Author’s Disclosure Block: Dominique Geoffrion,  OpenAI Employment/honoraria/consulting fees , OpenAI, Employment/honoraria/consulting fees; Joseph Hakim, none; Younes Agoumi, none; Mona Harissi-Dagher, none 

Abstract Body 

Purpose: Glaucoma is the most important vision-threatening complication after Boston keratoprosthesis (KPro) and its management is challenging. Experts have yet to agree on the safest and most efficient tube shunt for KPro eyes.The purpose of this study was to compare the two most frequently used aqueous shunts for the treatment of glaucoma in KPro eyes. Study Design: Retrospective comparative cohort study. Methods:  A total of 37 eyes (36 patients) were implanted with a KPro (2008-2015) and were diagnosed with poorly controlled glaucoma.After KPro implantation, 32 eyes received the Ahmed FP7 glaucoma valve (AGV) for immediate intraocular pressure (IOP) control, and5eyes received the Baerveldt 101-350 implant (BGI) for long-term IOP control. The primary outcome was tube shunt failure, defined as IOP outside 6-21 mmHg or reduced by <20% from baseline for 2 visits, severe vision loss,or need for additional glaucoma surgery. Secondary outcomes were IOP, best-corrected visual acuity (BCVA), medication use, and postoperative complications.Differences in outcomes were compared using parametric and non-parametric tests, and log-rank test to compare time-to-outcome events. Results: Averagefollow-up time was 8.8±3.3 years (mean±SD) from KPro implantation. The mean age at tube shunt surgery was 59.3±15.3 years. Baseline demographic characteristics and KPro indications were similar between groups (P>0.05). Mean IOP decreased in AGV eyes by 27% (22 to 16 mmHg) at 5 years, and in BGI eyes, by 21% (19 to 15 mmHg) (P=0.607).At 5 years, BCVA improved by 32% in the BGI group, compared to a decrease of 20% in the AGV group (P>0.05). Medications per eye increased by 2% at 5 years in AGV, while they decreased by 20% in BGI (P=0.281).The most common cause of failure in both groups was the need for additional glaucoma surgery at 19% for both groups(P=0.665). At 5 years, the cumulative probability of failure was 55% for AGV and 20% for BGI (Kaplan Meier, log rank P=0.204).The major tube shuntcomplication for both groups was elevated IOP (16%). Conclusion: Both tube shunts were effective at reducing the IOP in KPro eyes.The BGI group had a lower failure rate and a larger IOP-lowering effect on fewer glaucoma medications, compared to the AGV group.