Freeze-dried plasma enhances clot formation and inhibits fibrinolysis in the presence of tissue plasminogen activator similar to pooled liquid plasma

Abstract

BACKGROUND

Systemic hyperfibrinolysis is an integral part of trauma-induced coagulopathy associated with uncontrolled bleeding. Recent data suggest that plasma-first resuscitation attenuates hyperfibrinolysis; however, the availability, transport, storage, and administration of plasma in austere environments remain challenging and have limited its use. Freeze-dried plasma (FDP) is a potential alternative due to ease of storage, longer shelf life, and efficient reconstitution. FDP potentially enhances clot formation and resists breakdown better than normal saline (NS) and albumin and similar to liquid plasma.

STUDY DESIGN AND METHODS

Healthy volunteers underwent citrated blood draw followed by 50% dilution with NS, albumin, pooled plasma (PP), or pooled freeze-dried plasma (pFDP). Citrated native and tissue plasminogen activator (t-PA)-challenge (75 ng/mL) thrombelastography were done. Proteins in PP, pFDP, and albumin were analyzed by mass spectroscopy.

RESULTS

pFDP and PP had superior clot-formation rates (angle) and clot strength (maximum amplitude) compared with NS and albumin in t-PA-challenge thrombelastographies (angle: pFDP, 67.9 degrees; PP, 67.8 degrees; NS, 40.6 degrees; albumin, 35.8 degrees; maximum amplitude: pFDP, 62.4 mm; PP, 63.5 mm; NS, 44.8 mm; albumin, 41.1 mm). NS and albumin dilution increased susceptibility to t-PA-induced hyperfibrinolysis compared with pFDP and PP (NS, 62.4%; albumin, 62.6%; PP, 8.5%; pFDP, 6.7%). pFDP was similar to PP in the attenuation of t-PA-induced fibrinolysis. Most proteins (97%) were conserved during the freeze-dry process, with higher levels in 12% of pFDP proteins compared with PP.

CONCLUSION

pFDP enhances clot formation and attenuates hyperfibrinolysis better than NS and albumin and is a potential alternative to plasma resuscitation in the treatment of hemorrhagic shock.

Systemic hyperfibrinolysis is an integral part of trauma-induced coagulopathy (TIC) and is associated with high mortality due to uncontrolled bleeding.^1,2 Recent data suggest that plasma-first resuscitation attenuates hyperfibrinolysis, whereas crystalloid exacerbates it.³ The transport, storage, availability, and administration of fresh-frozen plasma (FFP) in austere environments, such as the battlefield and in field civilian settings, remain challenging and have limited its use. These limitations also prevent this product from being available in rural settings, where it could have a profound effect in the early management of TIC. In addition, when severely injured patients arrive at a hospital, less than one-half of the hospitals in the United States have thawed plasma immediately available, leading to an additional delay during the standard thawing of FFP (30 min in a 37°C water bath).⁴ Providers are then faced with a dilemma in treating the hypotensive patient: 1) use crystalloid infusions, which are known to exacerbate the coagulopathy; or 2) delay lifesaving interventions as the patient continues to bleed until plasma is available.

Freeze-dried plasma (FDP) is a potential alternative to FFP. FDP is easy to store (room temperature), transport, and reconstitutes quickly (3 min with injectable water) and efficiently. These logistic advantages have led to a renewed interest in FDP, especially in the trauma setting. FDP was first introduced in World War II by the US Army; however, the practice of pooled lyophilized plasma was subsequently abandoned due to the high risk of pathogen transmission.⁵ Although Germany, France, and South Africa currently use an FDP product,^6–8 the US Food and Drug Administration (FDA) has yet to approve an FDP product for human use in the United States.

