Proton Therapy for Uveal Melanoma: A Single-Institution Experience

Abstract

<h3>Purpose/Objective(s)</h3> Uveal melanoma (UM) is the most common intraocular malignancy in adults. Eye-preserving approaches like episcleral plaque brachytherapy are common but limited in patients (pts) with large tumors or those adjacent to the optic nerve. Proton therapy (PT) offers an alternative eye-preserving treatment option. We report a single-institution experience utilizing PT to treat UM. We hypothesized that PT would provide excellent local control (LC) and eye preservation (EP) with acceptable toxicity and visual outcomes across all patients, including those with large tumors in adverse locations. <h3>Materials/Methods</h3> We reviewed the records of treatment-naive pts who received PT for non-metastatic UM between June 2012 and December 2020. All pts had ≥ 6 months of follow-up. Treatment planning was performed in EyePlan (Clatterbridge, UK) based on eye/tumor measurements and 3-4 implanted scleral fiducial markers, registered fundus photographs, and confirmed by CT/MRI. Pts were set-up in a seated position with gaze fixation and real-time image guidance using a single en face beam to a total 50-60 Gy RBE in 4-5 consecutive daily fractions. We estimated disease control and survival, including LC, EP, distant metastasis-free survival (DMFS), cause-specific survival (CSS), overall survival (OS), and grade 3+ toxicity as defined by CTCAEv5 using the cumulative incidence method. Regression of multiple variables simultaneously was accomplished with competing-risks Cox regression. <h3>Results</h3> 120 pts met inclusion criteria; 33 had a biopsy of their lesion; 9.2% were class 2. Moreover, 47.5% had COMS medium while 35.8% had large tumors. Median tumor thickness was 5.0 mm (0.5 – 13.0) and median basal dimension was 13.0 mm (4.0 – 25.0). Median distance to the fovea and optic disc was 3.5 mm. 38% had retinal detachment with 47% having significantly impaired visual acuity at baseline (worse than Snellen 20/40). 98.3% received 60 Gy RBE over 4 daily fractions. Median follow-up was 31 months (6 – 104). The 2-year estimates (95% CI) for LC, EP, DMFS, CSS, and OS were 94% (88-97%), 95% (90-98%), 88% (81-93%), 94% (88-97%), and 90% (83-95%), respectively. 9 pts underwent enucleation following PT (5 for recurrent/persistent disease). Retinopathy (5%), neovascular glaucoma (3.3%), and optic nerve disorder (2.5%) were the most common severe toxicities. 60% of toxicity resolved during follow-up. 38% had functional vision (20/40 or better) at last follow-up. Pretreatment visual acuity had a substantial impact on post-treatment vision. Tumor size was an independent prognostic factor for OS (p=0.01), CSS (p<0.01), EP (p=0.02), and DMFS (p<0.01). No prognostic factor was associated with LC. <h3>Conclusion</h3> Our experience using PT for UM is consistent with previously reported outcomes demonstrating high rates of LC and EP with acceptable toxicity despite a significant percentage of patients with large tumors and/or adverse location near the optic nerve.