Reducing cardiac irradiation during bilateral whole lung irradiation with multiple co-planar or non-coplanar conformally shaped fields

Purpose: Previous studies suggest that bilateral whole lung irradiation (BWLI) might reduce the incidence of pulmonary relapse in patients with sarcoma (e.g., Breur et al., Burgers et al., Carlos et al., Nesbit et al.). However, BWLI is not commonly used because of concerns of radiation-induced cardiac effects, especially since adriamycin is commonly used in these patients. We herein describe the use of three-dimensional (3D) treatment planning tools to design coplanar and non-coplanar beam groups (with and without compensators) to reduce incidental cardiac irradiation during BWLI. Methods and Materials: Structures of interest (lungs, heart, liver, and kidneys) were identified on complete neck-thorax-abdominal CT scans of two pediatric patients. Using 3D treatment planning software, a variety of competing treatment plans were generated. A series of 'unusual' beam groups were studied including 3, 5, 7, and 9 coplanar beams in the transverse plane and 4, 6, 8, and 10 non-coplanar beams. For each beam in these groups, the heart was blocked using an 'auto-avoid' feature with varying 'negative margins' (i.e., the outer 10 or 15-mm shell of the heart was included within the beam field edge to provide adequate dose to the lung immediately adjacent to the heart). 'Conventional' AP/PA beam arrangements were also considered (with heart blocking on neither, both, or only one of the beams). Doses (reflecting density inhomogeneity corrections), dose volume histograms (DVH's), and normal tissue complication probabilities (NTCP's) for the lung, heart, liver, and kidneys were calculated. To facilitate inter-plan comparisons, each plan was designed to provide an equivalent 85% tumor control probability (TCP) for occult microscopic disease in the lung. NTCP's (Lyman) and TCP's (Webb and Nahum) were calculated using published DVH-reduction schemes (Kutcher and Burman). Results: Compared to the AP/PA beam arrangements, the 'unusual' coplanar and non-coplanar beam arrangements provided a reduction in the heart dose and a more heterogeneous lung dose (examples in Figs. 1 and 2). Some of the non-coplanar groups increased the doses to the liver and kidneys. Adding 'optimized' compensators and changing beam weights had minimal impact on the results. The results for the two patients were nearly identical. Conclusions: The more complex beam arrangements generally provide a 'lower' dose to the heart than the conventional AP/PA beams, while still providing equivalent therapeutic doses of elective RT to the lungs. Thus, the therapeutic ratio of BWLI might be 'better' with multiple non-coplanar or transverse coplanar beams than that achievable with traditional AP/PA beams. These complex beam groups can only be generated using sophisticated 3D treatment planning tools. This approach might be useful if 'elective' BWLI is to be studied further clinically in the 'adriamycin age'