Proefschrift_Katrijn_v10.indd 172 23-10-12 20:32

Proefschrift_Katrijn_v10.indd 172 23-10-12 20:32

10 Chapter 10 Summary and future perspectives Proefschrift_Katrijn_v10.indd 173 23-10-12 20:32

In this thesis we evaluated the involvement of insulin and IGF-I in the development of cardiovascular disease. Chapter 1 provides a general introduction on insulin and IGF-I. Both these hormones originate from a common ancestral precursor, which is nowadays reflected by the fact that these hormones still show structural similarities, and that their receptors share the same signal transduction pathways. However, the functions of these hormones have diverged over time with insulin predominantly influencing metabolism, and IGF-I growth and development. Whether insulin and IGF-I play a role in the development of cardiovascular disease is largely unknown and this question is further explored in this thesis. The first part of this thesis focuses on the role of insulin in the development of cardiovascular disease. Chapter 2 reviews the literature on clinical trials, population based studies and experimental studies pointing towards the hypothesis that high doses of insulin increase cardiovascular risk in patients with type 2 diabetes. Large-scale intervention trials have failed to show beneficial effects of strict glycemic control (target HbA1c < 48mmol/mol) on cardiovascular outcomes in patients with type 2 diabetes. The ACCORD trial even reported increased (cardiovascular) mortality among intensively treated patients. While the reason for this increased mortality rate remains unclear, it can be speculated that high doses of insulin that were administered to achieve low HbA1c levels, could be responsible for this unexpected outcome. Interestingly, in population based studies, both endogenous and exogenous insulin have been positively associated with cardiovascular disease. Additionally, a wide array of in vitro experiments has demonstrated that insulin has pro-atherosclerotic properties, like increasing adhesion molecule expression on endothelial cells, facilitating transendothelial migration of monocytes, increasing smooth muscle cell proliferation and stimulating metalloproteinase expression by macrophages. Although direct evidence for a pro-atherosclerotic role of insulin in vivo is lacking, we concluded that insulin has the potential to activate atherosclerotic processes within the vessel wall, and that this needs further exploration. In chapter 3 we evaluated whether the degree of exogenous insulin exposure in patients with type 2 diabetes is associated with risk of cardiovascular events. To this end, we performed a nested case-control study among new users of oral glucose lowering drugs, using the PHARMO Institute for Drug Outcomes Research database. Mean daily insulin exposure was calculated in 2,338 cases with a cardiovascular event and in 4,407 controls, who remained free of a cardiovascular event. Interestingly, patients within the highest tertile of daily insulin exposure had a 44% higher cardiovascular risk, whereas patients within the lowest tertile of daily insulin exposure had a 52% lower cardiovascular risk, compared to patients not using insulin. These data imply that high doses of exogenous insulin in patients with type 2 diabetes 174 Proefschrift_Katrijn_v10.indd 174 23-10-12 20:32

Summary and future perspectives might have adverse effects on the development or progression of atherosclerosis. However, whether high insulin exposure has a causal role in increasing cardiovascular risk remains unclear from these data. Therefore, in the following chapters of part I we tried to elucidate through which mechanism insulin could increase cardiovascular risk, using in vitro, ex vivo and in vivo experimental designs. One of our hypotheses was that insulin might increase atherosclerotic plaque vulnerability by enhancing intra-plaque angiogenesis. In chapter 4 we provide data that support this hypothesis. Using an in vitro angiogenesis assay, we showed that insulin enhances capillary-like tube formation of human microvascular endothelial cells. Furthermore, we demonstrated that insulin receptors are highly expressed on endothelial cells of small nascent microvessels, but not on larger, more mature microvessels, within human atherosclerotic plaques. This specific distribution pattern for the insulin receptor, as measured by immunohistochemistry, suggests that insulin signalling might be involved in outgrowth of microvessels within human atherosclerotic plaque. When we subsequently analysed microvessel density in atherosclerotic plaques of 165 patients with type 2 diabetes, we found that microvessel density tended to be higher among those patients using insulin as compared to those not using insulin. Collectively, these results imply that insulin may promote angiogenesis within atherosclerotic plaques. In chapter 5 we demonstrate that insulin can also be expected to increase angiogenesis within adenocarcinomas. We found that the endothelium of microvessels in tumoral stroma of breast, colon, pancreas, lung and kidney carcinomas stained consistently and strongly positive for the insulin receptor, as measured by double immunohistochemistry techniques. The microvasculature of surrounding non-tumoral stroma and parenchymal tissue showed reduced or absent insulin receptor staining. Hence, comparable to atherosclerotic plaques, insulin receptors also showed a specific distribution pattern on endothelial cells within human adenocarcinomas. This implies that insulin might also be involved in stimulating tumor angiogenesis and subsequent tumor progression. Because especially the insulin analogue glargine has been associated with the progression of adenocarcinomas in the literature, we next examined the angiogenic potential of distinctive insulin analogues in vitro. All four tested insulin analogues increased capillary-like tube formation significantly, suggesting that insulin glargine is not an exclusive candidate for stimulating angiogenesis. Chapter 10 In chapter 6 we show that insulin can directly elicit features of atherosclerotic plaque instability in vivo, irrespective of systemic effects of insulin on metabolism. Although exogenous insulin use has been associated with increased cardiovascular risk in population studies, it remains unknown whether insulin has a causal role in stimulating the atherosclerotic process. One of the problems with observational studies is indication bias, i.e. patients for whom insulin is prescribed are generally the patients with more severe insulin resistance and beta-cell failure. This may lead to confounding by disease severity. A second hurdle in studying the 175 Proefschrift_Katrijn_v10.indd 175 23-10-12 20:32

