is associated with an increased risk of morbidity and mortality.3 Severely injured patients present with a spectrum of fibrinolysis activity, quantified by thrombelastography (TEG), and those patients who have the pathologic extremes—hyperfibrinolysis (excessive) or fibrinolysis shutdown (impaired)—have increased mortality.4 Analysis of over 2,500 severely injured trauma patients demonstrated that hyperfibrinolysis (18%) is the least common, and fibrinolysis shutdown is the most common fibrinolysis group (46%), with the remaining third of patients having a normal physiological response.5 Within these three groups (hyperfibrinolysis, physiologic fibrinolysis, and fibrinolysis shutdown), subgroups exist based on fibrinolysis sensitivity to tissue plasminogen activator (t-PA).6 Appreciation of these subgroups is important as recent publications have lumped patients with fibrinolysis shutdown into one group, and their results suggest that viscoelastic assays without t-PA challenge information may lack discrimination for fibrinolysis.7,8 Approximately 27% of severely injured trauma patients form clots that are sensitive to exogenous t-PA.6 Identification of the subgroup of fibrinolysis shutdown patients lacking t-PA sensitivity is clinically relevant because they have a fivefold increase in mortality.6 While viscoelastic assays provide functional data on clot properties, structural differences in clot structure within proposed subgroups of fibrinolysis after trauma have not been previously described or analyzed.

Fibrin architecture plays a major role in the mechanical stability and fibrinolytic susceptibility of a clot. Fibrin is normally produced by thrombin-mediated fibrinogen cleavage, which removes N-terminal peptides and causes the formation of staggered, double-stranded protofibrils with subsequent lateral aggregation. The utilization of fluorescent-labeled, purified fibrinogen to monitor fibrin formation in plasma with confocal microscopy has demonstrated mechanisms which impact thrombin generation resulting in production of abnormally structured clots.9 Previous studies have demonstrated that thick, loosely woven fibrin fibers are prone to fibrinolysis, while thin, densely packed fibers resist breakdown.9–11 Patients with coagulopathy, including hemophilia, venous thromboembolism, stroke, and myocardial infarction form plasma clots with altered fibrin quality; microscopic monitoring of fibrin formation has not previously been studied in plasma from individuals with acute traumatic coagulopathy.

The objective of this study was to describe the patterns of fibrin architecture in trauma patients with abnormal fibrinolysis. To address this objective, we generated in situ fluorescent clots from the plasma of trauma subjects with evidence of abnormal fibrinolysis, evaluated by mechanical clot strength and t-PA sensitivity through TEG, imaged these clots with confocal microscopy, and quantified patterns of fibrin structure these clots.12,13 We hypothesize that patterns of fibrin polymerization in trauma patients will correlate with whole blood fibrinolysis and/or t-PA sensitivity subgroups.

METHODS

Study Design and Setting

We performed an observational, retrospective basic science study of previously collected plasma and clinical data from acutely injured adult trauma patients admitted to a single, urban Level I trauma center.

Selection of Participants

Samples were analyzed from the Trauma Activation Protocol (TAP) registry, which includes all trauma activation patients who sustained blunt or penetrating trauma from April 2014 to August 2018. TAP was approved by the Colorado Multiple Institution Review Board (COMIRB 13-3087) and performed under waiver of consent due to minimal risk. For this study, patients were included if they had an Injury Severity Score > 15 and had a citrated rapid TEG and t-PA-challenge TEG from within 1 hour postinjury. Patients taking anticoagulants, who had received antifibrinolytics prior to TEG or who were transferred from another hospital were excluded. Trained research staff recorded clinical information, including patient demographics, injury patterns, blood product use, and mortality. Fibrinolysis was categorized by initial TEG into one of the three as previously described phenotypes4 hyperfibrinolysis (LY30, ≥ 3%), fibrinolysis shutdown (LY30, < 0.9%), and physiologic fibrinolysis (LY30, 0.9%–2.9%), which has been validated in Houston,5 Miami,14 and Pittsburgh.15,16 Each of these three types of fibrinolysis was further differentiated by t-PA challenged TEG (TEG with 75 ng/mL t-PA added) and categorized as either t-PA hypersensitive (LY30, > 27.7%), t-PA–resistant (LY30, < 1.8%), or t-PA–sensitive (LY30, 1.8%–27.7%), resulting in nine potential subgroups. These subgroup cutoffs were previously described based on the distribution of t-PA TEG LY30 from healthy volunteer data, such that patients with a t-PA TEG LY30 < 5th percentile of healthy volunteers were classified as resistant and patients above the 95th percentile were classified as hypersensitive.17 We selected trauma patients based on their fibrinolysis subgroup for microscopic assessment of fibrin formation. Samples from five noninjured volunteers were selected for comparison.

