Dosimetry and cancer control after low-dose-rate prostate brachytherapy

To describe the relationship between two commonly used dosimetric quantifiers (dose received by 90% of the prostate [D(90)] and volume receiving 100% of dose [V(100)]) and biochemical disease-free survival (bDFS) in a cohort of men treated with low-dose-rate prostate brachytherapy (LDRPB).

The information in this report concerned the first 63 men treated with LDRPB alone at our institution between September 1997 and September 1998. All men had histologically confirmed, clinically localized prostate cancer. All men were treated with(125)I. The prescription dose was 144 Gy according to the Task Group 43 formalism. LDRPB was performed jointly by a radiation oncologist and urologist. Dosimetric quantifiers (D(90), V(100)) were calculated from a CT scan performed 1 month after LDRPB. Biochemical recurrence was defined according to the American Society for Therapeutic Radiology and Oncology consensus definition. Biochemical relapse-free survival (bRFS) was estimated using the product-limit method. D(90) and V(100) were examined as putative covariates for bRFS using the proportional hazards regression method. All p values are two-sided.

The median follow-up for the entire cohort was 62 months. The median D(90) was 122 Gy (range, 57-171Gy), and in 16 (25%) of 63 patients, the calculated D(90) was >140 Gy. The median V(100) was 81% (range, 51-97%). Nine men developed evidence of biochemical relapse at a median of 19 months (range, 6-38 months). The 5-year estimate of bRFS was 85% (95% confidence interval, 80-90%). The 5-year estimates of bRFS according to D(90) were as follows: D(90) > or =140 Gy, 86%; D(90) <140 Gy, 84% (p = not statistically significant). No threshold value of D(90) was predictive of the 5-year estimates of bRFS until the D(90) was <80 Gy (D(90) > or =80 Gy, 89%; D(90) <80 Gy, 50%; p = 0.02). The 5-year estimates of bRFS according to V(100) were as follows: V(100) > or =85%, 87%; V(100) <85%, 84% (p = not statistically significant). No threshold value of V(100) was predictive of the 5-year estimates of BRFS unless the dosimetry was particularly poor. The 5-year BRFS was 89% if the V(100) was > or =65% compared with 40% if the V(100) was <65% (p = 0.006).

The dosimetric or quantifiers described in this report did not predict for bRFS after LDRPB unless the dosimetry was very poor. This finding is not in complete agreement with those of previous reports. Possible reasons for this observation are (1) the study in underpowered, (2) inherent measurement error, (3) dosimetric quantifiers are poor surrogates of the dose received by the cancer, and (4) length of follow-up. Additional work in the area of quality assessment after LDRPB is required.