pro-atherosclerotic properties of insulin in vivo, is that insulin has systemic effects on lipid metabolism and weight, which both can affect cardiovascular risk as well. In order to overcome these confounding factors we used an in vivo model in which the direct effects of insulin on the progression of atherosclerosis could be studied. Atherosclerosis was induced in LDLr KO mice by collar placement around the carotid arteries. Thus formed atherosclerotic lesions were focally treated with pluronic gel containing insulin or PBS. While insulin treatment of atherosclerotic plaques did not cause systemic effects, we observed a 33% reduction of intra-plaque collagen content and a 50% increase in activated perivascular mast cells in mice that were treated with insulin compared with control mice. In vitro experiments subsequently revealed that the reduced collagen content may be due to complex effects of insulin on collagen homeostasis in the plaque, possibly in part mediated by its effects on resident mast cells. Stimulation of the insulin receptor with insulin leads to activation of two downstream signalling pathways. One of these pathways, the PI3K pathway, is defective during insulin resistance. In order to study the role of insulin resistance in the development and progression of atherosclerosis in vivo, an animal model in which both atherosclerosis and defective PI3K signalling are present, could be of great value. In chapter 7 we therefore characterized the phenotype of LDLr KO mice in which Akt2 (a kinase of PI3K signalling) is lacking. We found that these so called Akt2/LDLr dko mice develop insulin resistance, as reflected by raised glucose and insulin levels, and impaired glucose tolerance testing. Atherosclerotic plaques of these mice were significantly smaller compared with LDLr KO controls. However, plaque composition of Akt2/LDLr dko mice was more complex, showing reduced collagen content, larger necrotic cores and increased apoptosis, while ASMA staining in the media layer was decreased. In vitro experiments additionally showed that vascular smooth muscle cells together with macrophages are likely to play a central role in disturbing collagen homeostasis within atherosclerotic plaques. Taken together, these data imply that Akt2/LDLr dko mice can serve as an attractive mouse model to study the effects of insulin resistance on the development and progression of atherosclerosis. The second part of this thesis focuses on the role of IGF-I in the development of cardiovascular disease. Low circulating IGF-I levels in adulthood have been associated with increased risk of cardiovascular disease. In this context, an extremely low IGF-I status from childhood onwards might be expected to contribute to premature atherosclerosis and impaired cardiac function in early adulthood. Against this background, we evaluated, in chapter 8, cardiovascular disease profile in a 27 year old patient with extreme low IGF-I levels due to a homozygous IGFALS gene mutation. Besides a mild insulin resistance and lipid disturbances, we did not find any manifestation of subclinical cardiovascular disease, as measured by 64-slice coronary CT scanning, three dimensional-echocardiography and tissue deformation imaging of the heart. 176 Proefschrift_Katrijn_v10.indd 176 23-10-12 20:32

Summary and future perspectives These results suggest that low circulating IGF-I status, due to an IGFALS mutation, is not necessarily accompanied by increased cardiovascular risk. To further substantiate the findings in this case report, chapter 9 shows the results of a nested case control-study within the EPIC-Norfolk cohort. Among 1,013 cases with coronary artery disease and 2,055 matched controls, we studied the association between circulating IGF-I and IGFBP-3 levels, genetic variation at the loci encoding these proteins, and risk of cardiovascular disease. After adjustment for cardiovascular risk factors, we found no association between circulating IGF-I and IGFBP-3 levels and risk of coronary artery disease. We identified three tagging single nucleotide polymorphysms (tsnps) that showed independent association with either circulating IGF-I or IGFBP-3 levels. Finally, in an assessment of 31 SNPs spanning the IGF1 or IGFBP3 loci, none were associated with coronary artery disease in a meta-analysis that included EPIC-Norfolk and eight additional studies comprising up to 9,319 cases and 19,964 controls. Together, these results indicate that IGF-I and IGFBP-3 are unlikely to be involved in the aetiology of coronary artery disease in humans. Future perspectives Insulin In chapter 3, we show that high daily insulin exposure in patients with type 2 diabetes is associated with increased cardiovascular risk. Given the limitations of observational studies, we cannot plainly assume that this association has a causal basis. Therefore, the hypothesis that insulin is pro-atherogenic should ideally be tested in the setting of a randomised clinical trial. The ORIGIN Trial investigators recently reported on such a trial, in which patients with insulin resistance or type 2 diabetes were randomised to insulin glargine or standard care. In this trial, early use of insulin did not affect cardiovascular outcomes as compared to standard care. 1 When interpreting the results of this trial, it should be taken into account that the studied population consisted of patients with mild diabetes or pre-diabetes, that insulin doses were rather low (28 units/day in a 70kg person), and that median HbA1c levels during the trial were equal to or below 48 mmol/mol (6.5%). Repeating this trial in another study population might lead to different results. Since it seems that strict glycemic control may especially lead to increased (cardiovascular) mortality in patients that are older, have longer duration of diabetes and have concomitant macrovascular disease 2-4, it would be interesting to assess whether avoidance of high insulin doses in patients with persisting high HbA1c levels and a history of macrovascular disease may reduce cardiovascular risk. Chapter 10 In chapters 4, 5, 6 and 7 we searched for mechanisms that could explain the relationship between insulin and progression of atherosclerosis. In figure 1 an overview of proposed mechanisms, based on in vitro, ex vivo, and in vivo experiments conducted in this thesis, 177 Proefschrift_Katrijn_v10.indd 177 23-10-12 20:32

