Introduction: Breast cancer radiation therapy often employs deep inspiration breath-hold techniques to protect nearby organs at risk. AlignRT is an optical surface tracking system that allows patient position correction and control during radiation. Purpose: Our study investigates the accuracy and reproducibility of the AlignRT system during deep inspiration breath-hold radiation for breast cancer patients. We aim to assess the influence of patient characteristics on setup errors and optimize our institution's (OIL) geometric verification process, which combines AlignRT, orthogonal and tangential treatment verification imaging. Methods: We retrospectively analyzed the data of fifty patients, categorized into two groups: those with a simultaneous integrated boost to the tumor bed and those without. Data included shifts in six degrees of freedom, patient age, tumor location, and cup size. Results: Comparing patient groups receiving different radiation types, we found a statistically significant age difference (p = 0.022), with the simultaneous boost group being younger. The average total magnitude of shifts exceeded the predefined tolerances for translational and rotational shifts (5 mm and 3°) in ten patients. In patients without simultaneous boost, three on average exceeded the 5 mm tolerance in the LONG(+) direction after tangential control imaging. Age correlated significantly with the average total shift magnitude in the LONG(+) (p = 0.015, ρ = -0.342) and ROLL(+) (p = 0.045, ρ = -0.285) directions. In comparing the upper and lower breast quadrants, the lower quadrant group exhibited significantly larger average total shifts in the LAT(+) direction (p = 0.012). Similarly, in the comparison between the inner and outer quadrants, the inner quadrant group showed significantly larger shifts in the PITCH(-) direction (p = 0.041). Discussion and conclusion: In our study, we have confirmed the need for complementary use of AlignRT and treatment verification imaging that consists of both orthogonal and tangential imaging. We have additionally confirmed a connection between age and the extent of setup errors, as well as an impact of tumor location on the magnitude of setup errors.
|