Abstract
Purpose: :
To estimate the incidence of and identify risk factors for glaucoma drainage implant (GDI) erosions.
Methods: :
We retrospectively reviewed 1,013 charts of GDIs between 2006-2011 performed by five surgeons. Sixty eyes with erosions of GDIs were identified and 22 of these with complete data sets were selected. These eyes were matched by age, race, gender, and surgeon to eyes that did not erode. The outcome measures included diagnosis, patch graft material, implant type, tube and plate position, previous surgeries, systemic diseases, previous glaucoma lasers, intraoperative mitomycin C and/or triamcinolone, contact lens wear, smoking, and seasonal allergies. Matched statistical analysis of dichotomous and nominal data was analyzed using McNemar’s and stratified Mantel-Haenszel tests respectively.
Results: :
The overall incidence of erosion in GDIs over 5 years was found to be 5.8%. In the comparison of erosions to controls, the average follow up time was 2.95 + 2.1 yrs and 2.65 + 0.1 yrs respectively. Caucasians constituted 82% of both groups, with the remainder being African American. The erosion group consisted of 18% Baerveldts, 73% Ahmeds, and 9% Moltenos, while the control had 32% Baerveldts, 59% Ahmeds, and 9% Moltenos. There was a similar number of KeraSys (18% erosion and 23% control), Tutoplast pericardium (77% erosion vs. 68% control), Tutoplast sclera (5% each), and amniotic membrane (5% control) used. The percentage of eyes with mitomycin C was comparable in both groups (18% erosion vs. 14% control). None of the variables listed above were found to have a statistically significant association with GDI erosions. The prevalence of contact lens use (27% erosion vs. 9% control) and chronic obstructive pulmonary disease (COPD) and/or asthma (9% erosion vs. 4.5% control) was found to be higher in the erosion group. This prevalence was not found to be statistically significant (p= 0.16 each).
Conclusions: :
The incidence of erosions as reported above is similar to that in prior studies. This study in progress has not yet shown statistically significant associations of diagnosis, patch graft material, implant type, tube and plate position, previous surgeries, diabetes, hypertension, previous glaucoma lasers, intraoperative mitomycin C and/or triamcinolone, smoking, and seasonal allergies with GDI erosions. The higher, yet non-statistically significant prevalence of contact lens use in the erosion group is of interest. Recent studies have demonstrated frequent conjunctival breakdown in patients with keratoprostheses who have undergone tube surgery. This may be related to chronic contact lens use. Further research and follow up may help elucidate additional causes of erosion.
Keywords: clinical (human) or epidemiologic studies: prevalence/incidence • clinical (human) or epidemiologic studies: outcomes/complications • clinical (human) or epidemiologic studies: risk factor assessment