Figure 1 is presented. We believe that insulin may increase atherosclerotic plaque vulnerability by increasing intra-plaque angiogenesis (chapter 4) and by disturbing collagen homeostasis (chapter 6 and 7). Vascular smooth muscle cells and macrophages probably contribute to the latter process by increasing metalloproteinase (MMP) expression, resulting in increased collagen breakdown (chapter 6 and 7). Additionally, activated mast cells may increase inflammation by secreting Il-6 and TNF-α. It should be emphasized that this proposed mechanism requires confirmation by further research. The concept that insulin might contribute to intra-plaque angiogenesis should be further explored in-vivo. Unfortunately, in most atherosclerotic animal models (including mice) intra-plaque microvessels are scarce. 5 Indeed, we did not observe microvessels in the atherosclerotic plaques of mice that were treated with insulin or PBS in chapter 6. Consequently, larger experimental models of atherosclerosis, like hypercholesteraemic pigs or balloon injured dogs, in which microvessels are more frequently encountered, are required to study the effects of insulin on intraplaque angiogenesis in vivo. Interestingly, because imaging modalities for real time imaging of human plaque angiogenesis are improving 6, it may become also feasible in the future to study the effects of insulin on intraplaque angiogenesis directly in humans by using a prospective study design. Such studies should not only focus on the 178 Proefschrift_Katrijn_v10.indd 178 23-10-12 20:32

Summary and future perspectives association between insulin and intra-plaque angiogenesis, but also on the association with cardiovascular outcome. Collagen is an important contributor to atherosclerotic plaque stability. Impaired collagen synthesis and accelerated collagen breakdown cause thinning of the fibrous cap, which makes the plaque more prone to rupture. Our finding that insulin decreases collagen content within atherosclerotic plaques of LDLr KO mice therefore supports the hypothesis that insulin increases cardiovascular risk, but also requires further exploration. Above all, the mechanism through which insulin decreases collagen content should be elucidated. We propose that insulin may contribute to collagen degradation by activating MMPs, since we found that insulin increases MMP mrna expression of vascular smooth muscle cells and macrophages in vitro. Future research should address whether MMP mrna and protein levels are also increased in insulin treated atherosclerotic plaques in vivo. With the advent of targeted imaging markers, future clinical studies might address the MMP-activity within advanced plaque in diabetic patients using PET-MMP tracers. Additionally, the role of mast cells in this process needs to be investigated, for instance by using mice that are treated with a mast cell stabilizer, like cromoglicid acid. As outlined in chapter 2, we hypothesized that insulin might exert its pro-atherosclerotic properties especially during conditions of insulin resistance, when the PI3K pathway is impaired. The Akt2/LDLr dko mouse model, as presented in chapter 7, might therefore provide an interesting model to study the direct effects of insulin on atherosclerotic plaque progression during insulin resistance in vivo. To this end, it would be interesting to focally treat atherosclerotic plaques of these mice with pluronic gel containing insulin or PBS and subsequently study features of plaque instability. Taken together, the studies presented in the first part of this thesis support the hypothesis that insulin has pro-atherosclerotic potential. This raises the question whether clinicians should be cautious when prescribing insulin to their patients with type 2 diabetes, particularly those with marked insulin resistance and overt atherosclerotic disease. It is beyond dispute that exogenous insulin has significantly improved life expectancy in patients with type 2 diabetes and therefore this drug is indispensible for many diabetic patients. However, based on our results, and the results of large clinical trials, including ACCORD 2 and UKPDS 3, we feel that insulin dosages should not be excessively increased at all costs in order to achieve HbA1c levels of below 53mmol/mol (7%). Especially patients with longer duration of diabetes and concomitant cardiovascular disease might benefit from lower insulin dosages with respect to cardiovascular outcome, still randomised clinical trials are awaited to prove this opinion. In this respect, it should also be investigated whether addition of insulin sensitizing/insulin sparing compounds, such as the recently introduced GLP-1 analogues and DPP-4 inhibitors, will decrease insulin need and improve cardiovascular outcome. Chapter 10 179 Proefschrift_Katrijn_v10.indd 179 23-10-12 20:32

