Martin Sillesen

Assessment of coagulopathy, endothelial injury, and inflammation after traumatic brain injury and hemorrhage in a porcine model

BACKGROUND Traumatic brain injury (TBI) and hemorrhagic shock (HS) can be associated with coagulopathy and inflammation, but the mechanisms are poorly understood. We hypothesized that a combination of TBI and HS would disturb coagulation, damage the endothelium, and activate inflammatory and complement systems. METHODS A total of 33 swine were allocated to either TBI + HS (n = 27, TBI and volume-controlled 40% blood loss) or controls (n = 6, anesthesia and instrumentation). TBI + HS animals were left hypotensive (mean arterial pressure, 30–35 mm Hg) for 2 hours. Blood samples were drawn at baseline, 3 minutes and 15 minutes after injury, as well as following 2 hours of hypotension. Markers of coagulation, anticoagulation, endothelial activation/glycocalyx shedding, inflammation, complement, and sympathoadrenal function were measured. RESULTS The TBI + HS group demonstrated an immediate (3 minutes after injury) activation of coagulation (prothrombin fragment 1 + 2, 289 ng/mL vs. 232 ng/mL, p = 0.03) and complement (C5a, 2.83 ng/mL vs. 2.05 ng/mL, p = 0.05). Shedding of the endothelial glycocalyx (syndecan 1) was evident 15 minutes after injury (851.0 ng/ml vs. 715.5 ng/ml, p = 0.03) while inflammation (tumor necrosis factor &agr; [TNF-&agr;], 81.1 pg/mL vs. 50.8 pg/mL, p = 0.03) and activation of the protein C system (activated protein C, 56.7 ng/mL vs. 26.1 ng/mL, p = 0.01) were evident following the 2-hour hypotension phase. CONCLUSION The combination of TBI and shock results in an immediate activation of coagulation and complement systems with subsequent endothelial shedding, protein C activation, and inflammation.

Pharmacologic resuscitation for hemorrhagic shock combined with traumatic brain injury.

BACKGROUND We have previously demonstrated that valproic acid (VPA), a histone deacetylase inhibitor, can improve survival after hemorrhagic shock (HS), protect neurons from hypoxia-induced apoptosis, and attenuate the inflammatory response. We have also shown that administration of 6% hetastarch (Hextend [Hex]) after traumatic brain injury (TBI) decreases brain swelling, without affecting size of the lesion. This study was performed to determine whether addition of VPA to Hex would decrease the lesion size in a clinically relevant large animal model of TBI + HS. METHODS Yorkshire swine (42–50 kg) were instrumented to measure hemodynamic parameters, intracranial pressure, and brain tissue oxygenation. A custom-designed, computer-controlled cortical impact device was used to create a TBI through a 20-mm craniotomy: 15-mm cylindrical tip impactor at 4-m/s velocity, 100-millisecond dwell time, and 12-mm penetration depth. Volume-controlled hemorrhage was started (40% blood volume) concurrent with the TBI. After 2 hours of shock, animals were randomized to one of three resuscitation groups (n = 7 per group) as follows: (1) isotonic sodium chloride solution; (2) 6% hetastarch, Hex; and (3) Hex and VPA 300 mg/kg (Hex + VPA). Volumes of Hex matched the shed blood, whereas that of the isotonic sodium chloride solution was three times the volume. VPA treatment was started after an hour of shock. After 6 hours of postresuscitation monitoring, brains were sectioned into 5-mm slices and stained with 2, 3, 5-Triphenyltetrazolium chloride to quantify the lesion size (mm3) and brain swelling (percent change compared with uninjured side). Levels of acetylated histone H3 were determined to quantify acetylation, and myeloperoxidase and interleukine-1&bgr; (IL-1&bgr;) levels were measured as markers of brain inflammation. RESULTS Combination of 40% blood loss with cortical impact and a period of shock (2 hours) and resuscitation resulted in a highly reproducible brain injury. Lesion size and brain swelling in the Hex + VPA group (1,989 [156.8] mm3, and 19% [1.6%], respectively) were significantly smaller than the isotonic sodium chloride solution group (3,335 [287.9] mm3 and 36% [2.2%], respectively). Hex alone treatment significantly decreased the swelling (27% [1.6%]) without reducing the lesion size. The number of CD11b-positive cells as well as myeloperoxidase and IL-1 levels in the brains were significantly reduced by the VPA treatment. CONCLUSION In a combined HS and TBI model, treatment with artificial colloid (Hex) improves hemodynamic parameters and reduces swelling, without affecting the actual size of the brain lesion. Addition of VPA effectively reduces both the size of brain lesion and associated swelling by attenuating the inflammatory response.

