Effectiveness of Frequently Used TNF Inhibitors vs. Conventional Immunosuppressive Therapies for Noninfectious Uveitis.

Purpose: To compare the corticosteroid sparing efficacy of frequently used antimetabolites to tumor necrosis factor (TNF) inhibitors in the management of noninfectious ocular inflammation.

Methods: Retrospective analysis of patients with noninfectious uveitis on conventional antimetabolite (methotrexate, mycophenolate mofetil, azathioprine,or leflunomide, "CONV") or a TNF inhibitor (adalimumab or infliximab, "TNFi") with active inflammation or more than 7.5 mg daily prednisone. Eyes were assessed in three groups: CONV only, TNFi only and combination of both (COMB). Cox regression models compared treatment success, adjusted for age, race, smoking, anatomic location of uveitis, duration of uveitis and visual acuity. Corticosteroid sparing success was defined as: inactive or slightly active uveitis on <=7.5 mg daily oral prednisone and <=2 drops of prednisolone acetate 1%.

Results: There were 1475 eligible patients in the analysis. By 6 and 12 months, respectively, the Cox model-predicted, percentage success was 27.6% and 44.9% for the CONV group; 34.2% and 53.9% in the TNFi group and 39.9% and 61.1% for the COMB group. COMB was more likely than CONV to achieve success (adjusted HR 1.58 (95% confidence interval (CI), 1.28, 1.95, p < 0.0001). Factors associated with lower success were age under 18 years, smoking, visual acuity worse than 20/50 at cohort entry, over 4-year duration of uveitis and daily baseline prednisone 7.5 mg or higher (all p < 0.05).

Conclusion: Our results suggest COMB is more effective than CONV at achieving disease quiescence and corticosteroid sparing success among patients with active noninfectious uveitis. More research is needed to determine if TNFi alone is superior to CONV for uveitic corticosteroid-sparing.