IGF-I According to the results presented in chapter 8 and 9, circulating IGF-I and IGFBP-3 levels are not likely to be causally involved in ischemic heart disease. These findings are in line with other prospective studies that found no association between IGF-I, IGFBP-3 and ischemic heart disease 7-11, but in contrast with studies that did find a negative or positive association. 7,11-16 These divergent findings across studies may be due to heterogeneity in phenotype, age and sex of participants, heterogeneity among IGF-I and IGFBP-3 immuno-assays, and pleiotropic effects of these proteins that may result in different actions in different contexts. However, since we did not find an association between common SNPs at the IGF1 and IGFBP3 loci and cardiovascular risk, we are inclined to believe that IGF-I and IGFBP-3 are not causally involved in the development of cardiovascular disease. On the other hand, this does not exclude that locally secreted IGF-I can not be engaged in prognosis after cardiac ischemia. Paracrine/autocrine IGF-I in the heart may play an important role in this process, since IGF-I overexpression within hearts protects against apoptosis and ventricular dilatation after infarction, and local IGF-I therapy in the heart improves myocardial function after cardiac ischemia in animal models. 17-19 Therefore, we believe that future studies should focus on the contribution of paracrine/autocrine produced IGF-I in the heart on repair mechanisms after cardiac ischemia. In this context it might be interesting to measure IGF-I levels in blood samples taken from occluded arteries during acute coronary syndrome and correlate this to cardiac performance. Additionally, enrichment of the ischemic environment with target liposomes loaded with IGF-1/IGF BP-3 complex may lead to beneficial healing conditions of the heart. In conclusion, we believe that circulating IGF-I and IGFBP-3 levels are not causally involved in the development of cardiovascular disease. However, the role of paracrine/autocrine secreted IGF-I in cardiac repair mechanisms after myocardial infarction needs further exploration. 180 Proefschrift_Katrijn_v10.indd 180 23-10-12 20:32

Summary and future perspectives Reference List (1) The ORIGIN Trial Investigators. Basal Insulin and Cardiovascular and Other Outcomes in Dysglycemia. N Engl J Med 2012. (2) Gerstein HC, Miller ME, Byington RP et al. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008;358:2545-2559. (3) Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577-1589. (4) Ismail-Beigi F, Moghissi E, Tiktin M, Hirsch IB, Inzucchi SE, Genuth S. Individualizing glycemic targets in type 2 diabetes mellitus: implications of recent clinical trials. Ann Intern Med 2011;154:554-559. (5) Sluimer JC, Daemen MJ. Novel concepts in atherogenesis: angiogenesis and hypoxia in atherosclerosis. J Pathol 2009;218:7-29. (6) Zagorchev L, Mulligan-Kehoe MJ. Advances in imaging angiogenesis and inflammation in atherosclerosis. Thromb Haemost 2011;105:820-827. (7) Kaplan RC, McGinn AP, Pollak MN et al. Association of total insulin-like growth factor-i, insulin-like growth factor binding protein-1 (IGFBP-1), and IGFBP-3 levels with incident coronary events and ischemic stroke. J Clin Endocrinol Metab 2007;92:1319-1325. (8) Page JH, Ma J, Pollak M, Manson JE, Hankinson SE. Plasma Insulin-Like Growth Factor 1 and Binding-Protein 3 and Risk of Myocardial Infarction in Women: A Prospective Study. Clin Chem 2008;54:1682-1688. (9) Saydah S, Graubard B, Ballard-Barbash R, Berrigan D. Insulin-like growth factors and subsequent risk of mortality in the United States. Am J Epidemiol 2007;166:518-526. (10) Lawlor DA, Ebrahim S, Smith GD, Cherry L, Watt P, Sattar N. The association of insulin-like-growth factor 1 (IGF-1) with incident coronary heart disease in women: Findings from the prospective British Women s Heart and Health Study. Atherosclerosis 2008;201:198-204. (11) Yeap BB, Chubb SA, McCaul KA et al. Associations of IGF1 and IGFBPs 1 and 3 with all-cause and cardiovascular mortality in older men: the Health In Men Study. Eur J Endocrinol 2011;164:715-723. (12) Laughlin GA, Barrett-Connor E, Criqui MH, Kritz-Silverstein D. The prospective association of serum insulin-like growth factor I (IGF-I) and IGF-binding protein-1 levels with all cause and cardiovascular disease mortality in older adults: the Rancho Bernardo Study. J Clin Endocrinol Metab 2004;89:114-120. (13) Juul A, Scheike T, Davidsen M, Gyllenborg J, Jorgensen T. Low serum insulin-like growth factor I is associated with increased risk of ischemic heart disease: a population-based case-control study. Circulation 2002;106:939-944. (14) Friedrich N, Haring R, Nauck M et al. Mortality and Serum Insulin-Like Growth Factor I and Insulin-Like Growth Factor Binding Protein 3 Concentrations. J Clin Endocrinol Metab 2009. (15) Ruidavets JB, Luc G, Machez E et al. Effects of insulin-like growth factor 1 in preventing acute coronary syndromes: The PRIME study. Atherosclerosis 2011;218:464-469. (16) Chisalita SI, Dahlstrom U, Arnqvist HJ, Alehagen U. Increased IGF-1 levels in an elderly population in relation to heart failure and cardiovascular mortality. Impact of ACE-inhibitors. Eur J Endocrinol 2011. (17) Li Q, Li B, Wang X et al. Overexpression of insulin-like growth factor-1 in mice protects from myocyte death after infarction, attenuating ventricular dilation, wall stress, and cardiac hypertrophy. J Clin Invest 1997;100:1991-1999. (18) Davis ME, Hsieh PC, Takahashi T et al. Local myocardial insulin-like growth factor 1 (IGF-1) delivery with biotinylated peptide nanofibers improves cell therapy for myocardial infarction. Proc Natl Acad Sci U S A 2006;103:8155-8160. (19) Kotlyar AA, Vered Z, Goldberg I et al. Insulin-like growth factor I and II preserve myocardial structure in postinfarct swine. Heart 2001;86:693-700. Chapter 10 181 Proefschrift_Katrijn_v10.indd 181 23-10-12 20:32

