**1. Introduction: the role of biomarkers in disease prediction**

The prevalence of diabetes mellitus is increasing worldwide. The International Diabetes Federation estimated that 415 million people had diabetes in 2015, 90% of whom were

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

diagnosed with type 2 diabetes. With these trends continuing, 642 million patients with diabetes are expected by 2040 [1].

with type 1 diabetes, and DR progression by 54% over a period of 6.5 years [6, 7]. Patients with type 2 diabetes had a 25% reduction of DR with good glycemic control [8]. Even though HbA1c remains the most widely accepted biomarker nowadays [5], the "Joslin 50-Year Medalist Study," which focused on the identification of endogenous protective factors in patients with a diabetes duration of at least 50 years (therefore named "Medalists") showed that longitudinal glycemic control was unrelated to diabetic complications. However, the presence of specific advanced glycation end products (AGEs) (plasma carboxyethyl-lysine and pentosidine) was strongly associated with the develop-

Potential Imaging Biomarkers in the Development and Progression of Diabetic Retinopathy

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

11

Cytokines from aqueous humor or vitreous sample have also been considered in the search for a DR biomarker. Increased levels of vascular endothelial growth factor (VEGF), plateletderived growth factor (PDGF), transforming growth factor beta (TGF-β), and nitric oxide (NO) are commonly found in DR. However, these biomarkers can only be assessed using invasive methods. As tears are more accessible than serum and intraocular fluids (i.e., vitreous or aqueous humor), research has also started to focus on the presence of potential markers in this body fluid. Candidate biomarkers in tear fluid include nerve growth factor (NGF), lipocalin-1, lactotransferrin, lysozyme C, lacritin, lipophilin A, immunoglobulin lambda chain, heat shock

Ocular imaging biomarkers would offer the advantage of gaining an insight in the actual pathologic evolution of DR non-invasively. The most important candidates for such biomark-

Diabetic retinopathy is diagnosed clinically by the presence of microaneurysms and small hemorrhages visualized during fundoscopy. Assessing the presence and number of microaneurysms as well as their rate of formation and disappearance has been suggested to be an appropriate marker of retinal vascular damage and therefore DR progression [11]. However, there are microvascular changes that have been shown to antecede fundoscopically visible

Within the last decades and with the implementation of specialized computer software systems, grading of retinal vessel diameters to document generalized vessel narrowing or widening has become increasingly sophisticated, objective, and reliable. Multiple population-based studies have used these systems to calculate retinal vascular caliber in terms of the central retinal artery and vein equivalent (CRAE, CRVE), which summarizes the average diameter of the internal lumen of the vessel, reflecting the visualized erythrocyte column [12]. In sum, the results of these studies provide evidence for an association between larger venular caliber and DR in patients with type 1 [11, 13] as well as type 2 diabetes [14, 15], therefore being consistent with clinical experience. However, reported findings on arteriolar caliber remain contradictory.

ment of diabetic vasculopathy complications [9].

protein 27 (HSP 27), and tumor necrosis factor-α (TNF-α) [10].

**2. Microvascular changes in diabetic retinopathy**

ers will be discussed in the following chapter.

lesions of DR including microaneurysms.

**2.1. Retinal vessel caliber**

Patients with diabetes are at substantially increased risk of developing complications. In light of the cost of interventions implemented throughout the natural history of these complications, diabetes constitutes a tremendous clinical and public health burden that exceeds the resources of healthcare systems even in the most affluent countries. The International Diabetes Federation has reported that most countries spent 5–20% of their total healthcare budget on diabetes in 2015, which amounted to 673 billion US dollars in health expenditure worldwide. This figure is expected to increase to about 802 billion US dollars by 2040 [1].

The complications of diabetes are commonly divided into macrovascular complications including myocardial infarction, heart failure, and stroke, and microvascular complications including diabetic nephropathy, neuropathy, and retinopathy. Diabetic retinopathy (DR) is the most common microvascular complication of diabetes. The incidence of DR increases with the duration of diabetes. After 20 years, nearly all patients with type 1 diabetes and more than 60% of those with type 2 diabetes will develop signs of DR [2].

Current treatment guidelines target proliferative disease and macular edema, two sight threatening complications. The most common approaches are intravitreal injections of vascular endothelial growth factor (VEGF)-inhibiting agents or corticosteroids, laser treatments, and surgical interventions. These treatments are often sight saving, but are invasive and cost-prohibitive. Therefore, we need to shift our focus to targeting upstream events at earlier stages of non-proliferative DR (NPDR).

The major health economic burden caused by the increasing number of patients diagnosed with diabetes raises the need to identify patients at high risk of developing DR and sight threatening complications. Reliable biomarkers that help to predict the development and progression of the disease have to be defined.

A biomarker is traditionally defined as "a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacological responses to a therapeutic intervention" [3]. In order to be useful in the prevention of DR, a biomarker should (1) non-invasively detect early preclinical disease before the first clinical signs of the disease appear, (2) be causally linked or be an indicator of a causal mechanism that leads to the development of the disease, and (3) be consistently and strongly associated with the disease [3, 4]. Suitable biomarkers should identify patients at low risk to defer DR screening intervals facilitating cost-effective management and optimized resource allocation. Furthermore, biomarkers should help to predict the progression of DR to the vision-threatening stage, and may forecast the response to different treatment modalities, facilitating individualized care [5]. Important factors for valid biomarkers are reproducibility and validity in different populations. Furthermore, their measurements must be quick, cost-effective, and applicable in daily clinical decision-making [5].

To date, many serum variables have been proposed to be associated with DR incidence and progression. According to the Diabetes Control and Complications Trial (DCCT), a median glycated hemoglobin (HbA1c) of 7.2% reduced DR incidence by 76% in patients with type 1 diabetes, and DR progression by 54% over a period of 6.5 years [6, 7]. Patients with type 2 diabetes had a 25% reduction of DR with good glycemic control [8]. Even though HbA1c remains the most widely accepted biomarker nowadays [5], the "Joslin 50-Year Medalist Study," which focused on the identification of endogenous protective factors in patients with a diabetes duration of at least 50 years (therefore named "Medalists") showed that longitudinal glycemic control was unrelated to diabetic complications. However, the presence of specific advanced glycation end products (AGEs) (plasma carboxyethyl-lysine and pentosidine) was strongly associated with the development of diabetic vasculopathy complications [9].

Cytokines from aqueous humor or vitreous sample have also been considered in the search for a DR biomarker. Increased levels of vascular endothelial growth factor (VEGF), plateletderived growth factor (PDGF), transforming growth factor beta (TGF-β), and nitric oxide (NO) are commonly found in DR. However, these biomarkers can only be assessed using invasive methods. As tears are more accessible than serum and intraocular fluids (i.e., vitreous or aqueous humor), research has also started to focus on the presence of potential markers in this body fluid. Candidate biomarkers in tear fluid include nerve growth factor (NGF), lipocalin-1, lactotransferrin, lysozyme C, lacritin, lipophilin A, immunoglobulin lambda chain, heat shock protein 27 (HSP 27), and tumor necrosis factor-α (TNF-α) [10].

Ocular imaging biomarkers would offer the advantage of gaining an insight in the actual pathologic evolution of DR non-invasively. The most important candidates for such biomarkers will be discussed in the following chapter.
