Asymmetric dimethylarginine (ADMA): A novel risk marker in cardiovascular medicine and beyond

Abstract

There is abundant evidence that the endothelium plays a crucial role in the maintenance of vascular tone and structure. One of the major endothelium‐derived vasoactive mediators is nitric oxide (NO), an endogenous messenger molecule formed in healthy vascular endothelium from the amino acid precursor L‐arginine. Endothelial dysfunction is caused by various cardiovascular risk factors, metabolic diseases, and systemic or local inflammation. One mechanism that explains the occurrence of endothelial dysfunction is the presence of elevated blood levels of asymmetric dimethylarginine (ADMA) – an L‐arginine analogue that inhibits NO formation and thereby can impair vascular function. Supplementation with L‐arginine has been shown to restore vascular function and to improve the clinical symptoms of various diseases associated with vascular dysfunction.

• Atherosclerosis
• endothelium
• nitric oxide
• risk factors

Abbreviations
ACS	=	acute coronary syndrome
ADMA	=	asymmetric dimethylarginine
DDAH	=	dimethylarginine dimethylaminohydrolase
ICU	=	intensive care unit
L‐NMMA	=	NG‐monomethyl‐L‐arginine
NO	=	nitric oxide
PRMT	=	protein‐arginine methyltransferase
SDMA	=	symmetric dimethylarginine

Introduction

Methylation of the amino acid L‐arginine was discovered decades ago as a mechanism of post‐translational modification of proteins. It has been shown that within proteins, only L‐arginine residues but not free amino acid can be methylated, and that a family of enzymes named protein‐arginine‐methyltransferases catalyze this reaction 1,2. Methylation has been shown to affect diverse proteins such as histone proteins, myelin proteins, and many others 3, resulting in modified physicochemical properties and consequently, in modified functions of these proteins 3. Methylation of L‐arginine was shown to result in three different derivatives of this amino acid: the asymmetric isomers monomethyl‐L‐arginine (L‐NMMA) and asymmetric dimethylarginine (ADMA), as well as the symmetric dimethylarginine (SDMA). All of these L‐arginine analogues are endogenously present in the human body; however, due to its very low concentration, L‐NMMA probably plays no role in regulating NO production in vivo. Enzymes producing asymmetrically methylated arginines are called type 1 protein‐arginine methyltransferases (PRMTs), whereas enzymes that produce symmetrically methylated arginine are called type 2 PRMTs (for an overview of methyltransferases see 4,5). Figure 1 gives a schematic overview of ADMA biosynthesis and metabolism.

Figure 1. Schematic overview of biochemical pathways related to ADMA. Methylation of arginine residues within proteins or polypeptides occurs through N‐methyltransferases, which utilize S‐adenosylmethionine as a methyl group donor. After proteolytic breakdown of proteins, free ADMA is present in cytoplasm. It can also be detected circulating in human blood plasma. ADMA acts as an inhibitor of NO synthase by competing with the substrate of this enzyme, L‐arginine, and causes endothelial dysfunction and – subsequently – atherosclerosis. ADMA is eliminated from the body via urinary excretion and, alternatively, via metabolism by the enzyme dimethylarginine dimethylaminohydrolase (DDAH) to citrulline and dimethylamine. Reproduced from 17.

At the time when these biochemical pathways were discovered, it was not yet known that L‐arginine is the substrate of another important enzymatic pathway, namely the NO synthase pathway. L‐arginine acts as the natural precursor for the formation of nitric oxide (NO) in the endothelium and in other tissues 6. This reaction can be competitively inhibited by synthetic L‐NMMA, as first shown by Rees and co‐workers in 1989 7.

Only in 1992 did it become clear, by the groundbreaking discovery by Patrick Vallance and colleagues 8, that endogenous asymmetric dimethylarginine acts as a competitive inhibitor of NO synthase. These investigators detected ADMA (and, in much lower concentrations, L‐NMMA) in human plasma and urine, and they were able to show that it inhibits endothelium‐dependent vasodilation in vitro. Moreover, they also showed that intra‐arterial infusion of ADMA results in an increased arterial tone, supposedly by blocking vasodilation mediated by endogenous NO. Their most intriguing finding, however, was that ADMA accumulates in patients with renal failure by several‐fold, suggesting that ADMA may be responsible in part for the high prevalence of hypertension and atherosclerosis associated with chronic renal failure. We and others later confirmed this finding in larger clinical studies of patients with different types of renal function impairment (reviewed in 9).

