Thrombelastography indicates limitations of animal models of trauma-induced coagulopathy

Background: Thrombelastography (TEG) has been used to characterize the coagulation changes associated with injury and shock. Animal models developed to investigate trauma-induced coagulopathy (TIC) have failed to produce excessive bleeding. We hypothesize that a native TEG will demonstrate marked differences in humans compared with these experimental models, which explains the difficulties in reproducing a clinically relevant coagulopathy in animal models.

Methods: Whole blood was collected from 138 healthy human volunteers, 25 swine and 66 Sprague-Dawley rats before experimentation. Citrated native TEGs were conducted on each whole blood sample within 2 h of collection. The clot initiation (R-time, minutes), angle (degrees), maximum amplitude (MA; millimeter), and lysis 30 min after MA (LY30; percentage) were analyzed and contrasted between species with data represented as the median and 25^th to 75^th quartile range. Difference between species was conducted with a Kruskal-Wallis test with alpha adjusted with a Bonferroni correction for multiple comparisons (alpha = 0.016).

Results: Median R-time (clot initiation) was 14.65 min (IQR: 13.2-16.3 min) for humans, 5.7 min (4.9-8.8) for pigs, and 5.2 min (4.4-6) for rodents. Humans had longer R-times than both pigs (P < 0.0001) and rats (P < 0.0001); pigs were not different from rats (P = 0.4439). Angle (fibrin cross-linking) was 42.3° (interquartile range [IQR]: 37.5-50.2) for humans, 71.7° (64.3-75.6) for pigs, and 61.8° (56.8-66.7) for rats. Humans had reduced angle compared with both pigs (P < 0.0001) and rats (P < 0.0001); pigs were not different from rats (P = 0.6052). MA (clot strength) was 55.5 mm (IQR: 52.0-59.5) for humans, 72.5 mm (70.4-75.5) for pigs, and 66.5 mm (56.5-68.6) for rats. Humans had reduced MA compared with both pigs (P < 0.0001) and rats (P < 0.0001); pigs were not different from rats (P = 0.0161). LY30 (fibrinolysis) was 1.5% (IQR: 0.975-2.5) for humans, 3.3% (1.9-4.3) for pigs, and 0.5% (0.1-1.2) for rats. Humans had a lesser LY30 than pigs (P = 0.0062) and a greater LY30 than rats (P < 0.0001), and pigs had a greater LY30 than rats (P < 0.0001).

Conclusions: Humans, swine, and rodents have distinctly different coagulation systems, when evaluated by citrated native TEG. Animals are hypercoagulable with rapid clotting times and clots strengths nearly 50% stronger than humans. These coagulation differences indicate the limitations of previous models of trauma-induced coagulopathy in producing coagulation abnormalities associated with increased bleeding. The inherent hypercoagulable baseline tendencies of these animals may result in subclinical biochemical changes that are not detected by conventional TEG and should be taken into consideration when extrapolated to clinical medicine.