Murat Karacorlu MD, MSc Istanbul Retina Institute Inc. Hakki Yeten Cad. No: 8/7 Fulya Istanbul Turkey Tel: + 90 212 231 3121 Fax: + 90 212 233 2425 Email: email@example.com
Purpose: To define serous macular detachment in patients with diabetic cystoid macular oedema (CME).
Methods: This study involved 78 eyes of 58 patients with diabetic CME. The patients underwent complete ophthalmic examination, fluorescein angiography and optical coherence tomography (OCT). Eyes with epiretinal membrane or vitreo-macular traction were not included in the study. Optical coherence tomography-3 was used in all patients and fundi were scanned on the horizontal, vertical and four oblique planes through the centre of the fovea.
Results: In all cases the increased thickness of the retina was related primarily to the hyporeflective intraretinal cavities. With OCT, 24 of 78 eyes (31%) had serous macular detachment as shown by retinal elevation over a non-reflective cavity with minimal shadowing of the underlying tissues. Fluorescein angiography did not show serous macular detachment in any patient.
Conclusions: Our data showed that the incidence of serous macular detachment in diabetic CME was much higher than previously reported. Optical coherence tomography-3 allows an in vivo cross-sectional observation of very subtle serous macular detachment that is difficult to diagnose at the slit-lamp or by fluorescein angiography in patients with diabetic CME.
Macular oedema occurs in a wide variety of ocular diseases and is one of the most common causes of vision loss in patients with diabetic retinopathy (Patz et al. 1973). An alteration in the blood–retinal barrier leading to the accumulation of fluid in the retinal extracellular space and the development of fluid-filled (cystoid) spaces in the outer plexiform layer and the inner nuclear layer of the retina are common findings in cystoid macular oedema (CME). The term ‘cystoid’ rather than ‘cystic’ is used because the fluid-filled spaces do not possess the epithelial layer that is present in a true cyst. Although the histopathological features in CME have been well documented, the exact mechanism of the breakdown of the blood–retinal barrier in CME is not well understood (Cunha-Vaz & Travassos 1984).
In a few cases of CME, a serous retinal detachment, as detected by optical coherence tomography (OCT), has been described (Otani et al. 1999; Kokame et al. 2001; Spaide et al. 2003). Otani et al. (1999) were the first authors to publish three patterns of structural changes in diabetic macular oedema investigated by OCT, namely sponge-like swelling, cystoid macular oedema and serous retinal detachment. Kang et al. (2004) also categorized diabetic macular oedema and defined subfoveal fluid accumulation in diabetes patients. In the current literature the importance of serous macular detachment in diabetic CME is not well described. Our study attempts to address this issue using a prospective analysis of patients with diabetic CME.
Material and Methods
The study population comprised 78 eyes of 58 patients with diabetic CME. The eligibility criteria for this study included:
1the presence of clinically significant macular oedema in the fundus examination;
2the presence of angiographically confirmed diabetic macular oedema;
3the presence of CME and serous macular detachment documented by OCT;
4the absence of epiretinal membrane or vitreo-macular traction documented by OCT, and
5the absence of dense media opacity or pre-retinal haemorrhage that might prevent OCT examination.
Eyes with previous intraocular surgery or vitreo-retinal pathology other than diabetic retinopathy were excluded. In this series, no eyes had received previous grid laser photocoagulation or intravitreal triamcinolone acetonide injection for the treatment of CME.
The patients underwent complete ophthalmic examination, including corrected visual acuity (VA) measurement (logMAR) using the ETDRS chart, slit-lamp biomicroscopy with a 90 dioptre pre-corneal lens and a Goldmann three-mirror contact lens, colour fundus photography, fluorescein angiography and OCT. The interval between the OCT and fluorescein angiography examinations was required to be less than 1 week for entry into the study, with no therapeutic intervention applied to the eyes during the interval. Fluorescein angiograms were performed on a Heidelberg scanning laser ophthalmoscope (Heidelberg Engineering, Heidelberg, Germany). Optical coherence tomography examinations were performed using the OCT-3 (Carl Zeiss Ophthalmic System Inc., Humphrey Division, Dublin, California, USA). During OCT examination the fundi were scanned on the horizontal, vertical and four oblique planes through the centre of the fovea. Scan lengths of 6 mm were used in all eyes. The retinal thickness at the centre of the fovea was measured directly from the OCT image. We defined the retinal thickness as the length between the inner retinal surface and the outer retinal surface.
Fluorescein angiographic CME was thought to be present if typical oval or petaloid hyperfluorescent cystoid spaces radiating from the fovea were evident during fluorescein angiography. The OCT examination was thought to show CME if there were hyporeflective intraretinal cavities radiating from the centre of the macula in cross-sectional scans. Serous macular detachment was thought to be present if the posterior surface of the retina was elevated over a non-reflective cavity with minimal shadowing of the underlying tissues.