Since 1992, much progress has been made by an increasing number of researchers in the understanding of the functional roles of ADMA in the cardiovascular system and its role as a prognostic marker of morbidity and mortality. This review will give an updated overview of the current data on ADMA with a major focus on clinical data from prospective and cross‐sectional clinical studies.

Key messages

• ADMA is asymmetric dimethylarginine, an endogenous inhibitor of nitric oxide synthase.

• By reversibly blocking NO production from L‐arginine, ADMA induces endothelial dysfunction, which may result in the initiation and progression of atherosclerotic vascular disease.

• Prospective clinical studies in various patient populations have suggested a role for ADMA as a novel cardiovascular risk factor. In these studies, elevated ADMA was associated with a mean 3‐ to 5‐fold increase in risk of experiencing a serious cardiovascular event.

ADMA is a regulator of endothelial nitric oxide (NO) formation

NO is involved in a wide variety of regulatory mechanisms of the cardiovascular system, including vascular tone. In fact, it is regarded by many as the major mediator of endothelium‐dependent vasodilation. It also regulates vascular structure (inhibition of smooth muscle cell proliferation) and cell‐cell interactions in blood vessels (inhibition of platelet adhesion and aggregation; inhibition of monocyte adhesion). Thereby, NO plays a crucial role in the endothelium‐mediated regulation of vascular homeostasis 10. Dysfunction of the endothelial L‐arginine/nitric oxide pathway is a common mechanism by which several cardiovascular risk factors mediate certain deleterious effects on the vascular wall. Among these are hypercholesterolemia, hypertension, smoking, diabetes mellitus, homocysteine, and vascular inflammation 11–16.

Currently available experimental and clinical evidence suggests that even small modifications of circulating ADMA levels significantly change vascular NO production, vascular tone, and systemic vascular resistance (for review, see 17). This has recently been evidenced in a mouse model in which the gene for dimethylarginine dimethylaminohydrolase (DDAH‐1), the enzyme that inactivates ADMA 18, was genetically overexpressed 19. DDAH‐1 transgenic mice displayed decreased ADMA levels, enhanced vascular NO production and lowered arterial blood pressure and systemic vascular resistance. The opposite animal model, i.e. mice in which one of the DDAH genes is lacking, is currently being generated.

Interestingly, there are a few small experimental studies in humans in which ADMA was administered in vivo in order to investigate the effects on haemodynamic parameters. Even in their first report on ADMA as an endogenous inhibitor of NO synthesis, Vallance and co‐workers reported that intra‐arterial infusion of ADMA into the brachial artery led to significantly reduced forearm blood flow 8. This group of investigators later confirmed their findings in a more detailed study with a larger group of subjects 20. However, concentrations of ADMA reached in the local circulation after intra‐arterial infusion outreach by far those that can be measured in patients. For that reason, two studies in which ADMA was administered intravenously added data to the dose‐response relationship of ADMA in humans. In one study, intravenous infusion led to circulating plasma ADMA concentrations of about 2.6 µmol/L 21. At this concentration level, ADMA caused a significant elevation of arterial blood pressure and systemic vascular resistance and reduced cardiac output, heart rate and the vascular response to exercise. In another study by Kielstein and co‐workers 22, ADMA was administered intravenously to healthy human subjects at different concentrations. These investigators found that ADMA decreased plasma cyclic GMP concentrations at pathophysiologically relevant concentrations, whereas renal plasma flow and glomerular filtration rate were only reduced at very high concentrations.

In yet another study, Chan and co‐workers studied the adhesion of monocytes isolated from peripheral blood to cultured human endothelial cells. They found that the adhesiveness index was related to the circulating ADMA level of the patient from whom monocytes were isolated. Moreover, when they influenced the ratio of L‐arginine over ADMA by administering L‐arginine, they found that monocyte adhesion normalized during 12 weeks of treatment 23.

