Diabetic macular edema: Fundus autofluorescence and functional correlations

PURPOSE. Diabetic macular edema (DME) shows variable clinical characteristics with unpredictable results to local treatment, probably reflecting different phenotypes. The purpose of this study was to evaluate the role of structural and functional macular imaging in the characterization of DME patterns. METHODS. One hundred fifty-one eyes of 92 diabetic patients with untreated clinically significant macular edema (CSME) underwent best corrected visual acuity (BCVA) determination (logMAR), slit lamp biomicroscopy; fluorescein angiography; optical coherence tomography (OCT; mean central retinal thickness [CRT], volume, and DME pattern); fundus autofluorescence (FAF; absent or increased [i]FAF, single or multiple spots; iFAF area quantification); and microperimetry. Linear correlation, data agreement and three-way analysis of covariance were used for statistics. RESULTS. Thirty-five (23.2%) eyes had normal FAF; and 116 eyes had iFAF: 48 (31.8%) single-spot iFAF, 68 (45%) multiple-spot iFAF. Retinal sensitivity in areas with iFAF was 11.5 ± 5.3 dB (vs. 15.1 ± 3.9 dB in normal areas, P <0.005). Retinal sensitivity of the central field was 15.1 ± 3.9 dB in normal FAF, 12.4 ± 4.8 dB in single-spot iFAF and 11.4 ± 4.9 dB in multiple-spot iFAF (P <0.05). OCT CRT and volume were not significantly different between the FAF groups. OCT volume correlated to OCT CRT (r = 0.68), retinal sensitivity in iFAF (r=0.50) and BCVA (r = 0.42). Cystoid OCT pattern and FA edema patterns correlated with iFAF presence (P <0.0001). CONCLUSIONS. In CSME, FAF correlates better with OCT patterns and central field micro perimetry rather than with visual acuity. FAF is a rapid, noninvasive technique that may give new insight into the evaluation of DME. The validity of FAF in the follow-up and treatment outcomes in DME remain to be assessed.