Infectious and Sterile Endophthalmitis in Eyes with Glaucoma Drainage Device from Two Large Ophthalmic Institutions.

PURPOSE

To describe the predisposing factors, presentation, management, and outcome of glaucoma drainage implant (GDI)-associated endophthalmitis.

DESIGN

Retrospective chart review.

PARTICIPANTS

Eyes that developed GDI-associated endophthalmitis between December 1, 2011, and December 23, 2019, at the Duke Eye Center and Cole Eye Institute.

METHODS

Patient data search was performed on the basis of diagnostic codes for GDI and endophthalmitis. Endophthalmitis was defined clinically according to each physician's discretion. Eyes with infection source other than GDI were excluded. Worse vision was defined as a decrease of more than 2 Snellen lines. Data were collected on baseline demographics, systemic and ocular comorbidities, ocular surgical history, best-corrected visual acuity (BCVA), intraocular pressure (IOP), clinical presentation, eye culture results, and treatments performed. Statistical analysis included the paired t test and odds radio calculations.

MAIN OUTCOME MEASURES

Visual acuity and IOP at final follow-up.

RESULTS

Thirty cases (0.7%) of GDI endophthalmitis were identified among 4073 GDIs performed at the 2 institutions with active follow-up. Device exposure was identified in 20 eyes (67%) on presentation. The average follow-up after presentation was 22.4 ± 25 months. The most frequently identified organism on culture was Streptococcus pneumoniae. Same-day injection of intravitreal antibiotics was the universal first-line therapy. From baseline to final follow-up, the mean BCVA decreased from -0.84 ± 0.77 to -1.30 ± 0.93 (logarithm of the minimum angle of resolution, P = 0.02). Mean IOP did not change from baseline to final visit in the overall cohort (16.2 ± 8.2 mmHg to 14.6 ± 9.4 mmHg, P = 0.30) and in the subgroup that underwent tube explant (15.9 ± 5.5 mmHg to 15.2 ± 10.4 mmHg, P = 0.97). Eighteen of 20 tube exposure cases (90%) underwent tube explant, 1 underwent tube revision, and 1 re-epithelialized.

CONCLUSIONS

Glaucoma drainage implant-associated endophthalmitis was correlated with poor visual outcome. Immediate intravitreal antibiotic delivery was a universal first-line therapy. Tube exposure was a necessary risk factor for late-onset endophthalmitis and required surgical removal or repair.