Penetration of a topically administered anti-tumor necrosis factor alpha antibody fragment into the anterior chamber of the human eye

Abstract

Objective: To determine whether topically applied ESBA105, a single-chain antibody fragment against tumor necrosis factor (TNF)-α, could efficiently penetrate into the anterior chamber of the human eye.

Design: Multicenter, interventional cohort study.

Participants: Otherwise healthy patients undergoing cataract surgery (cohorts I-III) or combined cataract surgery and vitrectomy (cohort IV).

Methods: ESBA105 (n = 57) or placebo (n = 22) was preoperatively applied as eye drops to 1 eye in patients scheduled for cataract surgery (n = 73) or combined cataract surgery and vitrectomy (n = 6). ESBA105 was administered on the day of surgery at 1-hour intervals (last dose 1 hour preoperatively) as 1.6 mg in 4 drops for cohort I (n = 15) and as 3.2 mg in 8 drops for cohorts II (n = 15) and IV (n = 6). Cohort III (n = 43) was randomized 1:1 in double-masked fashion to receive either ESBA105 6.4 mg or placebo over 4 days using 4 drops per day at 4-hour intervals (last dose 12 hours preoperatively). Aqueous humor (all cohorts), vitreous humor (cohort IV only), and blood samples (all cohorts) were collected for measurement of ESBA105.

Main outcome measures: ESBA105 intraocular concentration.

Results: Both 4 times daily over 4 days dosing (cohort III) and 8 times daily dosing (cohorts II and IV) resulted in reliably high ESBA105 concentrations in aqueous humor. Mean molar excess of intraocular ESBA105 over its target (intraocular TNF-α) was calculated as 96-fold (cohort III) to 359-fold (cohorts II and IV). Results from the cohorts receiving 4 and 8 hourly drops per 1 day (cohorts I, II, and IV) indicated that dose-dependent intraocular concentrations of ESBA105 were achieved within hours of dosing. After 8 times daily dosing, 5 of 6 vitreous samples (cohort IV) had undetectable ESBA105 levels. ESBA105 was detected in 17 of 55 preoperative serum samples but no longer detectable in serum 1 day after surgery (0 of 19 samples). In cohort III, treatment-emergent adverse events were identical between ESBA105 and placebo groups (2 cases each of eye irritation).

Conclusions: These results demonstrate that the topically applied single-chain antibody fragment ESBA105 penetrated into the anterior chamber of the human eye at therapeutic levels.