Platelet activation and dysfunction in a large-animal model of traumatic brain injury and hemorrhage

BACKGROUND: Traumatic brain injury (TBI) and hemorrhage are the leading causes of trauma‐related mortality. Both TBI and hemorrhage are associated with coagulation disturbances, including platelet dysfunction. We hypothesized that platelet dysfunction could be detected early after injury, and that this dysfunction would be associated with early activation, as measured by circulating levels of platelet activation markers. METHODS: A total of 33 swine were allocated to TBI and hypotension (n = 27, TBI and volume‐controlled 40% blood loss) or controls (n = 6, anesthesia and instrumentation only). Animals in the TBI/Hemorrhage group were left hypotensive, defined as mean arterial pressure of 35 mm Hg, for 2 hours. Blood samples were drawn at baseline and 3 minutes and 15 minutes following injury as well as following 2 hours of shock. Samples were analyzed for platelet aggregation using impedance aggregometry with agonists collagen, arachidonic acid, and adenosine diphosphate (ADP) and thromboelastography (TEG) and circulating levels of platelet activation markers transforming growth factor‐[beta] (TGF‐[beta]), CD40 ligand, and sP‐selectin. RESULTS: Platelet ADP aggregation was significantly lower in the TBI/Hemorrhage group when compared with the control group 15 minutes following injury (62.4 vs. 80.4 U, p = 0.03) as well as following 2 hours of hypotension (59.9 vs. 73.5 U, p < 0.01). The latter was associated with lower TEG measured clot strength (TEG‐MA, 74.1 vs. 79.4 mm, p = 0.05). No difference in collagen or arachidonic acid aggregation was observed. TGF‐[beta] levels were significantly higher in the TBI/Hemorrhage group following 2 hours of hypotension (1,764 vs. 1,252 pg/mL, p = 0.01). No differences in CD40 ligand or sP‐selectin levels were observed. CONCLUSION: In this combined model of TBI and hemorrhage, a significantly lower ADP‐induced platelet aggregation was detected 15 minutes following injury that was further aggravated during the 2‐hour shock period. This dysfunction was associated with an increase in platelet activation marker TGF‐[beta].

Referral of patients with ST-segment elevation acute myocardial infarction directly to the catheterization suite based on prehospital teletransmission of 12-lead electrocardiogram.

BACKGROUND Time from symptom onset to reperfusion is essential in patients with ST-segment elevation acute myocardial infarction. Prior studies have indicated that prehospital 12-lead electrocardiogram (ECG) transmission can reduce time to reperfusion. PURPOSE Determine 12-lead ECG transmission success rates, and time saved by referring patients directly to primary percutaneous coronary intervention (pPCI) bypassing local hospitals and emergency departments. METHODS Prehospital 12-lead ECG was recorded in patients with symptoms suggesting acute coronary syndrome during a 1-year pilot phase and transmitted to the attending cardiologists mobile phone. Transmission success rates were determined, and prehospital and hospital delays were recorded and compared to historic controls. RESULTS Transmission was attempted in 152 patients and was successful in 89%. Twenty-seven patients were referred directly for pPCI. Median hospital arrival to pPCI was 22 vs 94 minutes in the control group (P < .01). CONCLUSIONS Transmission of prehospital ECG is technically feasible and reduces time to pPCI in ST-segment elevation acute myocardial infarction patients.

