Lack of Correlation Between Diabetic Macular Edema and Thickness of the Peripapillary Retinal Nerve Fibre Layer.

INTRODUCTION: We compared the thickness of the peripapillary retinal nerve fiber layer (RNFL) in patients with diabetic macular edema (DME) and/against the thickness in the normal population.
METHODS: This cross-sectional study compared the RNFL thickness in patients with DME (DME group) using optical coherence tomography (OCT) to a comparable group of healthy (nondiabetic) patients (control group). Measurements were performed in different/the four peripapillary quadrants and in the macula region for the fovea, parafoveal, and perifoveal areas. The mean RNFL thickness was compared between both groups.
RESULTS: There were fifty eyes of fifty nonglaucomatous diabetic patients with DME (29 with nonproliferative diabetic retinopathy [PDR] and 21 with PDR), and fifty eyes in the control group. The macular regions were significantly thicker in the DME group compared to the control group. The central foveal thickness was 149 μ thicker in eyes with DME compared to the control group (P < 0.001). The difference in total RNFL thickness between groups was not significant (4.4 μ [95% confidence interval: -3.1 to +12]). The between-group differences in peripapillary RNFL thickness by age group, glycemic control, history of intravitreal treatments, and refractive errors were not statistically significant (P > 0.05, all comparisons).
CONCLUSION: Peripapillary RNFL thickness measurements were not significantly influenced by DME. Hence, OCT parameters could be used to monitor/early detect glaucomatous eyes even in the presence of DME.

INTRODUCTION

Diabetic macular edema (DME) is a common complication of diabetes mellitus and a major cause of blindness worldwide.1 The pathogenesis is primarily due to changes in the retinal microvasculature.2 However, the presence of retinal neurodegeneration before the onset of clinically detectable microvascular damage has been suggested.3 Some have suggested that there is a decrease in retinal nerve fiber layer (RNFL) and an increase in inner nuclear layer/and outer plexiform layer thickness in diabetics even when in the absence of diabetic retinopathy (DR).3

Glaucoma is an optic neuropathy characterized by progressive loss of retinal ganglion cells and optic nerve damage that may result in visual field loss and irreversible loss of vision. Optical coherence tomography (OCT), an interferometry-based imaging modality, which generates high-resolution cross-sectional images of the retina.4 It is an important tool for detecting structural changes in early glaucoma, i.e., before the onset of clinically detectable visual field changes.5 Peripapillary RNFL, optic disc assessments, and macular ganglion cell complex (mGCC)/(MGCC) are the mainstay of glaucomatous structural measurements.6 Spectral domain OCT has a higher imaging resolution than time domain OCT, which makes it possible to study the segmentation of the inner macular layers.7

Diabetic patients are at a higher risk of glaucoma.89101112 Therefore, it is important to study the effect of diabetes and DME on the thickness of peripapillary and the different macular and retinal layers. Somfai et al.13 reported that the RNFL did not change in eyes with DME, but they did not evaluate the peripapillary RNFL or mGCC/MGCC thickness. Hwang et al.14 reported significant retinal layer thickness changes in and around the optic disc and retina excluding the macula in diabetic patients with DME. To address this clinically important question, we studied the effect, if any, that DME has on these measurements.

METHODS

The institutional ethical and research board of the Specialized Medical Center approved the study. Written informed consent was obtained from all participants before enrollment. The study was undertaken between June 2014 and October 2014. This was a cross-sectional study. Patients with DME were assessed for DR and retinal layer thickness using OCT (DME group). A comparable group of patients without glaucoma and without diabetes was also recruited (control group).

To calculate the sample size, we assumed that average nerve fiber layer thickness in nonglaucomatous eyes was 123 ± 5 μm and 135 ± 40 μm in diabetic patients with macular edema.1415 To achieve 95% confidence interval (CI) and 85% power of the study with 1:1 ratio of cases with DME and normal healthy controls, 44 eyes were required in each group. To compensate for subject dropout, we recruited fifty cases with DME (DME group and fifty healthy eyes (control group).

DME was defined as central macular thickness of 300 μm or diffuse parafoveal thickness. Parafoveal thickness was defined as having all four parafoveal OCT quadrants >300 μm, or focal edema when fewer than four quadrants are >300 μm. We excluded cases with glaucoma, dense media opacity that precluded OCT measurements, any neuroretinal disease other than DR, and intraocular surgery within 6 months from the onset of the study.

The control group was comprised nondiabetic healthy individuals that approached our institution for a vision check. The same exclusion criteria were applied to the control group as the DME group.

