Drug of Choice in the Management of Hypertension in Diabetes and Diabetic Nephropathy: Angiotensin-Converting Enzyme Inhibitors
Eberhard Ritz, MD, Nierenzentrum, Im Neuenheimer Feld 162, D-69120 Heidelberg, Germany
Hypertension is common in patients with diabetes mellitus and is a main cause of renal and cardiovascular complications. There has been recent controversy on what should be considered the optimal blood pressure goal and the optimal antihypertensive agent. It has become apparent that one blood pressure does not fit all in diabetic patients. Major confounders are preexisting cardiovascular disease and presence or absence of proteinuric kidney disease. In proteinuric diabetic nephropathy, renin-angiotensin system blockade is clearly indicated, but monotherapy is practically always insufficient to achieve target blood pressure values. J Clin Hypertens (Greenwich). 2011;13:285–289. © 2011 Wiley Periodicals, Inc.
The association of diabetes and hypertension has been known for a long time.1 Joslin and colleagues2 found that only 15% of diabetics between 25 and 50 years, yet 35% in patients older than 50 years, had systolic pressures >150 mm Hg, presumably as a result of delayed onset of diabetic nephropathy. With the prolongation of the lifespan, however, hypertension may be seen in type 1 diabetes even in the absence of diabetic nephropathy.
In patients with type 2 diabetes, hypertension is usually present even in the prediabetic stage. At the time diabetes is diagnosed, we find hypertensive blood pressure (BP) values by ambulatory BP monitoring in no more than 80% of type 2 diabetic patients.3
Prediabetes vs Diabetes
Elevation of BP precedes diabetes and frequently also prediabetes. In a population study, individuals without chronic kidney disease (CKD) at baseline were followed up to 20 years. The risk of CKD was progressively greater the higher the BP at baseline (even including BP values within the normal range). Compared with patients without diabetes at any given level of BP, the risk of future development of CKD was substantially higher in patients with diabetes.4
In the context of antihypertensive treatment it is also of interest that hypokalemia is a risk factor for the development of diabetes. It is unclear whether hypokalemia is the result of treatment with thiazides, the result of inappropriately high aldosterone concentrations, or the result of other factors.5
Albuminuria is not always the consequence of diabetes, as was thought in the past. Halimi and colleagues6 found that low grades of albuminuria are a predictor of subsequent diabetes. Patients with prediabetes and proteinuria, but without overt diabetes, may already have glomerular basement membrane thickening. They also have a high risk for developing diabetes in the course of 24 months.7 That the renal risk starts very early is also underlined by the recent finding that even in prediabetic patients, ie, with elevated values of 1-hour postprandial glycemia, renal function starts to deteriorate.8 This finding suggests that the definition of overt diabetes requires reassessment.
There are rare but potentially important observations of biopsy-confirmed diabetic glomerulosclerosis in patients with pathologic glucose tolerance but no overt diabetes.9,10 It is uncertain whether such patients had preceding episodes of overt diabetes (postdiabetic) or whether these observations are the peak of an iceberg documenting that renal damage may occur at values of glycemia that are lower than current definitions of diabetes (prediabetic).
Hypertension and Diabetic Kidney Disease
The classical scheme of the natural history of diabetic nephropathy11 suggested that in diabetes, hypertension was a relatively late result of renal dysfunction indicated by the presence of proteinuria. This paradigm of the obligatory link between hypertension and proteinuric diabetic kidney disease must to some extent be modified because a number of studies recently found that a high proportion of type 2 diabetic patients have major reduction of glomerular filtration, ie, <60 mL/min/1.73 m2, in the absence of proteinuria.12,13 Recent observations even suggest that this type of nephropathy is the result of small arterial vessel disease.13 Presumably, the response of this variant to BP lowering will differ from that of proteinuric diabetic kidney disease.
Office BP is less than optimal to assess cardiovascular and renal risk. A number of recent studies documented that nocturnal BP is the most important determinant for the increase in albuminuria in type 1 and in type 2 diabetes.14 In type 1 diabetic patients, comparison of the effect of BP and hemoglobin (HbA1c) on the development of microalbuminuria clearly indicated that BP, but not HbA1c, is the most powerful predictor.15 In type 2 diabetes, nocturnal and 24-hour BPs are also the best predictors for the onset of microalbuminuria,16 as well as for the progression of nephropathy.17 The risk of progression was also substantially higher in individuals with high pulse pressure, ie, above the median >57.5 mm Hg. This finding implies that high pulse pressure, ie, increased vascular stiffness, is associated with higher renal and cardiovascular risk.
