**5.2. Retinal neurodegeneration**

Signs of neurodegeneration were not visible in fundus examination in the era of the ETDRS. Therefore, these changes did not contribute to the characterization or diagnosis of the disease. However, retinal neurodegeneration has widely been accepted as part of DR over the last decades.

These abnormalities in retinal neural tissue lead to well-studied functional changes that typically precede the clinical diagnosis of DR, and in some cases occur even prior to the diagnosis of diabetes. Neurofunctional impairment becomes apparent as a dysfunction in dark adaption [119], abnormal contrast sensitivity [120], and altered microperimetry [105], as well as electroretinogram (ERG) results. The electroretinogram (ERG) is one of the most effective diagnostic tools in this context, with the oscillatory potential implicit time being the most consistent and widely reported aspect of the ERG that changes early in DR [121]. A delay in implicit time in multifocal ERG (mfERG) has been shown to be highly predictive (86% sensitivity and 84% specificity) of new retinopathy development at specific locations over 3 years in patients with early stages of DR at baseline [122, 123]. The European Consortium for the Early Treatment of Diabetic Retinopathy (EUROCONDOR) trial currently tests mfERG for its use and potential in DR prediction. However, while ERG is a very sensitive technique to detect neurofunctional deficits, it is also a quite burdensome and time-consuming examination.

association of venular dilatation and thinning of the RNFL along with deficits in the ERG was detected in adolescents with type 2 diabetes, showing that the structural changes are

Potential Imaging Biomarkers in the Development and Progression of Diabetic Retinopathy

http://dx.doi.org/10.5772/intechopen.71747

23

The mechanisms behind this neurodegeneration are not completely clear. Increased apoptosis in neuronal tissue may be caused by chronic hyperglycemia, when neuronal cells experience up to 4-fold increase in glucose uptake. If hyperglycemia is prolonged, nerves are damaged [136]. Additionally, glucose and glutamate accumulation in the extracellular space, increased oxidative stress, inflammation and imbalance in the production of neuroprotective factors are other factors thought to be involved in the development of neurodegeneration in the setting of DR [137]. Apoptosis of the retinal ganglion cells also tends to be accompanied by reactive changes in macroglial cells, known as "reactive gliosis." Apart from astrocytes, the predominant type of macroglia is the Müller cell, which is unique to the retina. One of the most prominent characteristics of reactive gliosis is that Müller cells overexpress glial acidic fibrillary protein (GFAP), which is considered a sensitive indicator of central nervous system injury, and is normally only expressed by retinal astrocytes [138]. Müller cells span the entire retina, surround all blood vessels, and produce molecules that contribute to the modulation of blood flow and vascular permeability. In addition, they are essential for the survival of neurons. Therefore, glial cells, and especially Müller cells, are thought to play a key role in the pathogenesis of both retinal microangiopathy and neurodegeneration. Unfortunately, Müller

Because neurons cannot be replaced, DR becomes irreversible with continuous disease progression. The identification of biomarkers that predict the development of neurodegeneration as well as mediators in the cross talk between neurodegeneration and microangiopathy is crucial for the development of new therapeutic strategies in DR. Safe and effective neuroprotective agents could possibly prevent neuronal apoptosis and vision loss but also impede the impairment of neurovascular coupling. Consequently, microvascular impairment and clinically apparent DR could be delayed. Evidence from the numerous studies mentioned above suggests that diabetic retinal neurodegeneration most likely precedes the microvasculopathy of DR. Functional examinations, like mfERG as well as structural evaluation of the inner retinal layers with SD-OCT may permit an early detection of the disease. However, further longitudinal studies are required to clarify the precise temporal relation between neurode-

Neurodegenerative changes occur outside the retina too. The cornea is one of the most densely innervated structures of the human body. A rich network of sensory nerves, known as the subbasal nerve plexus (SNP), derives from the ophthalmic division of the trigeminal nerve and lies between the corneal epithelium and Bowman's membrane [139]. This layer can be visualized with corneal confocal microscopy (CCM), a highly reproducible [140] in vivo imaging technique that provides diagnostic efficiency comparable to that of intra-epidermal nerve fiber density (IENFD) assessment [141, 142]. IENFD is the current gold standard for evaluating small nerve fiber damage, but is invasive, time-consuming and requires

accompanied by early vascular dysfunction [135].

cells can currently not be imaged *in vivo.*

generation and the microvascular alterations of DR.

