Malignant brain edema after MCA infarction is determined by the severity of ischemia within the infarction

Objective

Malignant brain edema is a serious complication after MCA-infarction. For reasonable implementation of invasive therapies like hemicraniectomy, probability of malignant brain edema has to be anticipated. However, pathophysiology of edema formation after MCA-infarction is poorly understood and identification of patients finally developing such edema is difficult. Experimental studies have shown maximal edema formation when CBF was deeply reduced to <6 ml/100g/min^{1, 2}. Primary endothelial damage and/or a blood brain barrier break-down have been discussed to result from such deep ischemia ³ . We semiquantitatively investigated CBF in patients with large MCA-infarction by using 11-C-Flumazenil-PET ⁴ in order to predict malignant brain edema.

Methods

We included 16 patients with clinical signs of MCA infarction and involvement >50% of MCA territory in early CT or MRI (<12 h). PET was performed within 15.9±6.4 h hours after symptoms onset. The ischemic core was defined as the volume of perfusional deficit <50% of the average within the contralateral hemisphere, corresponding to a CBF of approximately <14 ml/100 g/min. Within the ischemic core subregions with different ischemic CBF values were investigated. Subregions were defined as the volume of perfusional deficit <40, <30, <20, <10% of the average within the contralateral hemisphere, corresponding to a CBF of approximately <12, <8, <6, <4 ml/100 g/min. As a correlate of final edema formation, we measured maximal midline-shift on follow-up CTs at the level of the septum pellucidum. Regarding the clinical course and follow-up CT-findings, patients were post hoc divided into a malignant (n=7) and a benign group (n=9). We calculated correlations between ischemic volumes derived from subregional analysis and maximal midline shift in CT.

Results

Volumes of ischemic core showed significant differences between both groups (malignant group 149. 8±24.9 vs. benign group 51.5±35.8 ccm, p<0,01). Volumes of all subregions with deep ischemia were significantly larger in the malignant group than in the benign group (e.g. CBF<20% 83.1±29,8 vs. 15.0±26,1 ccm; CBF<10% 40.4±25,7 vs. 5,2±8,9 ccm; p<0,0001). The volume of these two subregions within the ischemic core was 55,5% respectively 26,9% of the total ischemic core volume in the malignant group and only 29,1% respectively 10,1% in the benign group. Volumes of regions with deep ischemia showed the best correlations with maximal midline shifts for CBF<20% Rho=0,74, p<0,005 and for CBF<10% Rho=0,76, p<0,005.

Conclusion

Subregions with deep ischemia seem to be relevant for massive edema formation in human stroke. Ischemic endothelial damage may lead to a primary blood brain barrier break-down and subsequent vasogenic edema. PET imaging is able to detect the magnitude of these deeply ischemic regions at an early stage and may thereby guide the indication for early hemicraniectomy.