Methods of Accurate 106Ru and 125I Eye Plaque Dosimetry Using Radiochromic Film in a Solid Water “Eye” Phantom and a Small Silicon Diode in a Water Tank

Purpose: The use of 106Ru eye plaques for the treatment of intraocular malignancies has produced inconsistent clinical outcomes and has even resulted in treatment failures. I hypothesized that inconsistent clinical results were attributable to high uncertainties in 106Ru eye plaque dosimetry. Furthermore, I hypothesized that more accurate methods for assessing radiation dose from eye plaques would lead to more reliable treatment planning and therefore better overall clinical outcomes. Methods: A Solid Water “eye” phantom with several novel features was developed for radiochromic film eye plaque dosimetry. Films perpendicular to the central axis of the eye plaques were sandwiched between inserts in the phantom. Small holes in the inserts enabled the film to be marked with respect to the eye plaques, assuring exact geometrical co-registration. In cooperation with the manufacturer, special thin radiochromic films were developed and utilized to permit dosimetric measurements almost at the eye plaque surface. Precise film punches were developed for the purpose of cutting films with diameters as small as 8.5 mm and making cutouts in films without damaging the cut edges. Findings from a secondary dosimetry system, utilizing a small silicon diode in a water tank, were compared to film data. In addition to testing the new dosimetry methods with 106Ru eye plaques, which utilize high energy (MeV) β emissions, this approach was also applied to 125I containing eye plaques, which due to their inherently lower energy (keV) γ emission spectrum, raised additional dosimetric complications. In the latter case dosimetry, films and the diode were calibrated for absolute dosimetry using calibrated 125I seeds in Solid Water and water, respectively, applying the TG-43 formalism. A novel calibration method of radiochromic film for low-energy photon dosimetry was introduced. Monte Carlo simulations were used to convert the results measured in Solid Water to liquid water, and to compare measured and simulated dosimetric results. Results: Dosimetric characterization of both 106Ru eye plaques and a novel concept 125I eye plaque are described. Furthermore, dosimetry of a 20 mm 125I Collaborative Ocular Melanoma Study (COMS) eye plaque validated the presumed substantial dose reduction resulting from its gold alloy backing and seed carrier insert predicted by Monte Carlo simulations. Dose distributions measured with radiochromic film were in good agreement with diode measurements and Monte Carlo simulations. Replicate film results were reproducible from 0.9% to 5.5%. As little as 4% non-uniformities in planar dose rates were easily detected using 106Ru eye plaques. The novel 125I eye plaques had uniform dose distributions. Dosimetric characterization of the 20 mm COMS plaque demonstrated that the plaque’s dose rate was 15% lower than that predicted by homogenous TG-43 calculations. Lastly, Monte Carlo simulations indicated dose conversion factors between water and film in Solid Water compared to water and Solid Water alone differed by as much as 16.8%. Change in the calcium content of Solid Water from 2.3% to 1.7% resulted in a 3.3% calculated difference in dose to film and in an 8.7% difference in dose to Solid Water. Conclusions: Precise and reproducible 106Ru and 125I eye plaque dosimetry was achieved utilizing radiochromic film in a water equivalent phantom and a small semiconductor diode in water. Co-registration of eye plaques and films permitted not only precise treatment planning calculations along the central axis of the plaque, but also made it possible to account for dosimetric non-uniformities using 2D or 3D methodologies. A calibrated 125I seed enabled calibration of the film and the diode for absolute dosimetry of 125I containing eye plaques. Dose measurements on the inner surface of the plaques provided precise assessment of the scleral dose, its homogeneity, and of the active area of the plaques for coverage determination. Monte Carlo simulations facilitated conversion of doses measured in various media to liquid water.