Hypoperfusion stimulates the release of prestored tissue plasminogen activator (t-PA) from endothelial cells.⁹ The mechanism of hyperfibrinolysis involves the release of excess t-PA, overwhelming the endogenous inhibitors, including plasminogen activator inhibitor-1 (PAI-1), resulting in the catalysis of plasminogen to plasmin and the breakdown of clots.¹⁰ In vitro and in vivo animal studies indicate that plasma can correct hyperfibrinolysis,^3,11 although the mechanism and proteins involved are unknown. Our group has previously described the use of a t-PA–challenge thrombelastogram (TEG),³ which replicates the increased t-PA seen in trauma patients and reliably increases the fibrinolysis profile on a whole blood (WB) sample. This in vitro assay allows us to determine the sensitivity and resistance of fibrinolysis with interventions. By using this assay, we hypothesize that normal saline (NS) and albumin dilution of WB results in deficient clotting and enhanced fibrinolysis in response to t-PA, whereas FDP dilution of WB results in enhanced clotting and attenuation of fibrinolysis similar to plasma.

MATERIALS AND METHODS

Materials

Human single-chain t-PA from Molecular Innovations was diluted in 5% bovine serum albumin in phosphate-buffered saline at 10 ng/μL, aliquoted, and immediately stored at −80°C. Type A-positive plasma was obtained from four healthy apheresis donors and pooled in a 3-L plasma-transfer bag. Pooled plasma was prepared by aliquoting pooled liquid plasma into 200-cc storage bags followed by freezing and storage at −80°C. FDP was prepared from 200-cc aliquots that were frozen and lyophilized using a Labconco FreeZone freeze dry system. Liquid plasma was subsequently thawed in a 37°C water bath for 30 minutes, aliquoted, and immediately stored at −80°C until the day of the experiment. FDP was stored in a 4°C refrigerator before reconstitution with 173 cc of US Pharmacopeia/European Pharmacopeia-grade water for injection (Rocky Mountain Bio). Reconstitution fluid was chosen based on previous work demonstrating decreased blood loss and improved coagulation properties of water compared with lactated Ringer’s (LR) and NS.¹² The volume of reconstitution was determined by calculating the water mass loss during lyophilization. Reconstituted FDP was aliquoted similar to pooled plasma (PP) and immediately stored at −80°C until the day of the experiment. Both liquid plasma and FDP underwent two freeze-thaw cycles, creating PP and pooled FDP (pFDP). Five percent human albumin was purchased from Grifols Biologicals, Inc. and stored at 1 to 4°C until use.

Healthy volunteer blood collection

Blood was collected from 10 healthy volunteers under a protocol approved by the Colorado Multiple Institutional Review Board. Volunteers were 60% male, ages 28 to 42 years, not pregnant, had no comorbidities, and were not taking anticoagulant or antiplatelet therapy.

Thrombelastography

All blood samples were collected in 3.3-mL buffered sodium citrate (3.2%) tubes (Vacutainer; Becton-Dickinson). To simulate a prolonged transport time, blood samples underwent a 50% dilution with NS, albumin, PP, or pFDP, as well as WB. Citrated native and t-PA–challenge (75 ng/mL) TEG were conducted for each dilution with the TEG 5000 Thrombelastograph Hemostasis Analyzer (Haemonetics). The methods for performing a t-PA– challenge TEG to quantify fibrinolysis sensitivity to t-PA were as previously described.³ TEG measurements included R-time, clot-formation rate (angle), maximum amplitude (MA), and percentage lysis at 30 minutes (LY30%).

QconCAT

QconCAT DNA constructs were synthesized de novo by Genscript and expressed in Escherichia coli auxotrophic strain BL21(l)DE3-LysA ArgA in minimal medium supplemented with 13C₆ arginine and 13C₆ lysine at 0.1 mg/mL (Sigma Aldrich). The cells were grown to mid-log phase (absorbance at 600 nm [A₆₀₀] = 0.6–0.8), at which point expression was induced by adding 1 mM isopropyl-D-1-thiogalactopyranoside. After 4 hours of growth at 37°C, the cells were harvested by centrifugation and processed as previously described with minor modifications.¹³ Briefly, the cells were lysed with the Bug Buster Protein Extraction Reagent (EMD Millipore). Inclusion bodies were suspended in 20 mM phosphate buffer, 6 M guanidinium chloride, 0.5 M NaCl, and 20 mM imidazole, pH 7.4. QconCAT proteins were desalted by affinity chromatography using a nickel-based resin. The purified QconCAT was desalted by three rounds of dialysis against 100 volumes of 10 mM ammonium bicarbonate, pH 8.5.