Measurements of Fibrinolysis and t-PA Sensitivity With TEG

Blood was collected in 2.7 mL citrated tubes (Becton-Dickinson, Franklin Lakes, NJ), re-calcified, and assayed with the TEG 5000 Hemostatic Analyzer (Haemonetics, Braintree, MA). The TEG values for activated clotting time (ACT), angle (degrees), maximum amplitude (MA, mm) and lysis 30 minutes after MA (LY30, %), were recorded.

Plasma Sampling

Platelet poor plasma was prepared within 30 minutes of collection. Citrated tubes with whole blood were spun at 1000g at 4°C for 15 minutes. The supernatant was decanted and respun at 12,600g for an additional 6 minutes, after which supernatant plasma was flash frozen with liquid nitrogen and placed in −80°C storage.

In Situ Fibrin Polymerization

We measured the polymerization of Alexa-488 labeled fibrinogen (Thermo Fisher Scientific, Waltham, MA), using a modification of previously published methods.18 For confocal microscopy, 0.5-mL frozen plasma samples were thawed in a 37°C water bath with 10 μL of corn trypsin inhibitor (CTI). A 4.66-μL Alexa-488-labeled fibrinogen (Thermo Fisher Scientific), 3 μL 0.5 M CaCl2 solution, and 3.6 μL tissue factor solution with phosphatidylcholine-phosphatidylserine were added to 93.4 μL of thawed plasma, pipetted onto glass chamber slides (Thermo Fisher Scientific) and allowed to clot for 30 minutes. Formed clots were washed multiple times with phosphate-buffered saline, fixed with HistoChoice (MilliporeSigma, Burlington, MA), and treated with an antiphotobleaching agent (Aigilent Technologies, Santa Clara, CA) prior to placement on a coverslip. Three-dimensional “Z-stack” images series were acquired at 0.125 micron steps using a Nikon C2 confocal microscope (Nikon Corporation, Tokyo, Japan) at 60 × 1.5 magnification. All trauma samples were prepared in duplicate, and during each plasma clot prep, an additional clot was made from pooled donor plasma as a day-to-day control to ensure consistent washing, fixing, fluorophore performance, and microscope consistency. Plasma samples from noninjured controls were also studied to provide a reference range for analysis. We validated our methods by establishing reproducibility and robustness on different days and in the hands of different technicians.

For each clot, we analyzed macroscopic fibrin structure using custom in-house software, for two quantifiable outcomes relevant to hemostatic potential: (1) fiber resolvability and (2) clot porosity. We established the “fiber resolvability index” as a measurement of the distinctness or clarity of fibrin polymerization. Fibrin fibers are most distinct when background fluorescence is low and fiber intensity is high. Slight variations in the brightness of each individual clot limit the suitability of absolute measurements of fiber intensity and background for comparison between samples. In contrast, analysis of fiber resolvability, as determined by standard deviation mapping, provides a reliable assessment of the clot architecture that is unaffected by the overall brightness of the fluorescence.19 The use of standard deviation as a measure of structural resolvability is a modification of common signal-to-noise or contrast-to-noise ratios, except we do not implicitly define a signal (i.e., a fluorescent fiber), as in some experiments they could not be reliably identified. As such, the use of just the standard deviation component of the signal-to-noise equations provides an relatively good estimate of how much fluorescent fibers are increasing the contrast range throughout the image—regardless of the absolute intensity of the background or the absolute intensity of the fiber fluorescence. It also unaffected by the regularity of spacing (which would be a problem with fast Fourier transformation) and does not require a stereotypical shape of fibers for vector weighting to work. Comparison of the variance of fluorescence intensity was, therefore, used to quantify the fiber resolvability for comparison between samples. We defined the fiber resolvability index as the standard deviation (STDEV) of fluorescence at every one of 262,144 pixels in the 512 × 512 pixel image. This standard deviation was calculated for each of the 262,144 pixels, across the depth of that column of pixels in the Z-stack of the three-dimensional microscopic image (Supplemental Digital Content 1, Figure 1, http://links.lww.com/TA/B515). The results for each clot were displayed visually as an STDEV map, in which the standard deviation of each pixel calculated in the Z-direction was colored, and as standard and cumulative histograms which quantify the area of the image that had a certain resolvability index. Clots with abnormal fiber formation had less distinct, “hazy” fibers, with little variation in signal intensity relative to the background fluorescence, relative to controls. This was mathematically expressed as a lower resolvability index (lower fiber definition). We measured the “clot porosity” by analyzing the size and distribution of the gaps between fibrin fibers in three-dimensional space. To examine whether the density of fibrin fibers in in vitro clots was sufficient to theoretically prevent passage of human red blood cells, we developed this measure of porosity by quantifying the size and distribution of the three-dimensional pockets in between fibrin fibers (negative space; see Supplemental Digital Content 1, Figure 1, http://links.lww.com/TA/B515).

