Vagus nerve stimulation improves coagulopathy in hemorrhagic shock: a thromboelastometric animal model study

Twenty-four (n = 24) animals were randomly divided in 3 groups (n = 8 per group) according to resuscitation regimen:

Animals were anesthetized with ketamine 60 mg/kg and xylazine 15 mg/kg (Fort Dodge Animal Health, Fort Dodge, IA) administered by intraperitoneal injection. Additional intravenous doses were given if needed. The right internal jugular vein and the right carotid artery were cannulated with 21G peripheral intravenous catheters previously filled with LR (Johnson & Johnson, Sao Jose dos Campos, Brazil). The vagus nerves were carefully dissected and tagged with a loop of 3–0 silk suture (Polysuture®, Sao Sebastiao do Paraiso, Minas Gerais, Brazil). Mean arterial pressure (MAP) was monitored continuously on the right carotid artery (Pro-Paq Protocol Systems, Beaverton, OR). A blood sample (1 ml) was obtained from the right carotid artery for baseline thromboelastometry (ROTEM® Coagulation Analyzer, Pentapharm, Munich, Germany). The sample was immediately transferred to a tube containing 3.2% sodium citrate as anticoagulant (MiniCollect®, Vacuette, Monroe, NC). Thromboelastometry was performed on temperature-corrected blood samples using NATEM® (Non-Activated TEM) and Star-tem® reagent to recalcify citrated blood (Ref. Number 503–01). Thromboelastometry parameters were calculated with the coagulation dynamics evaluation software (DyCoDerivAn; Avordusol, Rissov, Denmark). The following parameters were assessed: clotting time (CT), clot formation time (CFT), alpha angle (α), maximum clot firmness (MCF); at baseline and at the end of the experiment. A different blood sample was placed in a centrifuge tube (Eppendorf do Brasil Ltda, Sao Paulo, Brazil) with ethylenediaminetetra-acetic acid as anticoagulant for cytokine assay (IL-1β and IL-10). In brief, plasma was separated by centrifugation at 1600 × g for 15 minutes at 4°C. Samples were analyzed at a 1:5 dilution in potassium phosphate buffer. Enzyme-linked immunobsorbent assay plates (ELISA) (Neogen® Corporation, Indaiatuba, Brazil) were coated with sheep anti-rat IL-1β and IL-10 polyclonal antibodies (1–2 μg/ml) overnight. The plates were washed and blocked with 1% bovine serum albumin, and incubated overnight with samples of recombinant rat cytokine. Biotinylated polyclonal antibodies were used at a 1:1000 to 1:2000 dilution. The plates were washed and incubated with avidin-horseradish peroxidase (Dako North America Ltd., Carpinteria, CA) for 15 minutes. After a subsequent wash they were incubated with o-phenylenediamine and H₂O₂ for another 15 minutes. The reaction was stopped with sulfuric acid (150 μl/1.0 M); optical densities at 490 nm were determined. Cytokine assays were performed using blood samples obtained at baseline and at the end of the experiment.

Total blood volume was calculated as (TBV = 0.06 × Weight (g) + 0.77). Forty percent of the estimated TBV was taken from the right carotid artery. Thereafter, MAP was maintained at 45% of baseline level (MAP × 0.45) for 15 minutes (Figure 1). No fluids or blood products were given during this period to simulate Emergency Medical Services (EMS) arrival time. Resuscitation began with a bolus of LR (20 ml/Kg) infused over 3 minutes through the right internal jugular vein. Additional volumes of LR were infused to maintain MAP at baseline ± 5 mmHg. The resuscitation phase lasted for 45 minutes. Vagus nerves of group 3 animals were divided bilaterally. Subsequently, a total of seven stimuli were applied to the distal stump of the nerve bilaterally, during resuscitation, using a Grass Nerve Stimulator (Grass Technologies - Warwick, RI). Stimuli were 3.5 mA/5Hz and lasted for 30 seconds. A five minute interval was allowed between each stimulation (total time 35 minutes). Animals were euthanized with an anesthetic overdose after the final blood sample was obtained (Figure 1).

Average MAP was 91.1 ± 1.8 mmHg at baseline among the groups (p > 0.05). Bleeding provoked a significant decrease in MAP in groups 2 and 3 compared to sham (G1), and this difference persisted throughout the first 20 minutes of the experiment; respectively 62.3 ± 3.7 mmHg vs. 85.5 ± 3.5 mmHg (p < 0.05). Mean arterial pressure response to bleeding and resuscitation was unaffected by VNS (group 3) compared to group 2; p > 0.05 (Figure 2). There was no significant difference in the volumes of lactated Ringer’s used during resuscitation between groups 2 and 3, respectively 4.8 ± 0.3 ml and 4.6 ± 0.2 ml (p > 0.05).

Results showed that group 2 animals had a significant increase in plasma IL-1 after hemorrhage/resuscitation compared to baseline and final sham; respectively 8.5 ± 4.3 pg/ml vs. 22.6 ± 10.5 pg/ml, and 4.5 ± 1.3 vs. 22.6 ± 10.5 pg/ml (p < 0.05). In contrast, IL-1 levels decreased significantly when VNS was performed during resuscitation (group 3) compared to baseline and final group 2; respectively 7.7 ± 2.8 vs. 0.7 ± 0.5 and 22.6 ± 10.5 vs. 0.7 ± 0.5 pg/ml (p < 0.05). As expected, sham hemorrhage did not alter IL-1 levels significantly (Figure 3). Group 3 animals were also assessed for anti-inflammatory cytokine IL-10. There was a significant increase in plasma IL-10 levels with VNS during hemorrhage resuscitation compared to baseline; respectively 26.3 ± 3.5 pg/ml vs. 40.2 ± 6.0 pg/ml (p < 0.05).

Hemorrhage and resuscitation without VNS (group 2) resulted in a significant decrease in MCF compared to baseline; respectively 71.5 ± 1.8 mm vs. 64.0 ± 1.7 mm (p < 0.05). In contrast, when VNS was performed during resuscitation (group 3) MCF increased significantly compared to baseline 67.3 ± 2.7 mm vs. 74.5 ± 1.5 mm (p < 0.05) (Figure 4-A). Vagus nerve stimulation during resuscitation also improved other tromboelastometric parameters in group 3. The alfa angle increased significantly 72.7 ± 2.5^o vs. 78.7 ± 1.5° (p < 0.05), while the CFT decreased 90.3 ± 16.2 vs. 59.7 ± 8.9 seconds (p < 0.05); compared to baseline. Hemorrhage followed by resuscitation without VNS (group 2) did not improve those parameters (Figure 4-B and C).