Proefschrift_Katrijn_v10.indd 182 23-10-12 20:32

Appendix Proefschrift_Katrijn_v10.indd 183 23-10-12 20:32

Nederlandse samenvatting Meer dan 50% van de patiënten met type 2 diabetes mellitus (T2DM) overlijdt ten gevolge van hart- en vaatziekten. Om hart- en vaatziekten te voorkómen in deze patiënten is het van belang om te weten welke factoren het cardiovasculaire risico kunnen beïnvloeden. Omdat het momenteel ter discussie staat of patiënten met T2DM behandeld dienen te worden met hoge doses insuline heb ik onderzocht of, en hoe, hoge doses insuline het risico op harten vaatziekten kunnen verhogen. Tevens heb ik onderzocht of een hormoon dat erg lijkt op insuline, het insulin-like growth factor-i (IGF-I), geassocieerd is met een verhoogd risico op hart- en vaatziekten. In hoofdstuk 1 worden de overeenkomsten en verschillen tussen insuline en IGF-I beschreven. Deze hormonen hebben een gemeenschappelijke voorouder waardoor insuline en IGF-I tegenwoordig nog erg op elkaar lijken. Beide hormonen kunnen ook dezelfde effecten teweegbrengen. Echter, over de jaren heen heeft insuline zich ontwikkeld tot een hormoon met vooral metabole effecten, terwijl IGF-I met name betrokken is geraakt bij groei en ontwikkeling. Of beide hormonen betrokken zijn bij de ontwikkeling van hart- en vaatziekten is onderzocht in dit proefschrift. Deel 1. De rol van insuline bij het ontstaan van hart- en vaatziekten. In hoofdstuk 2 wordt een samenvatting van de literatuur gegeven waarbij studies besproken worden die erop wijzen dat insuline het risico op hart- en vaatziekten zou kunnen verhogen. In deze context worden ook studies besproken waaruit blijkt dat een hele intensieve behandeling van T2DM (waarbij HbA1c waarden van <48mmol/mol nagestreefd worden) niet leidt tot een verlaging van het risico op hart- en vaatziekten. Sterker nog, de ACCORD studie heeft laten zien dat intensieve behandeling het risico om te overlijden aan hart- en vaatziekten juist laat toenemen. Terwijl de redenen voor deze onverwachte verhoogde mortaliteit (nog) niet bekend zijn, is er gespeculeerd dat hoge insuline doseringen in de intensief behandelde patiënten, mogelijk bijgedragen kunnen hebben aan verhoging van het cardiovasculaire risico. Inderdaad blijkt er in de literatuur een associatie te bestaan tussen het gebruik van insuline en het risico op hart- en vaatziekten. Daarnaast hebben laboratorium experimenten aangetoond dat insuline eigenschappen bezit die atherosclerose ( aderverkalking ) kunnen stimuleren. Echter, direct bewijs dat insuline atherosclerose in mensen of dieren kan verergeren ontbreekt en is daarom de focus van het eerste deel van dit proefschrift. In hoofdstuk 3 hebben we onderzocht of de mate van insuline blootstelling in patiënten met T2DM geassocieerd is met hart- en vaatziekten. Hiertoe hebben we een studie gedaan onder mensen die voor het eerst orale antidiabetische middelen zijn gaan gebruiken. De gemiddelde dagelijkse insuline blootstelling werd berekend in 2338 patiënten die een cardiovasculair event hadden doorgemaakt en vergeleken met de insuline blootstelling in 4407 controles 184 Proefschrift_Katrijn_v10.indd 184 23-10-12 20:32

