A New Approach to Staging Diabetic Eye Disease

Topic The goal of this review was to summarize the current level of evidence on biomarkers to quantify diabetic retinal neurodegeneration (DRN) and diabetic macular edema (DME). Clinical relevance With advances in retinal diagnostics, we have more data on patients with diabetes than ever before. However, the staging system for diabetic retinal disease is still based only on color fundus photographs and we do not have clear guidelines on how to incorporate data from the relatively newer modalities into clinical practice. Methods In this review, we use a Delphi process with experts to identify the most promising modalities to identify DRN and DME. These included microperimetry, full-field flash electroretinogram, spectral-domain OCT, adaptive optics, and OCT angiography. We then used a previously published method of determining the evidence level to complete detailed evidence grids for each modality. Results Our results showed that among the modalities evaluated, the level of evidence to quantify DRN and DME was highest for OCT (level 1) and lowest for adaptive optics (level 4). Conclusion For most of the modalities evaluated, prospective studies are needed to elucidate their role in the management and outcomes of diabetic retinal diseases. Financial Disclosure(s) Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.


Parameter name
Retinal sensitivity Search terms "microperimetry and diabetes", "microperimetry and diabetic retinopathy" and "microperimetry and retinal neurodegeneration" from inception to 3/9/2021 Search results 330 articles

Pruning
We excluded repeated publications, letters, conference abstracts and those papers that were not relevant for the purpose of the study.A total of 38 articles in which full-text was available in English were reviewed.From these 38, 12 were excluded due to problems with the design of the study or absence of valuable information to answer the questions in the grid.
How is this parameter assessed?
Retinal sensitivity is assessed by microperimetry.MAIA microperimetry, similarly to standard automated perimetry (SAP), measures retinal sensitivity as the minimum light intensity that patients can perceive when spots of light stimulate specific areas of the retina.The examination can be customized with different number of stimuli covering a variable field of vision.The standard MAIA examination covers a 10° diameter area with 37 measurement points.
In MAIA, light stimuli are created by a white LED and projected directly onto the retina surface.The stimuli size are Goldmann III, background luminance is 4 asb and maximum luminance is 1000 asb, with a 36 decibels (dB) dynamic range.MAIA can work with 3 different projection strategies (software version 1.7.0,January 2013); the full threshold 4-2, the 4 Levels Fixed (4-LF) and the Scotoma-Finder (SF).In our group we have used the second strategy [1,2].
Has analytical validation been accomplished?If yes, give specifics of evaluation of parameter precision, accuracy, limit of detection, limit of quantitation, specificity, linearity and range, ruggedness and robustness.
Yes, but it is not currently used to measure retinal neurodysfunction in DR.
In eyes with AMD, it has been previously found the point-wise coefficient of repeatability of microperimetry (representing the location where 95% of the test-retest differences are expected to lie) to be ±4.37 dB or less [3].

Factors that should be taken into account:
• Pupil dilation: pupil dilation does not significantly influence microperimetry performance [4].[9] The values mainly depend on the following: age, device, the central degrees of the macula analyzed, and the number of stimulus points used.

Scientific Understanding of Relationship to DRD
What is the biological, anatomic and/or functional rationale or plausibility for the association of this parameter with DRD? (i.e., what is the degree to which diabetes triggers subsequent steps in a pathophysiologic pathway and the role of the parameter in that causal or outcome pathway) Retinal neurodysfunction and neurodegeneration are early events in the pathogenesis of DRD.
What is the current understanding of the molecular mechanism(s) underlying the alterations in this parameter in association with DRD? (specify whether mechanisms are physiologic, pathologic or pharmacologic) The reason by which there is a reduction of retinal sensitivity in the diabetic retina could be primarily due to the neural impairment/neural loss as has been reported in aging [10].It should be noted that a total reduction of 27 µm in retinal thickness results in approximately 1 dB of sensitivity loss in subjects with mild NPDR [11].In addition, it has been reported that retinal sensitivity is correlated to GCL-IPL thickness in diabetic subjects [12] and ganglion cell count [13].
The following confounding factors should be considered in clinical practice: Macular edema: In patients with DME a decrease of 0.83 dB for every 10% of deviation of retinal thickness from normal values has been reported [14].Macular edema could cause light to be blocked or scattered before it reaches the photoreceptors, suggesting that optical effects are a major cause of sensitivity loss.In addition, it has been recently reported a direct relationship between retinal sensitivity and macular thickness in the DM1 group [15].Nonperfused areas: Areas of capillary nonperfusion resulting from severe nonproliferative or proliferative diabetic retinopathy show morphologic changes of the retinal structure, which may lead to a loss of sensitivity [16].In addition, retinal sensitivity correlated inversely with superficial foveal avascular zone area measured by optical coherence tomography angiography (OCTA) [17].
Cognitive impairment: Retinal sensitivity has been found decreased in those patients with mild cognitive impairment and dementia [1].Therefore, cognitive status should be taken into account in the interpretation of the results or as exclusion criteria in clinical trials.
What is the outcome measure with which this parameter is associated?