ADMA is a novel cardiovascular risk factor

After the initial characterization of ADMA as endogenous inhibitor of NO synthase by Vallance et al. 8, we have been among the first to demonstrate that circulating ADMA levels are elevated in patients with systemic atherosclerosis 24, in clinically healthy human subjects with isolated hypercholesterolaemia 15, in patients with essential hypertension 25, and in patients with renal failure 26. The list of clinical conditions in which elevated ADMA levels were found has kept growing longer (Table I). Most recent data suggest a pathophysiological role for ADMA in limiting NO availability in sickle cell disease 27, and in pulmonary hypertension. Gorenflo et al. 28 were the first who reported elevated ADMA levels in patients with congenital heart disease and pulmonary hypertension. These findings were extended by showing that DDAH activity is dysregulated in chronic hypoxia‐induced pulmonary hypertension 29 and increased levels and reduced catabolism of ADMA in patients with idiopathic pulmonary hypertension. Kielstein et al. 30 demonstrated that elevated ADMA levels in idiopathic pulmonary hypertension might be of prognostic relevance.

Table I. Pathophysiological conditions in which ADMA was reported to be elevated.

Condition	x‐fold increase	Year of first report
Chronic renal failure	2 to 7‐fold	1992
Childhood hypertension	2.3‐fold	1997
Pregnancy‐related hypertension	2‐fold	1997
PAOD	2 to 3‐fold	1997
Hypercholesterolemia	2‐fold	1998
Chronic heart failure	2 to 3‐fold	1998
Preeclampsia	2‐fold	1998
Hypertension	2‐fold	1999
Hyperhomocysteinemia	2‐fold	2000
Pulmonary hypertension	2 to 3‐fold	2000
Stable CAD	2‐fold	2001
Stroke	2‐fold	2001
Diabetes mellitus type II	2 to 3‐fold	2001
Hyperthyreodism	2‐fold	2002
Insulin resistance	2‐fold	2002
Liver failure	2‐fold	2003
Erectile dysfunction	2‐fold	2004
Unstable coronary syndrome	2 to 3‐fold	2005

Data represent the x‐fold increase in patients with the clinical condition indicated as compared to healthy control subjects. Only conditions have been included for which data from clinical studies have been reported. CAD = coronary artery disease; PAOD = peripheral arterial occlusive disease.

In addition, several prospective studies in which surrogate endpoints for cardiovascular disease were used have supplied evidence of a pathophysiological role of ADMA in the pathogenesis of vascular dysfunction and cardiovascular disease in humans. High ADMA levels were found to be associated with carotid artery intima‐media thickness in a study with 116 clinically healthy human subjects 31. Taking this observation further in a prospective design, another study performed in haemodialysis patients reported that ADMA predicts the progression of intimal thickening during one year of follow‐up 32. In a nested, case‐control study involving 150 middle‐aged, non‐smoking men, high ADMA levels were associated with a 3.9‐fold elevated risk for acute coronary events 33.

In 2001, we published the first prospective clinical trial on ADMA as a predictor of cardiovascular outcome and total mortality in 225 haemodialysis patients during three years of follow‐up 34. In this study, patients whose ADMA levels were within the highest quartile at the beginning of the study had a three‐fold higher risk of death from any cause than patients with ADMA levels below the median.

Another study investigated factors related to outcome of patients undergoing intensive care unit treatment for multiple causes. Patients whose ADMA levels were in the highest quartile had a 17‐fold increase in mortality compared to patients with ADMA levels in the lowest quartile 35. In a third prospective study, the outcome of patients with stable angina pectoris after coronary balloon angioplasty was addressed 36. One hundred and fifty three patients with stable angina pectoris who underwent elective coronary angioplasty were included in the study. Patients were stratified into tertiles of pre‐procedural ADMA and followed for a median of 16 months. A total of 51 major cardiovascular events occurred during follow‐up. Interestingly, in a multifactorial Cox`s regression analysis (including age, smoking, hypercholesterolemia, use of stent and others), the risk of a cardiovascular event increased with ADMA levels and was independent of other risk factors.

We have recently added to available data by showing that ADMA is higher in patients with acute coronary syndrome than in patients with stable angina pectoris. In a prospective follow‐up, patients with unstable coronary syndrome whose ADMA declined within six weeks after the acute episode to levels similar to those of patients with stable coronary artery disease experienced fewer major cardiovascular events during one year of follow‐up than those whose ADMA had remained elevated 37.