Sample preparation for liquid chromatography-tandem mass spectrometry

The samples were digested according to the filter-aided sample preparation protocol using a 10-kDa molecular weight cutoff filter.¹⁴ The samples were supplemented with known concentrations of 13C₆-labeled QconCAT proteins. In brief, samples were mixed in the filter unit with 8 M urea and 0.1 M ammonium biocarbonate, pH 8.0, and centrifuged at 14,000 × g for 15 minutes. The proteins were reduced with 10 mM dithiothreitol for 30 minutes at room temperature, centrifuged, and alkylated with 55 mM iodoacetamide for 30 minutes at room temperature in the dark. After centrifugation, samples were washed three times with urea solution and three times with 50 mM ammonium bicarbonate, pH 8.0. Protein digestion was carried out with sequencing grade modified trypsin (Promega) at 1/50 protease/protein (weight/weight) in 0.02% ProteaseMax (Promega) surfactant at 37°C overnight. Peptides were recovered from the filter using 0.1% formic acid. The resultant peptide mixture was analyzed in triplicate by liquid chromatography-selected reaction monitoring (SRM).

Liquid chromatography-tandem mass spectrometry analysis

The SRM approach was performed using the QTRAP 5500 triple quadrupole mass spectrometer (ABSciex) interfaced with a UPLC system (Ultimate 3000; Thermo). Instrument parameters and run conditions were previously described.¹⁵

Data analysis

Raw SRM data files were imported to Skyline v3.1 software for data processing. Transition quality, peak shape, and peak area boundaries were manually validated. Integrated peak areas were calculated by the software after Savitsky-Golay smoothing, and quantification was based on the ratio of the 12C peptide representing the endogenous sample to the corresponding 13C peptide from QconCAT proteins.

Statistics

Sample size was calculated using PASS 14 (NCSS, LLC) for a noninferiority paired study for the main outcome LY30%. Based on 160 healthy volunteers, we have established that the median LY30% for healthy volunteers on a t-PA–challenge (75 ng/mL) TEG is 8.3% with a standard deviation of 9.9% and that the 95th percentile is 27%. Therefore, a t-PA–challenge LY30% of 27% would be the cutoff to declare hyperfibrinolysis and seek therapy. A sample size of 10 pairs (10 FFP and 10 FDP samples) would allow a noninferiority margin of 5 percentage points in the t-PA–challenge LY30% paired differences, with 90% power and 95% confidence.

Statistical analysis was completed using SPSS version 22 (IBM Corporation). Comparisons of WB, NS, albumin, pFDP, and PP dilutions were done using a nonparametric paired Friedman test followed by a post hoc paired Wilcoxon rank test for differences between groups. Based on a Bonferroni correction, assuming four comparisons per hypothesis, significance was set at 0.0125. Mass spectroscopy data are reported as the mean of three separate runs with the same sample as descriptive characteristics.

RESULTS

pFDP and PP enhance clotting in a citrated native TEG

The clotting capacity of WB was contrasted to a 50% dilution with NS, albumin, PP, and pFDP. In a citrated native TEG, R-time (p < 0.0001), formation (angle; p < 0.0001), and MA (p < 0.0001) differed significantly across dilution groups (Table 1A). PP and pFDP stimulated rapid clot formation, with shorter R-times and higher angles, as well as stronger maximal clot strength (MA). Citrated native lysis profiles at 30 minutes (LY30%) were not significantly different across dilution groups (p = 0.018).

TABLE 1.

