Soluble CD163 and glycated haemoglobin were independently associated with the progression of diabetic retinopathy in adult patients with type 1 diabetes
•,,,,.
...
Abstract
Objective High vitreous levels of soluble (s)CD163 have been demonstrated in severe diabetic retinopathy (DR). The aim of this study was to explore the predictive values of plasma sCD163 and glycated haemoglobin (HbA1c) for DR progression in adults with type 1 diabetes.
Methods and analyses The study design was prospective. Fundus photography performed in 2009 and at follow-up (≤12 years later) were compared after being categorised according to the International Clinical Diabetic Retinopathy Disease Severity Scale. ‘DR progression at least one level’ was calculated. In 2009, data collection (sex, age, diabetes duration, metabolic variables, serum creatinine, macroalbuminuria and lifestyle factors) and biochemical analyses were performed. Plasma sCD163 and HbA1c were divided into quartiles. Logistic regression analyses were performed.
Results The prevalence of DR in 2009 versus at follow-up in 270 participants (57% male) were: no apparent 28% vs 18%; mild 20% vs 13%; moderate 24% vs 26%; severe 11% vs 13%; and proliferative DR 17% vs 30% (p<0.001). DR progression occurred in 101 (45%) patients. HbA1c ≥54 mmol/mol (≥7.1%) (>1st quartile) (adjusted odds ratio (AOR) 3.8, p<0.001) and sCD163 ≥343 ng/mL (>1st quartile) (AOR 2.6, p=0.004) were independently associated with DR progression. The associations with DR progression increased significantly from the first to the fourth quartile for HbA1c (AORs: 1; 2.5; 3.6; 7.4), but not for sCD163 (AORs: 1; 2.9; 2.4; 2.4).
Conclusion Plasma sCD163 may constitute a valuable biomarker for DR progression in addition to and independent of the well-established biomarker HbA1c.
What is already known on this topic
Soluble CD163 concentration in the vitreous is increased in patients with proliferative diabetic retinopathy.
What this study adds
Increased levels of plasma soluble CD163 and HbA1c at baseline were independently associated with the progression of diabetic retinopathy in patients with type 1 diabetes.
How this study might affect research, practice or policy
Plasma sCD163 may be a useful biomarker in addition to HbA1c for risk evaluation of diabetic retinopathy progression.
Introduction
Diabetic retinopathy (DR) is a major cause of vision impairment and vision loss.1 Long-term hyperglycaemia, measured as high levels of glycated haemoglobin (HbA1c), contributes to the development of DR and progression into proliferative diabetic retinopathy (PDR).2–8 However, some patients with good glycaemic control develop vision-threatening stages of DR, while some with poor control do not.1 Other reported risk factors for the progression of DR are hypertension, dyslipidaemia, smoking, pregnancy, current level of DR, duration of diabetes and genetic predisposition.9 The pathophysiology of DR is complex, but there is increasing evidence that systemic and local inflammatory factors are involved in the development of DR.1 8 10–13
CD163 is an endocytic receptor for haptoglobin–haemoglobin complexes and is expressed exclusively on macrophages and monocytes.14 After ectodomain shedding, the extracellular portion of CD163 circulates in the blood as a soluble (s) protein, sCD163.14 Anti-inflammatory mediators induce the expression of CD163, and pro-inflammatory mediators stimulate shedding leading to increased sCD163 levels.15 The enzymes ADAM17/TACE and ADAM10 contribute to increased shedding of sCD163 from macrophages.16
Increased levels of sCD163 have been demonstrated in the vitreous in patients with PDR compared with non-diabetic controls.11 13 The sCD163 levels were particularly high in patients who had developed fibrovascular membranes on the neuroretinal surface which can cause vitreous haemorrhage and tractional retinal detachment.11 Vascular endothelial growth factor (VEGF) is a main driver of the development of severe DR.17 A positive correlation between vitreous sCD163 and VEGF has been demonstrated.13 We have previously shown in a cross-sectional study that plasma sCD163 were increased in patients with any level of DR compared with patients without apparent DR.12
Screening is necessary to identify patients who have treatable DR. Retinal laser photocoagulation, dexamethasone intravitreal implant, intravitreal agents which neutralise VEGF or vitrectomy are effective tools to prevent permanent vision loss in patients with PDR.2 8 18
We hypothesised that plasma sCD163, in addition to HbA1c, may be a valuable biomarker for the onset or progression of DR.
