Tracking target position variability using intraprostatic fiducial markers and electronic portal imaging in prostate cancer radiotherapy.

PURPOSE: Modern radiotherapy has achieved substantial improvement in tumour control and toxicity rates by escalating the total dose to the target volume while sparing surrounding normal tissues. It has therefore become necessary to precisely track tumour position in order to minimise geometrical uncertainties due to setup errors and organ motion. We conducted this prospective evaluation of prostate cancer patients treated with image-guided conformal radiation therapy at our institution. We implanted three fiducial markers (gold seeds) within the prostatic gland in order to quantify daily target displacements and to generate specific margins around the clinical target volume (CTV) to create an appropriate planned target volume (PTV).
MATERIALS AND METHODS: Between April and December 2009, ten patients affected with localised prostate cancer were transrectally implanted with three radio-opaque markers. Each patient underwent a computed tomography (CT) scan for planning purposes following proper bladder and rectum preparation. During treatment two orthogonal images were acquired daily and compared with previously generated digitally reconstructed radiographs. After manual localisation, comparison between the position of the gold seeds on the portal and reference images was carried out, and a set of extrapolated lateral-lateral (LL), anterior-posterior (AP) and cranial-caudal (CC) shift corrections was calculated and recorded. Couch corrections were applied with a threshold of 3 mm displacement.
RESULTS: Systematic and random errors for each direction were calculated either as measured according to displacement of the gold seeds prior to any couch movement and after couch position correction according to the radio-opaque markers. For skin marks, mean systematic and random errors were 0.12+2.94 mm for LL, 1.04+3.37 mm for AP, -1.14+2.71 mm for CC, whereas for seed markers, mean and systematic errors were 0.6+1.5 mm for LL, 0.51+2.45 mm for AP and -0.25+2.51 mm for CC. A scatter plot generated on all measurements after couch repositioning according to gold-seed displacement suggested a confidence range of shift distributions within 5 mm for LL, 8 mm for CC, and 7 mm for AP. The total systematic and random components were then used to calculate proper PTV in patients receiving conventional treatment (7 mm for LL and 9 mm for both AP and CC).
CONCLUSIONS: Prostate positional variability during a course of radiation treatment is strongly influenced by setup and organ motion. Organ tracking through fiducial markers and electronic portal imaging is able to reduce the spread of displacements, significantly contributing to improve the ballistic precision of radiation delivery.