Thrombelastography properties

Property*	Median (interquartile range)
	R-time, min	Angle, degrees	MA, mm	LY30%
A. Citrated native TEG
Whole blood	11.4 (9.6–12.6)	37.3 (32.0–50.2)	55.3 (50.1–62.6)†	0.4 (0–2.1)
Normal saline	10.5 (8.8–13.5)	40.6 (37.6–47.7)	44.8 (39.8–47.8)	0.9 (0.6–1.4)
Albumin	11.5 (10.7–13.5)	35.8 (65.9–70.2)	41.1 (34.9–44.4)	1.2 (0.1–4.1)
PP	7.5 (6.9–7.9)‡	65.9 (64.3–70.2)‡	63.5 (62.6–66.4)†	1.6 (0.9–2.7)
pFDP	8.0 (7.4–8.7)‡	65.4 (63.3–68.2)‡	62.4 (58.8–66.4)†	2.4 (1.5–3.2)
B. t-PA–challenge TEG
Whole blood	9.8 (8.7–10.3)	39.6 (36.9–50.0)	51.3 (47.3–60.4)	14.3 (10.4–17.5)§
Normal saline	8.5 (6.6–9.2)	48.1 (37.5–50.9)	36.3 (29.6–41.0)	62.4 (50.4–67.5)
Albumin	7.8 (6.8–9.6)	44.2 (34.8–49.5)	35.0 (27.5–39.0)	62.6 (57.4–67.0)
PP	6.5 (5.8–6.9)‖	67.8 (64.9–68.6)¶	62.5 (58.1–63.2)#	8.5 (5.7–11.6)§
pFDP	7.1 (5.8–7.5)	67.9 (67.1–70.8)¶	59.8 (57.3–61.4)#	6.7 (5.0–9.8)§

^*Whole blood 50% dilution using individual resuscitation fluids with (top) and without (bottom) the addition of exogenous t-PA (75 ng/mL). (A) In a citrated native TEG, PP and pFDP demonstrate increased clot initiation (R-time) and formation (angle) compared with whole blood, albumin, and normal saline (†p = 0.005 compared with normal saline and albumin; ‡p 0.011 compared with whole blood, normal saline, and albumin). (B) In the presence of t-PA (75 ng/mL), PP and pFDP stimulate significantly increased rates of clot formation and resistance to breakdown compared with normal saline and albumin (§p = 0.005 compared with normal saline and albumin LY30%; ||p = 0.005 compared with whole blood and normal saline R-time; ¶p = 0.005 compared with whole blood, normal saline, and albumin angle; and #p < 0.01 compared with whole blood, normal saline, and albumin MA; n = 10).

Citrated native TEG values were similar between WB, NS, and albumin, with the exception of MA (Table 1A). WB MA was significantly stronger than NS (p = 0.005) and albumin (p = 0.005). However, NS and albumin dilutions retained the rates of clot initiation (R-time) and formation (angle), and the clot did not breakdown (LY30%) in the absence of t-PA.

pFDP and PP stimulated faster clot initiation and formation compared with WB, NS, and albumin (Table 1A). R-time was shorter in pFDP and PP compared with WB (p = 0.007), albumin (p = 0.005) and NS (p < 0.012). The rate of clot formation (angle) was also significantly faster in pFDP and PP compared with the rate in NS, albumin, and WB (p = 0.005 for all). Finally, pFDP and PP had superior maximum clot strength compared with albumin and NS (p = 0.005 for both). pFDP and PP also trended toward a higher maximum clot strength compared with WB, although the difference was not significant (p > 0.02). PP and pFDP were similar in all TEG properties analyzed, with near-identical TEG tracings.

PP and FDP enhance clotting and prevent clot breakdown in a t-PA–challenge TEG

R-time, angle, and MA all differed significantly across dilution groups in the presence of t-PA (p < 0.0001 for all) (Table 1B). Contrary to the citrated native, the addition of t-PA created significant differences in the fibrinolysis profiles (LY30%) across dilution groups (p < 0.0001).

t-PA–challenge TEG characteristics differed significantly between NS, albumin, and WB dilutions. Although the rates of clot initiation (R-time) and formation (angle) were similar, NS and albumin dilutions in combination with exogenous t-PA resulted in decreased clot strength (MA) and a significantly higher rate of clot breakdown (LY30%) compared with WB (Table 1B). The most significant change was in the fibrinolysis profile, with NS and albumin dilutions resulting in significant increases in fibrinolysis susceptibility compared with WB (p = 0.005 for both). Representative t-PA–challenge TEG tracings for WB, albumin, and NS are shown in Fig. 1.

