CLINICAL INVESTIGATION INTEGRATION OF THREE-DIMENSIONAL ROTATIONAL ANGIOGRAPHY IN RADIOSURGICAL TREATMENT PLANNING OF CEREBRAL ARTERIOVENOUS MALFORMATIONS ALFREDO CONTI, M.D., PH.D.,* ANTONIO PONTORIERO, M.D., y GIUSEPPE F ARAG O, M.D., z FEDERICA MIDILI,PH.D., x CARMELO SIRAGUSA,PHD., x FRANCESCA GRANATA, M.D., z ANTONIO PITRONE, M.D., z COSTANTINO DE RENZIS, M.D., y MARCELLO LONGO, M.D., z AND FRANCESCO TOMASELLO, M.D.* Departments of *Neurosurgery, y Radiation Oncology, z Neuroradiology, and x Medical Physics, University of Messina, Messina, Italy Purpose: Accuracy in delineating the target volume is a major issue for successful stereotactic radiosurgery for arteriovenous malformations. The aim of the present study was to describe a method to integrate three- dimensional (3D) rotational angiography (3DRA) into CyberKnife treatment planning and to investigate its poten- tial advantages compared with computed tomography angiography (CTA) and magnetic resonance angiography. Methods and Materials: A total of 20 patients with a diagnosis of cerebral arteriovenous malformation were included in the present study. All patients underwent multislice computed tomography and 3D-volumetric CTA, 3DRA, and 3D magnetic resonance angiography. The contouring of the target and critical volumes was done separately using CTA and thereafter directly using 3DRA. The composite, conjoint, and disjoint volumes were measured. Results: The use of CTA or 3DRA resulted in significant differences in the target and critical volumes. The target volume averaged 3.49 ± 3.01 mL measured using CTA and 3.26 ± 2.93 mL measured using 3DRA, for a difference of 8% (p < .05). The conjoint and disjoint volume analysis showed an 88% volume overlap. The qualitative evaluation showed that the excess volume obtained using CTAwas mostly tissue surrounding the nidus and venous structures. The mean contoured venous volume was 0.67 mL measured using CTA and 0.88 mL (range, 0.1–2.7) measured using 3DRA (p < .05). Conclusions: 3DRA is a volumetric angiographic study that can be integrated into computer-based treatment plan- ning. Although whether 3DRA provides superior accuracy has not yet been proved, its high spatial resolution is attractive and offers a superior 3D view. This allows a better 3D understanding of the target volume and distribu- tion of the radiation doses within the volume. Additional technical efforts to improve the temporal resolution and the development of software tools aimed at improving the performance of 3D contouring are warranted. Ó 2011 Elsevier Inc. Three-dimensional rotational angiography, Stereotactic radiosurgery, CyberKnife, Robotic radiosurgery, Cere- bral arteriovenous malformations. INTRODUCTION Stereotactic radiosurgery (SRS) is an effective treatment mo- dality for properly selected cerebral arteriovenous malforma- tions (AVMs). This procedure can result in the obliteration of the AVM, after a latency of 24–36 months, in 60–90% of cases (1–3). Several factors will influence the outcome of SRS for AVMs, including the nidus volume, prescription dose, patient age, lesion location, angioarchitecture, and previous embolization (1–6). Accuracy in delineating the target volume has been a major issue (2, 7, 8). The entire nidus should be included in the target volume to achieve obliteration of the AVM; at the same time, highly conformal treatment should be delivered to reduce the risk of radiation-induced brain complications. Biplanar digitally subtracted angiography (DSA) remains the benchmark imaging modality to delineate cerebral AVMs in SRS treatment planning. At present, DSA is the only imaging study that can provide sufficient temporal res- olution to study the AVM hemodynamics (9). It also pro- vides the greatest spatial resolution. Nevertheless, when using DSA, the target volume is reconstructed from the con- tours drawn independently on two or more perpendicular or Reprint requests to: Alfredo Conti, M.D., Ph.D., Department of Neurosurgery, A.O.U. Policlinico Universitario, University of Mes- sina, Via Consolare Valeria, 1, Messina 98125 Italy. Tel: (+39) 090-221-2862; Fax: (+39) 090-693-714; E-mail: alfredo.conti@ unime.it Conflict of interest: none. Acknowledgments—The authors are particularly grateful to Dr. Dave Schaal for the scientific and editorial assistance. Received April 18, 2010, and in revised form Nov 5, 2010. Accepted for publication Dec 20, 2010. 1 Int. J. Radiation Oncology Biol. Phys., Vol. -, No. -, pp. 1–9, 2011 Copyright Ó 2011 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$ - see front matter doi:10.1016/j.ijrobp.2010.12.024