Comparison of dose length, area, and volume histograms as quantifiers of urethral dose in prostate brachytherapy

To determine the magnitude of the differences between urethral dose-volume, dose-area, and dose-length histograms (DVH, DAH, and DLH, respectively, or DgH generically).

Six consecutive iodine-125 ((125)I) patients and 6 consecutive palladium-103 ((103)Pd) patients implanted via a modified uniform planning approach were evaluated with day 0 computed tomography (CT)-based dosimetry. The urethra was identified by the presence of a urinary catheter and was hand drawn on the CT images with a mean radius of 3.3 +/- 0.7 mm. A 0.1-mm calculation matrix was employed for the urethral volume and surface analysis, and urethral dose points were placed at the centroid of the urethra on each 5-mm CT slice.

Although individual patient DLHs were step-like, due to the sparseness of the data points, the composite urethral DLH, DAH, and DVHs were qualitatively similar. The DAH curve delivered more radiation than the other two curves at all doses greater than 90% of the prescribed minimum peripheral dose (mPD) to the prostate. In addition, the DVH curve was consistently higher than the DLH curve at most points throughout that range. Differences between the DgH curves were analyzed by integrating the difference curves between 0 and 200% of the mPD. The area-length, area-volume, and volume-length difference curves integrated in the ratio of 3:2:1. The differences were most pronounced near the inflection point of the DgH curves with mean A(125), V(125), and L(125) values of 36.6%, 31.4%, and 23.0%, respectively, of the urethra. Quantifiers of urethral hot spots such as D(10), defined as the minimal dose delivered to the hottest 10% of the urethra, followed the same ranking: area analysis indicated the highest dose and length analysis, the lowest dose. D(10) was 148% and 136% of mPD for area and length evaluations, respectively. Comparing the two isotopes in terms of the amount of urethra receiving a given dose, (103)Pd implants were significantly cooler than (125)I implants over most of the range of clinical interest, from 100% to 150% of mPD.

Dose gradients in prostate implants result in the observed ordering of DAH, DVH, and DLH from higher to lower doses. The three histogram approaches remain in close agreement up to 100% of the mPD but diverge at higher doses. Although urethral point doses are the most easily determined, they underestimate the amount of urethra at risk at higher doses compared to dose area analysis. Because dosimetric parameters detailing high-dose regions such as D(10) show only slight differences between calculation methods, they are recommended over the corresponding geometric entities G(150) or G(175). The differences between the D(gg) entities are sufficiently small that they are unlikely to be of clinical significance or to confound analyses attempting to correlate urinary morbidity with urethral dosimetry.