Abstract
Purpose: :
To determine preoperative indicators of postoperative macular edema in diabetic patients undergoing phacoemulsification and intraocular lens implant.
Methods: :
Patients with diabetes mellitus (DM; type II insulin dependent and non insulin dependent) and controls (age–matched patients without diabetes mellitus) undergoing cataract extraction were consented to have serum and aqueous samples of vascular endothelial growth factor (VEGF) or insulin–like growth factor–1 (IGF–1) taken at the time of cataract surgery. All patients had full ophthalmologic examinations, visual acuity testing using the ETDRS chart (VA); and central macular thickness (CMT), percent change in preoperative and postoperative macular thickness (MTchange), and change in volume measured with the OCT–3 (Carl Zeiss Meditec). VEGF and IGF–1 levels were determined using enzyme linked immunosorbent assay (ELISA). Outcomes determined postoperatively (one month and 6 months) include VA, CMT, MTchange, and change in volume. Correlations were determined using the Pearson analysis. A linear discriminant analysis was performed to predict postoperative CMT using the variables: presence of DM, serum IGF–1, aqueous VEGF, serum VEGF, and preoperative CMT.
Results: :
At one month, 30 patients had been enrolled and completed testing (n=14 DM; 16 control). Among DM, serum VEGF level was significantly correlated with postoperative CMT and MTchange (r=0.649 p < 0.02; r = 0.75 p <0.003, respectively). However, there were no correlations among patients in the control group. The median postoperative CMT (215 micrometers) was used to create two groups: normal postoperative CMT (less than or equal to 215 micrometers) and high postoperative CMT (greater than 215 micrometers). The linear discriminant model correctly predicted 81% of patients as to postoperative CMT when all variables were included or when a reduced set was used (presence of diabetes, serum VEGF and preop OCT).
Conclusions: :
Our one month results show that postoperative macular edema determined as CMT or MTchange in diabetic patients undergoing phacoemulsification was correlated with serum VEGF, but not with aqueous VEGF or serum IGF–1. An informed decision rule using a reduced set of variables achieved a high prediction rate of patients who have high postoperative CMT one month following phacoemulsification. Further enrollment and analyses are ongoing.
Keywords: cataract • diabetic retinopathy • macula/fovea