Statistical Analysis

Data are expressed as means ± standard error of the mean (SEM). Ordinary one-way analysis of variance (ANOVA) with an alpha of 0.05 was used for comparison of fiber resolvability between fibrinolysis subgroups. A p value less than 0.05 was considered significant.

RESULTS

Overall, 810 patients were enrolled in TAP between April 2014 and August 2018, Among these individuals, we selected 35 subjects demonstrating the full range of fibrinolysis and t-PA phenotypes on admission TEG for further evaluation of clot structure using confocal microscopy (Fig. 1). Our cohort was selected to include different subpopulations of fibrinolysis shutdown, with five patients in each of seven fibrinolysis subgroups (no patients were identified with t-PA–sensitive fibrinolysis shutdown or t-PA–resistant hyperfibrinolysis) (Table 1). Samples were selected randomly from each fibrinolytic phenotype using R biostatistical software, and matched by age, sex, and injury mechanism.

Figure 1
Figure 1:
Flow diagram demonstrating allocation of enrolled subjects into cohort selected for study.
TABLE 1
TABLE 1:
Patients Selected for Microscopy by t-PA Sensitivity (Hypersensitive, Sensitive, or Resistant) and Fibrinolysis Group (Fibrinolysis Shutdown, Physiologic Fibrinolysis or Hyperfibrinolysis)

The study cohort consisted of severely injured patients with an age range from 18 years to 69 years and injury severity score (ISS) from 17 to 75, with 31% mortality. All samples were taken prehospital (in field en route to hospital) or upon arrival to the emergency bay and within 1 hour of presentation. Seven patients who died within 24 hours: 1 from TBI at 2.4 hours, 1 from TBI at 12.3 hours, 1 from TBI at 3.2 hours, 2 from hemorrhage at 3.1 hours, 1 from hemorrhage at 0.9 hours, and 1 from hypovolemic shock at 2.8 hours. The t-PA hypersensitive hyperfibrinolytic patients in our dataset had the highest injury severity, degree of shock (low systolic blood pressure), need for blood product, and mortality rate (80% mortality). These subjects had markedly abnormal TEG tracings, including several with very weak maximum clot strength and rapid clot breakdown (100% lysis) in a pattern that has been called a “death diamond” (Supplemental Digital Content 4, Figure 4, http://links.lww.com/TA/B518). Interestingly, individuals in the t-PA–sensitive hyperfibrinolytic group had less deranged clinical and laboratory results with a 0% mortality rate despite a New Injury Severity Score of 24. Additional demographic, TEG and clinical information for the study cohort can be found in Table 2.

TABLE 2
TABLE 2:
Demographic, TEG and Selected Clinical Information for All Seven Fibrinolysis/t-PA Sensitivity Subgroups Represented in This Study; t-PA Hypersensitive Fibrinolysis Shutdown (Hypersensitive SD), t-PA–Resistant Fibrinolysis Shutdown (Resistant SD), t-PA Hypersensitive Physiologic Fibrinolysis (Hypersensitive PL), t-PA–Sensitive Physiologic Fibrinolysis (Sensitive PL), t-PA–Resistant Physiologic Fibrinolysis (Resistant PL), t-PA Hypersensitive Hyperfibrinolysis (Hypersensitive HF), and t-PA–Sensitive Hyperfibrinolysis (Sensitive HF)