In an analysis of 1,874 patients from the AtheroGene study population elevated ADMA concentration was highly significantly and independently associated with increased risk of major cardiovascular events during a mean 2.6 years of follow‐up 38.

Figure 2 summarizes available data on ADMA as a cardiovascular risk factor from prospective clinical studies. In each of these studies, other cardiovascular risk factors and confounding variables were included in the analyses, and ADMA was found to be an independent predictor of cardiovascular events. Thus, it has recently been concluded that ADMA can be considered to be a novel cardiovascular risk factor 17,39,40.

Figure 2. Results from prospective clinical studies that have assessed ADMA as a risk marker. Data represent hazard ratios(boxes) and 95% confidence intervals (horizontal lines) for the comparison of the highest ADMA stratum with the reference group.

Mechanisms that link ADMA with cardiovascular disease

The main mechanism of action by which ADMA exerts its effects is thought to be competitive inhibition of NO synthases. Thereby, ADMA concentration‐dependently reduces NO formation and impairs the biological effects of NO in the vasculature, all of which are thought to be protective under healthy conditions. Besides direct interference with eNOŚs catalytic mechanism, experimental evidence suggests that in the presence of high ADMA levels, eNOS may be uncoupled. This means that the usual flow of electrons among the two domains of eNOS is disturbed, and eNOS becomes a generator of superoxide radicals instead of NO 41. A similar phenomenon has been described for eNOS under insufficient concentrations of its essential co‐factor, tetrahydrobiopterin, insufficient availability of its substrate, L‐arginine, or in the presence of the NOS inhibitor, L‐NMMA 42. This mechanism suggests elevated levels of oxidative stress in the presence of high ADMA concentrations, and may also explain ADMÀs detrimental effects on endothelial function. L‐arginine, the natural substrate of NO synthase, is taken up into cells by the cationic amino acid transporter system y⁺43. This amino acid transporter has been shown to be inhibited by high concentrations of ADMA 44. Although the concentrations that were found to be effective in vitro are extremely high in comparison to ADMA levels circulating in vivo, the in situ concentrations in the endothelial lining are not precisely known, and levels within the endothelial cells have been estimated to be 8‐ to 10‐fold higher than circulating levels 45,46. Thus, it may well be that cellular L‐arginine uptake is impaired in patients with certain cardiovascular diseases in which ADMA is elevated. In support of this hypothesis, cellular L‐arginine uptake has been shown to be reduced in hypertensive patients 47, who are characterized by high ADMA levels 25,48.

Inhibition of NO production or action, whichever mediates ADMA's effects, results in a shift of the homeostatic vascular balance between vasodilation and vasoconstriction, in activation of platelets and depression of cell‐cell interactions, in pro‐ and antiproliferative impacts, in pro‐ and antioxidant stimuli, and may contribute to a pro‐inflammatory state of the vascular wall, thereby promoting atherosclerosis.

Recently ADMA has been found to mediate the anti‐angiogenic effects of hypercholesterolaemia in a mouse model 49. In mice overexpressing human dimethylarginine, dimethylaminohydrolase reduced circulating levels of ADMA were associated with enhanced angiogenesis 50. These studies suggest a role for ADMA as a regulator of angiogenesis.

However, a recent study challenged the conception that all of ADMÀs effects are mediated via inhibition of NO action by showing that chronic administration of ADMA via osmotic minipumps induces similar vascular lesions in wild‐type and eNOS knockout mice 51. These authors reported upregulation of angiotensin converting enzyme activity during ADMA administration, and inhibition of ADMA‐induced lesion formation by angiotensin receptor blockade 51.

Association of ADMA with clinical conditions beyond cardiovascular medicine

Although the major focus in the research on the pathophysiological effects of ADMA has been in cardiovascular medicine, ADMA has been found to be associated with other clinical conditions as well.

ADMA has been shown to be elevated in severe liver failure. Tsikas and co‐workers demonstrated an about 10‐fold increase in circulating ADMA in eight patients with end‐stage liver disease compared with healthy controls 52. Siroen and colleagues 53 found elevated ADMA concentration in 42 patients with acute or chronic hepatic failure compared with healthy volunteers. After liver transplantation, ADMA levels immediately decreased; however, in patients who experienced a period of acute graft rejection, post‐transplantation ADMA levels were higher than in non‐rejectors. The latter group of investigators hypothesized elevation of ADMA due to impaired elimination by the kidney and the liver as one possible causative factor of multi‐organ failure 54.

