Dosimetric impact of inter-observer catheter reconstruction variability in ultrasound-based high-dose-rate prostate...

To investigate the dosimetric impact of interobserver catheter reconstruction variability in transrectal ultrasound-guided prostate high-dose-rate (HDR) brachytherapy.

Twenty consecutive patients with intermediate- or high-risk prostate cancer were treated with a single, 15-Gy HDR brachytherapy boost as part of this study. The treated plan was used as the study reference plan (P_R). Three expert treatment planners (observers) manually reconstructed the catheter paths on the static three-dimensional transrectal ultrasound images, and new plans were generated from the updated positions (P_OBS); subsequently, the dwell time and positions from the P_OBS plans were superimposed on the P_R catheter paths to evaluate the dosimetric effect of the interobserver variations (P_EVAL). Plans from each group were stratified by observer and by number of catheters (12 or 16) and then compared using a one-way Kruskal-Wallis H test with post hoc Mann-Whitney U tests reserved for significant variations (α = 0.05).

Greater than 98.9% of catheter reconstruction variations were <3 mm. When stratified by observer, there was a significant decrease (p << 0.05) in planning target volume (PTV) V_100% and increases in the urethral D_max between the P_OBS plans propagated to the P_R catheter paths and dosimetry evaluated and P_R plans only. Stratification of plans by catheter number showed nonclinically significant decreases in PTV V_100%, and D_90% and increases in urethral D_max for the 12-catheter plans.

Limiting interobserver variability, and its effects on prostate HDR brachytherapy plan quality, is critical to achieving good dosimetric outcomes; small variations in catheter reconstruction may translate to inadequate PTV coverage, excessive urethral dose, or both.