Fig. 1.

Sample tracing from t-PA–challenge thrombelastography for whole blood (black line), normal saline (dark gray line), and albumin (light gray line). Albumin and normal saline tracings are similar to the “death diamond” and greatly increase t-PA–induced fibrinolysis as well as decreasing the maximal clot strength.

pFDP and PP significantly enhanced clot formation as well as preventing clot breakdown in the presence of t-PA. pFDP and PP stimulated faster rates of clot formation (angle) compared with WB, NS, and albumin (p = 0.005 for all). pFDP and PP also had significantly increased maximum clot strength (MA) compared with WB (p = 0.009), NS (p = 0.005), and albumin (p = 0.005) (Fig. 2). The LY30% was significantly lower in pFDP and PP compared with NS and albumin (p = 0.005 for all) (Fig. 2, Table 1B).

Fig. 2.

t-PA–challenge TEG property box plots. pFDP and PP shorten the time to initial clot formation (R-time) compared with WB. PP and pFDP increase the speed of clot formation (angle) compared with WB, NS, and albumin. WB, pFDP, and PP demonstrate significantly increased MA and reduced clot breakdown (LY30%) compared with NS and albumin. PP and pFDP demonstrate similar t-PA–challenge TEG properties (*p = 0.005 compared with WB R-time; ^†p = 0.005 compared with WB, NS, and albumin angle; ^‡p < 0.01 compared with NS and albumin MA; ^§p = 0.005 compared with NS and albumin LY30%).

Finally, pFDP and PP exhibited nearly identical clotting properties in both citrated native TEG and t-PA– challenge TEG (Table 1). PP and pFDP demonstrated similar R-time (p = 0.139), angle (p = 0.241), MA (p = 0.386), and LY30% (p = 0.508) with the addition of exogenous t-PA (Fig. 2). Representative TEG tracings for WB, PP, and pFDP are provided in Fig. 3A. Figure 3B compares NS, albumin, PP, and pFDP dilutions.

Fig. 3.

Sample t-PA–challenge TEG tracings. (Top) In WB (black line), PP (light gray line), and pFDP (dark gray line) tracings, both pFDP and PP are similar to WB, with neither exhibiting an increase in t-PA–induced fibrinolysis or a change in MA. (Bottom) PP, pFDP, NS, and albumin t-PA– challenge TEG tracings. PP and pFDP have similar TEG tracings. NS and albumin demonstrate slower clot formation (angle), weaker clot strength (MA), and faster clot breakdown (LY30%).

Protein content of PP and pFDP

QconCAT was used to quantify protein concentrations in PP, pFDP, and albumin. Figure 4 is a heat map of the 125 quantified proteins. Most of the proteins had similar levels in PP and pFDP, with 19 proteins (15%) that had changes greater than 1.5-fold between the two groups. Fifteen of those 19 proteins (79%) had higher levels in pFDP (Table 2). Of these 15 proteins, four have a known role in coagulation: Carboxypeptidase B2 (thrombin-activatable fibrinolysis inhibitor [TAFI], 1.7-fold increase), coagulation Factor XIII B chain (1.6-fold increase), Factor XI (1.6-fold increase), and protein C (1.5-fold increase). Only four proteins (3.2%) had changes greater than or equal to 1.5-fold from PP to pFDP (Table 2), with only antithrombin known to have a role in coagulation.

Fig. 4.

Heat map demonstrating quantitative protein levels of PP, pFDP, and albumin (QconCAT: gray, zero; blue, low; red, high). PP and pFDP demonstrate similar protein levels, whereas albumin has several proteins with significantly increased levels, and the majority of clotting proteins are decreased compared with plasma.

TABLE 2.