Nederlandse samenvatting die vrij bleven van hart- en vaatziekten. Na analyse bleek dat patiënten met de hoogste insuline blootstelling (hoogste tertiel) 44% meer kans hadden op hart- en vaatziekten, terwijl patiënten met een lage insuline blootstelling (laagste tertiel) 52% lager risico hadden op hart- en vaatziekten, vergeleken met patiënten die geen insuline gebruikten. Deze resultaten suggereren dat hoge insuline blootstelling mogelijk negatieve effecten heeft op het proces van aderverkalking. Echter, of insuline daadwerkelijk een causale rol heeft bij verergering van aderverkalking kan niet afgeleid worden uit deze data. Daarom hebben we in de volgende hoofdstukken onderzocht via welke mechanismen insuline het cardiovasculaire risico kan verhogen. Een van onze hypotheses was dat insuline de stabiliteit van atherosclerotische afwijkingen ( plaques ) zou kunnen verstoren door angiogenese (vaatnieuwvorming) in deze plaques te bevorderen. In hoofdstuk 4 presenteren we data die deze hypothese ondersteunen. Met behulp van een in vitro angiogenese proef hebben we laten zien dat insuline de vorming van vaatachtige structuren kan stimuleren in een laboratorium setting. Om erachter te komen of insuline ook vaatnieuwvorming in de plaques bij mensen kan stimuleren hebben we vervolgens eerst onderzocht of de insuline receptor aanwezig is in deze plaques. Het bleek dat de insuline receptor met name aanwezig was op kleine net gevormde vaatjes, en niet op grotere meer uitgerijpte vaatjes. Dit specifieke expressie patroon van insuline receptoren suggereert dat insuline mogelijk betrokken is bij uitgroei van nieuwe vaatjes in atherosclerotische plaques in mensen. Om dit verder te ondersteunen hebben we vervolgens aangetoond dat het aantal vaatjes in plaques hoger is in T2DM patiënten die insuline gebruiken dan in T2DM patiënten die geen insuline gebruiken. Al deze resultaten samen wijzen erop dat insuline mogelijk betrokken is bij vaatnieuwvorming in atherosclerotische plaques van mensen. In hoofdstuk 5 laten we zien dat insuline waarschijnlijk ook betrokken is bij vaatnieuwvorming in tumoren. We hebben namelijk gevonden dat er ook een specifiek expressie patroon voor insuline receptoren bestaat in borst-, darm-, alvleesklier-, long- en niertumoren. In deze tumoren zagen we dat insuline receptoren vooral aanwezig waren op vaatjes in het tumorweefsel, terwijl deze op de vaatjes in het omringende weefsel niet of nauwelijks aanwezig waren. Mogelijk leidt stimulering van deze insuline receptoren op de vaatjes met insuline tot een toename van vaatnieuwvorming, wat vervolgens tumorgroei kan bevorderen. Omdat in de literatuur vooral de insuline analoog glargine geassocieerd is met kanker hebben we vervolgens onderzocht of verschillende insuline analogen een verschillende potentie hebben om vaatnieuwvorming te stimuleren in vitro. Alle vier de insuline analogen die we getest hebben bleken in dezelfde mate de vorming van vaatachtige structuren te kunnen stimuleren, wat erop wijst dat insuline glargine geen exclusieve kandidaat is om vaatnieuwvorming te stimuleren. In hoofdstuk 6 laten we zien dat insuline betrokken is bij het instabieler worden van atherosclerotische plaques in vivo, en dat dit onafhankelijk is van systemische effecten van 185 Proefschrift_Katrijn_v10.indd 185 23-10-12 20:32

