gms | German Medical Science

Objective: Decompressive craniectomy (DC) is a well established part in management of patients with malignant media infarction (MI) as well as in patients with traumatic brain injury (TBI). In this context midline shift and hemispheric swelling in pre- and postoperative CT-scans were analyzed to acquire further evidence in predicting the clinical course and outcome of MI and TBI patients.

Method: Pre- and postoperative CT-scans in 85 patients (50 with MI, mean age 52.7 years; 35 with TBI, mean age 54.4 years) were analyzed. The difference between post- and preoperative midline shift (DMLS) was measured at the septum pellucidum level. Additionally a coefficient of ipsi- and contralateral hemispheric swelling (CHS) at the Foramina of Monro level in postoperative CT-scans for both patient groups was calculated. DMLS and CHS were analyzed regarding significant differences in patients with favorable versus unfavorable outcome. Favorable outcome was defined as modified Rankin Scale (mRS) < 4 after rehabilitation (mean follow-up was 128 days). Furthermore Spearman’s rho (r) was used to correlate CT-findings with clinical outcome.

Results: In TBI-patients mean mRS was 4.74. CHS was significantly lower in patients with favorable versus unfavorable outcome (1.11 vs. 1.28; p < 0.01). There was no statistical difference regarding DMLS. In Spearman’s rho a significant dependence was detected between CHS and clinical outcome (r = 0.501; p < 0.01) yet there was no correlation between DMLS and mRS (r = 0.144; p = 0.41) In patients with MI mean mRS was 4.66. DMLS was significantly (p < 0.01) smaller in patients with favorable (-0.21 cm; negative value indicates postoperative decrease of MLS) compared with unfavorable outcome (0.24 cm) whereas CHS-analysis did not show any significant differences. In Spearman’s rho a significant correlation between DMLS and mRS (r = 0.469; p < 0.001) was found but none for CHS and mRS (r = 0.134; p = 0.39).

Conclusions: By DMLS and CHS in postoperative CT-scans predictive criteria are provided in TBI- and MI-patients. While DMLS proofs to be a reliable criterion in MI-patients it appears to be of poor quality in trauma patients whereas clinical outcome of TBI-patients can be well estimated with CHS. Although MI and TBI-patients may both be operated on in the same way by DC, our data clearly indicates that these two groups are different entities in which separate analyses have to be made.