QconCAT analysis of coagulation protein levels in albumin, PP, and pFDP^*

Protein	Gene	PP, fmol/1 μg	pFDP fmol/1 μg	Albumin fmol/1 μg	PP/FDP	PP/FFP
pFDP > PP
Carboxypeptidase B2	CPB2	72.8	121.6	0.0	0.6	1.7
Coagulation Factor XIII B chain	F13B	71.9	115.8	0.0	0.6	1.6
Coagulation Factor XI	F11	72.8	114.3	0.0	0.6	1.6
Protein C	PROC	127.8	191.6	0.0	0.7	1.5
PP > pFDP
Antithrombin-III	SERPINC1	16.0	10.7	0.0	1.5	0.7
PP = pFDP
Alpha-2-antiplasmin	SERPINF2	32.8	25.9	40.3	1.3	0.8
Coagulation Factor III (thromboplastin, tissue factor)	F3	2.4	2.1	85.5	1.1	0.9
Coagulation Factor IX	F9	167.2	198.7	0.0	0.8	1.2
Coagulation Factor VII	F7	2.8	3.1	160.9	0.9	1.1
Coagulation Factor X	F10	79.1	116.9	0.0	0.7	1.5
Coagulation Factor XII	F12	11.0	9.4	17.0	1.2	0.9
Coagulation Factor XIIIA	F13A1	1.8	2.0	0.0	0.9	1.1
Fibrinogen alpha chain	FGA	272.3	274.8	1.4	1.0	1.0
Fibrinogen beta chain	FGB	245.1	258.5	0.0	0.9	1.1
Fibrinogen gamma chain	FGG	256.5	279.3	1.1	0.9	1.1
Heparin cofactor II	SERPIND1	34.5	28.2	0.0	1.2	0.8
Plasma protease C1 inhibitor (C1 esterase inhibitor)	SERPING1	7.9	5.8	3.2	1.4	0.7
Plasminogen light chain A	PLG	42.2	32.0	17.1	1.3	0.8
Plasminogen light chain B	PLG	40.0	33.5	22.1	1.2	0.8
Platelet glycoprotein V	GP5	2.3	2.1	84.0	1.1	0.9
Protein S	PROS1	88.5	130.4	0.0	0.7	1.4
Prothrombin	F2	42.8	37.0	0.0	1.2	0.9
Tetranectin (PLG binding)	CLEC3B	4.2	4.8	0.0	0.9	1.2
Vitronectin	VTN	102.4	79.2	0.0	1.3	0.8

^*Four coagulation proteins are identified at significantly higher concentrations (1.5-fold) in pFDP compared with PP, whereas one protein is significantly higher in PP. Twenty coagulation proteins contained similar levels of PP and pFDP.

Interestingly, albumin had similar protein levels in 47 (37.6%) of the 125 proteins compared with PP and pFDP (Fig. 4). Of these proteins, 34 had significantly higher levels (defined as ≥1.5-fold change) compared with PP. Four of these 34 proteins have a clear function in the coagulation system (Table 3), with the most significant changes in coagulation Factor VII (58.0-fold increase), platelet glycoprotein V (36.7-fold increase), Factor III (tissue factor, 35.9-fold increase), and Factor XII (1.5-fold increase). Three of the coagulation factors have levels similar to those in PP (Table 3). As anticipated, our QconCAT was able to measure many coagulation factors in PP and pFDP that were not present in albumin, including TAFI; coagulation Factors XI, X, XI, and XIIA; and prothrombin, fibrinogen, plasminogen, proteins C and S, antithrombin, alpha-2-antiplasmin, and vitronectin (Table 3).

TABLE 3.

Coagulation factors in albumin^*

Protein	Gene†	PP, fmol/1 μg	pFDp, fmol/1 μg	Albumin fmol/1 μg	Albumin/PP
Coagulation Factor VII	F7	2.8	3.1	160.9	58.0
Platelet glycoprotein V	GP5	2.3	2.1	84.0	36.7
Coagulation Factor III (thromboplastin, tissue factor)	F3	2.4	2.1	85.5	35.9
Coagulation Factor XII	F12	11.0	9.4	17.0	1.5
Plasma protease C1 inhibitor	SERPING1	71.8	62.3	95.8	1.3
Alpha-2-antiplasmin	SERPINF2	32.8	25.9	40.3	1.2
Alpha-2-antiplasmin	SERPINF2	24.3	20.4	23.0	0.9
Prothrombin	F2	48.7	43.8	37.8	0.8

^*Albumin contains four coagulation proteins at significantly higher concentrations (≥1.5-fold) than PP and three coagulation proteins at similar concentrations.