The aim of this study was to explore the predictive values of sCD163 and HbA1c for the progression of DR at least one level in adult type 1 diabetes (T1D) patients with no apparent DR or non-proliferative DR (NPDR) at baseline.
Materials and methods
This prospective study was initiated in March 2009 and terminated in November 2021. All patients were recruited from one secondary care unit, specialised in diabetes mellitus, and all fundus photography was performed at the only specialist ophthalmology clinic in Växjö, Region Kronoberg, South East Sweden. Inclusion criteria were (1) patients with T1D who participated in our previous cross-sectional study of DR in 2009, where fundus photography and analyses of sCD163 were performed (n=287)12; and (2) at least one follow-up fundus photography was performed in Region Kronoberg between 2009 and 2021. Seventeen patients declined to participate in the follow-up, had died before any follow-up fundus photography was performed or could not be contacted and therefore excluded. Two hundred and seventy out of 287 (94%) patients were included. Exclusion criteria at baseline in 2009 were end-stage renal disease (ESRD), cancer, hepatic failure, pregnancy, psychotic disorders, severe substance abuse, cognitive deficiency or inadequate knowledge of Swedish.12 19 For the purpose of this prospective study, an ophthalmologist re-examined all fundus photography performed from 2009 to follow-up ≤12 years later, categorising the results according to the International Clinical Diabetic Retinopathy Disease Severity Scale.2 All data, except from follow-up fundus photography, was collected in 2009. We controlled for sex, age, diabetes duration, the duration of the follow-up period, metabolic variables, serum creatinine, macroalbuminuria, lifestyle factors, medication, foot complications and cardiovascular disease (CVD). Missing data is presented below table 1 and table 2. No imputation of data was performed.
Table 1
|
Baseline characteristics for all and separated into groups of different retinopathy disease severity among patients with type 1 diabetes
Table 2
|
Biochemical variables for all patients and comparisons between the five levels of DR for 270 T1D patients at baseline
Local screening program for DR
The screening with fundus photography was performed in a clinical setting. In Sweden, all patients with T1D (≥10 years old) are referred to eye clinics for regular screening of DR. In Region Kronoberg, patients with T1D with no DR are screened every second year. Patients with mild or moderate NPDR, independent of diabetes duration, are screened every year or every 6 months if the changes are close to the macula. Patients with severe NPDR will have a medical assessment within 3–6 months. Patients with PDR will have medical assessment within 2 weeks to 12 months depending on the severity and activity of the retinal disease.
Methods for screening
Patients with lower-grade DR were usually photographed using Scanning Laser Ophthalmoscopy (SLO) through an undilated pupil. The device used was an Optos California model: Panoramic Ophthalmoscope P200Dtx which produces an image of the retina in 200 degrees.
All patients with DR levels 3 or 4 were photographed using a Topcon fund camera through a dilated pupil. The SLO has a greater field of view and in cases when the wider field of view in the SLO photo would affect the level of DR, the assessed area was restricted to match the previously investigated area using the Topcon fundus camera. The SLO produces an image with slightly lower contrast and resolution than conventional fundus photography.
International Clinical Diabetic Retinopathy Disease Severity Scale
The grading of the International Clinical Diabetic Retinopathy Disease Severity Scale include level 0: no apparent DR; level 1: mild NPDR (microaneurysms and/or dot haemorrhages only); level 2: moderate NPDR (more than microaneurysms and/or dot haemorrhages but less than level 3); level 3: severe NPDR (more than 20 intraretinal haemorrhages in each of four quadrants or definite venous beading in at least two quadrants or prominent intraretinal microvascular abnormality in at least one quadrant and no PDR); level 4: PDR (neovascularisation (active or treated by pan-retinal photocoagulation) or vitreous/preretinal haemorrhage). The level of DR was determined by the eye with the most severe level of DR.
