SGRT-based stereotactic body radiotherapy for lung cancer setup accuracy and margin of the PTV.

Objective: Surface-guided radiation therapy (SGRT, AlignRT) was used to analyze motion during stereotactic body radiotherapy (SBRT) in lung cancer patients and to explore the margin of the planning target volume (PTV).

Methods: The residual errors of the AlignRT were evaluated based on grayscale cone-beam computed tomography registration results before each treatment. AlignRT log file was used to analyze the correlation between the frequency and longest duration of errors larger than 2 mm and lasting longer than 2 s and maximum error with age and treatment duration. The displacement value at the end of treatment, the average displacement value, and the 95% probability density displacement interval were defined as intrafraction errors, and PTV1, PTV2, PTV3 were calculated by Van Herk formula or Z score analysis. Organ dosimetric differences were compared after the experience-based margin was replaced with PTV3.

Results: The interfraction residual errors were Vrt , 0.06 ± 0.18 cm; Lng , -0.03 ± 0.19 cm; Lat , 0.02 ± 0.15 cm; Pitch , 0.23 ± 0.7°; Roll , 0.1 ± 0.69°; Rtn , -0.02 ± 0.79°. The frequency, longest duration and maximum error in vertical direction were correlated with treatment duration (r = 0.404, 0.353, 0.283, p < 0.05, respectively). In the longitudinal direction, the frequency was correlated with age and treatment duration (r = 0.376, 0.283, p < 0.05, respectively), maximum error was correlated with age (r = 0.4, P < 0.05). Vertical, longitudinal, lateral margins of PTV1, PTV2, PTV3 were 2 mm, 4 mm, 2 mm; 2 mm, 2 mm, 2 mm, 3 mm, 5 mm, 3 mm, respectively. After replacing the original PTV, mean lung dose (MLD), 2-cm chest wall dose (CD), lung V decreased by 0.2 Gy, 2.1 Gy, 0.5%, respectively (p < 0.05).

Conclusion: AlignRT can be used for interfraction setup and monitoring intrafraction motion. It is more reasonable to use upper and lower limits of the 95% probability density interval as an intrafraction error.