Dosimetric perturbations of linear array of beta-emitter seeds and metallic stent in intravascular brachytherapy

The radiation treatment with catheter-based beta-emitter sources is under c linical trials to prevent restenosis following interventional coronary proc edures. There are still large uncertainties in the dose calculation due to the complicated treatment geometry. We present the Monte Carlo simulations to account for the dosimetric perturbations due to neighboring trained seed s, proximal/distal gold markers, and a stainless steel stent. A catheter-ba sed beta-emitter system is modeled using the Monte Carlo code, MCNP4B. Dose distributions and dose rates are calculated in voxels (0.64 x 0.64 x 0.5 m m(3)) around the long cylindrical trains of Sr-90/Y source with and without the stent (at 1.92 mm from the source axis). For the total activity of 70 mCi (2.59 x 10(9) Bq). the dose around most of the source length (except fo r edge seeds and gold markers) varies from 40 to 0.23 cGy/s as the radial d istance from the source axis (r) increases from 0.63 to 6.4 mm. At the pres cription range of r = 1.5-4.0 mm, the dose gradient is very steep and the c ontribution of neighboring seeds to the dose is significant. The dose enhan cement due to neighboring seeds (the so-called "train effect") varies from 9% to 64% as r increases from 0.64 to 5.2 mm. The doses at r = 2 mm from th e last edge seed and the gold marker are about 80% and 40% of that of the n onedge seed (8.7 cGy/s), respectively. The dose enhancement due to the seco ndary electrons and the primary electrons scattered with the stent is shown to be about 9.3% in the voxel including the stent. However, as r increases beyond the stent (r = 2.0-6.4 mm), the dose is slightly reduced by 4%-12%, compared to that without the stent. (C) 2000 American Association of Physi cists in Medicine. [S0094-2405(00)01102-0].