Investigating the Metabolic and Lipidomic Profiles That Are Associated With Varying Grades of Diabetic Maculopathy and Retinopathy in South Wales

Introduction

Diabetic Retinopathy (DR) places a huge burden on ophthalmic clinical services throughout the world as one of the leading causes of reduced vision in working age adults [1,2]. This is only going to worsen as the population grows with estimates of 700 million patients worldwide by 2045, with increase in people with DR from an estimated 103 million in 2020 to 160 million in 2045 [3]. The pathophysiology of DR is poorly characterised, and although the investigators understand some aspects of the disease process, some remain unexplored.

Early in the development of DR there is a breakdown of the blood-retinal barrier with loss of pericytes. This allows for leakage of vascular constituents in the retina and also flow of metabolic markers from retinal tissue into the systemic circulation[4]. There are a number of aspects of this process that warrant further investigation from a perspective of the metabolic disturbances that occur, and what might be driving them.

Early in the development of DR there is disruption to the autoregulation of blood flow to the retina with an apparent disturbance in arginine metabolism as reflected in serum metabolomic analysis [5,6]. This evidence and evidence of subsequent alterations in fatty acid oxidation as well as complex lipid synthesis is derived from animal models and small population studies in very specific ethnic cohorts, such as the Pima Indians in the USA [5,7]. Furthermore, there appears to be limited evidence for an effect on purine metabolism which is crucial for phosphocholine and phosphatidylethanolamine synthesis [8].

Along with disruption in synthesis of these crucial lipids, ceramide metabolism is disturbed with a reduction in the "probarrier", very long chain ceramides and increased production of shorter chain "pro-apoptotic" ceramides [9-12]. This provides a link to the potential therapeutic benefit of fenofibrate which has been shown in large scale studies to have a beneficial effect on DR independent of its effects on serum Low Density Lipoprotein (LDL) levels [9,13]. This is postulated to potentially be due to altered ceramide metabolism and synthesis, a process that is well established to be disrupted in people with diabetes [14].

All of this evidence provides insights into some small areas of dysregulated lipid metabolism and function within the retina, but a more coherent picture is required to truly understand the role in potential biomarker development as well as understanding the potentially mechanistic roles disturbed lipid metabolism may play in the context of DR.

An emerging area of interest is the interplay between the gut microbiome and retinal disease. In this novel field there is some evidence that the constitution of the gut microbiome may influence the retinal pathology that individuals develop through a complex interplay of circulating factors. Once such factor is the metabolite of a bile acid tauroursodeoxycholic acid (TUDCA) [15-18]. This acts on the Takeda G protein-coupled receptor 5 (TGR5) receptor expressed in the retina, and has some evidence that it reduces the access of inflammatory cells into the retina protecting it from DR development and preserving photoreceptors [15-18]. However, the evidence is somewhat scant and there is not a complete understanding of how this effect may occur, and if it is indeed reproducible in different experimental environments.

As part of this study the investigators aim to investigate the changes in lipids within the blood that may reflect deleterious changes in retinal lipid metabolism not only resulting from but also contributing to the loss of integrity of the blood-retinal barrier, and how this might be affected by circulating markers associated with altered gut microbiome constituents, which has evidence to influence the development of DR and lipid metabolism.

Background

DR presents a great challenge to clinical practice and public health care provision from the low- and middle-income countries to high-income countries such as the United Kingdom [1]. As such any insight into mechanistic pathways that might be exploited to treat DR at an earlier stage potentially give great benefit in relieving healthcare systems of a vast quantity of patients requiring more intensive treatment with a greater burden of follow up care.

Vital to the development of any successful novel therapy or biomarker is a thorough understanding of the fundamental pathophysiological mechanisms by which the disease initiates and develops. Currently there is good understanding about certain aspects of DR such as the role of vascular endothelial growth factor (VEGF) in driving proliferative change and increased vessel permeability that underlies diabetic macular oedema (DMO) [19]. However, relatively little is understood about the metabolic changes that occur in the retina to drive the production of VEGF.

Fundamentally it is well established that poor glycaemic control predisposes to worse DR [20,21]. However, this variation in haemoglobin A1c (HbA1c) probably only accounts for approximately 6.6% of the risk of developing visions threatening DR [5,22]. It is vital that the research community better understand the unexplained risk to provide better early care for those with DR.