Quality assurance processes of fundus photography
Screening fundus photographs were identified from baseline in 2009 to follow-up (≤12 years later). For the purpose of this prospective study, an ophthalmologist with several years of experience of DR screening re-examined all fundus photographs. As all images were previously graded by ophthalmologists at the clinic, the two results were compared. The results did not match in 15 cases. These were re-examined by an additional ophthalmologist and a consensus decision was taken between the two examiners. The fundus images were graded using the International Clinical Diabetic Retinopathy Disease Severity Scale.2 The re-examining ophthalmologist, the author SH, was masked to the sCD163 and HbA1c levels, as well as all other data, when examining the photographs.2
Patients who had not followed the screening programme were invited to have a follow-up fundus photograph. Subanalyses were performed for deceased patients and for patients with shorter observation periods than 10 years.
Progression of DR by at least one level
The progression of DR by at least one level from baseline to follow-up was calculated for patients with no apparent DR, mild, moderate or severe NPDR.
Biochemical analyses
Blood samples for the analyses of sCD163 were collected in EDTA plasma tubes (BD, Franklin Lakes, New Jersey, USA). The analyses were performed using commercially available human ELISA DuoSet kits and supplementary ancillary kits (R&D Systems, Minneapolis, Minnesota, USA). Plasma samples from the participants were diluted 1:200 in Phosphate-buffered saline (PBS) with 1% Bovine serum albumin (BSA) and were run in duplicates according to the manufacturer’s instructions. Absorbance was measured at 450–580 nm in a FLOUstar optima plate reader (BMG Labtech GmbH, Ortenberg, Germany). A 7-point standard curve was used to create a 4-parameter logistic regression curve from which concentrations of unknown samples were calculated. The intra-assay coefficient of variation (CV) for the sCD163 analysis was 2.0%.12 sCD163 levels were first divided into quartiles, and second, dichotomised at the first quartile. Plasma sCD163 was analysed at the Diabetes Laboratory, Lund University, Lund, Sweden.
HbA1c was analysed with an Olympus automated clinical chemistry analyser (Olympus AU, Tokyo, Japan). The intra-assay CV was <1.2%. HbA1c levels were first divided into quartiles, and second, dichotomised at the first quartile.
Serum lipids were also analysed with Olympus AU. The intra-assay CV for total cholesterol was <2.1%; high-density lipoprotein-cholesterol <3.0%; low-density lipoprotein (LDL)-cholesterol <2.6%; and triglycerides <2.2%.
Serum creatinine was assayed by an AU2700 instrument (Beckman Coulter, Brea, California, USA). The intra-assay CV was <3%.
HbA1c, serum lipids and serum creatinine were analysed at the Department of Clinical Chemistry, Växjö Central Hospital, Växjö, Sweden.
Body mass index and blood pressure
Weight, length and blood pressure (BP) were measured by a nurse, according to standard procedures. Body mass index (kg/m2) was calculated.
Episodes of hypoglycaemia
A severe episode of hypoglycaemia was defined as needing help from another person. Episodes during the last 6 months prior to recruitment were registered.
Smoking and physical inactivity
Smokers were defined as having smoked any amount of tobacco during the last year.
Physical inactivity was defined as moderate physical activity, equivalent to 30 min of walking, performed less than once a week.12
Medication
Patients used either multiple daily insulin injections or continuous subcutaneous insulin infusion.
Lipid-lowering drugs (LLDs) used were hydroxy-methylglutaryl coenzyme A reductase inhibitors (statins) (ATC-code C10AA).
Indications for LLD were total cholesterol >4.5 mmol/L (>1.74 mg/dL) and/or LDL-cholesterol >2.5 mmol/L (>97 mg/dL) or previous CVD according to the Swedish national guidelines in 2009.20 The use of LLD was dichotomised into users and non-users.
Indications for AHD were systolic BP >130 mm Hg and/or diastolic BP >80 mm Hg or CVD according to the Swedish national guidelines in 2009.20 The use of AHD was dichotomised into users and non-users.
Kidney function
Serum creatine levels and the prevalence of macroalbuminuria, defined as an abnormal increase in albumin excretion rate in the range ≥300 mg albumin/g creatinine, were documented.
Foot complications
Foot complications were defined as neuropathy, angiopathy, past history of or present diabetic foot ulcer, foot infection, foot deformity, arthropathy or amputation of the lower limb. The results were dichotomised into foot complications or no foot complications.
