**3. Diabetic macular edema (DME)**

### **3.1 Pathogenesis**

#### *3.1.1 The normal blood retinal barrier (BRB) in healthy subjects*

The inner and outer retinal barriers keep the retina immune privileged and regulate fluid and molecular entry and drainage into and to the outside of the retina keeping the retina in a dehydrated state [9].

The inner blood retinal barrier (retinal blood vessel walls) is formed by tight junctions (zonula occludens (ZO)) between endothelial cells, adherens junctions between pericyte cytoskeleton and endothelial cells and glial cell processes wrap around retinal capillaries [10]. Astrocytes and Müller cells stabilize the tight junctions between endothelial cells and ensheath vascular plexuses [11]. Finally, microglia produce soluble factors important for vesicular communication necessary for the maintenance of the inner blood retinal barrier [12, 13].

The outer blood retinal barrier is formed of junctional complexes between retinal pigment epithelial (RPE) cells formed of tight, adherens and gap junctions separating the neurosensory retina from the fenestrated choriocapillaris. It controls the transport of fluid and solutes into and to the outside of the retina to maintain its integrity [14, 15].

#### *3.1.2 Pathologic alterations in diabetic macular edema*

Microglia monitor the physiological microenvironment in the retina and can detect early signs of hyperglycemia leading to their activation [16]. Activation of microglial cells is usually associated with perivasculitis with consequent release of inflammatory mediators, including vascular endothelial growth factor (VEGF), tumor necrosis factor alpha (TNF-α), interleukin 6 (IL-6) and monocyte chemotactic protein-1 (MCP-1), resulting in extension of the inflammation from the inner retina to all retinal layers with breakdown of the blood retinal barrier, increased vascular permeability and retinal neuronal damage [17–20].

Hyperglycemia upregulates intercellular adhesion molecule-1 (ICAM-1) which together with vascular cell adhesion molecule mediate leukocyte adhesion to the vascular endothelium (leukostasis), resulting in vascular damage and capillary nonperfusion. Additionally, leukocytes share in microvascular damage by the release of cytokines and superoxide [21].

Inflammation and hypoxia adversely affect the functions of Müller cells, altering their potassium channels with consequent accumulation of intracellular fluids [22, 23]. Both inflammation and hypoxia also stimulate retinal Müller cells to produce VEGF, tumor necrosis factor alpha (TNF-α), interleukin 1β (IL-1β) and prostaglandins, all of which contribute to the increased vascular permeability [24, 25]. On the other hand, Müller cells aggravate inflammation through stimulation of cluster of differentiation (CD) 40 and through the release of adenosine triphosphate (ATP) which promotes microglial inflammation [26].

Vascular endothelial growth factor, a main factor in the progression of DME and proliferative diabetic retinopathy (PDR) (**Figure 2**), is upregulated through several pathways, particularly the activation of hypoxia-inducible factor 1 (HIF-1) [27]

**Figure 2.** *VEGF in DME and PDR.*

and phospholipase A2 (PLA2) [28]. VEGF both increases vascular permeability by promoting phosphorylation of tight junction proteins, such as occludin and zonula occludens-1 (ZO-1) [29], in addition to its effect in promoting angiogenesis through the activation of mitogen-activated protein (MAP) [30].

Other angiogenic factors, particularly angioprotein 2 (Ang-2), an antagonist to endothelial receptor tyrosine kinase (Tie2), has been shown to promote vascular leakage in diabetic retina (**Figure 3**) [31].

Reactive oxygen species (ROS) are an important link between hyperglycemia and the main pathways responsible for hyperglycemic damage. Although mitochondrial production is an important source of ROS [32], yet more recently it was proved that ROS derived from nicotinamide adenine dinucleotide phosphate (NADPH) oxidase is involved to a higher extent in the process of pericyte apoptosis on exposure to high glucose levels [33]. High glucose could stimulate ROS production through protein kinase C (PKC) β-dependent phosphorylation of the p47phox (the phagocyte NADPH oxidase/NOX2 organizer) subunit (involved in the activation of NADPH oxidase) [34].