Currently the research community has a handful of studies into the changes in lipid metabolism in those with DR and several studies in murine models [7,7,9-12,18]. At present this does not form a coherent picture of which lipid pathways are truly disrupted and may well underpin the early changes in DR. In previous studies it has been postulated that complex sphingolipid and ceramide metabolism may in part reflect differences in those with and without DR [7,10,11]. This, however, has not been repeated in different groups of patients and not validated in a different population.

Likewise, there are some number of studies investigating the role of the gut microbiome in retinal diseases [16,23-26]. There is an even smaller number which focus on gut microbiome diversity and DR development [26,27]. Crucially there appears to be some good evidence that the gut microbiome may influence the development of DR via as yet ill-defined pathways [27].

Our proposed study aims to investigate the changes in lipid metabolism in people with different grades of DR, with this knowledge the research community will be able to better understand how these changes may predispose individuals to develop various grades of DR, and potentially provide novel therapeutic targets for future intervention.

Research Question

Are there blood plasma lipids that might reflect differing stages of DR, potentially act as a biomarker to guide screening and give mechanistic insight into the pathophysiology of DR?

Study Outline

Aims and Objectives

Primary objective

• To investigate the serum lipidome in people with differing severity of DR

Secondary objectives

• To compare the lipidomic markers in those with and without diabetic maculopathy to gain potential mechanistic insight
• To investigate the interaction of circulating markers associated with the microbiome and how it might influence lipid metabolism in the retina and the severity of DR expressed by an individual
• To map the lipid metabolism pathways that might contribute to development of differing grades of DR

Method

Study Type

Observational case-control study

Study Overview

Observational case-control study comparing lipidomic, and metabolomic profiles in people with diabetes and varying degrees of DR and DMO as outlined below. They will be screened and recruited via routine clinic attendance and identification from our current clinic database of DR patients. They will be provided with information regarding the study prior to consent.

Subsequent to this data will be collected from their case records including basic demographic data, medical history, current therapy for diabetes and details of their ophthalmic diagnosis including relevant parameters from optical coherence tomography (OCT) scans and clinical examination including Logarithm of the Minimum Angle of Resolution (LogMAR) visual acuity. They will provide serum, and urine samples. To minimise the impact of the study to participants and the additional costs implied, this sample collection will be taken at routine clinic visits where possible by members of the research team within Prince Philip Hospital (PPH).

Participants will have all data pseudo-anonymised on collection with a database of identifiable information and designated study IDs kept in the study file within HDdUHB in a locked secure office. As it is an observational case-control study there will be no randomisation, and blinding will only occur at the sample processing and analysis stage and the research team at HDdUHB will only have access to the pseudo-anonymised data.

Study Populations, Subject Selection, Recruitment and Study Schedule

Retrospective potential participant identification and recruitment:

Eligible patients will be identified from case records by the direct care team from the HDdUHB pre-existing DR database and other clinical sources including clinic attendance records. Inclusion of the research nurses as part of the research delivery team will allow for more efficient organisation of the study visit including informed consent and sample collection by the research nurses as part of the study around normal clinic attendance. This improves the efficiency of organisation of study visits around clinic visits. This was discussed with patients with diabetes who agreed this was an acceptable and preferable approach to make it as easy for them as possible to have the sample collection visit arranged.

These patients will be contacted on an ad hoc process via invitation letter and Participant Information Sheet (PIS) and the most up to date GDPR leaflet for patients from the HDdUHB R\&D department. This will be followed up by the research team with a telephone interview from the R\&D nurses. This will adhere to the telephone screening schedule (see appendix) should the participant communicate their willingness to participate. This "interview" will broadly comprise checking a potential participants willingness to participate, an assessment of their eligibility and organisation of the study visit - this will not be a part of the research data collection. If they wish to be included in the study, a visit to the R\&D department will be arranged around their standard clinic attendance. This visit will be for further assessment of eligibility prior to informed consent and sample collection.

Prospective recruitment:

Patients attending medical retina service clinics with newly diagnosed DR of varying degrees or DMO will be sent a PIS, the most up to date GDPR leaflet for patients from the HDdUHB R\&D department, and invitation letter prior to attendance at clinic. The reviewing clinician will then discuss the study with the patient and invite them to participate prior to clinical review. In the case of an eligible patient who is a new case identified on the day, the clinician may introduce the study and provide the PIS and invitation letter to the patient whilst awaiting OCT scans and clinical review. The clinician will then share the potential participants details with the research delivery team to arrange for the study visit including further eligibility assessment, informed consent and sample collection.

