Factors Contributing to Bladder Filling Variability in Patients Receiving Radiotherapy for Prostate Cancer

Abstract

Prostate cancer patients receiving radiotherapy (RT) are treated with full bladders to minimize bladder dose and displace bowel from the treatment field. While limitations in controlling bladder filling reproducibility during RT have been shown to affect bladder and bowel dosimetry, the factors affecting bladder volume variability have not been elucidated. We herein evaluate the impact of prostate cancer risk group, bladder volume (BV) at the beginning of treatment, and pre-treatment International Prostate Symptom Score (IPSS) on metrics quantifying BV changes during RT.We retrospectively reviewed 78 prostate cancer patients treated from 2016-19 with definitive RT 8,100 cGy in 45 fractions (n = 32), definitive stereotactic body RT 3625 cGy in 5 fractions over 2.5 weeks (n = 24), and adjuvant or salvage RT 7020 cGy in 39 fractions (n = 22). Thirty-six patients with high or unfavorable intermediate risk disease received androgen deprivation therapy (ADT) prior to RT, while 29 patients with high risk disease received pelvic nodal RT in addition to prostate or prostate bed RT. Each patient completed a pre-treatment IPSS survey and received daily cone-beam CT (CBCT) scans for positioning during treatment. The BV at beginning of treatment was defined as the patient's first attempt to fully fill his bladder without assistance. BVs were contoured on the CT simulation scan and CBCT images from the first and last days of radiotherapy. Age, risk group, IPSS and initial BV were compared with change in BV (ΔBV), change in BV relative to initial BV (%Δ) and absolute change in BV (|ΔBV|) between first and last day of RT.Median age and IPSS were 69 (range 50-84) years and 8 (0-30), respectively. Mean and standard deviation initial BV, ΔBV, %ΔBV and |ΔBV| were 314 ± 152 cc, -59 ± 123 cc, -8 ± 40%, 106 ± 85 cc, respectively. On multivariable analysis, initial BV was negatively correlated with ΔBV and %ΔBV independent of age, prostate cancer risk group, ADT, pelvic nodal RT, and duration of treatment (P < 0.05). Initial BV and prostate cancer risk group were positively, while IPSS was negatively, correlated with |ΔBV| (P < 0.05).Prostate cancer patients with large initial BV experienced larger total and relative declines in BV during RT. High initial BV, increasing prostate cancer risk group, and low IPSS correlated with greater absolute changes in BV (|ΔBV|) over the course of treatment. Hence, maximizing bladder volume during simulation appears to negatively impact reproducibility, which must be taken into consideration when optimizing bladder volume reproducibility using a non-invasive approach during prostate radiotherapy. We have already implemented these findings into our simulation and treatment protocols and look forward to reporting our clinical outcomes in the near future.