Not all types of diabetic target organ damage respond equally well to BP lowering. In type 2 diabetes, BP lowering does not significantly affect progression of retinopathy, possibly because the intervention comes too late for this type of target organ.18 In contrast, BP lowering is beneficial in diabetic patients with kidney disease. Even in type 2 diabetic patients with advanced nephropathy, the risk of progression to end-stage renal disease was—in addition to albuminuria—still clearly related to systolic BP.19
In all types of kidney disease, renin-angiotensin system (RAS) blockade with ACE inhibitors is superior to alternative antihypertensive agents if proteinuria is present.20 This is not the case in the absence of proteinuria.21 Therefore, in diabetic patients with proteinuric kidney disease, albuminuria/proteinuria are treatment targets and indications for RAS blockade. In diabetic patients, both ACE inhibitors22 and angiotensin receptor blockers (ARBs) (Haller and colleagues, N Engl J Med., in press) cause BP-independent reduction of the onset of microalbuminuria. In patients with proteinuric diabetic kidney disease, baseline proteinuria predicted both renal and cardiovascular end points;23,24 furthermore, lowering of proteinuria by ARBs was associated with significant reduction in renal and cardiovascular end points. In the Irbesartan in Diabetic Nephropathy Trial (IDNT), lowering proteinuria by more than 50% reduced the risk of adverse renal outcome by >50%.24 Assignment to irbesartan resulted in a 33% reduction in risk (P<.001) for reaching a renal end point beyond that achieved by lowering the systolic BP. Even in patients who achieved a systolic BP <134 mm Hg, adverse renal outcomes were significantly reduced with irbesartan (12%) compared with amlodipine plus placebo (20%; relative risk, 0.55). The effects of lower systolic BP and treatment with irbesartan were additive and completely independent,25 in line with the finding that in all types of proteinuric renal disease, both BP lowering and RAS blockade have beneficial effects on outcome.26
One aspect that may confound the issue of antihypertensive treatment in diabetes is the recent observation that a substantial proportion of patients with type 2 diabetes fails to develop albuminuria but nevertheless experiences progressive loss of estimated glomerular filtration rate. The natural history of this nonproteinuric type of diabetic nephropathy is not well-known, nor is there information on the response to BP lowering with different antihypertensive agents. Cerebral microinfarcts, ie, arteriolar disease, predicted doubling of serum-creatinine in such nonalbuminuric type 2 diabetic patients.13 This type of vascular lesion may even be adversely affected by aggressive BP lowering and patients are unlikely to derive BP-independent benefit from RAS blockade.
What is the relative efficacy of BP lowering vs RAS blockade in diabetic nephropathy? Although guidelines strongly emphasize the importance of RAS blockade, the IDNT trial25 showed that BP lowering per se had a greater impact on outcome than did RAS blockade. Even in patients who achieved a systolic BP <134 mm Hg, renal outcomes were reduced by irbesartan compared with amlodipine plus placebo (relative risk, 0.55; P=.034). These two effects, ie, BP lowering and RAS blockade, were completely independent. Obviously, RAS alone without lowering of BP does not provide optimal benefit and elevation of BP is the main culprit and target of intervention.
RAS blockade is fraught with one major problem, ie, the “escape” phenomenon. In a large proportion of diabetic patients, BP and proteinuria are initially lowered following blockade of the RAS with ACE inhibitors or ARBs. Unfortunately, in a substantial proportion of patients proteinuria tends to increase again after the initial decrease (“escape”). Such secondary increase of proteinuria responds well to spironolactone,27,28 pointing to the importance of mineralocorticoid receptor activation by aldosterone.
In patients with diabetic nephropathy and the escape phenomenon, a head-on comparison of losartan 100 mg/d and spironolactone (25 mg/d) as add-on therapy to lisinopril (80 mg/d), spironolactone was superior to losartan.29 High salt intake is a permissive factor for the adverse effect of mineralocorticoid receptor activation. The response to spironolactone is explained by higher concentrations of circulating aldosterone or possibly even by intrarenal local synthesis of aldosterone.30
In the presence of primary failure to reduce proteinuria with RAS blockade (nonresponse) or secondary increase (escape), several additional strategies are effective.
First, since the adverse effect of aldosterone is dependent on high salt intake,31 dietary salt intake should be reduced. An added benefit of lower salt intake is the facilitation of treatment with diuretics.
A second strategy is to increase the dose of ACE inhibitors or ARBs to levels above the limits licensed for BP reduction. The efficacy of this approach has been documented by many studies.32 The combination of ACE inhibitors and ARBs is controversial because of the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial (ONTARGET) findings.33
Endothelin receptor blockers were shown to further reduce proteinuria in diabetic patients on RAS blockade.34 They are definitely contraindicated, however, in diabetic patients with impaired renal function,35 and more information on safety is necessary.
Aliskiren36 effectively further reduces proteinuria in patients with diabetic nephropathy taking losartan.