Additionally, there will be posters publicising the study posted within PPH which will contain contact information for the study team so that interested potential participants can contact the study team. After this, they will receive the invitation letter and PIS, which will be followed by a telephone call from the research delivery team to assess eligibility and plan for their study attendance.

This is outlined in the following flow diagram:

Consent will be provided during the attendance of the R\&D department if the person agrees to participate and meets the inclusion and exclusion criteria.

The investigators will aim to recruit 14-20 patients in the groups defined below in section 2.2.4, this would satisfy the power calculations outlined below.

Eligibility Criteria

Inclusion criteria will be:

Diagnosis of type 2 diabetes mellitus Male or female aged 18 - 80 inclusive

Exclusion criteria will be:

Participant unable or unwilling to consent to inclusion in the study Prior treatment with intravitreal therapies for DMO Potential participant with a known infective disease that may put the study team at risk (eg. TB, HIV, hepatitis) Age 17 yo or less or 81 yo or older Known underlying genetic condition affecting lipid metabolism

Participants will be allocated into one of six groups, based on their varying grades of diabetic eye disease as detailed below:

Group 1 - No diabetic retinopathy or maculopathy Group 2 - Mild preproliferative DR Group 3 - Moderate Pre-proliferative DR Group 4 - Severe pre-proliferative DR Group 5 - Proliferative DR Group 6 - DMO with any degree of DR and no prior treatment

Power Calculation

Power Calculation

For primary outcome of lipidomic markers the investigators have based our calculations on three previous studies. The first is Yun et al., 2020, which demonstrated altered concentrations of acylcarnitine C16 (AC(C16:0). Here, AC(C16:0) was on average 0.35 μM lower in those with DR than those without (mean ± standard deviation: 0.125 ± 0.025 μM vs 0.09 μm). The second study is that of He et al., 2024, which compares various lipids between those with T2DM and DR versus those with T2DM and no DR. In thus study those with T2DM and DR demonstrated ceramide (d18:0/24:0) (Cer(d18:0/24:0)) of 22.36 Log2 transformed data to 23.00 in those without DR with a standard deviation of approximately 0.5 Log2 units.

Using the below formula for calculation of sample sizes:

n=2[(z_α+z_π )^2/{((μ_obs-μ_con ))/σ}^2 ]

Where, z is from the standardised normal distribution, representing significance and power. The investigators set the significance level at 5% (chance of a type I error), z_α = 1.96. The investigators set the power (π) at 90% (100% - β , the chance of a type II error) giving z_π = 1.282.

This generated a group size of 11 for AC(C16:0) and 13 for Cer(d18:0/24:0). Thus, groups of at least 14 should provide adequate power based on the above calculations.

Statistical Analysis Plan

Statistical Analysis Data Processing: Using standard protocols already established within JLG's group using standardised data extraction and normalisation for lipidomic and metabolomic techniques Software: PyChem, Umetrics SIMCA, R Studio, MetaboAnalyst 3.0, and ROCCET. Data analysis: suitable univariate two-way anova for parametric data and Kruskal-Wallis testing for nonparametric data will be used for statistical testing of suitable clinical parameters. Multivariate statistical analysis including PLS-DA and heirachical cluster analysis of metabolic networks will be performed for lipidomic and metabolomic data. Regression analysis will be used for the analysis of clinical parameters and potential biomarkers. Receiver operating characteristic curve analysis will be used for assessment of their performance at discriminating different severities of DR and DMO.

Significance Threshold: P < 0.05 with corrections for false discovery rates.

Study Outcomes

The study outcome will be completed analysis of circulating lipids to provide the insight as to whether there are any changes in systemic circulating lipids and metabolites that might be associated with DR and potentially provide some explanatory mechanism for the differential disease expression in different individuals.

Timescale

The study will apply for HDdUHB sponsorship via the Research Quality and Sponsorship Group (RQSG) by October 2024, with a view to proceeding to NHS REC and HRA/HCRW review in November/December 2024. Portfolio status will be assessed as part of the HRA/HCRW review. Following approvals and local setup, the target date for commencing recruitment is January 2025, with study visits to begin from mid-January 2025 for up to 12 months to January 2026. Analysis will occur from July 2025 and finalisation of results and submission for publication from April 2026 with ongoing analysis of blood plasma and urine metabolites and lipids up until January 2030 only for this study herein. The retention time is longer than the initial analysis so that any re-analysis can be performed should there be any discrepancy in results or to validate the findings of the study. Once this analysis has been completed samples will be destroyed, this may be prior to January 2030.