A total of 78 eyes of 58 patients with diabetic CME were evaluated. The subjects included 28 men and 30 women, ranging in age from 36 to 74 years (mean ± standard deviation 55 ± 9 years). The duration of diabetes was 11 ± 4 years. Nine patients (16%) had type 1 diabetes and 49 patients (84%) had type 2 diabetes mellitus. At the time of examination all the patients had good glycaemic control. The mean logMAR VA was 0.68 ± 0.18 (changed between 1.0 and 0.3). Fifty-six eyes (72%) had non-proliferative and 22 eyes (28%) had proliferative diabetic retinopathy. Twenty-eight patients (48%) had systemic hypertension and 22 patients (38%) were smokers.
At the time of initial examination no patient was suspected of having a serous macular detachment. On angiography, all patients had typical oval or petaloid hyperfluorescent cystoid spaces radiating from the centre of the fovea. Fluorescein angiography did not show serous macular detachment in any patient.
The mean foveal thickness, as determined by OCT, was 531 ± 138 µm (range 325–822 µm). In all cases the increased thickness of the retina was related primarily to the hyporeflective intraretinal cavities. There were at least two and up to eight cystoid cavities in the macular area within a scan length of 6 mm. Each cystoid cavity in the macula was walled off with septae. The cystoid cavities expanded in a round or oval configuration, resulting in loss of the normal depression of the anterior contour at the fovea (Fig. 1). Optical coherence tomography showed highly reflective hard exudates in the inner portion of the neurosensory retina in 23 eyes (29%).
With OCT, 24 of 78 eyes (31%) were shown to have serous macular detachment shown by retinal elevation over a non-reflective cavity with minimal shadowing of the underlying tissues. In all cases, the subretinal space was optically empty and the thickness of the subretinal space was greatest at the central fovea and declined peripherally (Fig. 2).
Detachment of the sensory retina in the absence of a retinal break is a well known clinical phenomenon. It may occur when fluid from the retinal or choroidal circulation leaks into the subretinal space and the compensatory mechanism for fluid removal is exceeded. Detachments such as these have been referred to as serous, exudative or non-rhegmatogenous, and occur in a variety of ischaemic, inflammatory, neoplastic and idiopathic conditions (Weinberg et al. 1990). The pathogenesis of serous retinal detachment is still being debated. Leakage from retinal or choroidal circulation into the subretinal space exceeding its drainage capacity is thought to be the main mechanism (Weinberg et al. 1990). According to (Ravalico & Battaglia 1992), it is linked not only to the limitations of the draining vascular system, but also to an impairment in the function of the retinal pigment epithelium. Kang et al. (2004) reported that in diabetic eyes the incidence of CME and serous macular detachment increases with the existence of retinal vascular hyperpermeability, and the pathology of these two phenomena might share a common pathogenesis in this regard. The external limiting membrane is not impermeable to fluid and albumin. Thus, with the disruption of the inner blood–retinal barrier, the excess fluid might reach the subretinal space in large amounts, might fail to be removed properly by the retinal pigment epithelium and might result in subfoveal detachment (Kang et al. 2004).
From OCT findings, serous retinal detachment associated with CME has been identified in patients with branch retinal vein occlusion and hypotony maculopathy (Kokame et al. 2001; Spaide et al. 2003). Wang et al. (1999) showed OCT to be a very useful technique for the detection of shallow foveal detachment in patients with central serous chorioretinopathy. The serous retinal detachment associated with CME in diabetic patients was first defined by Otani et al. (1999). According to this study, OCT-1 showed three patterns of diabetic macular oedema: retinal swelling, cystoid macular oedema and serous retinal detachment. Among the 59 eyes included in the study, only nine eyes (15%) showed serous retinal detachment (six eyes with retinal swelling and three eyes with both retinal swelling and cystoid macular oedema) (Otani et al. 1999). Kang et al. (2004) also categorized diabetic macular oedema into four types by using OCT-2 as follows: type 1 is shown by thickening of the fovea with homogeneous optical reflectivity throughout the whole layer of the retina; type 2 is shown by thickening of the fovea with markedly decreased optical reflectivity in the outer retinal layers, and type 3 is shown by thickening of the fovea with subfoveal fluid accumulation and the distinct outer border of a detached retina, and comprises type 3A, without foveal traction, and type 3B, with apparent vitreo-foveal traction. In this study, the incidence of each type of OCT was 55% for type 1, 30% for type 2 and 14% for types 3A and 3B.
During OCT examination of patients with diabetic CME, we found that the prevalence of serous macular detachment was much higher than previously reported. In the 78 eyes with diabetic CME examined in our study, OCT-3 detected serous macular detachment in 24 eyes. In our series no eyes had epiretinal membrane or vitreo-macular traction. Ophthalmoscopic examination and fluorescein angiography did not show serous macular detachment in any patient. It is obvious that CME prevents the detection of serous macular detachment by clinical examination and fluorescein angiography. Because of its improved resolution and image quality, OCT-3 allows an in vivo cross-sectional observation of very subtle serous macular detachment that is difficult to diagnose at the slit-lamp or by fluorescein angiography in patients with diabetic CME. Using OCT-3, we were able to show a much higher prevalence of serous macular detachment in diabetic CME than previously reported.