Summary and Introduction

Summary

We are currently confronted with an epidemic of renal failure caused by diabetic nephropathy. It has become apparent that blood pressure is a major determinant of the risk of developing diabetic nephropathy; individuals with a genetic predisposition to hypertension are at increased risk of developing diabetes and diabetic nephropathy. Antihypertensive medication has an impact on development of diabetes; beyond blood-pressure lowering, the risk of diabetes is further reduced by blockade of the renin–angiotensin system (RAS). In experimental studies, blockade of the RAS in the pre-diabetic stage ameliorates the severity of subsequent diabetic nephropathy. Guidelines recommend a target blood pressure of 130/80 mmHg for diabetic patients without proteinuria and some guidelines recommend a target of less than 125/175 mmHg for diabetic patients with proteinuria. Above a systolic blood pressure of approximately 110 mmHg, the risk of progression of diabetic nephropathy increases progressively with increasing blood pressure. Blood-pressure lowering and blockade of the RAS delays or prevents onset of microalbuminuria, slows worsening of microalbuminuria and attenuates progression of diabetic nephropathy, even in advanced stages. In addition to blood pressure, proteinuria is a treatment target and should be reduced to below 1 g/24 h.

Introduction

In recent years there has been a worldwide 'epidemic' of renal failure in diabetic patients. The majority of these patients have type 2 diabetes,^{[1, 2]} which is the focus of this article. So, what is a type 2 diabetic patient's risk of developing nephropathy? In the United Kingdom Prospective Diabetes Study, the risk of a patient with type 2 diabetes progressing from normoalbuminuria to microalbuminuria was 2% per year, risk of progressing from microalbuminuria to macroalbuminuria was 2.8% per year, and risk of progressing from macroalbuminuria to an elevated serum creatinine level (>175 µmol/l) was 2.3% per year. It is notable that the competing risk of death exceeded the risk of progression once macroalbuminuria had developed. Risk of death was 0.7% per year for normoalbuminuric patients, 3.5% per year for macroalbuminuric patients and 12.1% per year for patients with an elevated level of serum creatinine. Renal risk is strongly dependent on blood pressure. In the Kaiser Permanente Programme, follow-up observation of individuals with no proteinuria at baseline showed that the risk of developing end-stage renal disease increased with progressively higher baseline blood-pressure values. This blood-pressure-dependent risk was substantially greater in type 2 diabetics than in nondiabetic patients.^[3]

That parents of patients with type 1 diabetes plus diabetic nephropathy have higher blood pressures than parents of diabetic patients without diabetic nephropathy^[4] indicates that a genetic predisposition to hypertension increases the risk of developing this form of kidney disease. The same dependency of risk on family history of hypertension and pre-diabetic blood pressure has been observed in type 2 diabetes.^{[5, 6]} A hereditary predisposition to diabetic nephropathy is also indicated by the finding that nondiabetic offspring of patients with type 2 diabetes plus diabetic nephropathy have higher blood pressures and urinary albumin levels than offspring of diabetic parents without diabetic nephropathy.^[7] In light of the clustering of cardiovascular disease in families of diabetic patients who develop diabetic nephropathy,^[8] taking the family history—with particular attention to hypertension and cardiovascular events—is an important element in the management of diabetic patients.