The end of the study is defined as completion of all study visits, and completion of all metabolomic and lipidomic analyses on samples collected.

Publication Policy

The results will be submitted for peer reviewed publication and presentation at conferences for dissemination. In any instance of publication or presentation, only anonymised information will be published from which no individual participant will be identifiable from the grouped data. There will be no individual case studies or publication of any data that would allow for identification of any volunteer. For publication purposes, Francis Sanders will be the first author, Rebecca Thomas the second author, Colm McAlinden the third author, Julian Griffin penultimate author and Eirini Skiadaresi the final author. Any report will be shared with the study Sponsor, HDdUHB, HCRW and all of the collaborating team, and all participants will be given the option to receive a copy of published reports.

Ownership of the data arising from this study resides with the study team and their respective employers. Any intellectual property arising as a result of this study will be shared between the study contributors in accordance with their proportional contribution.

References

1. Teo ZL, Tham Y-C, Yu M, Cheng C-Y, Wong TY, Sabanayagam C. Do we have enough ophthalmologists to manage vision-threatening diabetic retinopathy? A global perspective. Eye Lond Engl 2020;34:1255-61. https://doi.org/10.1038/s41433-020-0776-5.
2. Liew G, Michaelides M, Bunce C. A comparison of the causes of blindness certifications in England and Wales in working age adults (16-64 years), 1999-2000 with 2009-2010. BMJ Open 2014;4:e004015. https://doi.org/10.1136/bmjopen-2013-004015.
3. Teo ZL, Tham Y-C, Yu M, Chee ML, Rim TH, Cheung N, et al. Global Prevalence of Diabetic Retinopathy and Projection of Burden through 2045: Systematic Review and Meta-analysis. Ophthalmology 2021;128:1580-91. https://doi.org/10.1016/j.ophtha.2021.04.027.
4. Wistrup Torm ME, Dorweiler TF, Fickweiler W, Levine SR, Fort PE, Sun JK, et al. Frontiers in diabetic retinal disease. J Diabetes Complications 2023;37:108386. https://doi.org/10.1016/j.jdiacomp.2022.108386.
5. Du X, Yang L, Kong L, Sun Y, Shen K, Cai Y, et al. Metabolomics of various samples advancing biomarker discovery and pathogenesis elucidation for diabetic retinopathy. Front Endocrinol 2022;13:1037164. https://doi.org/10.3389/fendo.2022.1037164.
6. Sumarriva K, Uppal K, Ma C, Herren DJ, Wang Y, Chocron IM, et al. Arginine and Carnitine Metabolites Are Altered in Diabetic Retinopathy. Invest Ophthalmol Vis Sci 2019;60:3119-26. https://doi.org/10.1167/iovs.19-27321.
7. Fort PE, Rajendiran TM, Soni T, Byun J, Shan Y, Looker HC, et al. Diminished retinal complex lipid synthesis and impaired fatty acid β-oxidation associated with human diabetic retinopathy. JCI Insight n.d.;6:e152109. https://doi.org/10.1172/jci.insight.152109.
8. Jian Q, Wu Y, Zhang F. Metabolomics in Diabetic Retinopathy: From Potential Biomarkers to Molecular Basis of Oxidative Stress. Cells 2022;11:3005. https://doi.org/10.3390/cells11193005.
9. Busik JV. Lipid metabolism dysregulation in diabetic retinopathy. J Lipid Res 2021;62:100017. https://doi.org/10.1194/jlr.TR120000981.
10. Kady NM, Liu X, Lydic TA, Syed MH, Navitskaya S, Wang Q, et al. ELOVL4-Mediated Production of Very Long-Chain Ceramides Stabilizes Tight Junctions and Prevents Diabetes-Induced Retinal Vascular Permeability. Diabetes 2018;67:769-81. https://doi.org/10.2337/db17-1034.
