Measurement of Lens Autofluorescence for Diabetes Screening

Abstract

Many efforts have been made lately to develop cost-effective, simple, and reproducible tests for diabetes screening besides the already established fasting plasma glucose, the oral glucose tolerance test, and the glycated hemoglobin A1_c. Several tests have been proposed lately, based on the measurement of the so-called advanced glycation endproducts (AGEs). AGEs production is exacerbated during hyperglycemia, and their accumulation in different tissues reflects the degree and duration of dysglycemia. The human lens represents a tissue where AGEs accumulation can be particularly well assessed. The present article comments on the article by Cahn et al. published in this issue of the Journal of Diabetes Science and Technology. Cahn and coauthors tested a new scanning confocal biomicroscope for its accuracy to detect noninvasively subjects with diabetes or at risk for developing diabetes.

Within the so-called Maillard reaction, advanced glycation endproducts (AGEs) result from the nonenzymatic reaction of reducing sugars with proteins, lipids, and nucleic acids.¹ Although AGEs are better known as products of hyperglycemia, they also form in food during heat-enhanced cooking.² Evidence has accumulated that dietary AGEs are partially absorbed and either retained in the body, or eliminated into the urine.^3,4

AGEs exert their deleterious effects by receptor-dependent or receptor-independent mechanisms. Glycation of proteins and lipoproteins can alter their normal function by several mechanisms: change in molecular conformation, alteration in enzyme activity, changes in clearance and interference with receptor recognition.⁵ AGEs receptors (e.g. RAGE) are present on the surface of different cells such as macrophages, adipocytes, endothelial cells, and vascular smooth muscle cells.⁶ The activation of RAGE (a member of the immunoglobulin multiligand receptor family) by AGEs triggers intracellular signal transduction, promoting inflammatory response, apoptosis, prothrombotic activity, expression of adhesion molecules, and oxidative stress.^6-8 This explains why the exacerbated generation and accumulation of AGEs in diabetes has been linked to increased cardiovascular risk and the risk for development of diabetes complications.^1,9

The measurement of HbA1c (a so-called Amadori product, actually an intermediate product of the Maillard reaction) as a parameter that mirrors the quality of glycemic control within the previous 8 to 12 weeks has markedly contributed to the guidance of diabetes therapy, but also to our understanding of the development of diabetes complications.¹⁰ The results of Diabetes Control and Complication Trial—Epidemiology of Diabetes Interventions and Complications have strengthened the “glycemic memory” concept, postulating that metabolic control over several years predicts the development of diabetic complications.¹¹ The fact that glycated proteins form under hyperglycemic conditions, accumulate within organs that are prone to diabetic complication, and have a very long persistence, suggests AGEs to be ideal candidates for the substrate of the “glycemic memory.”^12-14 In other words, by measuring AGEs one can assess the overall hyperglycemic burden of the last years.

Since AGEs accumulate even in prediabetic stages of dysglycemia, measuring AGEs represent a very tempting approach for diabetes screening and for the screening of patients at high risk for developing diabetes complications. The major problem is to identify methods that allow for a simple, rapid, reproducible, and cost-effective measurement of AGEs and thus qualify as a screening method.

Several methods are available for the measurement of AGEs. Circulating or tissue-bound AGEs can be measured (1) by ELISA (enzyme-linked immunosorbent assays) using monoclonal or polyclonal antibodies,¹⁵ (2) using the fluorescence properties of some AGEs and assessments based on fluorescence spectroscopy,¹⁶ or (3) using high-performance liquid chromatography and mass spectrometry (MS), with the later method being probably the most reliable.¹⁷ Further MS-based methods allow for the measurement of some AGEs in the urine by isotope dilution and gas chromatography/mass spectrometry analysis.^17,18 Moreover, methods exist that allow for quantitative measurement of protein glycation, oxidation, and nitration adducts by liquid chromatography with triple quadrupole mass spectrometric detection,¹⁹ liquid chromatography with tandem mass spectrometric (LC-MS/MS) detection,²⁰ or matrix-assisted laser desorption ionization-mass spectrometry with time-of-flight detection (MALDI-TOF/MS).^17,21

But all these methods have at least one of following drawbacks: they are complicated, are expensive, are not largely available, are mainly suitable for the measurement of circulating AGEs and not for tissue-bond AGEs, or are time-consuming. Besides, for diabetes screening purposes, the measurement of circulating AGEs is not reliable since circulating AGEs have a large diurnal variation depending on food intake.²² Therefore, the measurement of tissue-bond AGEs is preferable.