There is one report from a clinical study in patients with hyperthyroidism 55. These patients had elevated ADMA concentration and diminished L‐arginine/ADMA ratio compared with euthyroid controls, which reflected a decreased NO production. Hyperthyroidism may, by inducing a catabolic state, accelerate systemic protein turnover and thereby lead to increased release of methylated arginine residues from methylated proteins. However, this hypothesis has not yet been proven.

Insulin resistance and metabolic syndrome are two important clinical entities which will have a tremendous impact on population health in the Western world in the near future. Both pre‐diabetic subjects with impaired glucose tolerance 56 and patients with metabolic syndrome 57 have been demonstrated to have elevated ADMA levels. Thus, ADMA is suggested to promote the cardiovascular complications of these metabolic diseases.

Preeclampsia is another clinical condition that is associated with impaired endothelial function, poor prognosis and ill‐explained pathophysiology. Several studies in recent years have shown elevated ADMA levels in patients with preeclampsia compared with healthy pregnant women 58,59. However, this finding is again restricted to women from the Western countries, as in one study with Colombian women, ADMA was found not to be different between preeclamptic and healthy pregnancies 60.

Finally, early studies on the biochemistry of methylated arginines have shown that neuronal proteins are methylated to a large extent 4. However, there appears to be more SDMA than ADMA in the central nervous system, due to the higher expression of protein methyltransferase type II enzymes 4. Accordingly, data from clinical studies with patients suffering from neurodegenerative diseases have been inconclusive so far: in one study, ADMA levels in the cerebrospinal fluid were significantly higher in patients with Alzheimer`s disease than in neurologically normal controls 61. By contrast, Mulder and colleagues 62 found no difference in cerebrospinal fluid ADMA levels between 20 early stage Alzheimer`s disease patients and 20 age‐matched controls. Selley 63 reported a highly significant elevation of ADMA plasma levels in 25 Alzheimer`s disease patients compared with 25 control subjects. This was associated with reduced plasma nitrate and nitrite levels, suggesting reduced NO production in these patients. However, according to these studies, the relationship between neurodegenerative diseases and ADMA appears uncertain.

Measurement of ADMA in human subjects: methodological considerations

Research into ADMA has been hampered in the past by the insufficient comparability of values generated with different analytical methods, which were mainly based upon HPLC technology. ADMA levels determined with different HPLC methods sometimes yielded grossly different results 9. The main reason for this was the protein precipitation steps, which were performed by most but not all laboratories. Protein precipitation results in lower ADMA levels, probably because only the non‐protein‐bound portion of circulating ADMA is assessed. Indeed, circulating ADMA is bound to a considerable degree to plasma proteins 64. Since 2002, an international research initiative has undertaken a comparison of analytical methods, in order to allow the comparison of data measured in different laboratories around the world 65. Moreover, novel methods have been developed which allow quantitation of ADMA with higher precision and accuracy 65. Among these, mass spectrometric methods like LC‐MS/MS and GC‐MS may be the analytical ‘gold standard’ if they use a true (stable isotope‐labelled) internal standard 66; however, all of these methods are laborious, not available in many laboratories, and therefore not readily applicable for routine diagnostic use. More available, less time consuming and therefore more affordable methods to measure ADMA are clearly needed to make ADMA determination a routine clinical tool. With the development and validation of a novel competitive enzyme immunoassay that allows to sensitively and specifically assess ADMA levels, such technology has recently been developed and validated by us 67,68. This ADMA ELISA kit is based on the principles of a competitive immunoassay. In contrast to previously available methods for measuring ADMA, the new ELISA is easy to use and constitutes a high throughput technique. The combination of an acylation step and the competitive design of the ELISA resulted in a specific, highly sensitive and non‐isotopic immunoassay. The selected antiserum is specific for ADMA and results in negligible cross reactivities for L‐arginine (<0.02%) and other endogenous derivatives of L‐arginine. The precision of our ELISA has been demonstrated by low intra‐ and interassay coefficients of variation (interassay, 8.3%–10.3%; intraassay, 4.5%–7.5%). ADMA concentrations can be accurately measured across the full range of physiologically relevant concentrations (i.e. 0.05 µmol/L to 2 µmol/L). The values derived from the ELISA correlate well with expected values in recovery tests (mean recovery from all serum samples was 94.6%), and they display excellent linearity in dilution studies. The correlation with LC‐MS/MS is excellent (r = 0.984; P<0.0001). The ELISA has been validated for human serum (which is the preferential matrix) and plasma, as well as for rat and mouse plasma and cell culture supernatants.