11. Levitsky Y, Hammer SS, Fisher KP, Huang C, Gentles TL, Pegouske DJ, et al. Mitochondrial Ceramide Effects on the Retinal Pigment Epithelium in Diabetes. Int J Mol Sci 2020;21:3830. https://doi.org/10.3390/ijms21113830.
12. He M, Hou G, Liu M, Peng Z, Guo H, Wang Y, et al. Lipidomic studies revealing serological markers associated with the occurrence of retinopathy in type 2 diabetes. J Transl Med 2024;22:448. https://doi.org/10.1186/s12967-024-05274-9.
13. Chew EY, Davis MD, Danis RP, Lovato JF, Perdue LH, Greven C, et al. The Effects of Medical Management on the Progression of Diabetic Retinopathy in Persons with Type 2 Diabetes: The ACCORD Eye Study. Ophthalmology 2014;121:2443-51. https://doi.org/10.1016/j.ophtha.2014.07.019.
14. Croyal M, Kaabia Z, León L, Ramin-Mangata S, Baty T, Fall F, et al. Fenofibrate decreases plasma ceramide in type 2 diabetes patients: A novel marker of CVD? Diabetes Metab 2018;44:143-9. https://doi.org/10.1016/j.diabet.2017.04.003.
15. Mantopoulos D, Murakami Y, Comander J, Thanos A, Roh M, Miller JW, et al. Tauroursodeoxycholic Acid (TUDCA) Protects Photoreceptors from Cell Death after Experimental Retinal Detachment. PLOS ONE 2011;6:e24245. https://doi.org/10.1371/journal.pone.0024245.
16. Schmidt NS, Lorentz A. Dietary restrictions modulate the gut microbiota: Implications for health and disease. Nutr Res 2021;89:10-22. https://doi.org/10.1016/j.nutres.2021.03.001.
17. Li J, Huang Z, Jin Y, Liang L, Li Y, Xu K, et al. Neuroprotective Effect of Tauroursodeoxycholic Acid (TUDCA) on In Vitro and In Vivo Models of Retinal Disorders: A Systematic Review. Curr Neuropharmacol 2024;22:1374-90. https://doi.org/10.2174/1570159X21666230907152207.
18. Feng S, Guo L, Wang S, Chen L, Chang H, Hang B, et al. Association of Serum Bile Acid and Unsaturated Fatty Acid Profiles with the Risk of Diabetic Retinopathy in Type 2 Diabetic Patients. Diabetes Metab Syndr Obes Targets Ther 2023;16:2117-28. https://doi.org/10.2147/DMSO.S411522.
19. Stone J, Itin A, Alon T, Pe'er J, Gnessin H, Chan-Ling T, et al. Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia. J Neurosci 1995;15:4738-47. https://doi.org/10.1523/JNEUROSCI.15-07-04738.1995.
20. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet Lond Engl 1998;352:837-53.
21. Diabetes Control and Complications Trial Research Group, Nathan DM, Genuth S, Lachin J, Cleary P, Crofford O, et al. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-86. https://doi.org/10.1056/NEJM199309303291401.
22. Hirsch IB, Brownlee M. Beyond hemoglobin A1c--need for additional markers of risk for diabetic microvascular complications. JAMA 2010;303:2291-2. https://doi.org/10.1001/jama.2010.785.
23. Peng S, Li JJ, Song W, Li Y, Zeng L, Liang Q, et al. CRB1-associated retinal degeneration is dependent on bacterial translocation from the gut. Cell 2024;187:1387-1401.e13. https://doi.org/10.1016/j.cell.2024.01.040.
24. Labetoulle M, Baudouin C, Benitez Del Castillo JM, Rolando M, Rescigno M, Messmer EM, et al. How gut microbiota may impact ocular surface homeostasis and related disorders. Prog Retin Eye Res 2024;100:101250. https://doi.org/10.1016/j.preteyeres.2024.101250.
25. Floyd JL, Grant MB. The Gut-Eye Axis: Lessons Learned from Murine Models. Ophthalmol Ther 2020;9:499. https://doi.org/10.1007/s40123-020-00278-2.
26. Zhou Z, Zheng Z, Xiong X, Chen X, Peng J, Yao H, et al. Gut Microbiota Composition and Fecal Metabolic Profiling in Patients With Diabetic Retinopathy. Front Cell Dev Biol 2021;9:732204. https://doi.org/10.3389/fcell.2021.732204.
27. Wang R, Wang Q-Y, Bai Y, Bi Y-G, Cai S-J. Research progress of diabetic retinopathy and gut microecology. Front Microbiol 2023;14:1256878. https://doi.org/10.3389/fmicb.2023.1256878.