Retinal neurodysfunction
What is the link between the parameter and the accepted clinical outcome measure?Retinal sensitivity is a measurement of retinal neurodysfunction/neurodegneneration

Performance Expectations in DRD
What sensitivity to detect change does this parameter provide compared to the current standard (if available)?
The current gold standard to assess neurodysfunction is mfERG which is an objective measurement of suprathreshold responses at photopic adaptation levels and the first-order kernel responses originate from the cone photoreceptors and bipolar cells [18].However, this examination is cumbersome and time consuming and, therefore, it is reserved for clinical trials.
Microperimetry is a measurement of retinal sensitivity at mesopic adaptation levels, which may be mediated by both rod and cone photoreceptors.Measurements of retinal sensitivity are also not solely influenced by the physiological condition of the retina, but by the entire visual pathway.Retinal sensitivity measured by microperimetry has been found to correlate strongly with the integrity of the photoreceptor band on high-resolution OCT imaging in AMD [19,20].
It should be noted that no significant correlation between the measured functional deficit of microperimetric retinal sensitivity and mfERG implicit time nor response amplitude was observed in AMD subjects [21].This lack of correlation could be explained because they are giving us different information about retinal function and are measured under different conditions.It has been reported that microperimetry is even more sensitive that mfERG at detecting early functional changes [22] and changes in response to treatment [23].
In comparison with mfERG, microperimetry is a noninvasive and rapid test, with the total procedural time for each test in our study lasting for approximately 6 to 7 minutes.This contrasts with the total procedural time of at least 15 to 20 minutes for mfERG recording of one eye.The reported experience on this issue is very limited.
In early stages of DR, supplementation with high-dose DHA plus xanthophyll carotenoid multivitamin during 90 days was associated with a progressive and significant improvement of macular function measured by microperimetry (macular sensitivity increased from 25.9 ± 2.4 dB at baseline to 27.3± 2.3 dB in the DHA group; p <0.05) [24].
In advanced stages of DR, microperimetry retinal sensitivity was correlated with visual response in patients treated with undergoing intravitreal ranibizumab for DME.Thus, in good responders the mean intra-subject improvement after month 3 of treatment was 2.28 dB (p=0.049),whereas in poor responders was 1.07 db (p=0.28)[25].
Is there a temporal relationship between changes in this parameter and the clinical outcome?If yes, please give specifics of that relationship Impairment of retinal sensitivity is already the outcome (neurodysfunction).If we consider the clinical outcome microvascular progression there is a lack of experience.
What is the specificity of changes in this parameter for DRD?
It seems a very high specific method.However, the presence of cataracts [26], vascular leakage and cognitive impairment are potential confounding factors that should be considered.

Types of Data and Available for Evidential Evaluation
Are there preclinical studies that address the relationship of this parameter to outcomes in DRD?
N Are there literature reviews that address the relationship of this parameter to outcomes in DRD?
References for literature reviews N/A Please give the Level of Evidence available from these combined studies (use Tables 1 and 2 below to determine Level of Evidence.For this purpose, please substitute "DRD parameter" for "tumor marker" or "marker" in Table 1)

Statistical Considerations
What is the specific relationship of the parameter to clinical outcomes?Please specify effect sizes and measures of variability This is a parameter of neurodysfunction.
This point should be explored.

