|Home | About | Journals | Submit | Contact Us | Français|
The rationale for use of angiotensin-converting enzyme (ACE) inhibitors in patients with proteinuria or established renal impairment is in general twofold—first, to slow the progression of renal disease, by reducing blood pressure (BP) and proteinuria or by other renoprotective mechanisms; second, to lessen the risk of cardiovascular morbidity and mortality associated with chronic renal disease and with one of its leading causes, diabetes. This review highlights the evidence for and against targeting the renin-angiotensin system in the treatment of patients with different types of nephropathy. Findings from the major studies are summarized in Table 1.
Hypertension and the degree of proteinuria are the two most important factors influencing the rate of decline in renal function in most forms of chronic renal disease. There is evidence for renoprotective benefit from pharmacological intervention in types 1 and 2 diabetes as well as in non-diabetic nephropathy. Almost 20 years ago, Parving et al. reported that, in a small group of patients with type 1 diabetes and nephropathy, lowering of BP with combinations of beta-blockers, diuretics and vasodilators reduced albuminuria and slowed the decline in glomerular filtration rate (GFR)1. More recently his group published data from a prospective observational cohort study suggesting that aggressive BP control in some patients with type 1 diabetes may even improve the GFR as well as reduce proteinuria, thus challenging the widely held belief that diabetic nephropathy progresses relentlessly2.
The United Kingdom Prospective Diabetes Study (UKPDS)3 and the Modification of Diet in Renal Disease (MDRD) study4 provided evidence for the renoprotective effect of BP lowering in type 2 diabetes and in non-diabetic nephropathy respectively, as have numerous smaller studies. There is also evidence that the baseline level of proteinuria correlates with rate of progression of renal impairment in both type 1 and type 2 diabetes and in non-diabetic nephropathy4,5. Furthermore, the extent of reduction of proteinuria on initiation of antihypertensive therapy correlates with the degree of subsequent renoprotection6,7. Proteinuria is thus regarded as an independent risk factor for progression of nephropathy, rather than simply as a marker. There is at present much research interest in the mechanisms by which the protein filtered by the glomerulus damages renal function. Reabsorption in the proximal tubule may activate inflammatory pathways, leading to interstitial fibrosis8; thus, reduction of proteinuria to a minimum seems an important target in the management of nephropathy of any aetiology.
The issue of whether ACE inhibitors are more renoprotective than other modes of antihypertensive therapy has been clarified for some subgroups with renal disease but not others; the question of cardioprotection adds complexity to the debate.
In a landmark study published in 1993, Lewis et al. provided evidence for a renoprotective effect of captopril independent of BP lowering in type 1 diabetes9. Thereafter, ACE inhibitors were regarded as first-line antihypertensive agents in patients of the same kind—namely, those with type 1 diabetes, clinical grade proteinuria (> 500 mg/24 h) and mild to moderate renal impairment (creatinine < 221 μmol/L). Despite the possible confounding factors of slightly lower average BP and baseline proteinuria in the captopril group, the results are widely seen as robust. The use of ACE inhibitors in normotensive patients with type 1 diabetes has been proposed on the argument that hypertension was not an entry criterion for Lewis's study, although in practice over 75% of the patients were hypertensive and the target BP at < 140/ < 90 mmHg was higher than most current guidelines.
Meta-analyses incorporating studies of both types 1 and 2 diabetes supported these conclusions10,11 and generated further interest in ACE inhibition for such ‘normotensive’ patients. Bohlen and colleagues calculated that ACE inhibitors caused a 28% reduction in proteinuria when the benefit of blood-pressure reduction was allowed for, whereas the antiproteinuric effect of other classes of antihypertensive was thought to be blood-pressure related11. More recently Parving et al. suggested that ACE inhibition in normotensive type 1 diabetes provides long-term benefit through keeping BP within normal limits and preventing an increase in albuminuria12. However, this small non-blinded follow-up study could not show whether ACE inhibitors are superior in this respect to other antihypertensives.
Similarly, numerous studies have suggested a role for ACE inhibition in reducing the rate of progression of microalbuminuria (30-300 mg/24 h) to overt nephropathy (> 300 mg)13,14,15,16. A meta-analysis of individual patient data has pointed to an ACE-inhibitor effect beyond that of blood-pressure lowering17.
