Arterial Composition and Cardiovascular Outcome in DIabeteS

PhD project description

The arterial wall in patients with diabetes: Do changes in arterial basement membrane proteins predict future arterial disease? Are remodeling processes altered?

Applicant: Anne-Sofie Faarvang Thorsen1, MD

Main supervisor: Professor Lars Melholt Rasmussen1

Co-supervisors: Assoc. Professor Michael Hecht Olsen2

Assoc. Professor Martin Overgaard1

Collaborators: Director Morten Karsdal, Nordic Bioscience3

Consultant Lars Peter Riber4

Professor Peter Nilsson5

1: Department of Clinical Biochemistry and Pharmacology, Odense University Hospital. 2: Department of Internal Medicine, Holbaek Hospital, University of Southern Denmark. 3: Nordic Biosciences 4: Department of Heart, Vascular and Thoracic Surgery, Odense University Hospital 5: Malmö Diet and Cancer Study, Malmö, Sweden

Project description Cardiovascular diseases (CVDs) occur frequently among individuals with type 2 diabetes mellitus (T2DM), and the mortality after cardiac interventions and cardiovascular events is significantly increased in T2DM. We have recently identified a series of distinct molecular alterations in non-atherosclerotic parts of the arterial tree in diabetes, which represent elements of the generalized diabetic vascular pathology. Whether these generalized diabetic alterations are related to the development of clinical cardiovascular diseases is not known.

Overall hypothesis Our overall hypothesis is that specific arterial protein alterations, particularly related to basement membrane molecules as seen in T2DM, are followed by dysfunctional vessel remodeling (inward instead of outward). Such changes may provide the background for lumen reduction in relation to atherosclerosis build-up in diabetes and explain the high incidence of CVDs among these patients.

Aims and study designs

In this ph.d.-project, we will test two aspects of this hypothesis with the following aims:

1. We will determine if the presence of a diabetic arterial protein pattern predicts future incidence of cardiovascular events and occurrence of lumen reduction and ultimately occlusion in the repair vessel (left internal mammary artery (LIMA)) after by-pass operations.

We will quantitate specific arterial basement membrane proteins by targeted proteome analysis in non-atherosclerotic human arterial tissue (the repair artery from coronary by-pass operations) and in plasma by immunoassays from a large group of coronary by-pass patients with or without T2DM. We will determine if the concentration of these components in a follow-up study predicts occlusion (clinical proxy for lumen reduction) of the repair vessel itself, the incidence of cardiovascular events and/or the development of diabetic small- and large vessel disease.

2. We will determine if dysfunctional arterial remodeling occurs in T2DM or subjects with impaired fasting glucose in relation to plaque build-up in the carotid artery We will therefore identify the influence of T2DM on the relationship between the amounts of atherosclerosis, remodeling (luminal size and wall thickness) and blood pressure in previously scanned individuals from the Malmö Diet and Cancer Study.

Perspective The results will provide important information about whether arterial matrix changes in diabetes are involved in the development of cardiovascular symptoms and, moreover, determine if dysfunctional remodeling occurs in diabetes.

Such pathophysiological insight is a prerequisite for:

- better biomarkers for diagnosing and stratifying T2DM patients with risk of arterial disease

- development of new targets for treatment of large-vessel complications in diabetes

Background T2DM is a severe illness that results in affection of various organ systems, including the cardiovascular system (1). Diabetes is a global threat that affects every part of the world. The number of people suffering from T2DM is increasing and is expected to rise to more than 350 million people worldwide in the year 2030 (2). Patients with T2DM have increased risk of morbidity and mortality of CVDs such as stroke and acute myocardial infarction (AMI) and have a shorter life expectancy (3, 4). Diabetes also increases the risk of microvascular complications such as diabetic retinopathy (5).

Arterial tissue is affected by diabetes. The arterial pathology includes changes in the extracellular matrix, increased arterial stiffness, dysfunctional remodeling (6) , endothelial dysfunction and accelerated atherosclerosis. It is hypothesized that arterial stiffness in diabetic patients with hypertension correlates to endothelial dysfunction (7). Patients with T2DM that undergo coronary artery bypass graft surgery (CABG) have an increased risk of death compared to non-diabetic patients, as shown in the Swedish SWEDEHEART study (8). Diabetes affects long-term survival after CABG, and if the patient has diabetic kidney disease or other diabetic vascular complications the difference is even more profound (9). The exact mechanisms that link T2DM to CVDs are not fully understood. During recent years, we have identified molecular changes in the arterial wall in humans with T2DM. Using microarray techniques for RNA-analysis, we found that specific molecular pathways were altered in arterial tissue in diabetes. These dysregulated pathways included insulin signaling and matrix biosynthesis (10). We also observed that a basement membrane and elastin-related molecule called fibulin-1 is present in increased amounts in diabetes (11-13). Afterwards, we have been able to show that increased amounts of fibulin-1 in plasma correlates to cardiovascular morbidities and predict mortality (11, 13, 14). Recently, we have used proteome analysis of human arteries to find arterial proteins with different concentrations in diabetic arteries (15).

We observed that most basement membrane proteins, and certain proteins related to smooth muscle phenotypic states, in combination with other peptides are present in altered concentrations in diabetes.

As previously mentioned, we only understand some of the processes that link diabetes to the development of CVDs. We have yet to determine whether the above mentioned findings are indeed involved in the development of clinical CVDs.

The overall hypothesis of this PhD-project is depicted in the above figure 1: Metabolic and hormonal factors lead to arterial basement membrane changes diffusely in the arterial tree. This leads to altered ability of the arterial wall to adapt by outward remodeling as the normal consequence to atherosclerosis. On the contrary, inward shrinkage, leading to lumen obstructions, occurs. Thus, generalized diabetic arterial matrix changes may ultimately lead to arterial lumen obstructions and clinical symptoms coupling molecular changes to development of CVDs.

Patients with T2DM have elevated serum levels of fibulin-1, elevated expression of elements from lamina basalis in the left internal mammary artery, and increased glycation of collagen in the arterial wall due to hyperglycemia (13). This may all contribute to the increased central arterial stiffness observed in T2DM, which can be assessed by carotid femoral pulse wave velocity (cfPWV). Normal aging as well as hypertension is, in the carotid arteries, associated with hypertrophic remodeling with increase of both lumen diameter and wall thickness, assessed by intima-media thickness (IMT). We have recently demonstrated that progression of IMT is associated with the sum of cardiovascular risk factors but with different contributions of the individual risk factors in the common carotid artery (CCA) and the bifurcation, respectively. However, IMT alone might not fully reflect the remodeling of the carotid artery, because lumen may change differently compared to IMT in response to certain cardiovascular risk factors. IMT progression is positively associated with most of the cardiovascular risk factors, whereas increased lumen is primarily associated with aging and hypertension (16). Furthermore, coronary atherosclerosis has been demonstrated to lead to inward instead of outward remodeling in T2DM, which might be secondary to the stiffer arteries and explain the reverse outcome after percutane coronary intervention (PCI) in T2DM.

The PhD-study will consist of two independent substudies, as described below.

Study 1

Hypothesis We hypothesize that the amount of basement membrane proteins in arterial tissue and plasma from patients with T2DM can predict long-term clinical outcome.

Patients and methods From most coronary by-pass operations done at OUH for the last 8 years, we have secured immediate handling and storing of the repair artery (internal mammary artery). The mid-portion of each artery is freed from surrounding tissue, cut, short-time formalin fixed and embedded in paraffin. The rest of the tissue is frozen. The tissue is stored in "Odense Artery Biobank" where tissue and blood samples from patients undergoing CABG have been collected continuously since 2008. We will include all patients with available tissue in our biobank, which will be approximately 180 patients with T2DM and 420 patients without diabetes.

Ethical considerations All participants gave written informed consent, and the study was approved by the local ethics committees (S-20100044). The clinical studies will be approved by the Regional Scientific Ethical Committee for Southern Denmark and all projects will be reported to the Danish Data Protection Agency.

Targeted proteome analysis on arterial tissue Quantitative targeted proteome analysis based on mass spectrometry will be performed on small amounts of arterial tissue using modifications of our recently developed protocols. Four sections of paraffin-embedded tissue from each patient will be used and targeted proteome analysis will be performed based on the methods based both on proteome analysis of the arterial wall (15, 17, 18) and multiplex, targeted LC-MS analysis of different proteins in other studies (19, 20).Our proteome analysis is based on material from slices of the same formalin-fixed paraffin-embedded (FFPE) tissue as used for histology. We will quantitate peptides derived from abundant proteins, where we previously found changes in diabetes, i.e. type IV collagen alpha 1 and alpha 2 chain, nidogen, biglycan, laminin, and desmin and will furthermore use the amount of alpha-smooth muscle actin and vimentin as controls, since no alterations were previously observed in patients with T2DM.

From these results we will calculate a composite "Basement membrane risk score". This score is based on that each basement membrane proteins couints for 0, if the concentration is below the median and 1 if the concentration is above. The combined basement membrane score is calculated as the sum of each of the five proteins, i.e. it can have values between 0 and 5.

Quantitative histomorphometric calculations We will quantitate the tissue volume of fibrosis, elastin content, smooth cell content and number as previously described Preil et al (15).

Plasma measurements of basement membrane proteins We plan to measure the amounts of circulating fragments of several of the same proteins as mentioned above, i.e. type IV collagen (C4M3a, C4M), biglycan (BGN) and laminin as well as fibulin-1. These assays have been developed by Nordic Biosciences. Pilot data on a small group of our combined diabetic and non-diabetic patients (n=36) show that the plasma level of these components correlate well to HbA1c (C4M3a: r=0.39, p<0.05, C4M2: r=0.36, p<0.05, BGN: r=0.45, p<0.01, Signe Holm Nielsen et al, Nordic Bioscience, unpublished).

Outcome measurements

We will investigate possible associations between plasma and tissue concentrations of the specific proteins and several clinical parameters as registered in a 2-8 year follow up period after the harvest of arterial tissue. Follow-up parameters will include the following three types:

1. Parameters related to the repair vessel, i.e. LIMA lumen size in the latest coronary arteriography performed after the operation compared to LIMA lumen size at the operation

2. Parameters related to the heart, i.e. MI, angina, re-CAD, heart failure

3. Parameters related to occurrence of generalized cardiovascular events and diabetic complications in the diabetic group.

We have available a subfraction of the operated patients with renewed coronary arteriography (approximately 35 %). We are aware that this subgroup is biased due to clinical indications for renewed arteriography giving competing reasons for a smaller LIMA lumen at the renewed arteriography. We will measure lumen size of this internal mammary artery from the coronary arteriography and compare it to the size of the vessel at the time of harvest at the operation, where it can be measured by histology. Although the two methods may not be equally calibrated, we assume that the bias is approximately the same in all patients. In this subgroup we will measure correlations between lumen diameter change (or occlusion) and plasma and arterial measures of the selected proteins.

The following clinical follow-up parameters are registered: all-cause mortality, non-fatal myocardial infarction, coronary revascularization, cardiac arrest with resuscitation, hospitalization for heart failure or unstable angina pectoris, non-fatal stroke, progression of nephropathy or retinopathy. Progression of nephropathy is defined as renal failure (defined by the need for chronic dialysis), development of macroalbuminuria or doubling of s-creatinine (only above 200uM). Development of retinopathy is defined as proliferative retinopathy or macular edema that requires laser therapy, vitrectomy or diabetes related blindness (Snellen visual acuity below 0.1). Death is obtained from the Civil Registration System. The individual diagnoses, operations and procedure codes are obtained from The Danish National Patient Register. P-creatinine and U-albumin-creatinine-ratio is obtained through laboratory result database. Our primary end point is defined as a composite of cardiac events (MI, coronary revascularization, sudden death, hospitalization for heart failure or angina pectors). Event rate are expected to be 2 % yearly (ref), i.e. approximately 60 events out of all our 600 patients after the average follow-up time of 5 years.

Moreover, we will investigate the associations between plasma and arterial tissue levels of the measured components. We will also judge if specific proteins are related to the prevalence and incidence of diabetic microvascular disease as defined above (i.e. albuminuria, renal insufficiency, retinopathy).

Clinical data concerning diabetes debut, medication, biochemical data from time of operation and co-morbidities is stored in our database at the time of the operation.

Plan for analyses

1. Test whether the "Basement Membrane Protein Risk Score" is negatively associated with

1. Lumen size from the coronary arteriography performed in relation to the operation

2. The ratio between lumen size from the renewed coronary arteriography and the one performed in relation to the operation using univariate and multivariate regression analyses

2. Test whether the "Basement Membrane Proteins Risk Score" predicts the composite cardiac endpoint (MI, coronary revascularization, sudden death, hospitalization for heart failure or angina pectors) using univariate and multivariate Cox-regression analyses (primary end point).

3. Test whether the "Basement Membrane Proteins Risk Score" predicts the broad composite endpoint of all-cause mortality, major cardiovascular events, progression of nephropathy or retinopathy using univariate and multivariate Cox-regression analyses

Statistical analysis, bioinformatics, estimation of power and sample size considerations Data will be analysed using both simple correlation analysis, but also proportional hazard models with adjustments for follow-up time and relevant clinical parameters. Fisher's Exact test and Chi-square tests will be used to evaluate event rates in groups of patients with high or low amounts of specific proteins. The Benjamini-Hochberg method will be used to correct for multiple testing, where suitable. Analyses will be done primarily in the total group of individuals (n=600) testing for interactions by presence of T2DM, but subgroup analysis in diabetic and non-diabetic patients will also be done.

Sample size calculation for the primary end point (composite of cardiac events) is calculated as follows: We assume that the number and proportion of events will be: BM-score 0: n=50, events: 2 %, BM-score 1: n=100, events: 6 %, BM-score 2: n=150, events: 8 %, BM-score 3: n=150, events: 10 %, BM-score : n=100, events: 15 %, BM-score 5: n=50, events: 22 %. Calculation of the necessary number of individuals needs to find a statistical difference between the score 0 group and the score 5 group is 50 in each group.

Feasibility and project organization The project is done at the Center for Individualized Medicine in Arterial Diseases (CIMA), Center for Clinical Proteomics, and the Department of Clinical Biochemistry and Pharmacology at Odense University Hospital. All instruments and expertise are available. Methods for proteome analysis of arterial tissue have been developed.

Anticipated impact of results and perspective Less than ten previous papers, worldwide, have reported on the use of proteome analysis on human arterial tissue, most of them with the use of material from less than 10 individuals. Previous studies have shown that histological signs of dysfunctional remodeling (hypertrophy) in small vessels predict cardiovascular events very well (21). However, nothing is known about structural and molecular aspects in larger arteries.

Results from the proposed study may point out whether the amounts of arterial proteins, known to be altered in T2DM, are associated with the development of cardiovascular events. Therefore, we will generate new knowledge concerning important molecular pathways, which may be causally involved in the development of arterial diseases. The identification of such proteins is important for finding new treatment modalities and biomarkers for the development of cardiovascular diseases.

We expect to be able to show that quantitative alterations of specific proteins and groups of proteins in the vasculature and in plasma predict the development of different CVD-outcomes. Such findings will strongly support the overall hypothesis that general arterial alterations in diabetes are important for the development of cardiovascular diseases.

Study 2

Hypothesis We hypothesize that patients with T2DM have dysfunctional carotid remodeling with more inward remodeling and decrease of lumen diameter. The overall aim is to investigate the relative importance of different cardiovascular risk factors for carotid artery remodeling assessed by carotid lumen, IMT and cross-sectional area (CA) in CCA and the bifurcation, respectively.

The specific hypotheses are:

1. The bifurcation vs. CCA will present more inward remodeling as it is more prone to atherosclerosis

2. Glucose intolerance and to a higher degree T2DM are associated with a higher degree of inward remodeling

3. This higher degree of inward remodeling is associated with increased arterial stiffness assessed by cfPWV and changes in lamina basalis assessed by selected markers

Patients and methods This study is a substudy to the large Malmö Diet and Cancer Study in which carotid artery ultrasound was performed in 3426 middle-aged Swedish men and women in 1991-94 (baseline) and again in 2007-12 (re-examination).

Statistical analysis, bioinformatics, estimation of power and sample size considerations In multiple regression analyses we will test the association between

1. CV risk factors and IMT, lumen, CA and distensibility in CCA and the bifurcation at baseline

2. CV risk factors and IMT, lumen, CA and distensibility in CCA and the bifurcation at re-examination

3. Changes in CV risk factors and changes in IMT, lumen, CA and distensibility in CCA and the bifurcation between baseline and re-examination

4. Serum levels of selected markers of changes in lamina basalis and CV risk factors, cfPWV as well as IMT, lumen, CA and distensibility in CCA and the bifurcation at re-examination

5. Test for gender, fasting glucose, T2DM, CCA distensibility, cfPWV and selected markers of changes in lamina basalis interaction on the above associations.

Feasibility and project organization This project will be performed at the Center for Individualized Medicine in Arterial Diseases (CIMA), Center for Clinical Proteomics, and the Department of Clinical Biochemistry and Pharmacology at Odense University Hospital, Odense, Denmark. All instruments are available and expertise is accessible.

Anticipated impact of results and perspective The results from study 2 may generate new knowledge and a deeper understanding of how T2DM affects the carotid artery, the degree of inward remodeling and lumen reduction. This could change the way we treat diabetic patients with carotid atherosclerosis, as these patients are at high risk of developing total stenosis of the carotid artery. Knowledge of the pathogenesis behind is the key to earlier intervention and reduction of risk for the patient.

Regarding both studies 1 and 2

Practical performance of study 1 and 2 The applicant will take part in proteome analysis and histomorphometric analysis of the tissue. All expertise and equipment are available at the Department of Clinical Biochemistry and Pharmacology, Odense University Hospital. Data analysis will be performed by the applicant and publications will primarily be written by the applicant as first author. The applicant will measure lumen of the common carotid arteries and of the carotid bulb using stored cine-loop recordings.

Expected publications in study 1 and 2

1. Changes in arterial and plasma matrix proteins in relation to the development of LIMA-diameters after CABG (Authors: Thorsen, AF, Overgaard, M, Nordic Bioscience, Olsen, MH, Rasmussen, LM)

2. Changes in arterial and plasma matrix proteins in relation to cardiovascular mortality, Proteome analysis in relation to outcome after CABG (Authors: Thorsen, AF, Nordic Bioscience, Olsen, MH, Overgaard, M, Rasmussen, LM)

3. The impact of glucose intolerance and overt diabetes on arterial stiffening, atherosclerotic plaques and carotid artery remodelling (Authors: Thorsen, AF, Malmø-researchers, Rasmussen, LM, Olsen, MH) These suggestions may be subject to change depending on the development of the investigations.

Time schedule

June 2018 Start proteome analysis in study 1 November 2018 Start data analysis in study 2 June 2019 Complete study 2 December 2019 Complete proteome analysis in study 1 December 2019 Begin data analysis in study 1 December 2020 Complete study 1 January-June 2021 Complete PhD-study

References

1. Gu K, Cowie CC, Harris MI. Mortality in adults with and without diabetes in a national cohort of the U.S. population, 1971-1993. Diabetes Care. 1998;21(7):1138-45.

2. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27(5):1047-53.

3. Ergul A, Kelly-Cobbs A, Abdalla M, Fagan SC. Cerebrovascular complications of diabetes: focus on stroke. Endocrine, metabolic & immune disorders drug targets. 2012;12(2):148-58.

4. Franco OH, Steyerberg EW, Hu FB, Mackenbach J, Nusselder W. Associations of diabetes mellitus with total life expectancy and life expectancy with and without cardiovascular disease. Arch Intern Med. 2007;167(11):1145-51.

5. Ellis TP, Choudhury RH, Kaul K, Chopra M, Kohner EM, Tarr JM, et al. Diabetic retinopathy and atherosclerosis: is there a link? Current diabetes reviews. 2013;9(2):146-60.

6. Jensen LO, Thayssen P, Mintz GS, Maeng M, Junker A, Galloe A, et al. Intravascular ultrasound assessment of remodelling and reference segment plaque burden in type-2 diabetic patients. European heart journal. 2007;28(14):1759-64.