We validated previously published methods for monitoring the polymerization of fluorescently labeled fibrinogen, evaluating reproducibility and robustness in our laboratory.18 Aliquots of the same plasma sample prepared on different days yielded clot structures that were indistinguishable. Duplicate clots were produced in 26 patients and five noninjured controls; imaging and analysis were completed for 62 of 80 total samples. Some clots or their duplicate were unsuitable for analysis due to excessive photobleaching or mechanical errors such as failure of slip fixation during imaging or incorrect stepping of the microscope stage at 125-nm steps, causing smearing of the z-stack. In four instances, a fragile clot structure was observed in one of the clots but the clots lifted off the microscope slide while washing out excess fluorophore. Duplicate clots were visually similar in appearance, and quantitative analysis of fiber resolvability demonstrated a high degree of reliability for our measurement. The resolvability index calculated from duplicate experiments from the same patient was in close agreement (n = 10 patients), indicating that the pattern of fibrin clot formation of each patient was conserved. We immediately observed that clots from t-PA hypersensitive hyperfibrinolysis subjects were irregular, with a moth-eaten, ragged appearance that was dramatically different from previously published fibrin structure (Supplemental Digital Content 4, Figure 4, http://links.lww.com/TA/B518). We proceeded to develop analytic methods to quantify the visible differences that we observed. These methods included measures of porosity and fiber resolvability. Porosity measurements used particle analysis of three-dimensional structure. Using representative clots from a normal individual, we established proportion of three-dimensional pockets that had a volume greater than that of a human red blood cell (80 fL) was 0.18% of the total number of three-dimensional pockets, indicating normal fibrin fiber density was sufficient to severely restrict or impede the passage of red blood cells. Fiber resolvability, determined by standard deviation mapping, was selected because it provides a reliable assessment of clot architecture that is unaffected by the overall brightness.

Analysis of fiber resolvability revealed a significant difference (p < 0.05) between controls and trauma subjects (Fig. 2). Controls consistently demonstrated a robust meshwork of clearly defined, cross-linked fibrin fibers and had the highest fiber resolvability (342 ± 98.5 units). All trauma samples had significantly less fiber resolvability than controls. Median fiber resolvability was lowest in t-PA–hypersensitive samples (193 ± 41.8 units), followed by t-PA–sensitive samples (213 ± 66.0 units), and highest in t-PA–resistant samples (215 ± 41.1 units), but differences between these subgroups did not achieve statistical significance. Three samples were statistical outliers from their subgroups and excluded. A fourth sample was radically different from its duplicate sample and was also excluded.

Figure 2
Figure 2:
Fiber resolvability of trauma samples is statistically different from that of control samples, but does not statistically differ between different subtypes of trauma. These relationships are supported both when graphing the fiber resolvability for all analyzed pixels in each grouping (panel A) and when performing an ordinary one-way ANOVA; Bonferonni correction for multiple comparisons (p < 0.05), using the median pixel resolvability (panel B).

We observed a range of different patterns of fibrin polymerization (Fig. 3) with the most dramatic examples at the extreme ends of fibrinolysis spectrum. The t-PA–resistant fibrinolysis shutdown group appeared extremely dense and homogeneous relative to controls, while the t-PA hypersensitive hyperfibrinolysis subgroup appeared to have porous, weak clots (Fig. 4).

Figure 3
Figure 3:
Representative confocal images (Panel A) and corresponding computer generated standard deviation maps (Panel B) of clots for control (60× oil immersion), t-PA–resistant fibrinolysis shutdown, t-PA hypersensitive hyperfibrinolysis, and t-PA–sensitive with physiologic fibrinolysis, from left to right. Controls consistently demonstrated a robust fibrin network with high fiber resolvability. t-PA–resistant fibrinolysis shutdown samples were densely packed with extremely thin, almost hair-like fibers. t-PA hypersensitive hyperfibrinolytic samples showed widely spaced fibers, irregular structure and large holes or pores.
Figure 4
Figure 4:
Extreme subgroups of fibrinolysis demonstrate a difference in clot architecture porosity. (Top) Three individuals from the t-PA–resistant fibrinolysis shutdown subgroup. (Bottom) Three individuals from the t-PA hypersensitive hyperfibrinolysis subgroup. Confocal microscopy images (60×, oil immersion) of fluorescently labeled fibrinogen polymerized in the presence of trauma plasma were obtained in duplicate. For each clot, one cross-sectional image from the three-dimensional Z-stack of 40 slices is shown. t-PA–resistant fibrinolysis shutdown samples were almost devoid of pores. t-PA hypersensitive hyperfibrinolytic samples had pores as large as 2500 μm3 to 20,000 μm3. Pore sizes in noninjured controls ranged from 78 μm3 to 1250 μm3 (not shown).

