Purpose: To assess the influence of patient repositioning and organ motion on dose distribution within the prostate and the seminal vesicles (clinical target volume, (CTV)).
Material and methods: Nine patients were simulated and treated in the supine position, with an empty bladder, and without immobilization devices. While on treatment, patients underwent weekly pelvic computed tomography (CT) scans under conditions identical to those at simulation. Patients were aligned using lasers on anterior and lateral skin tattoos, onto which lead markers were placed. After each CT scan (n=53) the CTV was redefined by contouring, and a new isocenter was obtained. A six-field technique was used. The field margins around the CTV were 20 mm in the cranio-caudal axis, and 13 mm in the other axes, except in the lateral fields where a 10 mm posterior margin was used. Dose-volume histograms (DVHs) for each organ were compared with those determined at simulation, using the notion of the proportional change in the area under the CTV-DVH curve resulting from a change in treatment plan (cDVH).
Results: The reproducibility of the dose distribution was good for the prostate (%cDVH, mean+/-SD: -0.97+/-2.11%) and less than optimal for the seminal vesicles (%cDVH, mean+/-SD: -4.66+/-10.45%). When correlating prostate %cDVH variations with displacements of the isocenter in the Y axis (antero-posterior) the %cDVH exceeded (-)5% in only two dosimetries, both with an isocenter shift of >10 mm. For the seminal vesicles, however, ten out of 53 dosimetries showed a %cDVH exceeding (-) 5%. In nine of these ten dose distribution studies the posterior shift of the isocenter exceeded 8 mm.
Conclusions: Precise targeting of prostate radiotherapy is primarily dependent on careful daily set-up and on random changes in rectal geometry. Margins no less than 10 mm around the prostate and at least 15 mm around the seminal vesicles are probably necessary to insure adequate target coverage with a six-field technique.