Cardiovascular disease
CVD was defined as ischaemic heart disease, cardiac failure, stroke or transient ischaemic attack. The results were dichotomised into CVD or no CVD.
Statistical analysis
Histograms revealed that age, sCD163, HbA1c, serum lipids and serum creatinine were not normally distributed. Analyses were performed with Kruskal-Wallis test and data were presented as median (quartile (q)1, q3) or median (min–max). Linear-by-Linear Association was used for categorical data.
Crude odds ratios (CORs) were calculated with ‘progression of DR at least one level’ and ‘sCD163≥343 ng/mL’ as dependent variables. Variables with p values <0.10 for the CORs were included in multiple logistic regression analyses with ‘progression of DR at least one level’ and sCD163≥343 ng/mL as dependent variables. Nagelkerke R2 was presented. CIs of 95% were used. p<0.05 was considered statistically significant. SPSS V.23 (IBM) was used.
Results
In this prospective study, 270 patients with T1D participated (men 57%, women 43%). In 2009, the patients’ age ranged between 18 and 59 years and diabetes duration ranged between 1 and 55 years. The total prevalence of DR increased from 195 (72%) at baseline to 221 (82%) at follow-up. The follow-up period for 261 non-deceased patients was (median (q1, q3); min–max) 11 (11, 11); 1–12 years. Nine patients were deceased when the final data was collected and therefore had a shorter follow-up period of 5 (3, 8); 1–10 years. Five deceased patients had PDR at baseline and four had NPDR. One out of the four with NPDR acquired DR progression by at least one level. Subanalyses of 12 patients, still alive, who had a follow-up period shorter than 10 years, showed that 1 patient had PDR at baseline, 5 (46%) patients progressed at least one level of DR and 6 (54%) did not have any progression (p=0.003).
In table 1, baseline non-biochemical characteristics are compared between the five levels of DR. The 47 patients with PDR were the oldest, had the longest diabetes duration, highest systolic BP, as well as the highest prevalence of AHD, LLD and foot complications (all p<0.001).
In table 2, the results of the baseline biochemical analyses are compared between the five levels of DR. Exploration of the quartiles of sCD163 and HbA1c showed that the patients with PDR had the highest prevalence of plasma sCD163≥534 ng/mL (fourth quartile) (36%) (p=0.014) and HbA1c≥71 mmol/mol (fourth quartile) (40%) (p<0.001).
The distribution of the severity of DR at baseline and at follow-up, as well as the progression at least one level is presented in table 3 and with more details in the online supplemental material S1. The prevalence at baseline versus at follow-up was as follows: no apparent DR 28% vs 18%; mild NPDR 20% vs 13%; moderate NPDR 24% vs 26%; severe NPDR 11% vs 13%, PDR 17% vs 30% (p<0.001). Progression of DR at least one level occurred in 101 (45%) patients. Subanalyses showed that the progression of DR (at least one level) was in patients with no apparent DR at baseline 34.7%, in patients with mild NPDR 56%, moderate NPDR 44% and severe NPDR 57%.
Table 3
|
Distribution and progression of diabetic retinopathy from 2009 to follow-up ≤12 years later for all 270 patients
Associations with ‘progression of DR at least one level’ are presented in table 4. Model 1: plasma sCD163≥343 ng/mL (adjusted odds ratio (AOR) 2.6, p=0.004) and HbA1c≥54 mmol/mol (AOR 3.8, p=0.001 were independently associated with ‘progression of DR at least one level’. Model 2: for HbA1c, the associations with DR progression at least one level increased significantly from the first to forth quartile (AORs: 1; 2.5; 3.6; 7.4), which was not the case for sCD163 (AORs: 1; 2.9; 2.4; 2.4). The CORs were tried for all variables included in the study, but are not presented in table 4 (all p values were ≥0.24).
Table 4
|
High levels of HbA1c and sCD163 at baseline were independently associated with the progression of diabetic retinopathy
All variables included in the study were tried against sCD163≥343 ng/mL, but only ‘progression of DR at least level’ was significantly associated with sCD163≥343 ng/mL (all other p values>0.10).