However, among the subjects in the t-PA hypersensitive hyperfibrinolysis subgroup—and some others within the overall trauma cohort—we observed at least two different patterns of porous clots. In some cases, the “holes” between fibers appeared to be the result of fiber destruction, with findings including a thinning pattern between fibers, “sludge” between the fibers, or clumping of fluorescence suggesting abnormal fibrin aggregation. In other cases, holes may be due to abnormal fiber formation, with findings including long, axial segments of fibers lacking typical cross-linking patterns. On close inspection, control clots also demonstrated some “holes” but these appeared nearly spherical, not ragged, and occurred infrequently. Clots generated from healthy controls (n = 4) demonstrated holes smaller than 78 μm3; no hole was larger than 1,250 μm3 (a volume equivalent to 14 red blood cells at 90 μm each). t-PA–resistant fibrinolysis shutdown samples were almost devoid of pores. In contrast, some clots from the t-PA hypersensitive hyperfibrinolysis subgroup with poor fibrin architecture had pores as large as 2500 μm3 to 20,000 μm3 (30–220 red blood cells) (Supplemental Digital Content 2, Figure 2, http://links.lww.com/TA/B516 and Supplemental Digital Content 3, Figure 3, http://links.lww.com/TA/B517).

All subgroups of trauma samples demonstrate some degree of intragroup variability. Notably, in the t-PA hypersensitive hyperfibrinolysis group the one sample that differed the most from other clots in the group was from the single patient who survived (4/5, 80% mortality rate for this subgroup). Control samples exhibited less variability as a group than trauma subgroups.

DISCUSSION

Beyond the detriment to patient outcomes, trauma-induced coagulopathy (TIC) increases demand on hospital resources in the Emergency Department, Operating Room and Intensive Care Unit, and previous work has correlated increased fibrinolysis with need for blood products.20 While hyperfibrinolysis may lead to death by exsanguination, fibrinolysis shutdown may result in macrothromboses, resulting in stroke, deep vein thrombosis, and pulmonary embolism. Additionally, microthromboses can lead to multiple organ failure and eventually death.21

Despite the significant prevalence and mortality of TIC, underlying mechanisms are poorly understood. Standard laboratory measurements, such as D-dimer, fibrinogen, and prothrombin time are unreliable tools for predicting or diagnosing TIC.22 Viscoelastic assays, such as TEG and rotational thromboelastometry, measure the rate and strength of clot formation in real time and have the benefit of accounting for variables in whole blood versus plasma alone.23–25 While the assessment of fibrin polymerization by microscopy may lack utility as a point of care test, monitoring of fibrin clot structure provides novel insights into mechanisms underlying TIC.

Previous studies have used electron microscopy,13,18,26,27 or confocal microscopy12,18,27,28 to quantify altered fibrin structure. Methods have included pixel threshholding,28 counting fibers crossing superimposed gridlines,12,18 and measurements of individual fiber thickness.13,27 In the context of coagulopathy outside of severe trauma, altered fibrin structure has been described in coagulation factor deficiencies,28,29 deep vein thrombosis,30 stroke,31 myocardial infarction,32 and pulmonary embolism,33,34 where deficient clots are often composed of loosely woven thick fibers. In contrast, subjects with thrombotic pathologies demonstrate densely structured clots composed of thinner fibers. To our knowledge, our study represents the first attempt to evaluate fibrin patterns in trauma.

Using particle analysis of clot porosity, we distinguished clot subgroups that correlate with whole blood fibrinolysis subgroups, which have been linked to clinical outcomes. While we did not correlate clinical outcomes with clot structure, our results help explain the higher rate of massive transfusion in hyperfibrinolytic patients. The t-PA hypersensitive hyperfibrinolysis subgroup qualitatively demonstrated loosely packed fibrin structure, a pattern consistent with weak clot formation and likely to contribute to bleeding complications. The lone survivor in this subgroup demonstrated clot structure that was abnormal, but closer in appearance to controls than any of the t-PA hypersensitive hyperfibrinolytic patients that died. Conversely, the fibrinolysis shutdown cohort that was t-PA–resistant had a predominant fibrin matrix covering the slide with dense fibrin strands. This could contribute to thrombotic morbidity in fibrinolysis shutdown, as the dense clot structure would likely limit t-PA access to the fibers.

Importantly, our results support validity of these fibrinolysis subgroups as categorizations of similar clot architecture. One-way ANOVA analysis (Fig. 2, panel B) demonstrate that some heterogeneity exists within subgroups, including controls, but for all groups most data points fell within the upper and lower quartiles with the assigned trauma subgroupings.

