Dosimetric evaluation of volumetric-modulated arc therapy (RapidArc) for primary leiomyosarcoma in the spine

This study aims to investigate the suitability of volumetric-modulated arc therapy (VMAT) with RapidArc for primary leiomyosarcoma (LMS) in the spine, and present a new method to improve the target coverage and organs at risk (OAR) sparing. Five patients with LMS were retrospectively reviewed. The intensity-modulated radiotherapy (IMRT) with five coplanar beams (5b-IMRT) or seven coplanar beams (7b-IMRT), and VMAT using four quasi-quarter coplanar arcs (4q-VMAT) or two full coplanar arcs (2f-VMAT) were generated. Planning target volume (PTV) dose coverage, OAR dose sparing, conformity index (CI), and homogeneity index (HI) were evaluated. A hollow-cylinder model (HCM) was also used for feasible optimal beam arrangements. The mean doses to PTV were 95.2% ± 1.0%, 93.0% ± 1.0%, 97.9% ± 1.0% and 96.2% ± 1.5% for 4q-VMAT, 2f-VMAT, 5b-IMRT and 7b-IMRT respectively, while the mean maximum doses to spinal cord (SC) were 43.7 ± 0.9 Gy, 42.0 ± 0.8 Gy, 41.4 ± 1.2 Gy and 40.6 ± 1.4 Gy. Compared to 5b-IMRT, the mean doses delivered to kidneys decreased by about 35.1% (8.5 Gy), 2.5% (0.6 Gy) and 35.5% (8.6 Gy) for 4q-VMAT, 2f-VMAT, and 7b-IMRT, respectively. The CI proposed by Baltas et al. was twice as good with IMRT than with 4q-VMAT, and the numbers of monitor units were increased five- and threefold with 7b-IMRT and with 5b-IMRT compared to VMAT. The unexpected results we presented here show that VMAT technique can't achieve highly conformal treatment plans while maintaining SC sparing for LMS in the spine. An approach is proposed based on a hollow-cylinder model, but it is difficult to apply to clinical practice. In this case, VMAT is not superior to IMRT except for significant reduction in delivery time.