Purpose: A prostate brachytherapy program was initiated in 1990, when comparatively little was known of the relative importance of disease- and treatment-related factors on outcome. Patients treated during the first 6 years of the program were analyzed to determine the value of patient selection and implant quality on biochemical control. Methods and Materials: We treated 243 patients with clinically localized prostate cancer with radioactive seed implantation and underwent 1-month CT-based dosimetric analysis. Follow-up ranged from 61 to 135 months (median 75). The Gleason score was less than or equal to6 in 78% (n = 189),7 in 14% (n = 35), and 8-10 in 8% (n = 19). The initial prostate-specific antigen (PSA) level was less than or equal to10 ng/mL in 61% (n = 149),10.1-20 ng/mL in 26%,(n = 63), and >20 ng/mL in 13% (n = 31). The disease stage was T2a or less in 49% (n = 120), and Stage T2b-T2c in 51% (n = 123). A real-time ultrasound-guided technique was used with I-125 (n = 138) and Pd-103 (n = 105) isotopes. No patient underwent external beam radiotherapy as part of their primary treatment. Of the 243 patients, 60% also received hormonal ablation for at least 3 months before and 2-3 months after seed implantation. All patients included underwent a 1-month CT-based dosimetric analysis. The implant dose was defined as the dose delivered to 90% of the prostate volume on postimplant dosimetry (D-90). On the basis of prior dose-response analyses, patients were retrospectively grouped into optimal D-90 (I-125 greater than or equal to 140 Gy Task Group 43 or Pd-103 greater than or equal to100 Gy) and suboptimal D-90 (I-125 < 140 Gy or Pd-103 < 100 Gy) dose groups. Biochemical failure was defined using the American Society for Therapeutic Radiology Oncology definition. Results: Disease-related factors, including initial PSA level, Gleason score, and stage, were significant predictors of biochemical failure. The actuarial 8-year freedom from biochemical failure (bFFF) rate was 80% for those with a PSA level less than or equal to10 ng/mL, 86% for PSA 10.1-20 ng/mL, and 45% for PSA >20 ng/mL (p = 0.0019). Patients with a Gleason score of:56 had an 8-year bFFF rate of 81% vs. 67% for those with Gleason score 7 and 53% for those with Gleason score 8-10 (p = 0.0003). Patients with Stage T2a or less had an 8-year bFFF rate of 85% compared with 69% for those with Stage T2b-T2c (p = 0.013). The 8-year bFFF rate was 88% for low-risk patients (Stage T2a or less, Gleason score less than or equal to6, and initial PSA level less than or equal to10 ng/mL; n = 75),81% for moderate-risk patients (Stage T2b or Gleason score 7 or initial PSA level >10.1-20 ng/mL; n = 70), and 65% for high-risk patients (two or more moderate-risk features or Gleason score greater than or equal to8 or initial PSA level >20 ng/mL; n = 98; p = 0.0009). Patients with optimal dose implants (n = 145) had an 8-year bFFF rate of 82% compared with 68% for those with suboptimal dose implants (n = 98; p = 0.007). Hormonal therapy did not significantly affect biochemical failure (p = 0.27). In multivariate analysis, the statistically significant variables included initial PSA level (p < 0.0001), Gleason score (p = 0.024), and dose group (p = 0.046). Because our current practice limits implantation alone to low-risk patients, an analysis of this subgroup was undertaken to validate the importance of dose. In the optimal dose group, low-risk patients had an 8-year bFFF rate of 94% vs. 75% for the low-risk patients in the suboptimal dose group (p = 0.02). Conclusion: With minimal follow-up of 5 years, these data continue to support the use of implantation alone in low-risk prostate cancer patients and demonstrate the importance of implant quality (dose) in achieving optimal outcomes. Low-risk patients who receive an optimal dose implant have a 94% bFFF rate at 8 years. (C) 2003 Elsevier Inc.
