Improvement in dose escalation using off-line & on-line image feedback in the intensity modulated beam design for prostate cancer treatment

Abstract

Purpose: To test the capability of dose escalation in the IMRT process where the organ/patient temporal geometric variation, measured using either off-line or on-line treatment CT and portal images, are adapted for the optimal design of intensity modulated beam. Materials and Methods: Retrospective study was performed on five prostate cancer patients with multiple CT scans (14∼17/patient) and daily portal images obtained during the treatment course. These images were used to determine the displacements of each subvolume in the organs of interest caused by the daily patient setup and internal organ motion/deformation. The temporal geometric information was processed in order of treatment time and fed into an inverse planning system. The inverse planning engine was specifically implemented to adapt the design of intensity modulated beam to the temporal subvolume displacement and patient internal density changes. Three image feedback strategies were applied to each patient and evaluated with respect to the capability of safe dose escalation. The first one is off-line image feedback, which designs the beam intensity based on the patient images measured within the first week of treatment. The second is an on-line ’the target of the day’ strategy, which designs the beam intensity in daily bases by using ’the image of the day’ alone. The last one is also the on-line based. However, it designs the instantaneous beam intensity based on also dose distribution in each organ of interest received prior to the current treatment. For each of the treatment strategies, the minimum dose delivered to the CTV was determined by applying the identical normal tissue constraints of partial dose/volumes. This minimum dose was used to represent the treatment dose for each patient. Results: The off-line strategy appears feasible after 5 days of image feedback. The average treatment dose among the patients can be 10% higher than the one in the conventional IMRT treatment (Table) where the inverse planning was performed based on the predefined planning target volume (CTV + 1cm margin. The capability of safe dose escalation increases, on average, extra 15% in the first on-line strategy. The second on-line feedback strategy shows only a marginal improvement on the total treatment dose compared to the first one. However, it shows further improvement on the dose distribution in the normal organs. Conclusions: The capability of safe dose escalation can be significantly enhanced in both the off-line and the on-line image feedback strategies. In the off-line image feedback strategy, patient/organ geometric information obtained from the first week of treatment appears adequate to guide the inverse planning design for remaining treatment. In the on-line image feedback strategy, the knowledge of organ dose distribution received from the previous treatment improves the design of beam intensity. Supported in part by NCI-CA71785 & Deutsche Krebshilfe. Patient# Off-line vs Conventional IMR On-line vs Off-line 1 6.1% 16% 2 6.4% 23% 3 15.8% 14% 4 15.9% 5.9% 5 7.7% 14.5% MEAN 10.4% 14.7%