In experimental stroke, recurrent peri-infarct depolarizations (PIDs) are known to mediate the progressive deterioration of penumbral border zones into core infarction. PIDs are similar to cortical spreading depression (CSD), except that the hyperemia accompanying CSD is replaced by transient or permanent decrease in cerebral blood flow, exacerbating ischemic conditions. In rats and humans, CSD-like spreading depolarizations (SD) also occur following traumatic brain injury (TBI), although the hemodynamic correlates (i.e. CSD vs. PID) have not yet been determined. Nonetheless, Strong et al. have shown that SDs occurring over the initial days post-injury are associated with development of low brain glucose, suggesting a detrimental PID-like effect. Here we investigate clinically the correlates and incidence of SD in TBI. Patients requiring craniotomy for lesion/hematoma evacuation and/or decompression were enrolled in the Co-Operative Study on Brain Injury Depolarizations (www.cosbid.org) after obtaining research consent. Sub-dural linear electrode strips with six contacts were implanted near the contused region during the craniotomy procedure and four-channel recordings, AC amplified with a 0.02 Hz high pass cutoff, were then made in a sequential bipolar recording configuration. A microdialysis catheter was implanted intraparenchymally near the strip to monitor neurochemistry. Patients were ventilated, paralyzed, and sedated with morphine. Recordings were terminated immediately prior to withdrawal from sedation. SDs were identified by analyzing the amplitude of the electrocorticogram (ECoG) and shifts in the ECoG baseline potential which reflect the DC potential negativity caused by depolarization. Criteria for identification of SDs were an abrupt, >50% decrement of ECoG amplitude on at least one channel, followed by gradual recovery, and sequential baseline shifts on at least two channels. Six male patients aged 18–67 were enrolled with GCS 4–9 upon admission. ECoG strips were placed in either the left temporal (n=3) or right frontal lobe (n=3). Recordings were initiated from 0.6 to 5 days post-injury and recording durations thereafter ranged from 2.0 to 5.7 days, with a median of 5.0. Three of six patients (50%) experienced recurrent SDs numbering 4, 6, and 14. All SDs occurred between days 3–8 post-injury, but the majority of events (14/24; 58%) occurred on day six. In two patients, several (4 and 7) SDs recurred regularly at short intervals not exceeding 4 h and 2.5 h, respectively. During these periods, mean body temperatures were 38.5°C and 38.8°C, respectively, and always exceeded 37.8°C and 38.0°C. These results confirm the common occurrence of CSD-like depolarizations in the human brain as a sequela of TBI. Previously Strong et al. found, with similar recordings of a mean 36 h duration, that 4/11 (36%) patients with traumatic intracranial hemorrhages exhibited CSD-like events. The present results suggest that CSD-like events may have an even higher incidence and a delayed or prolonged time course. Results further suggest that episodes of fever may create a permissive environment for the initiation/propagation of depolarizations, consistent with the effectiveness of hypothermia to suppress PIDs in experimental stroke. Other factors influencing the occurrence of SD as well as effects on neurochemistry, hemodynamics, and outcome are under investigation by COSBID.
