AOS T HESIS Outcomes in Patients With Vogt-Koyanagi-Harada Disease From the First-Line Antimetabolites for Steroid-Sparing Treatment Uveitis Trial

Purpose: To compare the effectiveness of methotrexate (MTX) and mycophenolate mofetil (MMF) in achieving corticosteroid-sparing control of uveitis in patients with Vogt-Koyanagi-Harada (VKH) disease. Methods: A subanalysis of patients with VKH from the First-line Antimetabolites as Steroid-sparing Treatment Uveitis Trial, a randomized, observer-masked, comparative effectiveness trial, with comparisons by treatment (MTX vs MMF) and disease stage (acute vs chronic). Individuals with noninfectious uveitis were placed on a standardized corticosteroid taper and block randomized 1:1 to either 25 mg weekly oral MTX or 1.5 g twice daily oral MMF. The primary outcome was treatment success defined by corticosteroid-sparing control of uveitis at 6 months. Additional outcomes included change in best spectacle-corrected visual acuity (BSCVA), retinal central subfield thickness (CST), and resolution of serous retinal detachment (SRD). Results: Ninety-three out of 216 enrolled patients had VKH; 49 patients were randomized to MTX and 44 to MMF, of which 85 patients (46 on MTX, 39 on MMF) contributed to the primary outcome. There was no significant difference in treatment success by antimetabolite (80.4% for MTX compared to 64.1% for MMF; P = .12) or in BSCVA improvement (P = .78). MTX was superior to MMF in reducing CST (P = .003) and resolving SRD (P = .02). There was no significant difference in treatment success by disease stage (P = .25), but patients with acute VKH had greater improvement in BSCVA (P < .001) and reduction of CST (P = .02) than chronic VKH patients. Conclusions: MTX and MMF have comparable outcomes as corticosteroid-sparing immunosuppressive therapies for VKH. Visual acuity improvement was greater in acute vs chronic VKH. NOTE: Publication of this article is sponsored by the American Ophthalmological Society (c) 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.)