OCT in the Management of Diabetic Macular Edema

Authors: Nadia K Waheed Jay S Duker

Publish Date: 2013/07/24

Volume: 1, Issue: 3, Pages: 128-133

PDF Link

Abstract

Diabetic macular edema DME is the most common cause of mildtomoderate visual loss in diabetes With the introduction of antivascular endothelial growth factor therapies in addition to the previously available medical and laser therapies OCT has become the cornerstone in the diagnosis monitoring therapeutic selection and gauging response to therapy in eyes with DME A review of the recent literature shows numerous advancements in the way OCT scanning is used both to monitor DME and to guide the management of DMEDiabetic retinopathy is estimated to affect onethird of patients with diabetes The prevalence and severity are affected by the duration of diabetes glycemic control and the presence of concurrent hypertension Epidemiologic studies suggest that diabetic macular edema DME will affect up to 7  of patients with diabetes 1 DME is the most common cause of mildtomoderate visual loss in diabetics and studies confirm that its presence leads to a significantly diminished quality of life and disability 2 This impact is especially significant because many of the patients suffering vision loss in diabetes are in the working age group and the presence of macular edema results in significant productivity impact on the individual as well as that of society at largeThe Early Treatment Diabetic Retinopathy study defined the term ‘clinically significant macular edema’ CSME that represented the minimal level of edema based on clinical examination that warranted treatment with laser photocoagulation CSME is thickening of the retina at or within 500 μm of the center of the macula hard exudates at or within 500 μm of the center of the macula if associated with thickening of adjacent retina or a zone of retinal thickening 1 disc area or larger any part of which is within 1 disc diameter of the center of the macula 3 While this definition is still valid for eyes that have DME not involving the center of fovea in the era of OCTguided antivascular endothelial growth factor VEGF therapy the decision to treat DME also includes ‘center involving’ DME determined by OCT central subfield thickness of greater than or equal to 275 μm 4Commercial OCT machines provide several quantitative and automated measures that are useful in the evaluation and longitudinal monitoring of patients with DME These include retinal thickness and the average mean central subfield thickness CSMT or CST Although the center point retinal thickness can also be measured manually there is considerable interobserver variability depending on where the center point is assumed to be on scanning Thus central retinal thickness measurement in diabetics has been mostly replaced by the automated mean central subfield thickness measurement which has been shown to be closely correlated to the center point thickness in diabetic patients •5 ••6 7The mean central subfield thickness is often used as the critical measure in the diagnosis and management of patients with DME This is defined as the mean retinal thickness within a 1mm circle centered on the fovea As previously mentioned a CST of greater than or equal to 275 μm was used as a threshold to define “centerinvolving DME” However this value remains under debate Most large clinical trials including pharmaceutical sponsored trials and those conducted by the DRCRnet traditionally utilized time domain OCT TDOCT scanning Time domain OCT is now largely supplanted by spectral domain SD OCT in most clinical practices Although the pathology seen is comparable between the TD and SDOCT scans SD OCT provides better resolution and more rapid scan acquisition In addition it is important to understand that there are measurement differences between the SD and TDOCT devices Especially pertinent to DME retinal thickness is measured differently between TD and SDOCT devices and the numbers obtained are not comparable across systems neither are they comparable between different SDOCT systems This is primarily because of the different segmentation algorithms used in the commercial OCT devices which measure thickness from different points around the RPE layer •5 The Cirrus SDOCT Carl Zeiss Meditec for example reports significantly greater retinal thicknesses than the Stratus TDOCT Carl Zeiss Meditec for the same patient imaged within a few minutes on the two machines There have been recent attempts to derive ‘conversion factors’ that would allow for the conversion of retinal thickness obtained from one machine to another but these are not as yet well established enough to have gained widespread traction 8 Suffice to say that a value of 275 μm as a cutoff for treatment is too low for any SD OCT machineWhereas previous thresholds defined on the TDOCT typically accounted for a 10  margin of error there is now some evidence that measurements on the newer SDOCT may have better reproducibility A recent study by Comyn et al looking at the repeatability and reproducibility of OCT measures of retinal thickness in eyes with DME has found that on the Spectralis SDOCT device Heidelberg changes in central subfield thickness 8 μm can be considered more indicative of true clinical change rather than measurement variability 9 However since the study utilized only the Spectralis it is not clear whether the margin of error on other SDOCT devices are comparable since the Spectralis uses eye tracking Moreover the study did not simulate a real clinical scenario with different operators but was conducted by the same experienced operator Other studies demonstrate up to a 17  diurnal variation in retinal thickness in diabetics with the least OCTmeasured retinal thickness in the late afternoon and the greatest thickness in the early morning 10The retinal thickness map obtained on volumetric scanning is a useful adjunct in clinical practice providing a map of areas of retinal thickness as well as highlighting areas of thickening outside the central subfield but that may still be classified as CSME An important limitation however of the automated quantitative OCT scans is segmentation breakdown in which the inner or outer retinal boundary is incorrectly identified resulting in abnormal retinal thickness measurements This has been found to occur in up to 37  of scans of the central 1 mm in both the Cirrus and the Spectralis OCT machines 11 Thus careful evaluation of the OCT data is important in preventing these errors from making diagnostic mistakes in patients with DME Another limitation of thickness measurements is that they may not necessarily correlate well with visual acuity and visual prognosis Central subfield thickness has been shown to correlate only modestly with visual acuity in patients with DME •12 Thus subfield thickness by itself is not a good surrogate for posttreatment visual acuity and qualitative assessment of OCT scans is an important part of the evaluation of eyes with DME

Keywords: