Study of the relation between plasma level of von Willebrand factor and diabetic retinopathy in type 2 diabetes

Diabetic retinopathy (DR) is a major preventable health complication of type 2 diabetes (T2DM). Hyperglycemia represents the main mechanism in the pathogenesis of DR leading eventually to endothelial dysfunction. vWF is a marker of endothelial dysfunction, but its relation to DR is not yet well-established. The present study included 60 patients with T2DM divided into three equal groups according to the presence and the stage of retinopathy. Each patient was subjected to full history taking, physical examination including fundus examination, and laboratory investigations including glycemic and lipid profile, urinary albumin to creatinine ratio (UACR), and plasma vWF level. Plasma vWF was significantly higher in patients with DR than in patients without retinopathy. vWF was positively correlated to age, HbA1c, diabetes duration, and UACR in DR groups and the total sample. Univariate regression analysis revealed that HbA1c, vWF, BMI, diabetes duration, and UACR were independent risk factors for DR, while multivariate regression analysis showed none of them to be an independent risk factor for DR. A vWF cut-off value of > 180 could differentiate patients with DR from patients without DR with a sensitivity of 97.5%, specificity of 75%, and AUC 0.937*. vWF level is higher in patients with DR reflecting the role of endothelial dysfunction in the development of DR. Moreover, it is one of the risk factors affecting DR in univariate but not in multivariate analysis. A cut-off value of vWF level was set to differentiate between patients with and without DR.


Introduction
Diabetes mellitus (DM) represents a major health problem in Egypt due to its wide prevalence and the high burden of its chronic complications. Type 2 diabetes (T2DM) is the most prevalent type, representing 90-95% of cases [1]. Chronic hyperglycemia is the main mechanism of microvascular complications. Hyperglycemia causes vascular damage through different pathways, including the polyol pathway, advanced glycation end products and the interaction with its receptors, protein kinase C and hexosamine pathway. The end result is endothelial dysfunction and endorgan damage [2].
Diabetic retinopathy (DR) is one of the most common microvascular complications. Moreover, it is the leading cause of blindness worldwide making it a major health problem. DR is classified into two stages: non proliferative (NPDR) and proliferative (PDR) [3]. Previous data reported elevated markers of inflammation and endothelial dysfunction in DR suggesting their role in the pathogenesis of DR [4].
Von Willebrand factor (vWF) is an adhesive and multimeric glycoprotein synthesized exclusively in endothelial cells and megakaryocytes. It is found in plasma and platelets and plays a central role in hemostasis. vWF level is elevated in cases of endothelial dysfunction [5].
Hyperglycemia-induced endothelial cell hypoxia leads to upregulation of plasma vWF levels. In stressful conditions, a higher level of vWF induces platelet aggregation, which plays a central role in DR pathogenesis [6].
Previous studies detected higher levels of vWF in patients with diabetes [7], coronary artery disease [8][9][10][11], and diabetic kidney disease [12], but very few studies discussed the relation between vWF and DR [13][14][15][16]. This invited us to conduct this study to assess the relation between plasma vWF levels and DR in participants with T2DM and the possible link between high plasma vWF levels and the risk of DR.

Participants, materials, and methods
The present case-control study included 60 patients with T2DM divided into three groups according to the presence and stage of retinopathy: group A included 20 patients without DR, group B included 20 patients with NPDR, and group C included 20 patients with PDR. Patients were recruited from the outpatient clinics of the Diabetes and Metabolism Unit and Ophthalmology Unit of Alexandria Main University Hospital from June 2020 to June 2021.
Exclusion Criteria included Type 1 diabetes mellitus (T1DM) patients, malignancy, acute trauma, acute infection, pregnancy, kidney failure, liver disease, a blood disorder, recent hospitalization for any reason, immune system disorder, chemotherapy or immunosuppression, and age-related eye disorder.
All patients were subjected to full history taking with special emphasis on age and duration of diabetes, and complete physical examination, including vital signs, weight, height, and waist circumference [17]. Body mass index (BMI) was calculated as body weight (kg) divided by body height squared (m 2 ). Fundus examination was done in the ophthalmology outpatient clinic by an expert consultant using a slit lamp biomicroscope plus a fundus lens.

Samples collection and storage for vWF assay
Blood samples (7 ml each) were collected by clean venipuncture. Blood was withdrawn slowly without venous stasis and care was taken to avoid frothing and with minimal delay. Samples were gently mixed by inverting the tubes 4-6 times and centrifuged at 2000-2500g for 15 min at room temperature. Plasma was separated, liquated, and stored at -20 C until assay time.

Quantitative determination of vWF by ELISA
Antigenic assay for the quantitative determination of VWF was done by Asserachrom® vWF antigen kit from Diagnostica Stago (France) (according to the manufacturers' instructions.)

Principle of the test
The vWF to be measured is captured by specific rabbit antihuman vWF antibodies coated on the internal wells of a plastic microplate than rabbit anti-vWF antibodies coupled with peroxidase bind to the remaining free antigenic determinants of the bound vWF. The bound enzyme peroxidase is revealed by its action on the tetramethylbenzidine TMB substrate. After stopping the reaction with a strong acid, the intensity of the color is directly proportional to the concentration of vWF initially present in the plasma sample.

Calculation of the results
Duplicate readings for each standard and sera were performed. The optical density for the standards was plotted versus the concentration of the standards and curve drawn on log/log paper. The normal range is from 47-197% of normal.

Statistical analysis of the data
Data were fed to the computer and analyzed using IBM SPSS software package version 20.0. (Armonk, NY: IBM Corp). For continuous data, they were tested for normality by the Kolmogorov-Smirnov test. Quantitative data were expressed as mean and standard deviation. ANOVA was used for comparing the three studied groups. On the other hand, the Kruskal Wallis test was used to compare different groups for not normally distributed quantitative variables. Spearman coefficient was used to correlate between not normally distributed quantitative variables. Univariate and multivariate binary logistic regression was assessed for the parameters affecting diabetic retinopathy. Receiver operating characteristic curve (ROC) was used to determine the diagnostic performance of the markers, an area of more than 50% gives an acceptable performance, and an area of about 100% is the best performance for the test. The significance of the obtained results was judged at the 5% level.

Sample size calculation
The sample size was estimated using PASS program version 20. The minimal hypothesized total sample is 60 adult patients attending the Diabetes outpatient clinic at Alexandria Main University Hospital divided into three groups (20 per group) is required to determine the strength of association between vWF plasma level and the probability of developing DR taking into consideration 95% level of confidence and 80% power using chi-square test [22].