In view of the evidence that proteinuria is nephrotoxic and that microalbuminuria predicts progression to overt nephropathy in a substantial proportion of patients with type 1 diabetes, an obvious strategy is to delay the onset of microalbuminuria. This in turn may delay the development of hypertension—which in this group tends to correlate with the progression of proteinuria, creating a vicious circle. The Diabetes Control and Complications Trial (DCCT)18 suggested that tight glycaemic control has a role in primary prevention, but to date there is no convincing clinical evidence in favour of ACE inhibition or antihypertensives in general. The EUCLID study group concluded that lisinopril slowed the progression of renal disease in normotensive patients with little or no albuminuria, but the small benefit in the normoalbuminuric subgroup was not statistically significant, even before correction for BP lowering16. Data from streptozotocin-diabetic rats suggest a role for ACE inhibition in primary prevention19, and this important area is likely to receive further attention in clinical trials.
Various small studies have suggested equivalence of angiotensin II receptor blockers (ARBs) to ACE inhibitors in both overt nephropathy and microalbuminuria20, 21. However, in the absence of outcome data from large prospective randomized trials, ARBs (which selectively antagonize the type 1 receptor) have not as yet challenged ACE inhibitors as first-line antihypertensive or anti-proteinuric agents in these groups.
It is the increase in prevalence of type 2 diabetes that is largely responsible for the dramatic rise in end-stage renal failure in Europe and the USA. Yet convincing evidence for renoprotection from interruption of the renin-angiotensin system has only recently been acquired, far behind that for type 1 diabetes and non-diabetic nephropathy.
Comparisons of ACE inhibitors with other antihypertensives have yielded conflicting results with respect to progression of renal disease. In the UKPDS, BP reduction led to a clear reduction in the rate of progression of microalbuminuria, in addition to macrovascular benefits, over 9 years' follow-up3. No significant difference in microvascular or macrovascular outcomes was detected between the ACE-inhibitor and beta-blocker groups, suggesting that BP reduction was paramount22. However, this study was not powered to detect a large difference between these groups.
In the MICRO-HOPE study, ramipril reduced the risk of progression from microalbuminuria to overt nephropathy in diabetic patients (98% type 2) with high cardiovascular risk, the comparison being with placebo23. This result was consistent with data previously published by Ravid et al., but again there was the possible confounding influence of a small BP difference between groups24, 25. Italian workers suggested that the absence of clear benefit from ACE inhibition in type 2 diabetes could reflect the more heterogeneous nephropathy than that of type 1 diabetes, relating to the later onset of the disease and greater probability of coexistent renal injury from hypertension, dyslipidaemia and smoking26. This may be true, but we think that a more important reason for the lack of clarity is the dearth of large prospective studies with a primary renal outcome.
Some of the answers have now emerged from three studies investigating the benefits of ARBs rather than ACE inhibitors. In the Irbesartan Microalbuminuria Study 2 (IRMA-2), Parving et al. demonstrated a significant renoprotective effect, independent of blood-pressure lowering, of irbesartan in hypertensive type 2 diabetic patients with microalbuminuria27. Lewis et al., in the Irbesartan Diabetic Nephropathy Trial (IDNT), showed a similar renoprotective benefit of irbesartan, this time in hypertensive diabetic patients with established nephropathy, the primary outcome being a doubling of serum creatinine28. 300mg irbesartan per day was more renoprotective than either amlodipine or placebo, with no difference between the latter two groups. Cardiovascular morbidity and mortality did not differ between any of the groups, but these were secondary endpoints and the study was not necessarily powered to detect such differences.
Brenner et al. conducted a similar study, RENAAL, with the same endpoints but comparing losartan (50 mg or 100 mg per day) with placebo, other antihypertensives being used in each group29. In this study hypertension was not an entry criterion, although over 90% of patients were hypertensive. Again, a significant renoprotective effect was seen with the ARB, independent of BP, with no difference in the cardiovascular secondary endpoints between the groups.
