McNamara, K.

15639 Background: Permanent prostate brachytherapy (PPB) assesses implant quality but may suffer significant variability due to: 1) Prostate edema may occur during the procedure which usually subsides within several months, and 2) interobserver variability of CT- derived prostate volumes. An analysis is described, which uses the seeds as prostate fiducials to produce an implant reference volume (IRV). The ratio of the IRV volumes at day 0 and day 33 provides a unique way to quantify prostate edema; thus reducing this factor as a source of variability. Methods: Thirteen low risk prostate cancer patients, (Stage ≤ T2, PSA<13.9, Gleason Score ≤7) were implanted with 125Iodine seeds (77% stranded; prescription dose of 145 Gy). PPB was completed for each patient on day 0 and 33 ±3 days following the implant. The IRV was defined as the volume obtained by contouring the outer most peripheral seeds, circumferentially, on each axial image, resulting in a 3D volume. Contouring was performed by three separate investigators (KJM, DCG, EKM), who were blinded with respect to patient identification. The IRV's were compared both by investigator and by time of PPB (day 0 vs. day 33). The radiation dose encompassing 90% (D90) of the IRV and Prostate (PST) was calculated and compared . Results: The IRVs contoured were reproducible between the investigators with a maximum mean difference of .56 (SD .33) cc. There was no statistical difference between any of the investigators contoured volume (minimum Spearman correlation was 0.978). The average IRV was significantly larger on day 0 (Avg. 52.7;SD 19.7 cc) compared to day 33 (Avg. 44.6; SD 19.1 cc) (p<0.001). The day 0 IRV was an average of 1.18 times larger than the Day 33 PRV. There was no difference between either the day 0 D90PST (Avg.149.1; SD18.7 cGy) vs the day 0 D90IRV (Avg.149.8, SD 27.5 cGy) (p=.442) nor between the day 33 D90PST (Avg.163.7; SD22.7 cGy) vs the day 33 D90IRV (Avg.166.6, SD 30.3 cGy) (p=.257). Both the D90IRV and D90PST were significantly different at day 0 vs 33 (p=.0003, p=<.0001 respectively). Conclusions: The IRV analysis demonstrates no significant interobserver variability, provides accurate quantification of prostate swelling after PPB, and is an accurate objective surrogate analysis for PSI quality. No significant financial relationships to disclose.