To assess the frequency and characteristics of discrepant findings between fluorescein angiography (FA) and optical coherence tomography (OCT) in uveitic macular edema (ME).
Retrospective cross-sectional study of 112 eyes of 78 patients with uveitic ME on FA, OCT, or both.
ME was graded on OCT and FA of uveitis patients attending the University Medical Center Utrecht. The frequency and severity of discrepant findings were analyzed, and the clinical findings at the time of imaging were assessed. The imaging studies were compared with the clinical characteristics.
Positive results of both imaging methods (FA+/OCT+) were observed in 61 (54%) of 112 eyes, whereas discrepant results occurred in 51 (46%) of 112 eyes. The FA+/OCT− discrepancy occurred in 34 (30%) of 112 eyes, and the FA−/OCT+ discrepancy occurred in 17 (15%) of 112 eyes. No correlations between the discrepant imaging results and age, gender, duration of uveitis or ME, visual acuity, or cause of uveitis were identified. FA+/OCT− and FA−/OCT+ discrepancies comprised typically mild degrees of ME. The FA+/OCT− discrepancy occurred in 50% of eyes with birdshot chorioretinopathy (7/14), and the FA−/OCT+ discrepancy occurred more often in intermediate uveitis than in other anatomic locations. Although the FA+/OCT+ consistency was noted frequently in active uveitis, the FA−/OCT+ discrepancy was common in eyes with inactive uveitis (8/18; 44% of inactive eyes).
Our results emphasize that FA and OCT are complementary investigations, each revealing different aspects of the pathophysiology of uveitic ME.
Macular edema (ME) is a condition characterized by intraretinal accumulation of fluid in the macula, which is a major cause of decreased visual acuity in multiple ocular disorders, including uveitis. Clinical signs may manifest first when severe macular involvement has already developed. In contrast, therapeutic interventions are effective mostly in the early stages of ME. Until the introduction of the optical coherence tomography (OCT), fluorescein angiography (FA) was the most widely used method to evaluate the retinal vascular perfusion and the integrity of the inner blood-retinal barrier. OCT has become increasingly important in the assessment of ME by revealing quantitative data about retinal thickness and information on anatomic abnormalities of the retinal layers.
Although both FA and OCT are able to verify the presence of ME, discrepancies occasionally have been reported. Contradictory results from these examinations complicate clinical decision making. In this study, we investigated the incidence and nature of discrepancies between FA an OCT in a large group of patients with uveitis and attempted to clarify the factors causing these discrepancies.
In this study, we retrospectively included all 78 patients (112 eyes) with uveitis and ME who underwent both FA and OCT within an interval of 2 weeks at the University Medical Center Utrecht during a 21-month period from January 2007 through September 2008 and who did not undergo any therapeutic interventions between the FA and the OCT.
Excluded from study were 4 patients in whom poor-quality or misaligned OCT scans were obtained. Criterion was whether the software program could make a good fit on the inner and outer border of the retina. Medical charts of patients were reviewed for sex, age, duration of uveitis and of ME at the time of the imaging studies, anatomic classification, cause of uveitis, and best-corrected visual acuity. Our uveitis population included 22 patients with associated systemic diseases and 40 patients with an unknown cause of uveitis ( Table 1 ). Fluorescein angiograms and OCT scans were assessed in a masked fashion for characteristics of ME by 2 observers (J.O.-v.N., L.P.C.).
|FA+/OCT+ (n = 43)||FA+/OCT– (n = 26)||FA–/OCT+ (n = 9)||Total (n = 78)|
|Associated systemic diseases||14||7||1||22|
|Established clinical entities a||6||14||3||13|
|Mean age c (range), y||55 (13 to 83)||52 (18 to 77)||49 (26 to 73)||53 (13 to 83)|
|Median duration uveitis c (range), y||2.8 (0 to 24.9)||2.0 (0.03 to 31.9)||1.6 (0.01 to 6.2)||2.8 (0 to 31.9)|
|Median duration ME c (range), y||1.0 (0 to 16.1)||0.5 (0 to 9.9)||0.6 (0 to 6.0)||0.7 (0 to 16.1)|
|Median Snellen VA c (range)||0.5 d (0.01 to 1.2)||0.65 (0.1 to 1.2)||0.6 (0.01 to 1.2)||0.6 (0.01 to 1.2)|
a Includes birdshot chorioretinopathy, peripheral multifocal chorioretinitis, M. Eales.
b This group also includes all patients with pars planitis.
c Only the left eye is included, unless the uveitis was unilateral in the right eye.
