Background: Cerebral regional microcirculation is altered following severe head injury. In addition to tissue disruption, partial pressure of tissue oxygenation is impaired due to an increase in the oxygen tissue gradient. The heterogenic distribution of cerebral microcirculation is multifactorial, and acute anemia challenges further the delivery of oxygen to tissues. Currently, a restrictive transfusion threshold is globally applied; however, it is unclear how anemia modifies regional cerebral microcirculation; hence, it is unclear if by aiming to a global endpoint, specific anatomical regions undergo ischemia. This study aims to quantify the temporal changes in cerebral microcirculation after severe head injury, under the effect of anemia and transfusion. It also aims to assess its effects specifically at the ischemic penumbra compared to contralateral regions and its interactions with axonal integrity in real time. Twelve ovine models were subjected to a severe contusion and acceleration-deceleration injury. Normovolemic anemia to a restrictive threshold was maintained after injury, followed by autologous transfusion. Direct quantification of cerebral microcirculation used cytometric count of color-coded microspheres. Axonal injury was assessed using amyloid precursor protein staining.

Results: A mixed-effect regression model from pre-transfusion to post-transfusion times with a random intercept for each sheep was used. Cerebral microcirculation amongst subjects with normal intracranial pressure was maintained from baseline and increased further after transfusion. Subjects with high intracranial pressure had a consistent reduction of their microcirculation to ischemic thresholds (20–30 ml/100 g/min) without an improvement after transfusion. Cerebral PtiO2 was reduced when exposed to anemia but increased in a 9.6-fold with transfusion 95% CI 5.6 to 13.6 (p value < 0.001).

Conclusions: After severe head injury, the exposure to normovolemic anemia to a restrictive transfusion threshold, leads to a consistent reduction on cerebral microcirculation below ischemic thresholds, independent of cerebral perfusion pressure. Amongst subjects with raised intracranial pressure, microcirculation does not improve after transfusion. Cerebral oxymetry is impaired during anemia with a statistically significant increase after transfusion. Current transfusion practices in neurocritical care are based on a rigid hemoglobin threshold, a view that excludes cerebral metabolic demands and specific needs. An RCT exploring these concepts is warranted.