Radiotherapy of bladder cancer: relevance of bladder volume changes in planning boost treatment.

PURPOSE This study aims to evaluate tumor motion with controlled changes of the bladder volume, and to assess the reproducibility of bladder (and tumor) position using a urinary catheter balloon as an immobilization device. METHODS AND MATERIALS First, three patients with tumor growths in three different bladder regions (trigone, left lateral wall, anterior wall) were evaluated. Three-dimensional CT-based reconstructed images were used to measure the displacement of the tumors when 100 cc were removed from a bladder originally filled with 170 cc of contrast. The 3D calculated boost beam arrangements and field sizes for the three tumors in the partially emptied bladders were used to simulate treatment of the same tumors in the maximally filled bladders. Dose-volume histograms were obtained. Second, verification of an ellipsoid model for bladder volume changes was undertaken in 41 patients. Third, in eight additional patients a urinary catheter balloon filled with 80-cc sterile saline solution was used in an attempt to reproduce the shape and spatial coordinates of the bladder during the boost treatment. A pair of orthogonal films with the 80-cc balloon filled with contrast material were taken at simulation and repeated twice at weekly intervals during radiotherapy. The reproducibility was quantified by sequentially calculating the common surface of the bladder images in each orthogonal view. RESULTS Target motion, especially in the craniocaudal axis, appeared to be more relevant for tumors arising in the bladder walls (15 mm) than in the trigone (5 mm). Underdosage (<95% of the prescribed dose to the target volume) was observed in 20, 20, and 50% (with 1 cm margins around the tumor) and in 10, 10, and 15% (with 1.5 cm margins around the tumor) of the tumors arising in the trigone, left lateral wall, and anterior wall, respectively. The ellipsoidal model was validated with a strong correlation coefficient allowing to establish a predictive model for bladder wall displacements as a function of bladder volume. In the balloon reproducibility study, mean reproducibility factors of 0.84 (+/-0.06) and 0.82 (+/-0.07) were obtained for both anteroposterior and lateral views, respectively. CONCLUSIONS Changes in bladder volume and shape related to bladder filling can result in clinically significant displacements of the target volume. A minimum of 2-cm margins around the target may compensate for extreme bladder volume changes during boost treatment. An ellipsoidal model for the bladder is consistent with these observations. Although an 80-cc urinary catheter balloon helped to immobilize the bladder, reproducibility was less than perfect.