Diabetic macular edema is also associated with RPE dysfunction and impaired control of transport of water from the subretinal space into the choriocapillaris and vice versa [14].

#### **3.2 Pretreatment evaluation of DME**

#### *3.2.1 Screening*

All diabetic patients should perform baseline screening with full ophthalmological examination and imaging (fluorescein angiography (FA) together with OCT). Although FA has long been used for the diagnosis of diabetic retinopathy (DR), yet the intravenous dye injection can cause some side effects [35]. In addition, FA does

*Diabetic Macular Edema, Clinicopathologic and Keys for Management DOI: http://dx.doi.org/10.5772/intechopen.112974*

#### **Figure 3.**

*Role of angioprotein-2 in increasing vascular permeability and DME.*

not identify or separate the pathologies in the superficial and deep capillary networks of the retina and the leaking fluorescein dye may obscure the alterations in vascular morphology. All FA drawbacks are recently solved by the use of OCTA [36].

In type 1 diabetes mellitus (type 1 DM), it is usually advised to do the first ophthalmologic examination 5 years after the discovery of diabetes mellitus (DM), while in type 2 diabetes mellitus (type 2 DM), the examination is done once DM is diagnosed, then annually or earlier as needed [37].

#### *3.2.2 Classifications of DME*

Several attempts have been carried out to classify DME. A summary of these classifications is given as follows:

	- a.The vasogenic type was defined as retinal thickening with visible retinal vascular abnormalities detectable on biomicroscopy and fundus photography (focally grouped microaneurysm and dilated capillaries) at the macular region usually associated with lipid exudates.
	- b.The nonvasogenic type, on the other hand, was defined as retinal thickening with no visible vascular dilations and probably no lipid exudates detectable on biomicroscopy and fundus photography.

Three types of optical coherence tomography are recognized:

a.Diffuse thickening type (sponge-like diffuse retinal thickening) (**Figure 5**), which results from increased vascular permeability and breakdown of the inner blood retinal barrier secondary to inflammation and oxidative stress [14].

**Figure 4.** *Mixed type of DME.*

**Figure 5.** *Diffuse thickening type of DME.*

*Diabetic Macular Edema, Clinicopathologic and Keys for Management DOI: http://dx.doi.org/10.5772/intechopen.112974*

#### **Figure 6.** *Cystoid macular edema (CME) type.*

#### **Figure 7.** *Serous retinal detachment (SRD) type.*



#### **Table 1.**

*Optical coherence tomography (OCT) grading of diabetic macular edema (DME).*


#### **Table 2.**

*En face image-based classification of diabetic macular edema.*

*Diabetic Macular Edema, Clinicopathologic and Keys for Management DOI: http://dx.doi.org/10.5772/intechopen.112974*

The idea of this classification is to construct en face images from three-dimensional (3D) images to visualize and localize the extent of the area of fluid at a specific retinal depth in cases of DME (**Table 2**).

Conclusions drawn from this classification were as follows:

First, the presence of fluid in Segment 2 resulted in a significantly worse visual outcome as compared to cases without fluid in this area, which is probably because of the disturbed oxygenation and elimination of metabolites from the layer of photoreceptors.

Second conclusion was that the extent of fluid in Segment 1 did not affect the final visual acuity (VA) in DME [35].

#### *3.2.3 Baseline predictors for a good treatment response based on OCT findings*

Several pretreatment criteria were found to be of prognostic value as regards better vision gain posttreatment, and these predictors include; less subretinal fluid (SRF) and few intraretinal cystoid spaces (IRC) and no vitreomacular traction (VMT) (**Figure 8**) [39].

Better vision gain is also achieved if there was no disruption or disorganization of the inner retinal layers (DRIL) [40] and if there was preservation of the ellipsoid zone (EZ) and external limiting membrane (ELM) [41].

A thin subfoveal choroid at baseline may predict unfavorable posttreatment visual acuity, while eyes with a thicker baseline subfoveal choroidal thickness had better short-term anatomic and functional responses [42].

