Numerically Modeling the Dosimetric Consequences of Seed Position Uncertainties in COMS-Based Eye Plaques

Abstract

r I-125 10 2.5 (2.2-2.9) 2.9 (2.2-3.6) 16 2.0 (1.6-2.5) 2.4 (1.5-4.9) 22 1.5 (1.1-1.9) 1.9 (1.0-5.5) Pd-103 10 3.0 (2.6-3.3) 3.5 (2.6-4.2) 16 2.8 (2.1-3.5) 3.1 (1.9-6.4) 22 2.0 (1.4-2.5) 2.3 (1.2-6.2) 4 I-125 10 4.0 (3.3-4.8) 4.4 (3.1-7.8) 16 2.4 (2.0-2.9) 2.8 (2.0-4.2) 22 2.3 (1.9-2.8) 2.9 (1.8-4.6) Pd-103 10 5.8 (4.8-6.7) 6.2 (4.4-10.5) 16 3.2 (2.8-3.7) 3.7 (2.7-5.0) 22 3.1 (2.6-3.6) 4.0 (2.6-5.5) Ẑ I-125 10 4.9 (4.8-5.1) 5.1 (4.8-5.8) 16 4.2 (3.6-4.8) 4.9 (3.6-10.3) 22 5.7 (3.8-7.5) 9.0 (3.7-34.3) Pd-103 10 5.7 (5.6-5.7) 5.9 (5.6-6.2) 16 5.6 (4.7-6.5) 6.4 (4.6-13.4) 22 7.9 (5.3-10.5) 9.8 (4.8-28.7) Methods and Materials: Aneye plaque dose calculation routinewas created usingaTG-43 line-sourcemodel.Seedpositionuncertaintiesweremodeledby introducing Gaussian random displacements of each seed in predefined directions: parallel to the plaque central axis (z), parallel to the long axis of the seed groove (4), and in the direction of lateral displacements toward the groove sidewall (r). Distributions of dosimetric outcomes were obtained and fitted after 106 random seed displacement iterations were performed. A variety of clinical configurations were simulated, including I-125 and Pd103 seeds, three eye plaque sizes (10 mm, 16 mm, 22 mm), and eight representative plaque/eye orientations. Dose uncertainties were quantified at four anatomical sites (fovea, optic disc, lens center, lacrimal gland) and three central axis plaque depths (5 mm, 10 mm, 22.6 mm). A comparable analysis was performed for deterministic, systematic plaque displacements. The dose difference per unit plaque shift was calculated for shifts both along the eye surface and radially outward from theglobe center. This latter direction is the systematic analogof randomseeddisplacements along the plaquecentral axis. Results: Seed position uncertainty along the plaque central axis (z direction) typically produced the largest (followed by r then 4) dosimetric uncertainties. The standard deviation of dose outcomes across all the simulated clinical scenarios is rarely greater than 2% when assuming an RMS seed deviation amplitude of 0.2 mm in z. Dosimetric uncertainties generally increase with decreasing plaque size, and are systematically higher in Pd-103 plaques. This pattern is attributed to the presence of higher dose gradients, which magnify the effects of seed position uncertainties in these configurations. Systematic seed shifts produce larger fractional effects than random seed shifts for most scenarios. A linear fit to a systematic vs. random seed displacement (parallel to plaque central axis) scatter plot of dosimetric uncertainties rendered a slope of 2.45 for points in the small dosimetric uncertainty limit. Conclusions: For realistic random seed displacements, the magnitude of dose discrepancies is consistently less than 2%, only exceeding this value at locations close to the plaque. This result also applies to uncertainties between planned and fabricated seed positions in new eye plaque designs. Dosimetric uncertainties due to random seed position deviations are much smaller in magnitude than the differences associated with homogenous (TG-43) vs. heterogeneity-corrected dose calculations.