Metabolic Syndrome, Microalbuminuria and Diabetic Nephropathy

There is a close association between the metabolic syndrome^[9] and the risk of developing type 2 diabetes. The likelihood of a patient with the metabolic syndrome having microalbuminuria or chronic kidney disease increases with the number of features of the metabolic syndrome they possess,^[10] although the definition of the metabolic syndrome is somewhat imprecise and arbitrary^[11] (Box 1). Many slightly different and not necessarily concordant^[12] definitions of the metabolic syndrome have been proposed^{[13, 14]} (e.g. by the International Diabetes Federation). It has been argued that subsuming the cluster of cardiovascular risk factors related to insulin resistance under the definition of a 'syndrome' is not justified because there is no evidence for shared pathogenesis or for better predictive value from combined versus individual risk factors.^[11] Irrespective of these debates regarding definition, albuminuria is an independent predictor of the development of diabetes^[15] and the severity of albuminuria predicts progressive loss of renal function.^[16] In the absence of histological investigation, it is not known whether microalbuminuria in nondiabetic patients is caused by the same abnormality that underlies perturbed glomerular function in diabetes. At any rate, at the time of diagnosis of type 2 diabetes, microalbuminuria is present in approximately 15–20% of patients.^{[17, 18]} Whether microalbuminuria reflects the presence of the specific lesions of diabetic nephropathy, particularly abnormalities of the glomerular basement membrane and of podocytes, is uncertain.^[19]

The Renin–angiotensin System and Diabetic Nephropathy

In the context of selecting antihypertensive medication for use in hypertensive patients with the metabolic syndrome, it is of interest that the metabolic syndrome is a hyper-reninemic state.^[20] In an experimental model of metabolic syndrome, the angiotensin II type I receptor is upregulated.^[21] It is therefore not surprising that the high rate of cardiovascular events in microalbuminuric patients with metabolic syndrome is significantly reduced by inhibition of angiotensin-converting enzyme (ACE), as shown in the PREVEND study.^[22]

Blockade of the renin–angiotensin system (RAS) not only ameliorates microalbuminuria and associated cardiovascular endpoints^[23] but—at least according to experimental studies—might also minimize the severity of nephropathy when diabetes subsequently develops. In pre-diabetic Otsuka Long-Evans Tokushima fatty (OLETF) rats, transient RAS blockade with angiotensin-receptor blockers (ARBs), ACE inhibitors or a combination of the two drug types, attenuated the severity of diabetic nephropathy when type 2 diabetes ultimately supervened.^[24] The nonspecific antihypertensive agent hydralazine did not elicit this effect. Whether RAS blockade in pre-diabetic humans reduces the risk of subsequent diabetic nephropathy is a tantalizing issue that requires further investigation.

Antihypertensive Medication and Risk of Diabetes

Of particular interest in the pre-diabetic stage are new insights into the relationship between the RAS and the risk of developing diabetes. Hypertensive patients have, per se, a higher risk of developing de novo type 2 diabetes, and this is particularly true of patients taking β-blockers or diuretics. Conversely, the risk is reduced by blockade of the RAS, either using an ACE inhibitor (as indicated by the CAPP,^[25] ALLHAT,^[26] ANBP^[27] and SOLVD^[28] trials) or an ARB (as indicated by the LIFE,^[29] CHARM^[30] and ALPINE^[31] trials). It should be noted that in all of these studies, RAS blockade was assessed against antihypertensive agents that were not metabolically neutral. This caveat has been effectively eliminated, however, by the VALUE trial^[32] in which the ARB valsartan was compared with the metabolically neutral calcium-channel-blocker amlodipine. In this trial, valsartan reduced the rate of development of de novo diabetes. A recent meta-analysis^[33] showed that blockade of the RAS reduced the risk of de novo diabetes by an average of 22%.

The antidiabetogenic effect of RAS blockade can be explained by the fact that angiotensin II reduces insulin sensitivity and impairs insulin secretion. In the short term, angiotensin II interferes with glucose-mediated insulin secretion.^[34] In the long term (and presumably of greater importance) angiotensin II causes degeneration and fibrosis of islet cells.^[33] These actions probably have relevance to the progressive exhaustion of insulin secretion that occurs after diabetes has developed. Beta cells are particularly susceptible to angiotensin-II-induced damage because hyperglycemia activates the RAS in beta cells. Furthermore, beta cells have low antioxidant activity and are susceptible to injury mediated by oxidative stress, which might be involved in the metabolic syndrome.