Results
The present study included 60 patients with T2DM divided into three equal groups according to the presence and stage of retinopathy. Patients in group A had no retinopathy, group B had NPDR, and group C had PDR. A comparison between the three groups regarding different clinical and laboratory parameters is shown in Table 1. The mean plasma vWF level was significantly higher in patients with retinopathy compared to patients without retinopathy. However, this difference did not exist between retinopathy subgroups (p = 0.176).
Plasma vWF level was significantly positively correlated to diabetes duration, HbA1c, UACR, and BMI in all studied groups. Regarding age, plasma vWF level was positively correlated to age in retinopathy groups and total sample but not with group A (Table 2).
Univariate and multivariate regression analyses were done to detect factors affecting DR (Table 3). Univariate analysis showed that HbA1c, vWF, BMI, duration of DM, and UACR were independent risk factors for diabetic retinopathy, while none of them showed significance in multivariate analysis.
The present study assessed the validity of plasma vWF as a diagnostic biomarker of DR and set a cut-off point of plasma vWF level >180 IU/dl as a differentiating value between patients with and without DR in T2DM with an excellent sensitivity of 97.5%, specificity 75.0 % and AUC 0.937* ( Table 4, Fig. 1).
In the present study, in patients with T2DM, the mean plasma vWF was significantly higher in patients with NPDR, PDR, and both groups than in patients without retinopathy. In agreement with the results of the present study, El-Sersy et al. found significantly higher serum vWF levels in patients with T2DM than in healthy controls with the highest levels in those with PDR [15]. Oggianu et al. also found a On the other hand, Malecki et al. studied the relation between DR and vWF as a marker of endothelial dysfunction in patients with T2DM and found no significant difference in vWF level in patients with or without retinopathy [14]. Another study by Kadıköylü et al. studied serum vWF levels in diabetic complications and found no significant difference between patients with and without DR regarding vWF levels [25]. Siemianowicz et al. also found no significant relationship between markers of endothelial dysfunction (including vWF) and DR [26]. Moreover, Stehouwer et al. found no relation between plasma vWF level and DR in patients with type 1 DM and normal urinary albumin excretion [27]. Erem et al. also found no difference in vWF activity in patients with T2DM with and without vascular complications [28]. Laursen et al. also stated no relation between vWF and DR in type 1 DM [29]. This discrepancy may arise from using different cohorts, study designs, and ethnic groups.
In the present study, there was no significant difference in plasma vWF level between patients with NPDR and PDR, suggesting that although plasma vWF level is higher in patients with DR, it could not be a differentiating marker of DR severity. However, El-Sersy et al. found a strong direct positive correlation between vWF levels and different stages of DR with the highest level in patients with PDR [15].
In the current study, vWF was one of the independent risk factors for DR in univariate but not in multivariate regression analysis. Laursen et al., in concordance with the results of the present study, found that the multivariate model was not statistically significant for vWF and DR [29]. They concluded that vWF and DR are not associated with each other. Domingueti et al. also found that vWF was not an independent risk factor for DR in patients with T1DM in multivariate regression analysis [24].
The current study found a significant positive correlation between plasma vWF and HbA1c in all studied groups. Univariate regression analysis showed that HbA1c was one of the independent risk factors for DR but not in multivariate regression analysis.
In concordance with the results of the present study, Stehouwer et al. suggested a relation between poor glycemic control and endothelial dysfunction in patients with T1DM with and without DR [27]. On the other hand, the previously mentioned study by Laursen et al. found no significant correlation between serum vWF and HbA1C. They indicated the importance of other risk factors such as diabetes duration [29].
The present study demonstrated that there was a statistically significant positive correlation between duration of DM and plasma vWF level in DR. Univariate regression analysis showed that duration of DM was an independent risk factor for DR, whereas multivariate regression analysis showed that it was not. In agreement with the results of the present study, Laursen et al. found that there was a positive correlation between the duration of DM and serum vWF level [29].
In the present study, UACR was significantly higher in patients with than in patients without DR and also in patients with PDR than NPDR. Additionally, the mean plasma vWF level was positively correlated to UACR in all groups and the total sample. These results link the presence and severity of DR with DKD as both are microvascular complications sharing the same pathogenesis. Moreover, UACR was an independent risk factor for DR in univariate but not in multivariate regression analysis. These results are consistent with Jager et al. who investigated the prognostic implications of retinopathy and high plasma level of vWF in patients with T2DM and microalbuminuria [30]. They concluded that among patients with T2DM and microalbuminuria, the presence of retinopathy or a high plasma vWF level affects cardiovascular death risk. Their findings also support that the presence of generalized endothelial dysfunction is more common in the presence of high plasma vWF levels in diabetes. Similar results were observed by Stehouwer et al [27]. Another study by Ibrahim et al, who studied the relation between vWF, markers of oxidative stress, and microalbuminuria in patients with T2DM, found significantly higher serum vWF levels in patients with T2DM, particularly with microalbuminuria and history of CAD [31]. They also concluded that vWF levels were associated with markers of increased oxidative stress reflecting its contribution to diabetic vascular disease.
To the best of our knowledge, this is the first study identifying a cut-off value of plasma vWF level differentiating patients with DR from those without DR with excellent sensitivity and accepted specificity. This highlights the possible role of plasma vWF as a risk factor for DR. The limitations of this study include using a single center with the same cohort. In addition, the study design did not include follow-up for the study participants to detect the change in vWF level and its association with progression of DR. Further multicenter follow-up studies should be performed for this purpose.

Conclusions
DR is one of the most prevalent diabetic complications. Endothelial dysfunction is the core mechanism of microvascular complications. vWF is elevated in patients with T2DM and DR as a marker of endothelial dysfunctions but it is not the only factor affecting DR. Other factors may contribute to the development of DR such as diabetes duration, HbA1c, and UACR. This highlights the important role of diabetes control, duration, presence of coexisting complications, and endothelial dysfunction in the pathogenesis, diagnosis, and early detection of DR. The present study is the first study to set a cut-off value of >180 (IU /dl) for plasma vWF level that could differentiate between patients with and without DR in T2DM; therefore, plasma vWF level could be used as a diagnostic biomarker of DR.
Acknowledgment The authors would like to acknowledge the internal medicine department (Diabetes and Metabolism Unit) and ophthalmology department, Faculty of Medicine, Alexandria University where the study was done.
Funding Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB). This research did not receive any specific grant from funding agencies in the public, commercial, or notfor-profit sectors.
Data availability Data is available through the corresponding author upon justified request.

Declarations
Competing interests The authors have no relevant financial or nonfinancial interests to disclose.

Ethics approval
The study was conducted in concordance with the declaration of Helsinki and its revision (2013) and the approval of the ethical committee of the faculty was fulfilled.
Consent to participate Informed consent was obtained from all individual participants included in the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.