BP Targets in Diabetic Nephropathy
Presumably, target BP values differ for different end points. In the IDNT trial,25 achieved systolic BP values <120 mm Hg were associated with better renal outcomes, yet with higher mortality. Low diastolic BP values were also associated with a higher risk of myocardial infarction,37 similar to what is seen in patients with coronary heart disease.38 Many issues concerning renoprotective antihypertensive treatment in diabetes are currently unresolved.
A first point is the issue what BP measurement should be used, ie, office BP, self-measured BP, ambulatory 24-hour BP, or nighttime BP. It is clear today that self-measured BP and even more nighttime BP by ambulatory BP monitoring is far superior to clinic BP in predicting diabetic complications, ie, nephropathy, retinopathy, coronary heart disease.39
A second consideration is whether the most relevant treatment target is systolic BP, diastolic BP, pulse pressure, or a combination of these.
Furthermore, in diseases such as diabetes, which are characterized by major vascular remodeling, BP measured in the periphery may differ substantially from central BP. The most important targets of BP-induced damage (ie, heart, kidney, and brain) are exposed to central BP. Especially in coronary heart disease, which frequently goes unnoticed in diabetic patients, overshooting low central BP may cause adverse effects.40 Another condition that may be adversely affected by BP lowering is cerebral vascular disease of diabetes.13 Finally, autonomic polyneuropathy is frequent in patients with a long history of diabetes and exposes the patients to the risk of adverse coronary and cerebral damage.
Past guidelines on optimal BP values tended to extrapolate from observational data, but observational data do not necessarily allow conclusions with respect to interventional treatment targets. In the United Kingdom Prospective Diabetes Study (UKPDS),41 observational data suggested a correlation between BP and diabetic complications: the lower the BP, the less frequent the end points and mortality. This was true even for values <120 mm Hg systolic. The same was seen with respect to renal end points in the Action in Diabetes and Vascular Disease—Preterax and Diamicron MR Controlled Evaluation (ADVANCE) study.42 It had also been shown in a review of the literature that in all controlled BP-lowering studies that included diabetic patients, the group randomized for more aggressive compared with less aggressive BP treatment experienced benefit with respect to end points.43 In addition in the ADVANCE study on type 2 diabetic patients, BP lowering even to levels <120 mm Hg systolic reduced renal events significantly. In this group, with a median systolic BP of 113 mm Hg, the hazard ratio was 0.7.42
In 2003, Vijan and Hayward44 proposed a BP target of 135/80 mm Hg in type 2 diabetic patients and stated that “aggressive BP lowering is presumably the most important factor in the prevention of pathologic events in type 2 diabetic patients.” While this is presumably correct in type 2 diabetic patients without major target organ damage, we think that there is no target pressure that fits all diabetic patients and that a more individualized approach is necessary, specifically in diabetic patients with major target organ damage. This view is based on the following observations. In the IDNT study, lowering BP to values <120 mm Hg systolic reduced the frequency of renal events but was associated with increased mortality.25 In the same study, Berl and colleagues37 noted that the lower the diastolic pressure, the greater the risk of myocardial infarction. This is in line with observations in nondiabetic patients in the INVEST study.45 It has become increasingly clear that the underlying problem is the risk of diastolic malperfusion of the heart when diastolic BP is lowered in patients with pre-existing coronary disease or cardiac damage.45 Although strokes were not affected in the INVEST45 and IDNT trials, the relatively high proportion of patients without proteinuria and with microvascular disease in the brain13 argues for caution and individualization of antihypertensive treatment. The observations of Berl and colleagues37 in type 2 diabetics and of Messerli and colleagues45 in cardiac patients may also explain the observation in the ONTARGET study (which included a high proportion of type 2 diabetic patients) that patients with an initial systolic pressure >130 mm Hg benefited from BP lowering, whilst in patients with systolic pressure <130 mm Hg, the cardiovascular risk was increased.40
Limitations of Antihypertensive Treatment
In diabetic patients without coronary heart disease and heart disease in general, aggressive BP lowering, presumably <140 mm Hg systolic (personal opinion) is appropriate. In general, the shorter the duration of type 2 diabetes, the more aggressive should be BP lowering; however, in longstanding type 2 diabetes with vascular disease, BP lowering should be more cautious.
Absence of proteinuria with impaired renal function is presumably associated with cerebral microvascular disease in which aggressive BP lowering is also known to aggravate cerebral malfunction.
One point not well addressed in the past studies is the issue of orthostatic hypotension as a consequence of autonomic polyneuropathy. At least in the elderly diabetic patient and the patient with longstanding hypertension, measurement in the upright position should be the standard of treatment.
There is no single target pressure that fits all patients, and treatment must be carefully individualized (one size does not fit all).46