Hypertension, Diabetes and Diabetic Nephropathy

An abnormal diurnal rhythm of blood pressure, a feature of which is an attenuated decrease in blood pressure at night, precedes the onset of diabetic nephropathy in patients with type 1 diabetes^[35] and presumably also in patients with type 2 diabetes. The dramatic impact of relatively minor differences in blood pressure,^[36] particularly systolic blood pressure,^[37] on cardiovascular and renal outcomes in type 2 diabetes was illustrated by the United Kingdom Prospective Diabetes Study, which prompted Mogensen to state in an accompanying editorial: "Not only is antihypertensive treatment more effective than tight blood glucose control; the beneficial results also come sooner".^[38] No available study has assessed the effects of different target blood pressures on primary renal endpoints in patients with diabetic nephropathy. One such study in chronic kidney disease^[39] showed that intensive (91 mmHg mean arterial pressure) as compared to ordinary (107 mmHg mean arterial pressure) blood-pressure control slowed the annual decline in glomerular filtration rate (-1.9 ml/min/year versus -3.4 ml/min/year, respectively). This effect was particularly pronounced in proteinuric patients.^[40] There is little doubt that there is a similar relationship in patients with diabetic nephropathy. Post-hoc analysis of blood pressure and renal events in the IDNT trial showed that the risk of a renal endpoint (i.e. doubling of serum creatinine or end-stage renal disease) increased progressively with increasing achieved systolic blood pressures once they exceeded 110 mmHg.^[41]

Proteinuria, Diabetes and Diabetic Nephropathy

In addition to blood pressure, proteinuria is an important treatment target.^[42] In a post-hoc analysis of the LIFE study, Lindholm found that onset of microalbuminuria was observed less frequently in patients taking the ARB losartan (7%) than in those on the β-blocker atenolol (13%).^[43] The ongoing, prospective ROADMAP study has as its primary endpoint reduction of de novo microalbuminuria by the ARB olmesartan. Such a reduction has been clearly demonstrated for an ACE inhibitor. Ruggenenti randomized normoalbuminuric patients with type 2 diabetes of relatively short duration to placebo (i.e. alternative antihypertensive agents) or to the ACE inhibitor trandolapril.^[44] The frequency of de novo microalbuminuria was 10.0% in the placebo group and 5.7% in the trandolapril group, but there was a minor blood-pressure difference (139 ± 10 mmHg systolic versus 142 ± 12 mmHg). No significant effect of the calcium-channel blocker verapamil on development of microalbuminuria was observed.

The efficacy of ACE inhibitors and ARBs in reducing prevalent microalbuminuria is well documented and the effect was found to be highly significant in a recent meta-analysis.^[45] Particularly impressive is the result of the IRMA-2 (irbesartan in patients with type 2 diabetes and microalbuminuria) study. The ARB irbesartan dose-dependently reduced progression to overt proteinuria from 14.9% (placebo group) to 5.2% (300 mg irbesartan).^[46] The antiproteinuric effect of RAS blockade could be related, at least in part, to restoration of expression of nephrin in the glomerular slit membrane of diabetic patients, as documented in a renal biopsy study.^[47]

Controversy

The main issue under debate is whether retardation of progression of diabetic nephropathy by RAS blockade is explained by the concomitant reduction of blood pressure ("effect beyond blood-pressure reduction"). In the MDRD (Modification of Diet in Renal Disease) trial, intensive blood-pressure control had a highly significant impact on progression in mostly nondiabetic patients, compared with standard blood-pressure control (target blood pressures for patients <60 years of age were <107 mmHg mean arterial pressure for standard control and <92 mmHg mean arterial pressure for intensive control).^[39] In a meta-analysis of 11 studies that compared ACE inhibitors with placebo or other antihypertensive agents, the risk of progressing to end-stage renal disease was reduced by 37% and the risk of serum creatinine doubling by 35%.^[48]