7. Bruno RM, Penno G, Daniele G, Pucci L, Lucchesi D, Stea F, et al. Type 2 diabetes mellitus worsens arterial stiffness in hypertensive patients through endothelial dysfunction. Diabetologia. 2012;55(6):1847-55.

8. Holzmann MJ, Rathsman B, Eliasson B, Kuhl J, Svensson AM, Nystrom T, et al. Long-term prognosis in patients with type 1 and 2 diabetes mellitus after coronary artery bypass grafting. Journal of the American College of Cardiology. 2015;65(16):1644-52.

9. Leavitt BJ, Sheppard L, Maloney C, Clough RA, Braxton JH, Charlesworth DC, et al. Effect of diabetes and associated conditions on long-term survival after coronary artery bypass graft surgery. Circulation. 2004;110(11 Suppl 1):II41-4.

10. Skov V, Knudsen S, Olesen M, Hansen ML, Rasmussen LM. Global gene expression profiling displays a network of dysregulated genes in non-atherosclerotic arterial tissue from patients with type 2 diabetes. Cardiovascular diabetology. 2012;11:15.

11. Cangemi C, Skov V, Poulsen MK, Funder J, Twal WO, Gall MA, et al. Fibulin-1 is a marker for arterial extracellular matrix alterations in type 2 diabetes. Clinical chemistry. 2011;57(11):1556-65.

12. Cangemi C, Hansen ML, Argraves WS, Rasmussen LM. Fibulins and their role in cardiovascular biology and disease. Advances in clinical chemistry. 2014;67:245-65.

13. Hansen ML, Rasmussen LM. Associations between plasma fibulin-1, pulse wave velocity and diabetes in patients with coronary heart disease. Journal of diabetes and its complications. 2015;29(3):362-6.

14. Skov V, Cangemi C, Gram J, Christensen MM, Grodum E, Sorensen D, et al. Metformin, but not rosiglitazone, attenuates the increasing plasma levels of a new cardiovascular marker, fibulin-1, in patients with type 2 diabetes. Diabetes Care. 2014;37(3):760-6.

15. Preil SA, Kristensen LP, Beck HC, Jensen PS, Nielsen PS, Steiniche T, et al. Quantitative Proteome Analysis Reveals Increased Content of Basement Membrane Proteins in Arteries from Patients with Type 2 Diabetes and Lower Levels among Metformin Users. Circulation Cardiovascular genetics. 2015.

16. Rosvall M, Persson M, Ostling G, Nilsson PM, Melander O, Hedblad B, et al. Risk factors for the progression of carotid intima-media thickness over a 16-year follow-up period: the Malmo Diet and Cancer Study. Atherosclerosis. 2015;239(2):615-21.

17. Lyck Hansen M, Beck HC, Irmukhamedov A, Jensen PS, Olsen MH, Rasmussen LM. Proteome Analysis of Human Arterial Tissue Discloses Associations Between the Vascular Content of Small Leucine-Rich Repeat Proteoglycans and Pulse Wave Velocity. Arteriosclerosis, thrombosis, and vascular biology. 2015;35(8):1896-903.

18. Faarvang AS, Rordam Preil SA, Nielsen PS, Beck HC, Kristensen LP, Rasmussen LM. Smoking is associated with lower amounts of arterial type I collagen and decorin. Atherosclerosis. 2016;247:201-6.

19. Kristensen LP, Larsen MR, Mickley H, Saaby L, Diederichsen AC, Lambrechtsen J, et al. Plasma proteome profiling of atherosclerotic disease manifestations reveals elevated levels of the cytoskeletal protein vinculin. Journal of proteomics. 2014;101:141-53.

20. Ravnsborg T, Andersen LL, Trabjerg ND, Rasmussen LM, Jensen DM, Overgaard M. First-trimester multimarker prediction of gestational diabetes mellitus using targeted mass spectrometry. Diabetologia. 2016.

21. Mathiassen ON, Buus NH, Sihm I, Thybo NK, Morn B, Schroeder AP, et al. Small artery structure is an independent predictor of cardiovascular events in essential hypertension. Journal of hypertension. 2007;25(5):1021-6.