insuline op metabolisme. Ondanks dat observationele studies hebben laten zien dat het gebruik van insuline in patiënten met T2DM geassocieerd is met een verhoogd risico op harten vaatziekten, is het onduidelijk of insuline een causale rol heeft bij het stimuleren van aderverkalking. Een van de problemen van observationele studies is de zogenaamde indicatie bias. Dat wil zeggen dat de patiënten die insuline voorgeschreven krijgen vaak de patiënten zijn die ernstigere insuline resistentie en beta-cel falen hebben dan de patiënten die geen insuline voorgeschreven krijgen. Dit zou kunnen leiden tot verkeerde interpretaties ten gevolge van confounding door ernst van de diabetes. Een tweede probleem bij het bestuderen van de effecten van insuline op aderverkalking in vivo, is dat insuline ook systemische effecten heeft op lipiden metabolisme en gewicht, welke allebei ook het cardiovasculaire risico kunnen beïnvloeden. Om deze verstorende factoren te omzeilen hebben we een in vivo model gebruikt waarin de directe effecten van insuline op het verergeren van aderverkalking bestudeerd kunnen worden. In atherosclerose-gevoelige LDLr KO (low density lipoprotein receptor knockout) muizen werd aderverkalking geïnduceerd door een bandje te plaatsen rondom de halsslagaders. De gevormde atherosclerotische plaques werden vervolgens lokaal behandeld met insuline of met fysiologisch zout. Deze lokale toediening van insuline leidde niet tot systemische effecten, maar leverde wel plaques op die 33% minder collageen bevatten, een eiwit waarvan wordt verondersteld dat het de stabiliteit van plaques bevordert. Bovendien werden in het weefsel rondom deze lesies 50% meer geactiveerde mest cellen waargenomen. Uit in vitro experimenten bleek dat de verlaging van het collageen in de lesies het gevolg zou kunnen zijn van complexe effecten van insuline op collageen homeostase (balans van aanmaak en afbraak), deels via effecten van insuline op mest cellen in het omringende weefsel. Stimulatie van de insuline receptor met insuline leidt tot activatie van twee signaleringspaden in de cel. Een van deze paden, het PI3K-pad, werkt niet goed tijdens insuline resistentie. Om in een in vivo model te onderzoeken wat de rol van insuline resistentie is bij het ontstaan en verergeren van atherosclerotische plaques, zou een muismodel waarin aderverkalking optreedt en tegelijkertijd het PI3K-pad defect is, van grote waarde kunnen zijn. In hoofdstuk 7 karakteriseren we daarom het fenotype van LDLr KO muizen die geen Akt2 hebben (een eiwit betrokken in het PI3K-pad). Deze zogenaamde Akt2/LDLr dko muizen hadden hogere glucose en insuline waarden dan LDLr KO muizen, en bleken meer insuline resistent te zijn. Opvallend was dat de plaques significant kleiner waren en tegelijkertijd meer complex bleken doordat ze minder collageen bevatten, grotere gebieden van necrose en meer apoptose hadden. In vitro experimenten lieten vervolgens zien dat gladde spiercellen en macrofagen waarschijnlijk een belangrijke rol spelen bij het verstoren van de collageen homeostase in de lesies van de Akt2/ LDLr dko muizen. Tezamen suggereren deze resultaten dat Akt2/LDLr dko muizen kunnen dienen als een interessant model waarin de effecten van insuline resistentie op het ontstaan en verergeren van aderverkalking onderzocht kunnen worden. 186 Proefschrift_Katrijn_v10.indd 186 23-10-12 20:32

Nederlandse samenvatting Deel 2. De rol van IGF-I bij het ontstaan van hart- en vaatziekten. Verschillende studies bij volwassenen hebben laten zien dat lage IGF-I spiegels in het bloed geassocieerd zijn met een verhoogd risico op hart- en vaatziekten. In deze context zou verwacht kunnen worden dat patiënten die van jongs af aan lage IGF-I spiegels hebben gehad, op jong volwassen leeftijd al tekenen van aderverkalking en/of hartfalen kunnen vertonen. Daarom hebben we, in hoofdstuk 8, het cardiovasculaire risico profiel bepaald bij een 27 jarige man met extreem lage IGF-I spiegels ten gevolge van een homozygote IGFALS gen mutatie. In het bloed van de patiënt werden extreem lage IGF-I spiegels en milde insuline resistentie en dyslipidemie gemeten. 64-slice CT scan en drie-dimensionale echografie van het hart lieten geen enkele tekenen van subklinisch hart- en vaatlijden zien. Deze resultaten suggereren dat lage IGF-I spiegels, veroorzaakt door een IFGALS mutatie, niet per definitie gepaard gaan met een verhoogd cardiovasculair risico. Om de bevindingen in de bovengenoemde casus verder te onderbouwen laten we, in hoofdstuk 9, de resultaten zien van een studie binnen het EPIC-Norfolk cohort. In 1013 cases met coronair lijden en 2055 controles hebben we de associatie bestudeerd tussen IGF-I en IGFBP-3 spiegels in het bloed, genetische variatie in de chromosomale gebieden (loci) die coderen voor deze eiwitten, en het risico op hart- en vaatziekten. Na correctie voor de aanwezigheid van cardiovasculaire risicofactoren bleek er geen associatie te zijn tussen IGF-I en IGFBP-3 spiegels en het risico op coronair lijden. Vervolgens hebben we drie tagging SNPs (single nucleotide polymorphysms) geïdentificeerd die geassocieerd waren met IGF-I of IGFBP-3 spiegels. Tot slot bleek, in een meta-analyse van de EPIC-Norfolk en 8 andere studies, dat geen van de 31 onderzochte tagging SNPs die de gehele IGF1 en IGFBP3 loci omvatten, geassocieerd was met coronair lijden. Samen wijzen deze resultaten erop dat IGF-I en IGFBP-3 waarschijnlijk niet betrokken zijn bij het ontstaan van hart- en vaatziekten in de mens. 187 Proefschrift_Katrijn_v10.indd 187 23-10-12 20:32