^†For gene names, see Table 2.

DISCUSSION

Renewed interest of FDP in trauma has prompted recent animal studies.^16–18 These data suggest pFDP is equivalent to PP in restoring clot formation.^16–18 However, animal models are consistently unable to produce systemic hyperfibrinolysis, which is the most deadly form of TIC. Therefore, the role of FDP in systemic hyperfibrinolysis remains undefined. The t-PA–challenge TEG approximates systemic hyperfibrinolysis and produces a reliable in vitro assay.

With this assay, albumin and NS potentiate the effects of t-PA in WB, resulting in significantly altered clot properties (Table 1B, Fig. 2). In contrast, pFDP and PP preserve these clot properties (Fig. 2). In the face of hypotension and increased t-PA levels in the severely injured trauma patient, FDP may be a superior resuscitation fluid compared with NS or albumin in the correction of TIC and is similar to liquid plasma. The perceived trend toward improved clot properties (decreased R-time and LY30%; increased angle and MA) of plasma (PP and pFDP) compared with WB may be due to the increased plasma protein concentration observed with the exclusion of red blood cells during dilution of WB with plasma (Fig. 2). These differences are not as significant in the citrated native TEG (Table 1A).

Increased interest in plasma resuscitation stems from the Prospective, Observational, Multicenter, Major Trauma Transfusion (PROMMTT) study, which demonstrated a clear mortality benefit with early plasma resuscitation, whereas it was more difficult to extrapolate any benefits from early platelet transfusions, which were uncommon.¹⁹ Cotton and coworkers² previously reported an increased risk of hyperfibrinolysis in trauma patients with each liter of prehospital, crystalloid fluid. This same group demonstrated in vitro that a 15% dilution of WB with either crystalloids, albumin, or plasma with t-PA resulted in increased fibrinolysis compared with WB; however, the concentration of t-PA was three times that used in our experiment.²⁰ We previously examined saline and plasma dilutions of WB and observed similar exacerbation of hyperfibrinolysis in the presence of t-PA with saline while platelet poor plasma attenuated this response.³ That study used fresh plasma, whereas our study demonstrates conservation of the ability to inhibit fibrinolysis after frozen storage as well as after the freeze-dry process.

A previous study examining the French FDP product with an in vitro TEG employed a 30% dilution of WB with LR followed by an additional 30% dilution in FDP, FFP, or LR.²¹ Similar to our studies, those investigators observed a decrease in clot strength (MA) in the LR group and the preservation of MA in the FDP and FFP groups. Those authors also reported a slight increase in clot initiation (R-time) and formation (angle) of FFP and FDP compared with LR. We also observed increases in the angle and the reported R-time in the current study, although the differences were not clinically significant (8.8 vs. 7.5 min). Finally, the French group was unable to report a change in the lysis profile, because they did not add t-PA to their samples.²¹

