Efficacy of virtual block objects in reducing the lung dose in helical tomotherapy planning for cervical oesophageal cancer: a planning study.

Background: Intensity-modulated radiotherapy is useful for cervical oesophageal carcinoma (CEC); however, increasing low-dose exposure to the lung may lead to radiation pneumonitis. Nevertheless, an irradiation technique that avoids the lungs has never been examined due to the high difficulty of dose optimization. In this study, we examined the efficacy of helical tomotherapy that can restrict beamlets passing virtual blocks during dose optimization computing (block plan) in reducing the lung dose.

Methods: Fifteen patients with CEC were analysed. The primary/nodal lesion and prophylactic nodal region with adequate margins were defined as the planning target volume (PTV)-60 Gy and PTV-48 Gy, respectively. Nineteen plans per patient were made and compared (total: 285 plans), including non-block and block plans with several shapes and sizes.

Results: The most appropriate block model was semi-circular, 8 cm outside of the tracheal bifurcation, with a significantly lower lung dose compared to that of non-block plans; the mean lung volumes receiving 5 Gy, 10 Gy, 20 Gy, and the mean lung dose were 31.3% vs. 48.0% (p <  0.001), 22.4% vs. 39.4% (p <  0.001), 13.2% vs. 16.0% (p = 0.028), and 7.1 Gy vs. 9.6 Gy (p <  0.001), respectively. Both the block and non-block plans were comparable in terms of the homogeneity and conformity indexes of PTV-60 Gy: 0.05 vs. 0.04 (p = 0.100) and 0.82 vs. 0.85 (p = 0.616), respectively. The maximum dose of the spinal cord planning risk volume increased slightly (49.4 Gy vs. 47.9 Gy, p = 0.002). There was no significant difference in the mean doses to the heart and the thyroid gland. Prolongation of the delivery time was less than 1 min (5.6 min vs. 4.9 min, p = 0.010).

Conclusions: The block plan for CEC could significantly reduce the lung dose, with acceptable increment in the spinal dose and a slightly prolonged delivery time.