All FA photographs were recorded using a Zeiss digital fundus camera (Zeiss FF 450 plus fundus camera; Carl Zeiss AG, Oberkochen, Germany). Fluorescein angiograms were graded according to a modified grading system ( Table 2 ), that is, a combination of the criteria reported by Yannuzzi and the proposed grading system of the Angiography Scoring for Uveitis Working Group. In addition, we registered the number of affected clock hours separately. ME was assessed on images obtained at least 5 minutes after intravenous injection of 5 mL 20% sodium fluorescein. In this study, only the cases in which there was obvious leakage in the center (grades 2 through 4) were classified and analyzed as ME.
|0||No perifoveal hyperfluorescence|
|1||Faint perifoveal hyperfluorescence: specific localization of hyperfluorescence too difficult because of very minimal leakage|
|2||Evident perifoveal hyperfluorescence in an area centered on the fovea of less than 1 optic disc diameter|
|3||Evident perifoveal hyperfluorescence in an area centrad on the fovea of between 1 and 1.5 optic disc diameter(s)|
|4||Evident perifoveal hyperfluorescence in an area centrad on the fovea of more than 1.5 optic disc diameters|
OCT was performed with the StratusOCT (Stratus OCT 3000, software version 4.0.1; Carl Zeiss Meditec, Inc, Dublin, California, USA). Mean central retinal thickness (CRT; inner, 1000 μm;) was recorded, and 6 radial line images were reviewed on screen for the presence of ME, intraretinal cysts, associated central serous retinal detachment (CSRD), epiretinal membranes, vitreoretinal attachment with or without traction (vitreoretinal traction vs −adhesion), or a combination thereof. Poor quality or misaligned OCT scans were excluded from the analysis (n = 12). The following characteristics of ME were registered: diffuse macular edema, cystoid macular edema, and CSRD. Diffuse ME presented as diffuse thickening of the retina, disturbance of the orderly layered retinal structure, spongelike low reflective areas, or a combination thereof. Cystoid macular edema was characterized by clearly defined intraretinal cystoid spaces. In a CSRD, a separation between neurosensory retina and retinal pigment epithelium was present in addition to ME.
Based on a meta-analysis of 3 large studies investigating retinal thickness in northern and northwestern Europeans using the StratusOCT, we chose an upper limit value of 249 μm for CRT, which is 2 standard deviations above the mean central retinal thickness (pooled mean of 205.5 ± pooled standard deviation of 21.7 μm). ME on OCT was considered in cases with CRT higher than 249 μm or when evident cystoid spaces could be observed on radial lines (n = 5).
The activity of the uveitis was classified as active or inactive, based on the presence or absence of aqueous or vitreous cells, or both, signs of activity on fluorescein angiography (such as vasculitis, optic disc leakage, and capillary leakage outside the perifoveal area), or a combination thereof.
An analysis of variance was used to test for differences in mean age, duration of uveitis, duration of ME, visual acuity, and CRT between eyes with ME on both examinations, ME solely on FA, and ME solely on OCT. The Dunnett post hoc test was applied to determine any differences between separate groups. A chi-square equation was used to calculate differences between groups for frequencies of cause, location of uveitis, FA grade, presence of CSRD, presence of epiretinal membranes, and presence of vitreoretinal traction. All statistical analyses were performed using SPSS for Windows version 15.0 (SPSS, Inc, Chicago, Illinois, USA). We included all eyes with ME in our analyses because, in bilateral cases, no correlation was found in the occurrence of discrepancies and CRT measurements between the left and the right eyes ( P = .78).
One hundred twelve eyes of 78 patients (mean age, 53 years; range, 13 to 83 years) fulfilled the diagnostic criteria of ME either on FA or on OCT. Our study cohort comprised 21 eyes (19%) with severe ME that was FA leakage group 3 and 4 and 16 eyes (14%) with a CRT of more than 400 μm.
ME present in both imaging examinations (FA+/OCT+ group) was present in 61 (54%) of 112 patients, and discrepancies were documented in 51 (46%) of 112 eyes ( Table 3 ). ME visible solely on FA (FA+/OCT− discrepancy) was noted in 34 (30%) of 112 eyes (34/51; 67% of those with discrepancies) and ME was identified solely on OCT (FA−/OCT+ discrepancy) in 17 (15%) of 112 eyes (19/51; 37% of those with discrepancies). In Figures 1 and 2 , 1 illustration is given of each type of discrepancy. The CRT of the FA+ group (FA+/OCT+ and FA+/OCT−) did not differ from that of the FA− group (269 vs 274 μm; P = .98). A CSRD was present in 15% of the eyes with ME and was not correlated with 1 of the 3 groups.
|FA+/OCT+ (n = 61)||FA+/OCT– (n = 34)||FA–/OCT+ (n = 17)||Total (n = 112)|
|Anterior||1 (50%)||1 (50%)||0||2 (100%)|
|Intermediate||6 (37.5%)||4 (25%)||6 (37.5%) a||16 (100%)|
|Posterior||22 (61%)||9 (25%)||5 (14%)||36 (100%)|
|Pan||32 (55%)||20 (34%)||6 (10%)||58 (100%)|
|Median CRT (range), μm||318 (188 to 615)||215 b (187 to 248)||274 c (160 to 464)||270 (160 to 615)|
|0 or 1 d||17||17|
|Median clock hours, n (range)||12 f (1 to 12)||10 (1 to 12)||0 (0 to 7)||10 (0 to 12)|
|Epiretinal membrane||31 g||8||5||44|