Therefore, a wide array of reliable methods to quantify ADMA is now available which all have advantages and disadvantages: Modern HPLC techniques have been shown to produce reliable results 69; HPLC and LC‐MS/MS techniques allow to simultaneously measure ADMA, SDMA, and L‐arginine in one sample; however, these techniques are not available in every laboratory. ELISA is rapid, easy to perform and widely available; it produces data comparable to other methods, but allows to only measure ADMA.

Assessing ADMA in humans: in whom, when?

Considering the steadily accumulating data from prospective clinical trials that suggest a role for ADMA as a marker of cardiovascular disease, one may speculate about a potential role of ADMA measurement in clinical routine. One important question to ask therefore is: which patient should be tested, and when?

ADMA is elevated in association with almost all traditional cardiovascular risk factors, and in patients with established cardiovascular or metabolic diseases. However, its concentration shows significant variability even within these patient populations. Moreover, multivariate regression analyses that have taken into consideration all of the traditional and some additional novel risk factors have shown that ADMA remains independently associated with cardiovascular risk or total mortality even after such multivariate adjustment has been carried out.

A steadily increasing number of prospective clinical trials show that ADMA does more than merely discriminating between health and cardiovascular disease (for which, by the way, there are other and established diagnostic criteria). These data point out that even within a group of patients that have previously been regarded as homogeneous for their similar risk factor profile according to traditional risk factors, ADMA is useful to identify individuals with a comparatively high probability of experiencing a major event.

Cardiovascular risk assessment is currently re‐oriented towards assessment of total risk from an array of different risk markers, as opposed to the traditional approach to assess risk factors individually. This is effected by risk factor scores such as the Framingham risk score 70 (http://www.nhlbi.nih.gov/about/framingham/riskabs.htm), the PROCAM risk score 71 (http://chdrisk.uni‐muenster.de/calculator), or the risk charts developed by the European Society of Cardiology 72 (http://www.escardio.org/initiatives/prevention/SCORE+Risk+Charts.htm).

Where in this spectrum is there a place for novel risk factors such as ADMA? Novel risk markers may be of little help in assessing the total cardiovascular risk of patients who clearly fall into the high‐risk (i.e. greater than 20% risk for a major event during the subsequent 10 years) or low‐risk (i.e. below 10% risk) categories. However, what about the many patients who fall within the ‘grey zone’ of intermediate risk? For example, in a patient with a 17% risk of experiencing a major adverse cardiovascular event within the next 10 years, novel risk markers like ADMA levels may be helpful in estimating the real risk. Similarly, with an ADMA level well above the normal range (which has been determined as 0.36 – 1.17 µmol/L (73)), an elevated high‐sensitive CRP level, or an elevated level of another novel risk marker may still identify this patient as a candidate for pharmacotherapeutic intervention. This conception has led to the development of the ‘multimarker’ approach for cardiovascular risk assessment, which is currently put forward by many experts in the field 74. Indeed, a recent study based upon data from the first prospective study on ADMA as a cardiovascular risk factor in haemodialysis patients 34 showed that the combined use of ADMA, BNP and CRP explained 9%–10% of the mortality variance of standard Cox regression models, as opposed to 3%–5% when these factors were added individually 75.

When ADMA is elevated: Therapeutic consequences?