A greater evidence base has therefore been established for ARBs than for the older and cheaper ACE inhibitors for renoprotection in the potentially huge population of patients with type 2 diabetes and proteinuria. Important though these studies are, they have not proved that ARBs are superior to ACE inhibitors, since there are no similar ACE-inhibitor studies and no large studies have directly compared the two. With several ACE inhibitors approaching end of patent, there may be little commercial interest in funding comparative studies of this kind.
Furthermore, the relative importance of renal and cardiovascular endpoints is not clear. The lack of significant cardiovascular protection in the IDTN and RENAAL studies may reflect inadequate statistical power and the exclusion of high-cardiovascular-risk patients, since the prime purpose was to investigate renoprotection rather than cardioprotection. Recently, the LIFE (losartan intervention for endpoint reduction in hypertension) study has provided useful information on cardioprotection in hypertensive patients with left ventricular hypertrophy (LVH), comparing atenolol and losartan30. In a substudy involving 1195 hypertensive diabetic patients (>80% type 2) with LVH, about 11% of whom had clinical albuminuria at baseline, losartan was significantly more effective than atenolol in reducing cardiovascular and all-cause mortality31. Notably, the intended mean follow-up time of RENAAL was shortened by about one year because of the publication of a subgroup analysis from the HOPE study suggesting cardiovascular benefit from ACE inhibition in patients with renal impairment and other cardiovascular risk factors, a proportion of whom were diabetic32. Indeed from MICROHOPE (the diabetic subgroup of HOPE) it is argued that any patient with type 2 diabetes aged over 55 with microalbuminuria or hypertension should, in the absence of contraindications, be given an ACE inhibitor. These criteria include a large proportion of the type 2 diabetic population, and it might be reasonable also to include those with overt nephropathy, since cardiovascular risk increases with advancing renal impairment.
Functionally significant bilateral renal artery stenosis (RAS) is recognized as a contraindication to the use of ACE inhibitors and ARBs. This is because reduction of pressure in the afferent arteriole caused by the stenosis renders maintenance of adequate intraglomerular filtration pressure dependent on constriction of the efferent arteriole by angiotensin II. In view of the high prevalence of atherosclerosis in type 2 diabetes, the possible presence of RAS is of concern when ACE-inhibitor or ARB therapy is contemplated. Using magnetic resonance angiography Valabhji et al. detected a prevalence of 17% in 117 hypertensive patients with type 2 diabetes33. The functional significance of the stenoses in this and other studies was not clear, but 1% of patients in the captopril group in the UKPDS were withdrawn from the study because of a rise in creatinine (degree unspecified)22. Valabhji et al. suggested from their study that a femoral bruit was predictive of RAS, and markers suggested by others include advanced age, hypertension, and history of peripheral vascular or coronary artery disease. High-risk ‘vasculopaths’ should have their renal function monitored with great care if an ACE inhibitor or ARB is prescribed.
The MDRD study demonstrated that BP-lowering reduces proteinuria in non-diabetic nephropathy, with greater benefit the higher the proteinuria at baseline. This baseline level was also a prognostic indicator for the rate of decline in renal function4.
A meta-analysis of eleven randomized controlled trials (two of which had not been published) suggested that antihypertensive regimens containing ACE inhibitors are more efficacious than regimens without ACE inhibitors in reducing progression of renal impairment, as well as in limiting proteinuria, in patients with baseline proteinuria of 0.5 g per day or more34. Yet again, BP was a possible confounding factor, with the mean fall in systolic BP 4.5 mmHg greater in the ACE inhibitor group. After correction for the predicted benefits of reduction in BP and level of proteinuria, the ACE inhibitor remained more effective. However, in a separate analysis of the same eleven trials, it was concluded that the greater benefit of ACE inhibition on decline in renal function in patients with high baseline proteinuria was due to the greater reduction in proteinuria35. It was suggested that the degree of proteinuria at any given time after treatment has begun is a better prognostic indicator than baseline level.