Several OCTA biomarkers have been found to be valuable in the determination of the degree of retinal ischemia and these include two FAZ biomarkers, namely foveal avascular zone area (FAZ-A) and the irregularity in the contour of the foveal avascular zone contour irregularities (FAZ-CI) and three vessel biomarkers, namely the tortuosity of retinal vessels (VT), the average vessel caliber (AVC) and the density of retinal vessels (VD) [43].

#### *3.2.4 Pretreatment systemic monitoring*

Systemic risk factors for progression of diabetic macular edema should be controlled to provide a suitable environment for ocular treatment to give better and longer lasting favorable results;

#### **Figure 8.**

*Base line predictors for an unfavorable treatment response based on OCT findings.*

Multiple randomized clinical trials demonstrated the benefits of both controlled blood sugar level and arterial blood pressure in the reduction of retinopathy progression [44–49]. The American Diabetes Association mentioned that the glycated hemoglobin (HbA1c) level should not exceed 7%, while the arterial blood pressure level should be kept under 130/80 mmHg and total lipids under 100 mg/dL [50], in addition, however, treatment should not be delayed to correct all systemic parameters [51–53].

Other parameters that are also to be taken into consideration are smoking cessation [54, 55], weight loss if required for a normal body mass index [55], renal impairment [54, 55] and sleep apnea [55], as all of those factors have an impact on the results of management of DME.

#### **3.3 Treatment options**

#### *3.3.1 Laser treatment*

For diabetic macular edema, laser photocoagulation was the gold standard for treatment until the introduction of anti-VEGF therapy.

#### *3.3.1.1 Focal/grid laser*

According to the Early Treatment Diabetes Retinopathy Study (EDTRS), focal/ grid macular laser was shown to be effective in marked macular edema reduction or cure, in addition to the reduction of the risk of moderate visual loss by 50% at the termination of the 3-year follow-up trial [56]. More recent clinical trials following ETDRS showed a similar result of reduction of moderate visual loss by 50% with a gain of ≥10 letters of visual acuity in 28% of DME eyes [57].

Focal laser applied to leaking microaneurysms at least 300–500 μm from the center of the macula and guided by FA in noncenter-involving DME remains a gold standard for treatment of this subset of DME [43, 58].

In grid laser, mild power laser marks are made with a spot size of 50 μm to 200 μm, to treat widespread and diffuse edema [59], particularly in cases of resistance or contraindication to the use of anti-VEGF drugs [60].

Combining laser photocoagulation with anti-VEGF injections for DME has been described in several studies. Combined treatment was found to be more effective in improving visual acuity in DME patients. With combined treatment, 10 to 40% of patients gained ≥15 letters in their VA [61–63].

The mechanism of action of focal/grid photocoagulation is not clear, although direct closure of leaking microaneurysms in focal laser and the destruction of the high oxygen consuming photoreceptors, reduction of retinal tissue and improvement of oxygenation together with restoration of the function of retinal pigment epithelial cells in grid laser are postulated mechanisms [64–66] (**Figure 9**).

#### *3.3.1.2 Micropulse laser*

The basis of micropulse laser (diode laser 810 μm and yellow laser 577 μm) unlike the conventional continuous-wave laser is to apply the minimum laser irradiance (watts per square meter). The aim is to raise the temperature of the RPE cells leading

**Figure 9.**

*Possible mechanism of action of grid laser in the management of DME.*

to their activation, but without exceeding the protein denaturation threshold in the neural retina which is thus not injured by the laser application.

In the traditional continuous-wave mode, the preset laser energy is delivered totally in a single laser pulse of 0.1–0.5 s. In the micropulse mode, a train of repetitive short laser pulses (each pulse is 100–300 μs) is delivered within an envelope of laser energy having a width of 0.1–0.5 s [67].

The Diabetic Macular Edema and Diode Subthreshold Micropulse Laser (DIAMONDS) trial found that the subthreshold micropulse laser was noninferior to the traditional continuous-wave laser (focal/grid) in terms of functional (VA improvement) and anatomical (optical coherence tomography central macular thickness (OCT CMT)) improvement [68].
