All Movements are Not Created Equal in the Course of Prostate Cancer Treatment

Abstract

Purpose/Objective: IGRT technology can be used to correct for prostate motion which, based on data from multiple studies, can be as much as 2 cm. Both deformation and dislocation of the prostate can occur, with target day-to-day dislocation having the dominant effect. The same amount of movement may create different degradation of planning target volume (PTV) coverage for different movement directions when patients are treated with IMRT. This study quantifies dosimetric deviations for prostate movements in individual directions. Materials/Methods: A total of 10 prostate cases with the GTV ranging from 15 cc to 100 cc were selected for this study. IMRT plans that have 95% of the PTV (1 cm uniform margin from GTV) volume covered by the prescription dose of 75.6 Gy were generated. The dose objectives for the rectum and bladder were that less than 20% of each volume would receive more than 65 Gy. Eight gantry directions were used. Two IMRT plans using collimator angles 0o and 90o were generated to study the effect of different intensity distribution. To simulate prostate shifts, virtual systematic movements of the targets of 3 mm, 5 mm and 10 mm were introduced in the anterior (ANT), posterior (POST), superior (SUP), inferior (INF), left (LT) and right (RT) directions. The same IMRT plans were applied to the displaced targets. Dosimetric analysis was performed for the PTV coverage before and after the displacement. Results: For patients whose GTV was relatively small (<20 cc), a 3 mm displacement created the same amount of PTV coverage degradation as 5 mm displacement for larger GTVs (table 1): movement in the LT or RT direction brought 1–2% decrease in the volume covered; for ANT movement, PTV coverage fell 1–3%, POST: 3–16%; the SUP/INF movement decreased the volume covered by 1–10%. For a 10 mm displacement the degradation of PTV coverage are 3–20%, 3–20%, 4–18%, 12–27%, 3–24% and 4–24% for LT, RT, ANT, POST, SUP and INF directions respectively. From table 1, LT and RT movements caused similar dosimetric effects. Same amount of movement in POST direction may cause significantly more degradation of PTV coverage than in the ANT direction. This observation was most likely explained by the difference in the target shape ANT and POST. The effect from the same amount of INF or SUP movement varied, from causing identical effect to having significant difference (up to 10% deviation in PTV coverage). Some of the differences were caused by the different intensity distribution in either direction. When the lighter intensity pattern (lower fluence) was at the SUP edge and darker pattern (higher fluence) was along the INF edge, the same amount of movement in INF direction caused very little PTV coverage degradation while the superior movement may cause significant degradation. Conclusions: The relative dosimetric impact of target movement depends not only on the amount of movement, but also on the movement direction, individual target characteristic and IMRT treatment plan parameters. The target motion has to be evaluated in the context of dose variation. When we set action levels for IGRT, we have to consider not just the amount of the movement, but all the other parameters that affect the treatment delivery quality. Tabled 1