Long-term follow-up of an overexposure radiation incident in a cohort treated with linear accelerator–based stereotactic radiosurgery for intracranial arteriovenous malformations

Abstract

OBJECTIVE Dosimetric radiosurgery incidents are rare and probably insufficiently reported in scientific publications. After a long follow-up (FU), the authors studied the outcomes of patients treated with overexposure radiation for arteriovenous malformation (AVM) administered via stereotactic radiosurgery (SRS) at their department. METHODS Between May 2006 and June 2007, 22 patients were treated for AVM with SRS. The mean (range) patient age was 43.5 (11.8–78) years. Previous treatments were embolization (n = 10), SRS (1), and surgery (1). The average (range) volume was 2.1 (0.2–6.4) cm3. The median prescribed minimal dose was 18.0 Gy. An initial error in the estimation of scatter factors led to overexposure to radiation. Due to this incident, the median delivered minimum dose was 25.0 Gy. All patients were prospectively followed with clinical examination and imaging. RESULTS The mean (range) clinical FU was 14.5 (12.0–15.2) years. AVM obliteration after SRS was completed in 90.9% of patients at a mean (range) of 39.4 (24.4–70.4) months. No patient had post-SRS AVM bleeding. Three patients (13.6%) had new permanent deficits due to radiation-induced changes (RICs). Obliteration without new deficits was achieved in 18 patients (81.8%). Two patients had new epilepsy that was probably due to RIC but well controlled. The median (range) MRI FU was 13.8 (2.5–14.9) years. During MRI FU, two RIC periods were observed: one classic period during the first 3 years showed T1-weighted annular irregular enhancement (13%), and the other period between 5 and 15 years after SRS showed the occurrence of cystic and hemorrhagic lesions (22.7%). There were no cases of radiation-induced tumor. CONCLUSIONS The present long-term report showed that this overexposure incident probably increased the AVM obliteration rate. This overexposure seems to have induced RIC and in particular a higher rate of cystic and hemorrhagic late lesions with nevertheless moderate clinical consequences. Long-term FU for AVM is mandatory due to the risk of late RIC.