Purpose: Three-dimensional (3D) geometric conformation of the therapeutic dose volume to the shape of a target tissue volume is the motivation for both conformal radiotherapy and radiosurgery. Although noncoplanar arcs have a clear physical and geometric advantage over fixed fields for small spherical targets, those advantages are reduced for large or irregularly shaped targets where static fields can be individually shaped. We have developed a system that allows efficient and flexible design and reliable delivery of customized "bouquets" of fixed nonopposed coplanar or noncoplanar shaped fields, resulting in highly uniform dose distributions. This report describes our initial experience using beam bouquets to treat intracranial lesions.
Methods and materials: Patients with primary (11) or metastatic (4) intracranial lesions with a maximum diameter less than approximately 6 cm, most of whom candidates for single-fraction radiosurgery, were treated with beam bouquets of four to eight nonopposed coplanar or noncoplanar beams. Doses ranged from 16-20 Gy in four fractions for recurrent lesions (8) to 45 to 68 Gy in 25 to 34 fractions for primary lesions (7). The patients were immobilized with custom foam head supports and face masks attached to a fixed base plate. Planning computed tomography scans were acquired, from which the physician developed the custom beam bouquet using 3D treatment-planning tools. The bouquet was designed based primarily on geometric concerns. The bouquet was subsequently modified to add wedge filters chosen by vector analysis of dose gradients to achieve uniform dose over the volume of beam crossfire. At the time of treatment, the isocenter was placed using the instructions provided by the treatment-planning system and pretreatment orthogonal port films were compared to digitally reconstructed radiographs (DRR) to assure proper isocenter placement. For several situations, the 3D dose distributions resulting from alternative coplanar and noncoplanar plans were compared.
Results: Each patient was treated without incident. Daily pretreatment port films showed excellent reproducibility of isocenter placement in 87% of setups. With short follow-up (0-12 months), two patients with recurrent glioblastoma experienced clinical deterioration 2 to 4 weeks following treatment. One had increased edema on scans and responded to steroids. Six patients clinically improved following radiation therapy. Review of alternative treatment plans reveals that the relative utility of coplanar vs. noncoplanar beams is likely dependent on the location of the lesion. Noncoplanar beam bouquets are likely preferable to coplanar beams when the target is located in the central regions of the head. Coplanar beams are likely adequate, and possibly preferable, for peripherally located targets.
Conclusion: The biological advantages of fractionation and the physical advantages of radiosurgery are exploited with this approach. The use of multiple nonopposed coplanar or noncoplanar conformal wedged fields provides a uniform dose to the target and acceptable dose gradient at the target edge. This technique may prove to be an alternative to arc-based radiosurgery in some settings and has the potential advantages that fractionation should improve the therapeutic ratio, and each beam can be individually shaped to conform to irregularly shaped targets. Additional studies are underway to improve this system and better define its utility.