The impact of daily shifts on prostate IMRT dose distributions

It is becoming common clinical practice to shift prostate patients daily based on transabdominal ultrasound (US) or imaging of implanted fiducial markers. This paper investigates the dosimetric impact of these shifts by looking at five patients shifted using an optically guided 3D US localization system and treated with IMRT. Treatment plans were generated for each patient for the following 3 cases: (1) the initial preplan, which represents the ideal case in which no shifts are necessary; (2) a postplan incorporating each day's actual shifts; and (3) a postplan in which no shifts were made but the internal organs move by the amounts indicated by daily US imaging. Results show that when daily shifts are made, doses to the target, rectal wall, and bladder wall are nearly identical to those in the preplan. Equivalent uniform dose (EUD) and tumor control probability (TCP) for these plans were also the same as for the preplans. When no shifts were made, however, the dose distributions were degraded, and the computed target EUD and TCP were lower for all five patients. The magnitude of these differences varied: for three patients, the TCP was only 1%-2% lower than for the preplan. For the other two patients, however, the EUD was reduced by more than 10%, resulting in TCP reductions of 6% and 11%. These results indicate that for a symmetric beam arrangement and properly chosen margins, shifting the patient each day and treating without recalculating the dose is unlikely to affect local control or the sparing of normal tissues. (C) 2004 American Association of Physicists in Medicine.