**5.3. Neurodegeneration outside the retina**

Anatomical evaluation of retinal neurodegeneration has become possible with the implementation of SD-OCT. In OCT, the most useful measure for identifying diabetes-induced neurodegeneration is the thickness-reduction of the retinal nerve fiber layer (RNFL) and the ganglion cell complex, consisting of the ganglion cell layer (GCL) and the inner plexiform layer (IPL). Retinal ganglion cells (RGCs) are the retinal neurons in which the apoptotic process related to diabetes is first detected [124]. An impaired integrity of these cells compromises information processing and the transmission of visual signals to the brain. The damage primarily affects the RGC's nuclei and dendrites, as shown by a diffuse thinning of the combined retinal ganglion cell-inner plexiform layer (GCIPL). Secondarily, their axons are affected too, as indicated by a reduction of the retinal nerve fiber layer (RNFL) thickness [125]. A significant thinning of the GCIPL complex alone [126] or in combination with thinning of the RNFL has already been shown in patients with type 1 diabetes even without any fundoscopically manifest signs of DR [127, 128]. A longitudinal analysis in patients with type 1 diabetes depicted an average progressive thickness loss of 0.25 μm/year and 0.29 μm/ year in the RNFL and the GCL + IPL, respectively, over a 4-year follow-up period in patients with no or minimal DR, independent of age, sex and even Hb1Ac. Intriguingly, the extent of thickness loss was similar to that of patients with severe glaucoma [129]. Research results are also consistent in finding reduced RNFL and GCIPL thicknesses in patients with type 2 diabetes [130–132].

Further, relation between structural signs of diabetic retinal neurodegeneration and functional deficits has been investigated thoroughly. Reduced GCIPL complex thickness has been shown to significantly correlate with impaired visual function assessed by contrast sensitivity and pattern ERG amplitudes in patients with diabetes without DR [131]. In patients with type 1 diabetes and no or minimal DR, GCL thickness was an important predictor of loss of macular visual function measured by the Rarebit perimetry [133].

Research has also started to focus on the temporal and causal relationship of neurogenic and vascular changes in DR. Preliminary results of the EUROCONDOR study suggest that in patients with no or mild DR, retinal vessel caliber is independently associated with structural changes of the neuroretina. Specifically, CRAE was statistically significantly associated with macular GCL thickness and CRVE with RNFL thickness at the optic disc [134]. An association of venular dilatation and thinning of the RNFL along with deficits in the ERG was detected in adolescents with type 2 diabetes, showing that the structural changes are accompanied by early vascular dysfunction [135].

The mechanisms behind this neurodegeneration are not completely clear. Increased apoptosis in neuronal tissue may be caused by chronic hyperglycemia, when neuronal cells experience up to 4-fold increase in glucose uptake. If hyperglycemia is prolonged, nerves are damaged [136]. Additionally, glucose and glutamate accumulation in the extracellular space, increased oxidative stress, inflammation and imbalance in the production of neuroprotective factors are other factors thought to be involved in the development of neurodegeneration in the setting of DR [137]. Apoptosis of the retinal ganglion cells also tends to be accompanied by reactive changes in macroglial cells, known as "reactive gliosis." Apart from astrocytes, the predominant type of macroglia is the Müller cell, which is unique to the retina. One of the most prominent characteristics of reactive gliosis is that Müller cells overexpress glial acidic fibrillary protein (GFAP), which is considered a sensitive indicator of central nervous system injury, and is normally only expressed by retinal astrocytes [138]. Müller cells span the entire retina, surround all blood vessels, and produce molecules that contribute to the modulation of blood flow and vascular permeability. In addition, they are essential for the survival of neurons. Therefore, glial cells, and especially Müller cells, are thought to play a key role in the pathogenesis of both retinal microangiopathy and neurodegeneration. Unfortunately, Müller cells can currently not be imaged *in vivo.*

Because neurons cannot be replaced, DR becomes irreversible with continuous disease progression. The identification of biomarkers that predict the development of neurodegeneration as well as mediators in the cross talk between neurodegeneration and microangiopathy is crucial for the development of new therapeutic strategies in DR. Safe and effective neuroprotective agents could possibly prevent neuronal apoptosis and vision loss but also impede the impairment of neurovascular coupling. Consequently, microvascular impairment and clinically apparent DR could be delayed. Evidence from the numerous studies mentioned above suggests that diabetic retinal neurodegeneration most likely precedes the microvasculopathy of DR. Functional examinations, like mfERG as well as structural evaluation of the inner retinal layers with SD-OCT may permit an early detection of the disease. However, further longitudinal studies are required to clarify the precise temporal relation between neurodegeneration and the microvascular alterations of DR.