Early treatment with lyophilized plasma protects the brain in a large animal model of combined traumatic brain injury and hemorrhagic shock

BACKGROUND Combination of traumatic brain injury (TBI) and hemorrhagic shock (HS) can result in significant morbidity and mortality. We have previously shown that early administration of fresh frozen plasma (FFP) in a large animal model of TBI and HS reduces the size of the brain lesion as well as the associated edema. However, FFP is a perishable product that is not well suited for use in the austere prehospital settings. In this study, we tested whether a shelf-stable, low-volume, lyophilized plasma (LSP) product was as effective as FFP. METHODS Yorkshire swine (42–50 kg) were instrumented to measure hemodynamic parameters, intracranial pressure, and brain tissue oxygenation. A prototype, computerized, cortical impact device was used to create TBI through a 20-mm craniotomy: 15-mm cylindrical tip impactor at 4 m/s velocity, 100-millisecond dwell time, and 12-mm penetration depth. Volume-controlled hemorrhage was induced (40–45% total blood volume) concurrent with the TBI. After 2 hours of shock, animals were treated with (1) normal saline (NS, n = 5), (2) FFP (n = 5), and (3) LSP (n = 5). The volume of FFP and LSP matched the shed blood volume, whereas NS was 3 times the volume. Six hours after resuscitation, brains were sectioned and stained with TTC (2, 3, 5-Triphenyltetrazolium chloride), and lesion size (mm3) and swelling (percent change in volume compared with the contralateral, uninjured side) were measured. RESULTS This protocol resulted in a highly reproducible brain injury, with clinically relevant changes in blood pressure, cardiac output, tissue hypoperfusion, intracranial pressure, and brain tissue oxygenation. Compared with NS, treatment with LSP significantly (p < 0.05) decreased brain lesion size and swelling (51% and 54%, respectively). CONCLUSION In a clinically realistic combined TBI + HS model, early administration of plasma products decreases brain lesion size and edema. LSP is as effective as FFP, while offering many logistic advantages. LEVEL OF EVIDENCE Therapeutic study, level V.

Fresh frozen plasma resuscitation attenuates platelet dysfunction compared with normal saline in a large animal model of multisystem trauma.

BACKGROUND Platelet dysfunction following trauma has been identified as an independent predictor of mortality. We hypothesized that fresh frozen plasma (FFP) resuscitation would attenuate platelet dysfunction compared with 0.9% normal saline (NS). METHODS Twelve swine were subjected to multisystem trauma (traumatic brain injury, liver injury, rib fracture, and soft tissue injury) with hemorrhagic shock (40% of estimated blood volume). Animals were left in shock (mean arterial pressure, 30–35 mm Hg) for 2 hours followed by resuscitation with three times shed volume NS (n = 6) or one times volume FFP (n = 6) and monitored for 6 hours. Platelet function was assessed by adenosine diphosphate (ADP)–induced platelet aggregation at baseline, after 2 hours of shock following resuscitation, and 6 hours after resuscitation. Fibrinogen levels and markers of platelet activation (transforming growth factor &bgr; [TGF-&bgr;], sP-Selectin, and CD40L) as well as endothelial injury (intercellular adhesion molecule 1 [ICAM-1], vascular cell adhesion molecule 1 [VCAM-1]) were also assayed. Thromboelastography was used to measure clotting activity. RESULTS ADP-induced platelet aggregation was significantly higher in the FFP group (46.3 U vs. 25.5 U, p < 0.01) following resuscitation. This was associated with higher fibrinogen levels (202 mg/dL vs. 80 mg/dL, p < 0.01) but lower endothelial activation (VCAM-1, 1.25 ng/mL vs. 3.87 ng/mL, p = 0.05). Other markers did not differ. After 6 hours of observation, ADP-induced platelet aggregation remained higher in the FFP group (53.8 U vs. 37.0 U, p = 0.03) as was fibrinogen levels (229 mg/dL vs. 153 mg/dL, p < 0.01). Endothelial activation was lower (ICAM-1, 21.0 ng/mL vs. 24.4 ng/mL, p = 0.05), whereas TGF-&bgr; levels were higher (2,138 pg/mL vs. 1,802 pg/mL, p = 0.03) in the FFP group. Other markers did not differ. Thromboelastography revealed increased clot strength in the FFP group at both postresuscitation time points. CONCLUSION Resuscitation with FFP resulted in an immediate and sustained improvement in platelet function and clot strength compared with high-volume NS resuscitation. This was associated with an increase in fibrinogen levels and an attenuation of endothelial activation.