Our results suggest possible mechanisms that drive altered fibrin structure in trauma patients. Abnormal thrombin generation is one possible explanation. Increased thrombin is associated with denser clots that resist lysis, and decreased thrombin generates more sparsely structured clots prone to breakdown.9–11 Alternatively, a shifted balance of the t-PA/plasminogen activator inhibitor-1 (PAI-1) system could explain altered fibrin structure in trauma. Massive t-PA release overwhelms PAI-1 reserves in hyperfibrinolytic trauma patients, possibly as a protective mechanisms to ensure microvascular patency.35 Thrombin generation also increases t-PA production.36 Future mechanistic work will focus on thrombin generation and t-PA/PAI-1 activity in trauma patients, particularly those with fibrinolysis shutdown.

Analysis of clot structure, as well as viscoelastic evaluation of coagulation status, suggests that some trauma patient populations, specifically those in fibrinolysis shutdown and those with physiologic fibrinolysis, may be harmed by tranexamic acid. More recent studies in the United States have demonstrated a mortality increase with tranexamic acid use. Prophylactic administration, or administration without clear evidence of hyperfibrinolysis, is potentially detrimental to individual patient outcomes.37,38 Additionally, analysis of fibrin architecture can give possible mechanistic insight into tPA sensitivity and fibrinolysis phenotypes in trauma. Differential responses to antifibrinolytic therapy in other patient populations may also be explained by fibrin structure.

Limitations exist in our study design. The relatively small sample size (n = 35), with only five subjects in each of seven subgroups, limits statistical comparisons between different subtypes of fibrinolysis. Nevertheless, the automated, computational analyses of fiber resolvability and porosity we developed may prove useful when analyzing larger cohorts. Changes in overall signal intensity secondary to alterations in microscope settings or photobleaching create a potential source of bias in other visual, nonautomated measures of fibrin network density. To reliably identify a fluorescent object in an image, it needs to have sufficient contrast above background in order to threshold/segment it. Second, it needs to be spatially restricted such that there is sufficient background around the object and the pixel representation of the object needs to be great enough so that noise does not corrupt the object (i.e., like a fiber that is 1 pixel wide). Usually the cutoff for identifying an object is two to three standard deviations above background intensity. Normally, any part of the image not meeting these criteria would be discarded, or considered as background. This, of course, biases the analysis to bright, well-spaced and well defined structures on dark backgrounds. Clots in which fibrin fibers were not fully assembled often had small patches alternating fluorescence intensity—that could not be resolved as individual fibers per se. Instead of discarding those regions, the standard deviation method includes them as part of the continuum of contrast (variance), such that all fluorescence features of the clot are measured during the analysis. This reduces bias toward and gives a more accurate account of the nature and composition of clots.

Additional computational analyses may provide other quantitative metrics for discrimination between samples, including fiber thickness, branching, interfiber spacing, and total fiber density. Additional work into the time course of clot formation and degradation using video microscopy may provide further insight. We also hope to improve resolution of images up to 100×, with deconvolution or superresolution microscopy, to further define these abnormalities in severely injured patients.

In conclusion, we successfully characterized fibrin architecture in human trauma patients for the first time. Our observation of different patterns of fibrin polymerization in the most extreme ends of fibrinolysis spectrum suggests that multiple different processes can produce highly porous clots, with different patterns suggesting abnormalities in fiber polymerization, thickening, and/or degradation. Demonstration of abnormal fibrin structure in this investigation, and the computational methods we developed to quantify these differences, provides the critical and necessary foundation for additional mechanistic studies that will further elucidate the pathogenesis of different fibrinolysis subgroups in trauma.

AUTHORSHIP

N.E.D. contributed to study design, experimental series, data collection and analysis, data interpretation, writing, and critical revision. J.R.C. contributed to study design, experimental series, data collection, writing and critical revision. H.M. contributed to study design, experimental series, data collection, writing, and critical revision. Z.O. contributed to data collection and critical revision. A.M.S. contributed to data interpretation, writing, and critical revision. G.H. contributed to data analysis, interpretation, and critical revision. S.B. contributed to study design, data collection, analysis, and interpretation. M.T.N. contributed to data interpretation and critical revision. E.E.M. contributed to study design, data collection, analysis, interpretation, writing, and critical revision. K.F. contributed to study design, data collection, analysis, interpretation, writing, and critical revision.

ACKNOWLEDGMENTS

Funding for this study provided by the National Institutes of Health (UM1HL120877, R01GM123010, R35HL140027, and T32GM08315).

DISCLOSURE

The authors have no conflicts of interest to disclose. This article has not been previously presented or published.