Tissue-bound AGEs are usually measured in the skin because of its accessibility. The golden standard is represented by the biochemical analysis of skin biopsies²³ but rapid, noninvasive methods have also been made available measuring skin autofluorescence (AGE-Reader, DiagnOpics, Groningen, Netherlands or SCOUT DS®, VeraLight, Albuquerque, NM) or corneal or lens autofluorescence (Fluorotron, Ocumetrics, San Jose, CA; or ClearPath DS-120® Lens Fluorescence Biomicroscope, Freedom Meditech, Inc, San Diego, CA).^16,24,25 Measures of corneal autofluorescence are particularly suitable for patients who underwent a cataract surgery, in whom lens autofluorescence (AF) measurements make no sense.²⁶

One physiological aspect has to be particularly taken into account when measuring lens AF: The age-related opacification of the lens is part of the nonenzymatic glycation of the lens proteins and the degree of the glycation process (that can be assessed by measurements of the lens AF) reflect partly this physiologic process and can be exacerbated in the presence of hyperglycemia.^27-30

In this issue of the Journal of Diabetes Science and Technology, Cahn et al present data from a study aiming at investigating possible uses of a new scanning confocal biomicroscope to identify subjects with undiagnosed type 2 diabetes.

The study recruited 53 subjects physician-diagnosed with type 1 or type 2 diabetes mellitus or prediabetes and 180 subjects self-reported as normal (as control group) in whom measurements of lens AF (as a surrogate of AGEs accumulation) were performed. Since lens AF physiologically increases with age,³⁰ all comparisons were performed with an age-adjusted lens AF value of healthy subjects. At a defined threshold (fluorescence deviation of 2500 counts above the value of age-matched healthy controls), the authors report for their method a sensitivity and a specificity (67% and 94%, respectively) for identifying subjects with type 2 diabetes that is comparable to that of glucose threshold tests. Therefore, the authors highlighted the feasibility of lens AF to screen subjects for undiagnosed type 2 diabetes. These data are in line with those of Koefoed et al, who reported that measurements of the lens AF (Fluorotron) contributed with a sensitivity of 79% and a specificity of 100% to the diagnosis of type 2 diabetes.³¹

In the study by Cahn et al, the lens AF continuously increased from normal subjects to subjects with prediabetes, type 2 respectively type 1 diabetes mellitus in a manner that suggests that this value is dependent on the overall glycemic exposure. However, these data have to be considered with caution due to the small number of subjects investigated in some of the groups (e.g. only 9 subjects with prediabetes).

There seems to be a good reproducibility of the method, with an approximately 7.5% coefficient of variation for the fluorescence intensity measurements.

The data by Cahn and coworkers are interesting because they suggest that scanning with a confocal biomicroscope for the assessment of lens AF represents a sensitive, noninvasive, and rapid method for type 2 diabetes screening that also does not require fasting of subjects. Further studies are warranted to strengthen these findings.

What confers further interest to the investigation of lens AF is the fact that it can be performed in ophthalmologic units. Many subjects with undetected diabetes attend an ophthalmologic unit for the clarification of their visual impairment as one of the first symptoms of their diabetes. Therefore, the performance of a lens AF measurement might represent an easy-to-do screening for diabetes in these settings. Further technical advantages are that dilation of the pupils is not required and the rapidity of the test (6 seconds). The investigation is suitable for most patients without cataract or lens replacement surgery.

Conclusion

Accumulation of AGEs in tissues seems to parallel the development of diabetes and diabetic complications. Therefore, especially the measurement of tissue-bond AGEs might be used for diabetes screening or the screening of patients at high risk for developing diabetic complications. Noninvasive measurements of AGEs accumulation by assessment of skin, lens, or corneal AF emerged lately and are able to generate rapid and reliable data. Some major advantages are that they do not require fasting and their sensitivity and specificity seems to be comparable to that of classical diabetes screening methods like fasting plasma glucose, HbA1c, and the oral glucose tolerance test.

Therefore, noninvasive measurements of tissue AF might be particularly suitable for type 2 diabetes mellitus or prediabetes screening. The availability of different methods that allow for measurements at the level of the skin, lens, or cornea bears the advantage that the most suitable method can be chosen (e.g. lens AF can be measured in an ophthalmologic unit, skin AF in a general practitioner setting) but bears also the challenge to validate each of these devices. The validation process for screening purposes not only in subjects with known diabetes but especially on a population-based manner is paramount for the future acceptability of these devices.