Data were collected on patients’ age, gender, and in the DME group, the eye involved, degree of glycemic control, stage and grade of DR, and previous treatment of DR (panretinal photocoagulation, focal laser, and intravitreal injections).

Each subject underwent a comprehensive ophthalmic examination including a review of the medical history, best-corrected visual acuity, and distance visual acuity of each eye was tested using Early Treatment Diabetic Retinopathy Study chart placed at 3 m distance. Anterior segment eye examination was performed with slit-lamp biomicroscopy (HAAG-STREIT AG, Koeniz, Switzerland). Intraocular pressure was measured with Goldmann applanation tonometry, attached to a slit-lamp biomicroscope (Topcon Corp, Tokyo, Japan). The pupil was dilated using 0.5% tropicamide. The posterior segment and peripheral retina were assessed using binocular indirect ophthalmoscope and +20 D lens. The central retina was assessed using slit-lamp and +90 D Volk lens (Volk Optical Inc., Mentor, OH, USA).

The retinal thickness measurements were performed with three-dimensional (3D) OCT-2000 (Topcon Corporation, Tokyo, Japan). Each eye was examined after pupillary dilatation. The following OCT scans were used: 1-macular 3D scan protocol (6 mm × 6 mm area centered on the fovea with a scan density of 512 [vertical] × 128 [horizontal] scans). The mean and standard deviations of retinal thickness were calculated in central fovea, superior, nasal, inferior, temporal parafoveal, superior, nasal, inferior, and temporal perifoveal. 2-disc 3D scan protocol was performed (6 mm × 6 mm area centered on the optic disc with a scan density of 512 vertical × 128 horizontal scans). The mean and standard deviations of RNFL thickness in four quadrants and clock-hour sectors at a circular distance of 3.4 mm in diameter from the optic disc were calculated. 3-macular 3D V scan protocol was performed (scanning of a 7 mm × 7 mm area centered on the fovea with a scan density of 512 [vertical] × 128 [horizontal] scans). The mean and standard deviations of the macular RNFL (mRNFL), ganglion cell layer + inner plexiform layer (GCL + IPL), and GCL + IPL + nerve fiber layer (NFL) (GCL++) thicknesses were calculated in superior, inferior hemiretina. These measurements were performed out to one disc diameter inferiorly and superiorly to the fovea. The unit of measurement for retinal thickness was microns (μ).

An Excel (Microsoft Corp., Redmond, WA, USA) spreadsheet to collect the data, which were then transferred to statistical package for social studies (SPSS 22; IBM Corp., Armonk, NY, USA). Univariate analysis was performed to compare the mean and standard deviations for RNFL and GCL between groups and then calculate the difference of mean and the 95% CI. The DME group was subdivided based on glycemic control (HbA1c ≥7 and HbA1c <7) and compared. Statistical significance was indicated by P < 0.05.

RESULTS

In the DME group, fifty eyes of fifty nonglaucomatous subjects were enrolled. There were fifty eyes of healthy nonglaucomatous nondiabetic subjects in the control group. The demographic profile of both groups is presented in Table 1.

Table 1

Profile of persons with diabetic macular edema and healthy Arabs

There were 29 eyes with nonproliferative DR (NPDR) and 21eyes with PDR in the DME group. In the DME group, 18 eyes underwent panretinal photocoagulation, six eyes underwent focal laser treatment, and 24 eyes underwent intravitreal injections. Nine eyes were pseudophakic, and five eyes had early cataract in the DME group. There were 27 myopic eyes (54% of cases), 14 emmetropic eyes (28% of cases), and nine eyes were hyperopic (18% of cases) in the DME group. Glycemic control was adequate (HbA1c mean 7.8 ± 1.8) in the DME group.

The RNFL thickness in each quadrant and the difference between groups are presented in Table 2. The peripapillary RNFL parameters of the quadrants were not different between the DME and control groups. Macular OCT parameters (mRNFL, total thickness of GCL + IPL, total thickness of GCL + IPL + NFL, total foveal thickness, parafoveal thickness, and perifoveal thickness) were significantly thicker in the DME group compared to the control group [Table 3].

Table 2

Retinal nerve fiber layer thickness in eyes with diabetic macular edema and healthy Arabs

Table 3

Retinal thickness at macula of eyes with diabetic macular edema and healthy Arabs

To determine whether the apparent lack of significant intergroup differences in peripapillary RFNL thickness may mask differences due to specific patient characteristics, we performed subgroup analysis by age (younger and older than fifty years of age), and refractive status (myopia, emmetropia, and hypermetropia) and found no statistically significant differences [Tables 4 and 5].

