Hypertension in diabetes: diuretic, ACE inhibitor or AIIRA?

Abstract

Several studies provide evidence for the aggressive treatment of hypertension in diabetes. The review by Perry et al. (Hypertension in diabetes: what's new, what's true, what's next, page 247) provides a timely review of this evidence, including thresholds for intervention and targets for treatment. They emphasise the importance of blood pressure reduction per se as the most important consideration, with choice of antihypertensive agent a secondary consideration. The recent studies of angiotensin II-receptor antagonists (AIIRAs) in patients with proteinuria1-3 or left ventricular hypertrophy (LVH)4 are described in detail, along with the recently reported ALLHAT study,5 the largest reported study of hypertension to date. As they indicate, ALLHAT has laid to rest any doubts about the safety of amlodipine in people with diabetes, but does it help us to decide which agents to use first and which agents to add when treating people with diabetes? The ALLHAT study compared a regimen based on a diuretic, chlorthalidone, with regimens based on lisinopril or amlodipine (the fourth arm with treatment based on doxazosin was discontinued early). No major differences were found in the primary end-point, and commentators have seized on diuretics as the treatment of first choice on cost grounds rather than ACE inhibitors, amlodipine or other newer agents. To the surprise of many, the diuretic appeared to be better than lisinopril in reducing the secondary end-point of cardiac failure. At first sight this is a surprising finding, as in cardiac failure diuretics are of symptomatic benefit only, yet ACE inhibitors are of clear prognostic benefit. Comparative studies in non-diabetic subjects also indicate that ACE inhibitors are better than AIIRAs for the treatment of cardiac failure, and AIIRAs should be reserved for patients who cannot tolerate ACE inhibitors due to side-effects. Closer analysis reveals that the end-point ‘cardiac failure’ was defined by the individual investigator. There was no corroboration of the end-point by an independent adjudicator, and left ventricular ejection fraction was not measured. Many of the investigative centres that took part in ALLHAT worked in a primary care setting, and many had not previously taken part in a study of this kind. For patients in the lisinopril arm, the second-line agent was either atenolol, clonidine, or reserpine, and diuretics could not be used. Thus it seems likely that chlorthalidone was better at reducing the development of ankle oedema than lisinopril, which is perhaps not so surprising. Even if we wish to use a diuretic as the drug of first choice based on a cost/benefit analysis, the majority of diabetic patients will require two or more antihypertensive agents. ACE inhibitors and AIIRAs can both be usefully combined with diuretics, so which drug should be used? For type 2 diabetic patients with microalbuminuria, in the IRMA-2 study irbesartan reduced the development of overt proteinuria.1 In other studies in patients with overt proteinuria irbesartan and losartan reduced the progression to end-stage renal disease or doubling of creatinine,2, 3 but had no significant effect on cardiovascular mortality over a two and a half year period. There are similar but less data available for the use of ACE inhibitors in type 2 diabetes. As mentioned by Perry et al, in the MICRO-HOPE study ramipril reduced the development of microalbuminuria and decreased the risk of overt nephropathy over five years of follow-up.6 More importantly, ramipril significantly reduced myocardial infarction, cardiovascular death, and all-cause mortality in these subjects. So should we be using ramipril in hypertensive diabetic patients who have microalbuminuria for definite cardioprotection and probable renoprotection, or should we be using AIIRAs for definite renoprotection and possible cardioprotection? Microalbuminuria and albuminuria are now routine measurements in people with type 2 diabetes. A routine ECG to look for LVH is performed less frequently. Based on the results of the LIFE study,4 should this now be a routine investigation in the hypertensive diabetic patient to target therapy with losartan? And what if the patient has LVH and microalbuminuria, or microalbuminuria without LVH? Only time and the results of further studies will tell.