REFERENCES

1. King DR. Initial care of the severely injured patient. N Engl J Med. 2019;380(8):763–770.
2. Kornblith LZ, Moore HB, Cohen MJ. Trauma-induced coagulopathy: the past, present, and future. J Thromb Haemost. 2019;17:852–862.
3. Hardaway RM, Drake DC. Prevention of “irreversible” hemorrhagic shock with fibrinolysin. Ann Surg. 1963;157:39–47.
4. Moore HB, Moore EE, Gonzalez E, Chapman MP, Chin TL, Silliman CC, Banerjee A, Sauaia A. Hyperfibrinolysis, physiologic fibrinolysis, and fibrinolysis shutdown: the spectrum of postinjury fibrinolysis and relevance to antifibrinolytic therapy. J Trauma Acute Care Surg. 2014;77(6):811–817; discussion 7.
5. Moore HB, Moore EE, Liras IN, Gonzalez E, Harvin JA, Holcomb JB, Sauaia A, Cotton BA. Acute fibrinolysis shutdown after injury occurs frequently and increases mortality: a multicenter evaluation of 2,540 severely injured patients. J Am Coll Surg. 2016;222(4):347–355.
6. Moore HB, Moore EE, Huebner BR, et al. Fibrinolysis shutdown is associated with a fivefold increase in mortality in trauma patients lacking hypersensitivity to tissue plasminogen activator. J Trauma Acute Care Surg. 2017;83(6):1014–1022.
7. Gall LS, Vulliamy P, Gillespie S, et al. The S100A10 pathway mediates an occult hyperfibrinolytic subtype in trauma patients. Ann Surg. 2018.
8. Cardenas JC, Wade CE, Cotton BA, George MJ, Holcomb JB, Schreiber MA, White NJ. Teg Lysis shutdown represents coagulopathy in bleeding trauma patients: analysis of the Proppr cohort. Shock. 2019;51:273–283.
9. Wolberg AS. Thrombin generation and fibrin clot structure. Blood Rev. 2007;21(3):131–142.
10. Wolberg AS, Campbell RA. Thrombin generation, fibrin clot formation and hemostasis. Transfus Apher Sci. 2008;38(1):15–23.
11. Wolberg AS. Plasma and cellular contributions to fibrin network formation, structure and stability. Haemophilia. 2010;16(Suppl 3):7–12.
12. Pan X, Gong YY, Martinelli I, Angelici L, Favero C, Bertazzi PA, Mannucci PM, Ariens RA, Routledge MN. Fibrin clot structure is affected by levels of particulate air pollution exposure in patients with venous thrombosis. Environ Int. 2016;92-93:70–76.
13. Barua RS, Sy F, Srikanth S, Huang G, Javed U, Buhari C, Margosan D, Aftab W, Ambrose JA. Acute cigarette smoke exposure reduces clot lysis—association between altered fibrin architecture and the response to t-PA. Thromb Res. 2010;126(5):426–430.
14. Meizoso JP, Dudaryk R, Mulder MB, Ray JJ, Karcutskie CA, Eidelson SA, Namias N, Schulman CI, Proctor KG. Increased risk of fibrinolysis shutdown among severely injured trauma patients receiving tranexamic acid. J Trauma Acute Care Surg. 2018;84(3):426–432.
15. Leeper CM, Neal MD, McKenna C, Sperry JL, Gaines BA. Abnormalities in fibrinolysis at the time of admission are associated with deep vein thrombosis, mortality, and disability in a pediatric trauma population. J Trauma Acute Care Surg. 2017;82(1):27–34.
16. Leeper CM, Neal MD, Billiar TR, Sperry JL, Gaines BA. Overresuscitation with plasma is associated with sustained fibrinolysis shutdown and death in pediatric traumatic brain injury. J Trauma Acute Care Surg. 2018;85(1):12–17.
17. Coleman JRM, Sauaia EE, Samuels S, et al. Sex dimorphisms in coagulation: implications for sex-specific clinical practice [unpublished report]. In press 2019.
18. Campbell RA, Overmyer KA, Selzman CH, Sheridan BC, Wolberg AS. Contributions of extravascular and intravascular cells to fibrin network formation, structure, and stability. Blood. 2009;114(23):4886–4896.
19. Quinn KP, Georgakoudi I. Rapid quantification of pixel-wise fiber orientation data in micrographs. J Biomed Opt. 2013;18(4):046003.
20. Nardi G, Agostini V, Rondinelli B, et al. Trauma-induced coagulopathy: impact of the early coagulation support protocol on blood product consumption, mortality and costs. Crit Care. 2015;19:83.