Table 4

Age-group and retinal layer thickness in eyes with diabetic macular edema and healthy Arabs

Table 5

Refractive status and retinal layer thickness in eyes with diabetic macular edema and healthy Arabs

The total RNFL thickness in eyes with DME of 21 diabetic patients with HbA1c <7 was 99 ± 16.3 μm, and it was 101 ± 10.9 μm in 29 eyes of 29 diabetic patients with HbA1c ≥7 (difference of mean −2.0 (95% CI: −10.4–6.2); P > 0.05).

The potential effect of prior treatment was reviewed in the DME group. There were 24 eyes in the DME group that had a history of at least one intravitreal injection. The RNFL thickness in these 24 eyes was 98 ± 15.9 μm. The remaining 26 eyes in the DME group had no history of intravitreal injections. The RNFL thickness in these 26 eyes was 102 ± 10.3 μm. The difference in RNFL thickness in these two groups was not statistically significant (difference of mean = 4.4 [95% CI: − 3.1; +12]; P > 0.05).

We finally considered the degree of DR as a potential modifier. There were 29 eyes with DME and severe NPDR. The RNFL thickness in these 29 eyes was 99.3 ± 10.9 μm. The other 21 eyes with DME had PDR. The RNFL thickness in these 21 eyes was 101 ± 16.4 μm. The difference in RNFL thickness in these two groups was not significant (difference of mean = 1.6 [95% CI: − 6.7; +10]; P > 0.05).

DISCUSSION

Early detection of glaucoma in its early preclinical stages represent a clinically proven preventive strategy that ameliorates the irreversible damage to vision associated with this disease. Diabetics are at high-risk group for developing glaucoma; hence, early detection techniques are even more pressing in this patient population.89101112 However, it has been suggested that the classical “readout” of glaucoma, i.e., RFNL thickness by OCT, may be influenced by the development of DME such that an increase in measurement may not reflect actual early glaucomatous changes. Others have challenged this suggestion, and it remains unclear in the literature whether DME truly influences RFNL in a significant manner.

In this study, we show that although DME is associated with increased thickness of the macular retinal layers, it is not associated with a significant change in the peripapillary RNFL thickness. Factors such as glycemic control, age, refractive status, grades of DR, and history of intravitreal injection treatment did not significantly influence the RNFL thickness measured by OCT. Thus, peripapillary RNFL thickness could be measured with OCT to study the glaucomatous changes even in an eye with DME.

Our results concur with Somfai et al.13 who evaluated 11 eyes with diffuse DME and eight control eyes with the retinal layer segmentation method and found that the RNFL layer did not change in eyes with DME. However, they did not evaluate the peripapillary RNFL thickness.13 In contrast, Hwang et al.14 reported a significant effect of DME on peripapillary RNFL thickness. Although their study compared diabetics with DME and diabetics without DME, whereas our study compared diabetics with DME and healthy controls, we note that this difference in design should not have masked a true effect of DME on peripapillary RNFL in our cohort.

In our study in eyes with all three types of refractive status (myopic, emetropic and hyperopic), we did not find significant variation in RNFL thickness between cases and controls. RNFL thickness measured by the HRT algorithms are not significantly influenced by the refractive status of eye among Singaporean children.15 Even among diabetics, while assessing retinal status using OCT, one should be careful as myopic refractive error influences the study outcomes.16

OCT is a useful tool to monitor functional and morphological changes in DME before and after intravitreal injections. Unfortunately, we did not collect the information of intravitreal injections given in the past for the treatment of diabetic macular edema in our study. However, in another study influence of intravitreal injections on peripapillary RNFL thickness was not significant to interfere in early diagnosis of glaucoma in eye with DME.1718

In eyes with DME, we did not observe the difference in RNFL thickness in young and old people with diabetes. However, age was noted to influence the RNFL in different areas.19 Diseases influencing retinal thickness like age related macular degeneration (AMD) are less common in Arab population.20 This could be the explanation of fewer changes in RNFL by age in our study targeting Arabian patients with resilient retina to age related factors.

Our study also compared the effect of DME on the GCC thickness at the macula among nonglaucomatous persons with DME and the normal nonglaucomatous healthy population. The macular GCC thickness is a sensitive marker of early glaucomatous change, allowing detection of structural changes associated with glaucoma even in the apparently normal hemifield.21 We found that the thickness of the different retinal layers at and around macula in the DME group (mRNFL, total thickness of GCL + IPL, total thickness of GCL + IPL + NFL, total retinal thickness at fovea, parafovea, and perifovea) was significantly greater than the control group. Therefore, it appears that applying the same normative data for these layers in nondiabetics to diabetics with DME can impede detection of early glaucomatous changes, and we recommend peripapillary RNFL thickness be used instead in this patient population.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.