In a post-hoc analysis of the Ramipril Efficacy in Nephropathy (REIN) trial, the original part of which was included in the above meta-analyses, ACE inhibition was shown to be renoprotective for all levels of renal function, down to a GFR of 10 mL/min, although the maximum benefit occurred when treatment started with a GFR of >50 mL/min36. This result may be of considerable interest to clinicians managing patients with advanced renal failure. There is often a dilemma as to whether or not to start an ACE inhibitor in the patient who presents late with low GFR, or whether to continue the drug in an individual approaching end-stage renal failure, given that ACE inhibition initially causes a small fall in GFR (resulting from a decrease in intraglomerular capillary pressure). As with most ACE inhibitor studies designed to investigate renal endpoints, no cardiovascular benefit was seen in REIN.
Several experimental studies and small clinical trials have suggested that, in this group of patients, ARBs have similar effects to ACE inhibitors in reduction of proteinuria and renoprotection37. There is some weak evidence that ARBs cause less hyperkalaemia than ACE inhibitors in patients with chronic renal disease; this advantage, if it exists, might be due to less suppression of plasma aldosterone38.
Since angiotensin II can be produced by pathways other than ACE—for example, by chymase39—addition of an ARB to an ACE inhibitor should give more complete blockade of the renin-angiotensin system than ACE inhibition alone. It is less obvious why this combination should be more efficacious than an ARB alone, if the latter causes complete blockade of the angiotensin II type 1 receptor (AT1, Figure 1). Reduction in circulating angiotensin II by ACE inhibition might provide some synergy, as might the increased levels of the vasoactive peptide bradykinin, which is also broken down by ACE40.
So, is combination therapy likely to yield further improvement in renal (and cardiovascular) outcomes? From the small studies published to date, there does seem some promise. The combination of enalapril with losartan reduced proteinuria more than either drug alone in a group of 10 normotensive patients with normal GFR and biopsyproven IgA nephropathy41. Similarly, Ruilope et al. demonstrated a synergistic antiproteinuric effect between benazepril and valsartan in non-diabetic renal disease42. Whilst they did not report any serious adverse effects, the possibility of hyperkalaemia, particularly in patients with severe renal impairment, requires further investigation. In the CALM study, involving 199 hypertensive patients with type 2 diabetes and microalbuminuria, the combination of lisinopril and candesartan was more effective than either agent alone in reducing BP43. A reduction in albuminuria was also seen, but may have been attributable to the BP effect rather than to the combination.
Moving one step further down the renin-angiotensin system, the addition of spironolactone to enalapril in a small study of 8 patients with mild renal impairment of various aetiologies and persistent proteinuria (> 1g/day) resulted in a 54% decrease in proteinuria with no effect on BP44. This raises the question of the relative importance of aldosterone and angiotensin II in glomerular haemodynamics.
Should all hypertensive renal patients be treated with an ACE inhibitor? From current evidence, patients with type 1 diabetes who have microalbuminuria or overt proteinuria should be so treated—as should non-diabetic nephropaths with > 0.5g/day proteinuria. Patients with type 2 diabetes and microalbuminuria or worse should be treated with an ACE inhibitor or an ARB, and ‘normotensive’ diabetic patients of both types will probably benefit, the evidence being less comprehensive. Less clear still is the optimal therapy for non-diabetic nephropaths with < 0.5g/day protein urea. The most important confounding factor in most of the studies comparing ACE inhibitors or ARBs with other antihypertensives is the BP difference between groups. Without this difficulty, blockade of the renin-angiotensin system might have become more widely accepted as first-line renoprotective therapy in most forms of nephropathy.
A possible treatment algorithm is shown in Figure 2. There is no clear consensus at present on blood-pressure targets for these groups of patients. The Joint British recommendations on prevention of coronary heart disease in clinical practice45 suggest a target of < 130/80 mmHg for type 1 diabetic patients, and we feel that on current evidence it is reasonable to use this as a target in type 2 diabetes and non-diabetic nephropathy, with respect to both renoprotection and cardioprotection. Where proteinuria is > 1g/day, the suggested target of < 125/75 mmHg is based on the MDRD study.
Note This review follows a meeting of the Nephrology section of the RSM in Cambridge, which included a debate entitled ‘This house believes that all renal patients with hypertension should receive an ACE Inhibitor’. MAJD and FEK are funded by the Children's Kidney Care Fund and the Wellcome Trust, respectively. The meeting was supported by an unrestricted educational grant from AstraZeneca UK Ltd.