Discussion
The main finding of this prospective study of DR in 270 patients with T1D was that increased levels of plasma sCD163 and HbA1c at baseline were independently associated with the progression of DR by at least one level after a follow-up period of up to 12 years. No other variables included in the study were independently associated with DR progression. The higher the levels of HbA1c, the stronger the associations were with DR progression. All levels of sCD163 above the first quartile were associated with the progression of DR, but the associations did not grow stronger by increasingly high sCD163 levels. Additionally, at baseline in 2009, patients with PDR had the highest levels of both plasma sCD163 and HbA1c.
High levels of sCD163 have previously been demonstrated in the vitreous in patients with PDR undergoing vitrectomy.11 13 Hitherto it has not, to the best of our knowledge, been shown that increased plasma sCD163 levels both reflect disease severity and have a prognostic value for the progression of DR. In our previous cross-sectional study, we had not explored whether increasing sCD163 levels were linked to increasing severity of DR.12 Our findings are in keeping with previous research suggesting that inflammatory factors are implicated in the development of DR.1 10–13 Impaired glycaemic control, expressed as high levels of HbA1c, is a well-established risk factor for the incidence and progression of DR,2–7 which was also demonstrated in our study. However, HbA1c is not a perfect indicator of glycaemic control.21 22 Sudden variations with high peaks of blood glucose contribute to the development of DR, but may not be reflected by very high HbA1c levels.21 22 Thus, plasma sCD163 may be a valuable predictive biomarker in addition to HbA1c.
Several quality assurance processes were performed. All fundus photographs were first examined by an ophthalmologist as part of the ordinary screening programme, followed by a re-examination by one ophthalmologist who was masked to all other data in the study. In a few dubious cases, an additional ophthalmologist was consulted. The severity of DR was graded according to the International Clinical Diabetic Disease Severity Scale, which is clinically useful and validated.2 Subanalyses were performed for patients with a shorter follow-up period less than 10 years and the results were controlled for the length of follow-up period. Precise ELISA techniques were used, and the commercial ELISA assay showed a low intra-assay CV for sCD163. The study population was well-defined. Pregnant women and patients with severe somatic comorbidities or psychiatric disorders with psychotic symptoms at baseline were excluded.12 High levels of sCD163 have been demonstrated in several types of diseases such as cancer and severe liver disorders where macrophages are activated.23 Therefore, patients with severe comorbidities were excluded in order to avoid confounding results. We systematically controlled for potential confounders, such as sex, age, diabetes duration, diabetes-related metabolic disturbances, lifestyle factors, medication, kidney function, CVD and foot complications. Diabetic nephropathy has in previous research been linked to increased levels of sCD163,24 but neither serum creatinine nor macroalbuminuria were associated with high plasma sCD163 (≥343 ng/mL), which could be due to the exclusion of patients with ESRD.
One limitation of our study was that plasma sCD163 was not analysed during the follow-up period. Yet, we were interested in the predictive value of sCD163. Higher sCD163 levels may indicate that it would be wise to have shorter DR screening intervals. Another weakness of this study was that two colour fundus imaging modalities were used. The resolution and contrast are slightly lower in SLO imaging, when acquiring photographs with the Optos California device, compared with conventional fundus photography. This might have underestimated the number of patients with progression from level 0 to level 1. This potential error would make the association between high levels of sCD163 and the progression of DR weaker. However, the SLO device was only used in level 0 and level 1 and did therefore not affect the level of DR in the higher levels of DR. A third weakness was that some patients had shorter follow-up periods as they had deceased or had moved out of Region Kronoberg. The majority of the diseased patients had PDR already at baseline, and just a few individuals could potentially have acquired DR progression after a longer follow-up period. The prevalence of DR progression was the same in patients with an observation period shorter than 10 years as in the whole population. We adjusted for the duration of the follow-up period, but there was no association between the follow-up period and the progression of DR.
In future research, it would also be of interest to explore whether sCD163 production indirectly or directly could be regulated by medication in order to decrease the risk of DR progression and to explore whether sCD163 could be used for therapeutic drug monitoring. There are drugs for other diseases that regulate sCD163 production.16 It would also be of interest to explore the association between sCD163 and the progression of DR in patients with type 2 diabetes.
In conclusion, increased levels of sCD163 and HbA1c were the only variables independently associated with DR progression at least one level in adult patients with T1D. Thus, in addition to HbA1c, sCD163 may be a valuable risk biomarker for DR progression.
