Cerebral Regional Tissue Oxygenation as Surrogate for Blood Loss in Nonhuman Primate Models of Shock

Hemorrhage is the leading cause of preventable death, with a majority of mortalities in the prehospital setting. Current hemorrhage resuscitation guidelines cannot predict the critical point of intervention to activate massive transfusion (MT) and prevent cardiovascular decompensation. We hypothesized that cerebral regional tissue oxygenation (CrSO₂) would indicate MT need in nonhuman primate models of hemorrhagic shock.

Nineteen anesthetized male rhesus macaques underwent hemorrhage via a volume-targeted (VT) or pressure-targeted (PT) method. VT animals were monitored for 30 min following 30% blood volume hemorrhage. PT animals were hemorrhaged to mean arterial pressure (MAP) of 20 mmHg and maintained for at least 60 min until decompensation. Statistics for MAP, heart rate (HR), end tidal carbon dioxide (EtCO₂), and CrSO₂ were analyzed via one- or two-way repeated-measures analysis of variance, Pearson's R, and receiver-operator curve. A P < 0.05 is considered significant.

Following initial hemorrhage (S0), there were no significant differences between groups. After cessation of hemorrhage in the VT group, MAP and EtCO₂ returned to baseline while CrSO₂ plateaued. The PT group maintained model-defined low MAP, suppressing EtCO₂, and significantly decreased CrSO₂ compared to the VT group by S25. Linear regression of CrSO₂versus shed blood volume demonstrated R² = 0.7539. CrSO₂ of 47% was able to detect >40% blood loss with an area under the curve of 0.9834 at 92.3% (66.7%-99.6%) sensitivity and 95.5% (84.9%-99.2%) specificity.

Regardless of hemorrhage modality and compensatory response, CrSO₂ correlated strongly with shed blood volume. Analysis demonstrated that CrSO₂ values below 49% indicate Advanced Trauma Life Support class IV shock (blood loss>40%). CrSO₂ at the point of care may help indicate MT need earlier and more accurately than traditional markers.