Abstract
Abstract: :
Purpose:Scleral sutured tantalum markers are used for tumor localization in proton beam radiotherapy (PBRT) of uveal melanomas. This study examines how varying the number of markers placed affects treatment outcomes. Methods:Seventy-four patients with uveal melanoma were treated with PBRT between April 1995 and August 2001. Tantalum marker placement was varied in a non-randomized fashion with patients receiving a minimum of 3 and a maximum of 7 markers per eye. Tumor size was designated as small, medium, large, or extra-large. Patients received between 48 and 56 Gy. Median age was 61.2 years. Median follow-up was 18.2 months. Results:Thirty patients received 3 markers, 22 patients received 4 markers, and 22 patients received 5 or more markers. Neovascular glaucoma (NVG) rates were similar in all groups, ranging from 9.5% to 14%. Cataract formation was similar in all groups, 16% to 20%. Enucleation was performed in 17% of patients with 3 markers, 9.5% with 4 markers, and 24% with 5 or more markers. Dry eye symptoms were less frequent in patients with 3 markers, 10%, compared to 4 or more markers, 31%. Lash loss, 27%, and radiation retinopathy, 13%, were more frequent in patients with 3 markers compared to those with 4 or more markers, 9.3% (lash loss) and 7% (retinopathy). The subset of small and medium sized tumors was compared further. None of these patients had NVG or underwent enucleation. In this subset, patients with 3 markers experienced lash loss of 29% and radiation retinopathy of 19%, compared to those with 4 or more markers, 4.7% (lash loss) and 4.5% (retinopathy). Dry eye symptoms were seen in 10-26% of those with 3 or 4 markers and none of those with five or more markers. Conclusion:In patients with small and medium sized tumors, increased marker number correlated with a decrease in complications such as lash loss, retinopathy, and dry eye symptoms. Complications such as NVG and enucleation were limited to patients with large or extra-large tumors. Cataract formation was similar in all groups and may be related to the median age of the patient population.