Synergistic effects of fresh frozen plasma and valproic acid treatment in a combined model of traumatic brain injury and hemorrhagic shock

INTRODUCTION Traumatic brain injury (TBI) and hemorrhagic shock (HS) are major causes of trauma-related deaths and are especially lethal as a combined insult. Previously, we showed that early administration of fresh frozen plasma (FFP) decreased the size of the brain lesion and associated swelling in a swine model of combined TBI+HS. We have also shown separately that addition of valproic acid (VPA) to the resuscitation protocol attenuates inflammatory markers in the brain as well as the degree of TBI. The current study was performed to determine whether a combined FFP+VPA treatment strategy would exert a synergistic effect. METHODS Yorkshire swine (42-50 kg) were instrumented to measure hemodynamic parameters, intracranial pressure, and brain tissue oxygenation. TBI was created through a 20-mm craniotomy using a computer-controlled cortical impactor: 15-mm cylindrical tip impactor at 4 m/s velocity, 100 ms dwell time, and 12-mm penetration depth. The TBI was synchronized with the initiation of volume-controlled hemorrhage (40 ± 5% of total blood volume). After a 2-hour period of shock, animals were randomized to 1 of 3 resuscitation groups (n = 5 per group): (1) 0.9% saline (NS); (2) FFP; and (3) FFP and VPA 300 mg/kg (FFP+VPA). The resuscitative volume for FFP was equivalent to the shed blood, whereas NS was 3 times this volume. VPA treatment was started 1 hour after hemorrhage. Animals were monitored for 6 hours post-resuscitation. At this time the brains were harvested, sectioned into 5-mm slices, and stained with 2,3,5-triphenyltetrazolium chloride to quantify the lesion size (mm(3)) and brain swelling (percent change compared with the uninjured side). RESULTS The combined TBI+HS model resulted in a highly reproducible brain injury. Lesion size and brain swelling (mean value ± standard error of the mean) in the FFP+VPA group (1,459 ± 218 mm(3) and 13 ± 1%, respectively) were less than the NS group (3,285 ± 131 mm(3) [P < .001] and 37 ± 2% [P < .001], respectively), and the FFP alone group (2,160 ± 203 mm(3) [P < .05] and 22 ± 1% [P < .001], respectively). CONCLUSION In a large animal model of TBI+HS, early treatment with a combination of FFP and VPA decreases the size of brain lesion and the associated swelling.

Detection of Extracellular Genomic DNA Scaffold in Human Thrombus: Implications for the Use of Deoxyribonuclease Enzymes in Thrombolysis

PURPOSE Mechanisms underlying transition of a thrombus susceptible to tissue plasminogen activator (TPA) fibrinolysis to one that is resistant is unclear. Demonstration of a new possible thrombus scaffold may open new avenues of research in thrombolysis and may provide mechanistic insight into thrombus remodeling. MATERIALS AND METHODS Ten human thrombus samples were collected during cases of thrombectomy and open surgical repair of abdominal aortic aneurysms (five samples < 3 d old and five samples > 1 y old). Additionally, an acute murine hindlimb ischemia model was created to evaluate thrombus samples in mice. Human sections were immunostained for the H2A/H2B/DNA complex, myeloperoxidase, fibrinogen, and von Willebrand factor. Mouse sections were immunostained with the H2A antibody. All samples were further evaluated after hematoxylin and eosin and Masson trichrome staining. RESULTS An extensive network of extracellular histone/DNA complex was demonstrated in the matrix of human ex vivo thrombus. This network is present throughout the highly cellular acute thrombus. However, in chronic thrombi, detection of the histone/DNA network was predominantly in regions of low collagen content and high cell density, which were mostly near the lumen. These regions of high cell density contained neutrophils and monocytes. Similarly, sections from the acute murine hindlimb ischemia model also exhibited extensive immunoreactivity to the histone antibody in the extracellular space within murine thrombi. CONCLUSIONS Extensive detection of genomic DNA associated with histones in the extracellular matrix of human and mouse thrombi suggest the presence of a new thrombus-associated scaffold.

