First Author: C.Azrak SPAIN
Co Author(s): M. Baeza Díaz J.J. Martínez Toldos C. Hernández Martínez A. Palazón Bru D. Orozco Beltran V. Gil Guillen 0 0 0 0 0 0 0 0 0
Back to previous
Validation of Optical Coherence Tomography (OCT) Topcon 3D 2000® as a diagnosis tool for diabetic macular edema (DME), based both on the numerical value of macular thickness and on retinal morphology, in order to rule out or confirm the disease.
Cross-sectional observational study in the Ophthalmology department, General University Hospital of Elche (Elche, Alicante, Spain), between June 2012-october 2013.
Diagnosis test: OCT. Images were acquired using a 3D pattern, 512 linear vertical scans and 128 horizontal scanners centered on the visual fixation point. Mean retinal thickness was calculated automatically. We measured the thickness of the central circle, which was centered in the fovea and had a radius of 500µm The existence of cysts and / or intraretinal liquid (IRL) in B-scan corresponding to the horizontal line passing through the foveal depression was recorded. Gold Standard: macular Biomicroscopy exam by an experienced retina specialist. The presence of clinically significant macular edema was defined according to ETRDS criteria. Inclusion criteria: Patients diagnosed with diabetes. Exclusion criteria: -Bedridden patients, dementia. -Cataract surgery in the last 3 months. -Laser treatment within the macula. -Antiangiogenic drug treatment. -Vitreoretina surgery. -Other macular pathology. -High myopia. Area under the Receiver operating characteristic curve (ROC) was calculated and the following breakpoints were determined: 1) the best cutoff, 2) the highest LR-<0.1 and 3) minimum LR +> 10. A ROC curve of macular thickness adjusted for the presence of cysts or IRL was performed again and Multivariate analysis was performed by risk scale calculation.
The final sample included 130 eyes of 70 diabetic patients. Macular retinal edema was detected in 36 eyes by the gold standar. The AUC was 0.887 (95% CI: 0808-0966) as the OCT is discriminative to classify DME. The optimal cutoff was 270 microns where the sensitivity of the test was 83.3% (95% CI: 66.53-93.04) and specificity of 89.4% (95% CI: 80.88-94.50). The point of maximum point that produces a LR-<0.1 was 221, LR-was 0.09 (95% CI: 0.01-0.66), the sensitivity of the test was 97.22 (95% CI: 83.80-99.85) and specificity of 29.79% (95% CI: 21.02-40.23). The minimum point resulting in a LR +> 10 was 275 microns. The sensitivity was 80.6% (95% CI: 63.43-91.20) and specificity was 93.6% (95% CI: 86.09-97.38) The LR + was 12.63 (95% CI: 5.72-27.83). AUC: 0.912 (95% CI: 0847-0978) is obtained adjusting the ROC curve to the existence of intra retinal cysts or liquid.
Perhaps there is not a cutoff point which allows definition of macular edema. There is a range of foveal thickness measurement which can not establish the presence or absence of DME. The lower limit of this range would be set to 221 microns which allows ruling out the DME conclusively, and an upper limit of 275 microns which allows conclusive confirmation of DME In the range between these 2 limits the qualitative assessment of the B scan for the presence of cysts or IRL allows us to improve the diagnosis of the disease. In summary, if the foveal thickness is equal to or greater than 275 microns, or if cysts / LIR are observed in the B-scan, macular edema can be conclusively confirmed. Otherwise DME can be conclusively discarded. OCT can possibly get the diagnose of DME in a stage when the macular dysfunction cannot be detected by biomicroscopy, and would condition a change in treatment protocols for diabetic maculopathy, considered superior to the gold standard diagnostic test.