RESUMO
PURPOSE: Comparison of inverse optimization (IO) to modified peripheral (MP) and geometric optimization (GO) intraoperative computer planning options for permanent seed implantation (PSI) of the prostate. METHODS AND MATERIALS: One hundred ten patients underwent PSI with iodine-125. Three computer planning options were compared including MP loading, GO, and IO. Preimplant dose goals (prescribed dose [PD] of 144 Gy) and normal tissue constraints were determined at the outset by the participating physicians before intraoperative computer planning. A single computer planning system was used for this comparison. Postimplant dosimetry was performed at 4-5 weeks and compared for V(100) and D(90), urethral V(150), and rectal V(110) of the PD. Acute urinary morbidity was evaluated and compared. RESULTS: All three options achieved a similar preimplant median V(100) (97%). The median number of needles and seeds implanted was greater with GO (29, 75) compared to MP (16, 66) and IO (17, 66) (p<0.0001 and p=0.0024, respectively). Postimplant dosimetry showed that IO achieved a higher percentage with V(100) >95% of the PD in multivariate analysis (p=0.04) and a lower percentage postimplant D(90) <140 Gy (7%) than for MP/GO (26%) (p = 0.01). IO predicted for lower urethral dose (p=0.0169), despite a higher median D(90) (169 Gy) than either MP (159 Gy) or GO (151 Gy) (p = 0.0025). The median percentage V(150) urethra for IO was 8% vs. 16% for MP and 23% for GO (p = 0.0005). With a median followup time of 6 months, acute Grade 2 urinary symptoms were higher with GO (81%) vs. MP (36%) and IO (53%) (p = 0.0019). CONCLUSIONS: Dosimetric outcomes for IO compare favorably to either MP or GO when performed in real time for PSI. In contrast to GO, IO and MP demonstrated excellent correlation between the intraoperative and postoperative plans while using fewer total and interior placed needles and seeds. IO appears feasible as an alternative intraoperative planning solution for PSI.