Pharmacologic modulation of cerebral metabolic derangement and excitotoxicity in a porcine model of traumatic brain injury and hemorrhagic shock

BACKGROUND Cerebral metabolic derangement and excitotoxicity play critical roles in the evolution of traumatic brain injury (TBI). We have shown previously that treatment with large doses of valproic acid (VPA) decreases the size of brain lesion. The goal of this experiment was to determine whether this effect was owing to metabolic modulation. METHODS Yorkshire swine (n = 9) underwent a protocol of computer-controlled TBI and 40% hemorrhage and were resuscitated randomly with either fresh frozen plasma equal to the volume of shed blood (FFP; n = 4) or VPA (300 mg/kg) and FFP (FFP+VPA; n = 5). Hemodynamics, brain oxygenation, and blood glucose were monitored continuously for 6 hours after resuscitation. Cerebral microdialysis was used to measure glucose, lactate, pyruvate, glutamate, and glycerol levels at baseline, 1 and 2 hours post-shock, post-resuscitation (PR), and at 2, 4, and 6 hours PR. Brain samples from the injured side were then separated into mitochondrial and cytosolic fractions, and activity of pyruvate dehydrogenase complex (PDH) was measured using a dipstick assay kit. RESULTS At baseline, there was no difference in brain lactate, pyruvate, glycerol, and glutamate concentrations between the groups. At all time points, there were no differences between the groups in brain oxygenation, cerebral perfusion pressure, or blood and brain glucose concentrations. After VPA infusion (PR time point), however, there was sustained decrease in lactate (0.91 ± 0.47 vs 2.54 ± 0.59 mmol/L; P < .01) and pyruvate (12.80 ± 4.89 vs 46.25 ± 9.22; P < .001) concentrations compared with the FFP alone group, implying superior glucose utilization for ATP production. There was also a decrease in concentrations of glutamate (6.64 ± 3.68 vs 42.25 ± 27.07 mmol/L; P = .02) and glycerol (19.20 ± 6.76 vs 69.75 ± 30.07 mmol/L; P = .01), in the FFP+VPA group, signifying lesser degree of excitotoxicity and brain damage, respectively. Brain PDH activity was greater in the mitochondrial fractions (5,984 ± 504 adjusted volume intensity [INT] × mm(2) vs 4,332 ± 1,055 INT × mm(2); P = .04) and lower in cytosolic fractions in the FFP+VPA group (1,597 ± 1,395 vs 4,026 ± 1,067 INT × mm(2); P = .03), indicating better mitochondrial membrane function and enhanced mitochondrial PDH retention. CONCLUSION VPA treatment attenuates perturbation of post-traumatic cerebral metabolism by mitigating mitochondrial dysfunction, and decreases glutamate-mediated excitotoxic damage. These properties could explain its effectiveness in decreasing lesion size and post-traumatic cerebral edema.

Statins improve the resolution of established murine venous thrombosis: Reductions in thrombus burden and vein wall scarring

Despite anticoagulation therapy, up to one-half of patients with deep vein thrombosis (DVT) will develop the post-thrombotic syndrome (PTS). Improving the long-term outcome of DVT patients at risk for PTS will therefore require new approaches. Here we investigate the effects of statins—lipid-lowering agents with anti-thrombotic and anti-inflammatory properties—in decreasing thrombus burden and decreasing vein wall injury, mediators of PTS, in established murine stasis and non-stasis chemical-induced venous thrombosis (N = 282 mice). Treatment of mice with daily atorvastatin or rosuvastatin significantly reduced stasis venous thrombus burden by 25% without affecting lipid levels, blood coagulation parameters, or blood cell counts. Statin-driven reductions in VT burden (thrombus mass for stasis thrombi, intravital microscopy thrombus area for non-stasis thrombi) compared similarly to the therapeutic anticoagulant effects of low molecular weight heparin. Blood from statin-treated mice showed significant reductions in platelet aggregation and clot stability. Statins additionally reduced thrombus plasminogen activator inhibitor-1 (PAI-1), tissue factor, neutrophils, myeloperoxidase, neutrophil extracellular traps (NETs), and macrophages, and these effects were most notable in the earlier timepoints after DVT formation. In addition, statins reduced DVT-induced vein wall scarring by 50% durably up to day 21 in stasis VT, as shown by polarized light microscopy of picrosirius red-stained vein wall collagen. The overall results demonstrate that statins improve VT resolution via profibrinolytic, anticoagulant, antiplatelet, and anti-vein wall scarring effects. Statins may therefore offer a new pharmacotherapeutic approach to improve DVT resolution and to reduce the post-thrombotic syndrome, particularly in subjects who are ineligible for anticoagulation therapy.