Abstract
Abstract: :
Purpose: Oscillatory components (OPs) of the conventional electroretinogram are generated within the inner retina and are affected by diabetes, diabetic retinopathy and other diseases. Recently, we introduced a technique to enhance and measure local OPs of the mfERG (mfOPs) in normal eyes (ARVO 2003, abstract # 2696). Here, we use the technique to measure local mfOP abnormalities and compare them to retinopathy sites in patients with early non–proliferative diabetic retinopathy (NPDR). Methods: Slow flash mfERGs (sf–mfERGs) were recorded from one eye of each of 26 normal subjects and 12 diabetics with early, predominantly mild, NPDR. 103 locations within the central 45 deg were stimulated by 100 cd/m2 pseudorandom flashes separated by 3 dark video frames (75 Hz frame rate). At each of the 103 locations, the induced first order and second order sf–mfERG components were extracted and combined to enhance the mfOPs. These waveforms were then combined into 34 triplets to produce, with the central response, 35 responses from 35 contiguous retinal areas. The 35 waveforms were filtered digitally 90–225 Hz to isolate the local mfOPs. In each retinal area the signal–to–noise ratio (SNR) of the mfOPs was calculated and abnormality was defined as a SNR below the fifth percentile of the normal subject group (P<0.05). The retinal distributions of mfOP abnormalities and NPDR (obtained from fundus photographs) were then compared. Results: In the diabetics, 25% of the 420 (35 areas * 12 eyes) retinal areas contained NPDR. Each instance of NPDR was much smaller than an individual examined retinal area. Abnormal mfOPs occurred in 36% of the diabetics’ retinal areas. These mfOP abnormalities were found in 49% of the 72 retinal areas with microaneurysms, 41% of the 17 areas with hard exudates, 50% of the 6 areas with soft exudates, and 70% of the 10 areas with edema. MfOP abnormalities were associated with the retinal locations of NPDR (Chi–square test, p<0.001). The mfOPs were, however, also abnormal in many retinal areas without NPDR (97 out of the 420 areas). Conclusions: The cells contributing to the generation of mfOPs function abnormally in diabetic eyes with early NPDR, and local mfOP abnormalities are associated with (but not restricted to) the retinal sites of NPDR. The existence of mfOP abnormalities in retinal areas that do not have signs of NPDR suggests that functional abnormalities of the inner retina are more widespread than NPDR, and might precede its appearance.
Keywords: diabetic retinopathy • electroretinography: clinical • diabetes