Master Theses UMCG - University of Groningen
English | Nederlands

“Risk of hemorrhage after Gamma Knife radiosurgery versus natural history for hemorrhage in brain arteriovenous malformations: a retrospective descriptive study

(2016) Govindarajan, L.P.

Background: Gamma Knife Radiosurgery (GKRS) is frequently used for the treatment of brain arteriovenous malformations (BAVM). Patients with unruptured BAVM have a risk of 2.2% per year for intracranial hemorrhage with severe morbidity and mortality[16]. Patients with ruptured BAVM have 4.5% risk for a second haemorrhage[13]. Adverse radiation effects (ARE’s) are seen post treatment with variable incidence (1.8-22.9%)[39]. The latency period post GKRS is a crucial time frame for hemorrhage and ARE’s to occur. Timely follow-up and treatment are essential during this period.
Objective: The purpose of this report was to evaluate the risk of hemorrhage post GKRS for unruptured BAVM versus the natural history, in order to conclude that treatment in unruptured BAVMs with GKRS is a good approach in preventing hemorrhages. Secondly the goal was to evaluate the risk of a second hemorrhage post GKRS in ruptured BAVM versus the rebleeding risk. The incidence of ARE’s were evaluated to analyse the safety of GKRS treatment. Lastly, to evaluate the effect of treatment, obliteration rates for hemorrhage and non-hemorrhage BAVM were analysed.
Methods: The investigation was carried through a retrospective descriptive study of patients with BAVM who underwent GKRS between 2002 and 2015 at the Gamma Knife Center of the Elisabeth-Tweesteden hospital in Tilburg, the Netherlands. Patients were included based on the inclusion criteria ( age > 18 years, angiographically established BAVM and with a minimal follow-up of ≥ 3 years). Univariate and multivariate analyses were performed. The Chi square test was used to calculate the significance level (p<0.05) of hemorrhage risk post GKRS in our study versus the bleeding risk in the natural history of BAVM. Ruptured BAVMs were compared to the rebleed risk (4.5%). Incidence rates for ARE’s were calculated. Survival analysis (Kaplan-Meier) was performed for the obliteration rates.
Results: From a total of 443 treatments, we yielded 373 BAVM treatments fulfilling the selection criteria. Out of 373 treatments, 74 were double or triple treatments and 6 patients refused further treatment. Overall we included 293 patients for further analysis of which 275 had symptomatic BAVM and 18 non-symptomatic BAVM. In the symptomatic BAVM population, the non-hemorrhage group (n =140), 2.1% (n =3) experienced hemorrhage post GKRS. In the hemorrhage group 5.9% (n=8) experienced hemorrhage post GKRS. The incidence of ARE’s is between 0.27-2%. 84 patients in the hemorrhage group and 109 patients in the non-hemorrhage group achieved full obliteration in a follow up(FU) time of 5 years.
Conclusion: We cannot conclude that in our study population the risk of hemorrhage post GKRS is significantly less, higher or the same compared to the natural history. This is due to the short follow up time (until obliteration), unequal number of patients in each group and different AVM grading system compared to other studies [16][13][39]. Also no control group could be formed due to the limited time frame for this study. What we do conclude is that patients with prior hemorrhage have a higher risk of hemorrhage post treatment; this is in line with other studies[20]. There is no difference in obliteration rate between hemorrhage and non-hemorrhage BAVM patients. GKRS could be a good and safe treatment option for symptomatic BAVM with moderate side-effects (0.27-2%) and a 60-70% complete obliteration after single treatment. For more accurate post treatment risk assessment a longer FU-time, equal distribution of patients in the comparing groups and more profound categorization of the BAVM angioarchitecture is needed.

To top