Previous comparisons of liquid plasma and FDP suggest equivalent coagulation factor levels after the freeze-dry process.^7,17 To date, the majority of research has focused on coagulation factors, demonstrating adequate levels in FDP compared with FFP. Although these proteins are important in clot formation, and relative levels are significant in transfusing hemophiliacs, the major plasma proteins involved in the correction of TIC (and specifically hyperfibrinolysis) are unknown and likely involve additional proteins. A previous proteomic analysis of FDP indicated a reduction in 38 of the 600 plasma proteins detected, including alpha-1-antiplasmin levels as well as plasmin-inhibitor activity,²² whereas subsequent studies produced mixed results regarding plasmin-inhibitor activity.^7,21,23 We observed changes greater than 1.5-fold in 19 of the 125 proteins analyzed, and 15 of those proteins had higher levels in the pFDP sample (Table 2). Four of those 15 proteins are known to have a role in coagulation. Increased TAFI levels in pFDP did not impact the rate of clot breakdown. Factor XIIIB is the inhibitory portion of Factor XIII heterodimer, dissociating from Factor XIIIA and leading to XIIIA activation and linking of fibrin. Factor XIIIA levels were similar in the plasma groups. High free Factor XIIIB levels in pFDP, although recently shown to bind fibrinogen,²⁴ did not impact the TEG angle. Increased levels of Factor XI and protein C in FDP did not impact the rate of clot initiation (R-time) or formation (angle) in the TEG tracings, and their functions are known to oppose each other. The only coagulation protein that was higher in PP was antithrombin, and this did not impact the R-time. Although there were significant differences in several of the coagulation proteins, the overwhelming majority was conserved during the freeze-dry process (Table 2), resulting in near-identical TEG tracings (Fig. 3).

Human albumin is FDA-approved for the restoration and maintenance of circulating blood volume.²⁵ Albumin is prepared from human PP by a cold ethanol fractionation process followed by filtration, with the final product containing more than 96% albumin protein. Although our albumin sample contained nearly 40% of the proteins measured as well as several key coagulation proteins, it was unable to influence clotting and had a coagulation profile similar to that of NS (Table 3, Fig. 4). Albumin may be stored at room temperature or refrigerated, leading to a decrease in the activity of the coagulation proteins, likely contributing to the inability to affect clotting. In addition, albumin lacked many of the necessary coagulation factors, including Factors I, II, IX, X, XI, XIII, and TAFI, present in plasma, also contributing to clotting insufficiencies (Table 3). Albumin received significant interest as a volume expander in trauma and intensive care unit patients; however, the Saline and Albumin Fluid Resuscitation (SAFE) trials demonstrated no significant benefit of albumin over saline in critically ill patients,²⁶ with increased mortality in those who had traumatic brain injury.²⁷

The major limitation of this study is the in vitro environment in which the experiment was carried out. Unfortunately, animal models of both severe shock and extensive tissue injury have been unable to produce a reliable hyperfibrinolysis model, which would be the ideal environment for studying the effects of FDP on this lethal coagulopathy phenotype. It is also worth noting that PP/pFDP samples underwent two freeze-thaw cycles before the experiment. Coagulation factor activity levels, undoubtedly, are lower after a second freeze-thaw cycle, although previous studies have demonstrated the retention of normal factor activity levels as well as normal conventional coagulation tests (prothrombin time, partial thromboplastin time) after a second freeze-thaw cycle.^28–30 Another potential limitation is the relatively high concentration of t-PA used to provoke coagulation changes. Plasma from trauma patients contains 30 to 50 ng/mL of total t-PA, whereas the activity levels are much lower, in the range from 1 to 6 IU/mL.¹⁰ Although our exogenous concentration of 75 ng/mL t-PA (total, not activity) is higher than both of these published values, the local tissue concentration of t-PA at the site of injury is likely much higher than that in the circulating plasma. In addition, our group has observed that hypercoagulable trauma patients (fibrinolysis shutdown) are resistant to changes in their fibrinolysis profile, with the addition of this t-PA concentration demonstrating relevance.³¹

The in vitro results from this experiment demonstrate that pFDP is not inferior to PP and is superior to albumin and NS in the inhibition of t-PA–induced clot changes. Although additional in vivo studies are needed to validate these findings, FDP has known logistical advantages (shelf-life, reconstitution time, storage) over liquid plasma and may be the answer to the challenges of plasma-first resuscitation.

ABBREVIATIONS

FDP

freeze-dried plasma

LR

lactated Ringer’s

LY30%

percentage lysis at 30 minutes

MA

maximum amplitude

NS

normal saline

pFDP

pooled freeze-dried plasma

PP

pooled plasma

SRM

selected reaction monitoring

TAFI

thrombin-activatable fibrinolysis inhibitor

TEG

thrombelastogram

TIC

trauma-induced coagulopathy

t-PA

tissue plasminogen activator

WB

whole blood.