gms | German Medical Science

Objective: About 10% of intracerebral hemorrhages comprise cerebellar bleedings mainly caused by arterial hypertension with a peak incidence between 60 and 80 years. Associated with high mortality cerebellar hemorrhages can quickly lead to a life threatening loss of consciousness due to herniation and brainstem compression. Evidence concerning the choice of therapy lacks.

Method: In this retrospective study 92 patients (47 female, 45 male) with an average age of 69 years suffering from spontaneous cerebellar hemorrhage were evaluated from 2005 to 2014. Considering four therapy groups (suboccipital decompression including hemorrhage evacuation, only hemorrhage evacuation, external ventricular drainage (EVD), conservative) neurology was assessed using GCS and mRS preoperative, at discharge, 2-3 months and 4-6 months following hemorrhage. Also pre- and postsurgical hemorrhage volumes, sizes of decompression and oral anticoagulation were examined.

Results: Decompression was performed in 28 patients, hemorrhage evacuation in 33, EVD in 12 and 19 patients were treated conservatively. Overall mortality was 16%. The initial hemorrhage volumes of the EVD and conservative group were significant smaller than those of the decompression and evacuation group (ANOVA p<0.001, post hoc test Bonferroni p<0.005). A significant positive correlation was detected between an oral anticoagulant therapy and the size of the initial volume. Comparing pre- and postsurgical hemorrhage sizes decompression and evacuation as well as evacuation-only could significantly reduce volumes (t-test each group p<0.001). Although preoperative GCS was significant lower in the decompression group compared to the other groups (ANOVA p<0.001, post hoc test Bonferroni p<0.05), this difference abated from discharge until 4-6 months. In the decompression group 58% of the patients had an initial GCS 3-8, only 30% at discharge. Compared to the preoperative GCS there was a significant improvement of GCS in this group at discharge and 2-3 months (t-test discharge p<0.05, 2-3 months p<0.001). A significant increase of GCS in the evacuation group was found after 2-3 months (t-test p<0.001). The other two groups showed ameliorations without significance.

Conclusions: In summary, suboccipital decompression including hemorrhage evacuation leads to a significant neurological improvement comparing pre- and postoperative neurology. It might be beneficial for patients with spontaneous cerebellar hemorrhage, further prospective clinical studies have to follow.