Emptying the rectum before treatment delivery limits the variations of rectal dose-volume parameters during 3DCRT of prostate cancer

Purpose: To investigate the impact of rectum motion on dose-volume histograms of the rectum including filling and of the wall (DVH and DWH, respectively), during 3D-conformal radiotherapy (3DCRT) for localized prostate cancer. Materials and methods: Ten patients received a planning CT scan (CT 0) and 11-14 CT during 3DCRT for prostate cancer (total CT scans = 126). CT images were 3D matched using bony anatomy. A single observer drew the external contours of rectum and rectum wall and the CTV (prostate + seminal vesicles) on CT 0. Patients were asked to empty their rectum before every CT, as generally performed at the Institute for Cancer Research and Treatment (IRCC) before treatment delivery. Bladder was kept full by drinking 500 cm 3 of water 60 min before the scan, according to our protocol. A 4-field box 3DCRT technique was planned and dose statistics/dose-volume histograms of the rectum were calculated for each contour referred to CT 0, CT 1, ..., CT n for each patient. Average DVHs during treatment were calculated along with their standard deviation (SD rand) and compared to the planned DVH. The analyses on the patient population included the assessment of systematic deviation (average difference and SD, named SD sys) as well as the average SD rand value expressing the random component of organ motion. Rectum shifts were also assessed by anterior and lateral BEV projections. Results: As to the rectum, 8/10 patients showed a "better" average DVH than DVH on CT 0. Wilcoxon test showed a statistically significant reduction when correlating the difference Δ between the average DVH during therapy and planning DVH at CT 0: for instance V 70 Δ = -3.6% and p = 0.022, V 50 Δ = -5.5% and p = 0.022, D med Δ = -3.2 Gy and p = 0.007. Average values of DVH systematic difference (average difference between planning scan and treatment), standard deviations (SD sys) and average standard deviations of the random fluctuation (SD random) were -4.0%, 4.7% and 6.6%, respectively. Whilst the fluctuation results were slightly smaller for DWH. Volume analysis showed a slight systematic variation of the rectal volume between planning and treatment BEV. The average rectal volume during therapy was larger than at the planning CT in 8/10 patients. The systematic shifts of the rectal wall between the planning phase and the treatment were rather small, both below and above the flexure. The larger random fluctuation of the rectum shape was found to be in the cranial half (1 SD = 4.4 mm). Conclusions: The practice of carefully emptying the rectum during simulation and therapy for prostate cancer, which is a safe and simple procedure, reduces the impact of organ motion on dose-volume parameters of the rectum.