Arteriovenous malformations in the basal ganglia region: Gamma Knife radiosurgery as first choice treatment in selected cases.

BACKGROUND

The authors report the results of Gamma Knife (GK) radiosurgery on a clinical series of selected patients with brain basal ganglia arteriovenous malformations (BG AVMs). Furthermore, clinico-epidemiological and anatomo-functional pictures of BG AVMs and supratentorial cortical AVMs are comparatively analyzed, and their influence on the radiosurgical outcome is discussed.

METHODS

At our Department, 33 BG AVMs (21 with FU >2 years) and 209 cortical AVMs (110 with FU >2 years) with a radiosurgical volume 10 cc underwent GK from February 1993 to July 2001. Mean age, male/female ratio (M/F), and bleeding rate at clinical onset in the 2 subgroups of patients were as follows: 25.5 years (5-62 years), 16 M/17 F, and 30/33 (91%) in BG AVMs, respectively; 34.4 years (8-74 years), 124 M/85 F, and 107/209 (51%) in cortical AVMs, respectively. Statistical analysis performed with contingency table method and deviance analysis according to generalized linear models showed that the differences concerning age at onset and bleeding rate were highly significant.

RESULTS

The rates of complete obliteration, permanent morbidity, bleeding/rebleeding during latency period, and unsuccessful embolization attempt in the 2 subgroups of patients were as follows: 81%, 4%, 8%, and 41.7% in BG AVMs, respectively; 85.5%, 2%, 2%, and 2% in cortical AVMs, respectively.

CONCLUSIONS

In our experience, the different clinico-epidemiological and anatomo-functional characteristics between BG AVMs and cortical AVMs do not seem to influence the radiosurgical outcome as complete obliteration and permanent neurological sequelae rates are very similar in these 2 subgroups of patients. GK radiosurgery may be considered the first choice treatment modality in BG AVMs with <10 cc volume (very high cure rate, very low permanent morbidity, and zero mortality). Although the haemorrhagic onset in BG AVMs is much more frequent than in cortical AVMs, the difference in the bleeding/rebleeding rate between the 2 subgroups of patients during the latency period is not statistically significant and this occurrence seems to be limited within the first year after GK.