What is the usefulness of the parameter or its thresholds for clinical or research decision making?
A large study in order to obtain a normative data base is needed.This will be crucial to standardize the presence of neurodysfunction and monitoring any intervention that could have an impact on neurodysfunction.Are there covariates that should be adjusted for when considering this parameter?Age.It has been reported a loss in visual field sensitivity with increasing age (approximately 0.8 dB per decade) by Humphrey Field Analyzer [10].Are there any additional statistical considerations for the use of this parameter?No

Gap Analysis
What are the gaps in the literature to prove or disprove the utility of this parameter?
The main gaps are: 1) Lack of standardization.2) Altered values in patients with cognitive impairment* *There is evidence that retinal sensitivity assessed by microperimetry is able to discriminate normocognition, mild cognitive impairment and dementia in subjects with type 2 diabetes.In addition, a significant correlation was found between retinal sensitivity and the MRI and 18FDG-PET parameters related to brain neurodegeneration [1].This is because, fundus-driven microperimetry assesses not only the functional status of the retina but also the entire visual system, and it is a dynamic test that requires short-term memory and adequate perceptual speed and executive function.Moreover, the impairment of gaze fixation (which can also be assessed by microperimetry) is strongly related with cognitive decline [2].In your opinion, what clinical research study/studies could address these gaps?
A large study in order to obtain a normative data base is needed.
Are there currently available datasets that could be used for these validation efforts?
A large prospective study (RECOGNISED.H2020.Grant agreement: 847749) is ongoing.In this study the correlation between retinal sensibility measured by MAIA and other parameters of DRD will be evaluated in type 2 diabetic patients.

Miscellaneous Questions
Is this parameter currently employed in clinical use?Yes, but it has not been generally used for the assessment of neurodysfunction in the setting of DR.Is assessment instrumentation needed to measure this parameter currently: available commercially, available but not FDA approved, not readily available, or not available?

Available commercially
What is the ease of implementation in the following environments: high resource academic center, high resource community practice, low resource/underserved environment?High resource academic center and high resource community practice.However, this could be extended to low-income countries when the cost would be reduced due to competitiveness among the companies.The reason by which there is a reduction of retinal sensitivity in the diabetic retina could be primarily due to the neural impairment/neural loss as has been reported in aging.
Overall, retinal sensitivity assessed by microperimetry seems a very high specific method.However, the presence of cataracts, vascular leakage and cognitive impairment are potential confounding factors that should be considered.Among them, the presence of cognitive impairment represents a significant limiting factor in patients >65 years old.The current gold standard to assess neurodysfunction in diabetic patients is mfERG.However, this examination is cumbersome and time consuming and, therefore, it is reserved for clinical trials.In comparison with mfERG, microperimetry is a noninvasive and rapid test, with the total procedural time for each test lasting for approximately 6 to 7 minutes.This contrasts with the total procedural time of at least 15 to 20 minutes for mfERG recording of one eye.
The reported clinical experience on retinal sensitivity to evaluate neurodysfunction due to diabetes is very limited.However, in early stages of DR, supplementation with high-dose DHA plus xanthophyll carotenoid multivitamin was associated with a progressive and significant improvement of macular function measured by microperimetry.
There is a lack of consensus regarding the normal values of retinal sensitivity that mainly depend on the following variables: age, type of device, central degrees of the macula analyzed and the number of stimulus points used.Therefore, a large study in order to obtain a normative data base is needed.This will be crucial to standardize the screening of neurodysfunction and monitoring any intervention that could have an impact on neurodysfunction/neurodegeneration.
It is unknown whether neurodysfunction assessed by microperiemtry is a predictor of microvascular impairment.A large prospective study (RECOGNISED.H2020.Grant agreement: 847749) is ongoing.In this study, the correlation between retinal sensibility measured by MAIA and other microvascular parameters of DRD will be evaluated in type 2 diabetic patients.In addition, the RECOGNISED clinical trial will determine whether retinal sensitivity is a predictor of cognitive decline and dementia.This study will be finished in 2024.

•
Adaptation: mesopic adaptation can influence microperimetry performance.For this purpose, 20 min of mesopic adaptation prior to testing is required.
[5]re is a lack of consensus regarding the normal values.This is because the different devices used and the characteristics of subjects included.Please see below the most important series: • Midena et al.Macular automatic fundus threshold versus standard perimetry threshold.Eur J Ophthalmol. 2007;17:63-8.[5]