Brain temperature but not core temperature increases during spreading depolarizations in patients with spontaneous intracerebral hemorrhage

Abstract

Spreading depolarizations (SDs) are highly active metabolic events, commonly occur in patients with intracerebral hemorrhage (ICH) and may be triggered by fever. We investigated the dynamics of brain-temperature (T_brain) and core-temperature (T_core) relative to the occurrence of SDs. Twenty consecutive comatose ICH patients with multimodal electrocorticograpy (ECoG) and T_brain monitoring of the perihematomal area were prospectively enrolled. Clusters of SDs were defined as ≥2 SDs/h. Generalized estimating equations were used for statistical calculations. Data are presented as median and interquartile range. During 3097 h (173 h [81-223]/patient) of ECoG monitoring, 342 SDs were analyzed of which 51 (15%) occurred in clusters. Baseline T_core and T_brain was 37.3℃ (36.9-37.8) and 37.4℃ (36.7-37.9), respectively. T_brain but not T_core significantly increased 25 min preceding the onset of SDs by 0.2℃ (0.1-0.2; p < 0.001) and returned to baseline 35 min following SDs. During clusters, T_brain increased to a higher level (+0.4℃ [0.1-0.4]; p = 0.006) when compared to single SDs. A higher probability (OR = 36.9; CI = 36.8-37.1; p < 0.001) of developing SDs was observed during episodes of T_brain ≥ 38.0℃ (23% probability), than during T_brain ≤ 36.6℃ (9% probability). Spreading depolarizations - and in particular clusters of SDs - may increase brain temperature following ICH.

Keywords: Cortical spreading depolarizations; body temperature; brain temperature; brain temperature regulation; spontaneous intracerebral hemorrhage.

Figure 1.

(a) Percentage of monitoring time ≥38℃/24 h of brain and body temperature following ICH; day 0 indicates the admission day. (b) Temperature and temperature variability after ICH; the black arrow indicates the time within 90% of patients had surgery; TBrain: brain temperature; TCore: core body temperature; ICH: spontaneous intracerebral hemorrhage. Error bars indicate one standard error of mean.

Figure 2.

(a) A higher probability (OR = 36.9; CI = 36.8–37.1; p < 0.001) of developing SDs was observed during episodes of brain temperature ≥38.0℃ (23% probability), than at a brain temperature ≤36.6℃ (9% probability). (b) Brain temperature but not core body temperature significantly increased 25 min preceding the onset of SDs by a median of 0.2℃ (0.1–0.2; p < 0.001) and returned to baseline 35 min following SDs; SDs: spreading depolarizations; error bars indicate one standard error of mean.

Figure 3.

(a) Brain temperature (+0.4℃ [0.1–0.4]; p = 0.006) was higher during clusters (≥2 SDs/h) than during single SDs. (b) Unfavorable outcome (mRS > 3) at three months was associated with a higher SDs frequency (unfavorable: 4 SDs per hour vs. good: 1 SDs per hour; p = 0.013, OR = 4.7 CI = 1.1–20; adjusted for ICP and MAP); mRS: modified Rankin scale; SDs: spreading depolarization; error bars indicate one standard error of mean.

Figure 4.

(a) Schematic 3D Reconstruction of a patient with right frontal hematoma; (b) Note the nearest distance between the subdural ECoG strip (electrode 6) and the tip of the ICP-temp probe in the perihematomal region (b); white arrow = subdural ECoG strip; triangle = tip of the ICP-temp probe; asterix = hematoma location.