188 Proefschrift_Katrijn_v10.indd 188 23-10-12 20:32

List of publications List of publications Endothelial insulin receptor expression in human atherosclerotic plaques: linking micro- and macrovascular disease in diabetes? Rensing KL, von der Thüsen JH, Weijers EM, Houttuijn Bloemendaal FM, van Lammeren GW, Vink A, van der Wal AC, van Hinsbergh VWM, van der Loos CM, Stroes ES, Koolwijk P, Twickler ThB. Atherosclerosis 2012; 222:208 215 Prospective study of insulin-like growth factor-i, insulin-like growth factor-binding protein 3, genetic variants in the IGF1 and IGFBP3 genes and risk of coronary artery disease Ricketts SL, Rensing KL, Holly JM, Chen L, Young EH, Luben R, Ashford S, Song K, Yuan X, Dehghan A, Wright BJ, Waterworth DM, Mooser V, GEMS Investigators, Waeber G, Vollenweider P, Epstein SE, Burnett MS, Devaney JM, Hakonarson HH, Rader DJ, Reilly MP, Danesh J, Thompson SG, Dunning AM, van Duijn CM, Samani NJ, McPherson R, Wareham NJ, Khaw KT, Boekholdt SM, Sandhu MS. Int J Mol Epidemiol Genet. 2011;2:261-85 Reducing cardiovascular disease risk in diabetic patients with established macrovascular disease: can insulin be too much of a good thing? Rensing KL, Reuwer AQ, Arsenault BJ, von der Thüsen JH, Kastelein JJP, Twickler ThB. Diabetes Obes Metab. 2011;13:1073-87 Case-report: Low circulating IGF-I levels due to Acid-Labile Subunit deficiency in adulthood is not associated with early development of atherosclerosis and impaired heart function. Rensing KL, van Duyvenvoorde HA, Cramer MJM, Teske AJ, Prokop M, Stroes ES, Wit JM, Hermus ARMM, Twickler ThB. Growth Horm IGF Res. 2011;21:233-7 Could recombinant insulin compounds contribute to adenocarcinoma progression by stimulating local angiogenesis? Rensing KL, Houttuijn Bloemendaal FM, Weijers EM, Richel DJ, Büller HR, Koolwijk P, van der Loos CM, Twickler TB, von der Thüsen JH. Diabetologia. 2010; 53:966-70. Is insulin the preferred compound in lowering glucose levels in patients after a myocardial infarction? Rensing KL, Kastelein JJ, Twickler M. Arch Intern Med. 2009; 169:1636-7 189 Proefschrift_Katrijn_v10.indd 189 23-10-12 20:32

Abdominal aorta aneurysm: an exceptional expression of atherolscerotic disease in type II diabetes. Rensing KL, Twickler ThB. European Heart Journal, 2008; 29:581-582 Lessons from France. The Cardiovascular Prevention Clinic in the La Pitié- Salpêtrière Hospital in Paris. Rensing KL, Timmerman MK, Giral P, Cramer MJM, Bruckert E, Twickler ThB. Netherlands Heart journal. 2007;15:22-26 High daily insulin exposure in patients with type 2 diabetes is associated with increased risk of cardiovascular disease. Rensing KL, Nieuwdorp M, Twickler ThB, Holleman F, von der Thüsen JH, Zwinderman AH, Stroes ES, Currie CJ, Hutten BA. Submitted Insulin elicits features of atherosclerotic lesion instability in LDLr KO mice. Rensing KL, Bot I, de Jager SCA, Twickler ThB, Stroes ES, Kuiper J, von der Thüsen JH. Submitted Akt2/LDLr double knockout mice display impaired glucose tolerance and develop more complex atherosclerotic plaques than LDLr knockout mice due to altered collagen homeostasis. Rensing KL, de Jager SCA, Stroes ES, Vos M, Twickler ThB, Dallinga-Thie GM, de Vries CJM, Kuiper J, Bot I, von der Thüsen JH. Submitted 190 Proefschrift_Katrijn_v10.indd 190 23-10-12 20:32

Authors and affiliations Authors and affiliations Arsenault, BJ Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands Ashford, S Department of Public Health and Primary Care, Strangeways Research Laboratory, University of Cambridge, Cambridge, UK Boekholdt, SM Department of Cardiology, Academic Medical Center, Amsterdam, The Netherlands Bot, I Division of Biopharmaceutics, LACDR, Leiden University, Leiden, The Netherlands Büller, HR Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands Burnett, MS Cardiovascular Research Institute, MedStar Research Institute, Washington Hospital Center, Washington, DC, USA Chen, L Division of Cardiology, University of Ottawa Heart Institute, Ottawa, ON, Canada Cramer, MJ Department of Cardiology, University Medical Center Utrecht, Utrecht, The Netherlands Currie, CJ Department of Primary Care and Public Health, School of Medicine, Cardiff University, Cardiff, United Kingdom Dallinga-Thie, GM Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands Danesh, J Department of Public Health and Primary Care, Strangeways Research Laboratory, University of Cambridge, Cambridge, UK Dehghan, A Department of Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands 191 Proefschrift_Katrijn_v10.indd 191 23-10-12 20:32