Two studies specific to type 2 diabetes detected blood-pressure-independent additional renal protection in type 2 diabetic patients on ARBs. Reduced rates of serum creatinine doubling, end-stage renal disease or death were achieved with the ARB irbesartan in the IDNT study^[49] and with losartan in the RENAAL trial.^[50] These results are remarkable because average clinical blood-pressure values were almost identical in the different arms of the studies. Nevertheless, if one compares the rate of progression in untreated nephropathic diabetic patients in historical series (loss of glomerular filtration rate approximately 10 ml/min/year) with the placebo arm of the RENAAL study (i.e. patients on antihypertensive agents other than losartan, who showed a loss of glomerular filtration rate of approximately 5 ml/min/year) it is obvious that blood-pressure lowering per se has a major impact. Consequently both blood-pressure lowering and blockade of the RAS are necessary.

The discussion about the relative impact of blood-pressure lowering versus blockade of RAS is largely academic because, in diabetic patients with impaired renal function, an average of 4.5 different classes of antihypertensive agents are needed to achieve target blood pressure. Attaining blood-pressure goals is virtually impossible if the RAS is not blocked. There is a further caveat; circadian blood-pressure profiles were not monitored during any of the studies mentioned above. In a normotensive model of renal damage, Griffin found no difference in the degree of glomerulosclerosis between animals on the ACE inhibitor enalapril and those administered a combination of three antihypertensive agents ('triple therapy'), but noted a striking correlation between radiotelemetrically measured blood pressure and renal damage.^[51] In type 1 diabetes, amelioration of the renal endpoints by administration of the ACE inhibitor captopril was documented in the seminal prospective controlled trial of Lewis.^[52]

What is the Right Dose of ACE Inhibitors or ARBs?

In the IRMA-2 study, 300 mg irbesartan was significantly more effective than 150 mg.^[46] Ongoing studies clearly show that increasing the dose further reduces proteinuria but not blood pressure. This has been well documented in animal experiments^[53] and by clinical observation.^[54] Presumably, the explanation is that even doses of an ACE inhibitor that lower blood pressures maximally fail to lower intrarenal concentrations of angiotensin II,^[55] possibly reflecting the activity of local intrarenal RASs in deep compartments that are not easily penetrated by ACE inhibitors or ARBs. High doses of ARBs cause drastic further reduction of proteinuria, glomerulosclerosis and other renal damage indices in the renal ablation model^[56] and in proteinuric diabetic patients.^[57] The extreme example is that massive doses of an ACE inhibitor not only halt progression, but even induce regression of glomerulosclerosis, in the renal ablation model.^[58] An alternative strategy is the combination of ACE inhibitors and ARBs.^[59]

Conclusions

It is obvious that both lowering blood pressure to target values and blockade of the RAS are necessary to ameliorate onset and progression of diabetic nephropathy. On the basis of data in nondiabetics, it has recently been concluded that blockade of the RAS is the more important component of this management strategy.^[60] Our belief—albeit based on retrospective data—is that lowering blood pressure is equally indispensable, as shown in the IDNT trial.^[41] What has not been resolved is the issue of the optimal dose of ACE inhibitors or ARBs,^[57] and whether combination of the two is superior to maximal doses of monotherapy.^[59] The most worrying aspect of this field, however, is the observation in many studies that the antiproteinuric effect of RAS blockade lessens over time. This phenomenon has been related to 'aldosterone escape' and angiotensin-II-independent effects of aldosterone on the kidney.^[61] The safety and efficacy of aldosterone blockade plus RAS blockade remains to be proven, but initial observations are encouraging.^[62]

Sidebar: Box 1. Definition of the metabolic syndrome according to the National Cholesterol Education Program.^[12]

Sidebar: Key Points

E. Ritz. Department of Internal Medicine, Ruperto Carola University Heidelberg, Nierenzentrum, Im Neuenheimer Feld 162, D-69120 Heidelberg, Germany. Email: prof.e.ritz@t-online.de