Clinical and experimental evidence cited above suggests that elevation of ADMA causes NO deficiency, even in the presence of ‘normal’ L‐arginine plasma concentrations. As ADMA is a competitive inhibitor of NO synthase, its inhibitory action can be overcome by increasing the concentration of this enzymès natural substrate. Therefore, dietary L‐arginine supplementation may be one means of specifically antagonizing the deleterious effects of ADMA on endothelial function in primary and secondary prevention. Nutraceuticals containing high levels of L‐arginine are widely available around the globe, and dietary intervention with L‐arginine has been shown to improve the symptoms of cardiovascular disease in several studies (for a review of L‐arginine studies in the context of elevated ADMA, see 76).

In a more general sense, elevated ADMA indicates an elevated risk of experiencing a major adverse cardiovascular event and represents one element of a multimarker risk profile. Thus, any pharmacotherapeutic intervention that has been proven to be effective in lowering cardiovascular morbidity and mortality such as β blockers, ACE inhibitors, statins, and platelet function inhibitors, can be expected to contribute to risk reduction in patients with elevated AMDA as well.

Nonetheless, a few small sample sized clinical studies suggested that certain drugs may decrease ADMA levels. This has been shown for ACE inhibitors and angiotensin receptor blockers in two studies 77,78. However, patient numbers were 20 and 19, respectively, and we were recently unable to reproduce these data in a study involving 24 hypertensive patients treated with the angiotensin receptor blocker telmisartan (R. Benndorf and R.H. Böger, unpublished data). The data from the large multinational VIVALDI trial that are expected for 2006 will more clearly show whether angiotensin receptor blockers exert an effect on ADMA levels 79.

Certain antidiabetic drugs like rosiglitazone 56 and metformin 80 have also been suggested to lower ADMA. The former is believed to exert this action via activation of PPARγ receptors, while the latter has been shown to have some structural similarities with ADMA, and may thus interfere with ADMA metabolism via currently unknown mechanisms.

Although a close relationship has been established between ADMA and LDL cholesterol levels in some studies 15, the only statin that has been shown to reduce ADMA levels in one study is rosuvastatin 81. Simvastatin, pravastatin, and atorvastatin had no effect on ADMA levels 82,83. Notwithstanding this, ADMA levels may be important for statin treatment. The beneficial vascular effects of statins depend in part on their ability to upregulate endothelial NO synthase gene expression (Figure 3a) 84. However, there are about as many clinical studies in which statins failed to improve endothelium‐dependent vasodilation as there are in which they failed. ADMA may help to resolve this discrepancy: upregulation of eNOS gene expression by statins may be functionally useless when the enzymès catalytic activity is blocked by ADMA (Figure 3b) 85. Indeed, Janatuinen et al. 82 recently found that six months of treatment with pravastatin resulted in enhanced myocardial blood flow in patients with low ADMA levels, whereas the drug was ineffective in patients with elevated ADMA. We took up the idea that ADMA may block the NO synthase despite its upregulated gene expression after statin treatment, and that this blockade may be overcome by L‐arginine supplementation (Figure 3c) 85. Indeed, we found, in a randomized controlled trial, that in patients with elevated ADMA concentration, simvastatin only enhanced endothelium‐dependent vasodilation when it was combined with supplemental nutritional L‐arginine 86.

Figure 3. ADMA may modify the statins' therapeutic effects. Statins are suggested to exert at least part of their vascular protective effects via upregulation of NO synthase gene expression (a). In the presence of high circulating ADMA levels, upregulation of the eNOS's gene by statins may still result in functionally inactive enzyme, since it is inhibited by ADMA (b). Excess exogenous L‐arginine may overcome inhibition by ADMA and thus restore the statins' effects on endothelial function (c). Reproduced from 75.

In conclusion, the role of ADMA as a novel cardiovascular risk factor is becoming increasingly clear with the publication of additional prospective clinical studies in various patient populations. However, data are still too sparse to draw definite conclusions as to how to incorporate ADMA into a multimarker concept of cardiovascular risk assessment. More and yet larger clinical trials are necessary to make it possible. As a tool for experimental studies in which endothelial regulation of vascular function is addressed, as a means to understand the pathophysiological regulation of the vascular regulator NO, and as a diagnostic approach in patients with unexplained cardiovascular disease, ADMA can be helpful at the moment. As far as therapeutic interventions to prevent the deleterious cardiovascular effects of ADMA are concerned, the long road from small, experimental clinical studies to large prospective clinical trials needs to be taken before we ultimately understand how to specifically deal with elevated ADMA levels.