21. Pfeiler S, Massberg S, Engelmann B. Biological basis and pathological relevance of microvascular thrombosis. Thromb Res. 2014;133(Suppl 1):S35–S37.
22. Maegele M, Gu ZT, Huang QB, Yang H. Updated concepts on the pathophysiology and the clinical management of trauma hemorrhage and coagulopathy. Chin J Traumatol. 2017;20(3):125–132.
23. Gorlinger K, Dirkmann D, Solomon C, Hanke AA. Fast interpretation of thromboelastometry in non-cardiac surgery: reliability in patients with hypo-, normo-, and hypercoagulability. Br J Anaesth. 2013;110(2):222–230.
24. Coleman JR, Moore EE, Chapman MP, Banerjee A, Silliman CC, Ghasabyan A, Chandler J, Samuels JM, Sauaia A. Rapid TEG efficiently guides hemostatic resuscitation in trauma patients. Surgery. 2018;164(3):489–493.
25. Gonzalez E, Moore EE, Moore HB, et al. Goal-directed hemostatic resuscitation of trauma-induced coagulopathy: a pragmatic randomized clinical trial comparing a viscoelastic assay to conventional coagulation assays. Ann Surg. 2016;263(6):1051–1059.
26. Davies NA, Llwyd O, Brugniaux JV, et al. Effects of exercise intensity on clot microstructure and mechanical properties in healthy individuals. Thromb Res. 2016;143:130–136.
27. Gersh KC, Nagaswami C, Weisel JW. Fibrin network structure and clot mechanical properties are altered by incorporation of erythrocytes. Thromb Haemost. 2009;102(6):1169–1175.
28. Gidley GN, Holle LA, Burthem J, Bolton-Maggs PHB, Lin FC, Wolberg AS. Abnormal plasma clot formation and fibrinolysis reveal bleeding tendency in patients with partial factor XI deficiency. Blood Adv. 2018;2(10):1076–1088.
29. He S, Blomback M, Jacobsson Ekman G, Hedner U. The role of recombinant factor VIIa (FVIIa) in fibrin structure in the absence of FVIII/FIX. J Thromb Haemost. 2003;1(6):1215–1219.
30. Undas A, Zawilska K, Ciesla-Dul M, Lehmann-Kopydlowska A, Skubiszak A, Ciepluch K, Tracz W. Altered fibrin clot structure/function in patients with idiopathic venous thromboembolism and in their relatives. Blood. 2009;114(19):4272–4278.
31. Undas A, Podolec P, Zawilska K, et al. Altered fibrin clot structure/function in patients with cryptogenic ischemic stroke. Stroke. 2009;40(4):1499–1501.
32. Fatah K, Silveira A, Tornvall P, Karpe F, Blomback M, Hamsten A. Proneness to formation of tight and rigid fibrin gel structures in men with myocardial infarction at a young age. Thromb Haemost. 1996;76(4):535–540.
33. Undas A. Prothrombotic fibrin clot phenotype in patients with deep vein thrombosis and pulmonary embolism: a new risk factor for recurrence. Biomed Res Int. 2017;2017:8196256.
34. Mazur P, Sobczynski R, Zabczyk M, Babiarczyk P, Sadowski J, Undas A. Architecture of fibrin network inside thrombotic material obtained from the right atrium and pulmonary arteries: flow and location matter. J Thromb Thrombolysis. 2013;35(1):127–129.
35. Chapman MP, Moore EE, Moore HB, et al. Overwhelming tPA release, not PAI-1 degradation, is responsible for hyperfibrinolysis in severely injured trauma patients. J Trauma Acute Care Surg. 2016;80(1):16–23; discussion -5.
36. Van Hinsbergh VW, Sprengers ED, Kooistra T. Effect of thrombin on the production of plasminogen activators and PA inhibitor-1 by human foreskin microvascular endothelial cells. Thromb Haemost. 1987;57(2):148–153.
37. Moore HB, Moore EE, Huebner BR, Stettler GR, Nunns GR, Einersen PM, Silliman CC, Sauaia A. Tranexamic acid is associated with increased mortality in patients with physiological fibrinolysis. J Surg Res. 2017;220:438–443.
38. Myers SP, Kutcher ME, Rosengart MR, Sperry JL, Peitzman AB, Brown JB, Neal MD. Tranexamic acid administration is associated with an increased risk of posttraumatic venous thromboembolism. J Trauma Acute Care Surg. 2019;86(1):20–27.
Keywords:

Trauma; coagulopathy; fibrin; fibrinolysis; tissue plasminogen activator; thrombelastography

Supplemental Digital Content

© 2020 Lippincott Williams & Wilkins, Inc.