LIFE and ARBITER

Summary. LIFE1 was a double-blind, randomized, parallel-group trial in 9193 hypertensive patients with electrocardiographic (ECG) evidence for left ventricular hypertrophy (LVH) with 4-year follow-up to determine whether there was a difference in primary cardiovascular events between an angiotensin II receptor blocker (losartan), and a β blocker (atenolol). A substudy evaluated these differences in diabetics. Results. With a similar reduction in blood pressure, primary events occurred in fewer participants in the losartan group (24/1000 patient-years) than the atenolol group (28/1000 patient-years) (p=0.021). Similar results were found in the diabetic substudy (p=0.03). Comment. The LIFE study was undertaken to determine whether a representative of one of the relatively newer classes of antihypertensive agents could be demonstrated to be more effective than a member of the β blocker class in decreasing morbidity and mortality in a relatively high risk hypertensive group. It was an investigator-initiated study with support from an unrestricted grant from Merck & Co. The use of a β blocker was based on its efficacy in trials with placebo in hypertension and after myocardial infarction. Losartan represents one of the newer classes of antihypertensive agents, similar to angiotensin-converting enzyme (ACE) inhibitors, with potential analogous efficacy based on evidence of reduction in cardiovascular events with ACE inhibitors and further, the evidence for a possibly greater effect in reduction in hypertensive LVH by ACE inhibitors than some other class of agents.2 Additional antihypertensive drugs could be used for blood pressure control, if necessary. The risk in this hypertensive population was heightened by the age range for inclusion (55–80 years) and the presence of ECG LVH. The ECG-LVH criteria used have a sensitivity of 50% compared with echocardiography or necropsy.3,4 Approximately 35% of the entire group had a history of vascular disease, and 24% had a history of coronary artery disease. The presence of diabetes in a subset of participants added to the already high risk for outcome events. A large substudy of this trial included echocardiographic evaluation, which will not be reported here. The study was undertaken in the United States and Scandinavian countries with a 4.7-year follow-up. The primary composite end point evaluated was cardiovascular death, stroke, or myocardial infarction. The results of the main study indicated a significant benefit in the losartan group, with the primary composite end point reached in 23.8% of the losartan group and 27.9% of the atenolol group (Figure 1).1 Adjusted for the degree of LVH at baseline and Framingham risk score at randomization, the hazard ratio for the losartan group was 0.87 (95% confidence interval [CI], 0.77–0.98; p=0.021). In a follow-up response to a letter to the editor, the authors reported event-free survival in 85.5% of the losartan group and 83.6% of the atenolol group (p=0.04).5 Blood pressure, initially 174/98 mm Hg in both groups, was equally reduced (144/81 mm Hg and 145/81 mm Hg at last visit). Analysis for specific primary events showed a benefit for losartan for stroke (adjusted risk reduction, 25%; p=0.001) but no significant difference in myocardial infarction. The losartan group had an 11% reduction in cardiovascular mortality but this was not statistically significant (p=0.2). Quantitative ECG-LVH evaluation showed a significantly greater reduction in LVH criteria in the losartan group. Adverse events causing dropout from the study were more common in the atenolol group (18% vs. 13%). New onset diabetes developed in 6% of the losartan group vs. 8% of the atenolol group (p=0.001). Kaplan Meier curves for the primary composite end point in the main Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) trial (cardiovascular mortality, stroke, and myocardial infarction). The losartan group had significantly fewer first events. Adjusted risk reduction accounted for Framingham risk score and electrocardiographic left ventricular hypertrophy. Reprinted with permission of Elsevier Science, from Lancet. 2002;359:995–1003.1 In the diabetic subgroup of 1195 patients, blood pressure fell similarly in both groups.6 The primary composite end point for cardiovascular events was achieved in 24.5% fewer of the losartan group (adjusted hazard ratio, 0.76 [95% CI, 0.58%–0.98%]) (Figure 2). For specific primary outcome events, no differences were found for myocardial infarction or stroke, although the losartan group trended toward fewer events. However, cardiovascular mortality was significantly lower in the atenolol group by 37% (p=0.028), as was total mortality. Event-free survival was 79% in the losartan group and 72% in the atenolol group.5 Quantitative ECG-LVH decreased and in general, adverse effects were similar to results of the main study group. Serum creatinine increased by 24%–26% in both groups. Albuminuria was reported more commonly in the atenolol group (13% vs. 7%; p=0.002). Kaplan Meier curves for the primary composite end point in the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) diabetes substudy. Composite end point and adjusted risk reduction as in Figure 1. The losartan group had significantly fewer first events; risk reduction was greater than for the entire study group, although the proportion of patients with first events was 20% in the diabetic patients compared with 12% in the entire study group. Reprinted with permission of Elsevier Science, from Lancet. 2002;359:1004–1010.6 The overall study results differ from those of other reported studies between antihypertensive drug classes that showed no differences in outcome events between treatment groups.7,8 The results cannot be explained by blood pressure control, which was similar in both groups. Reduction in ECG-LVH was greater in the losartan group but this did not affect the outcome of fatal or nonfatal myocardial infarction. As indicated, there were some differences in outcomes in the diabetics vs. the entire group, the main difference being the significantly lower cardiovascular mortality in the losartan group among the diabetics. It is possible that a relatively greater difference in ECG-LVH reduction with losartan in the diabetics vs. the entire group could have influenced this result. There was virtually no change from baseline in the most sensitive ECG indicator of LVH, the Cornell voltage criteria, in the atenolol group, but there was an 8% decrease in the losartan group. However, the possibility remains that angiotensin II-inhibiting effects, independent of effects on blood pressure or regression of LVH, may have contributed to the beneficial results in the losartan group. One of the concerns of intention-to-treat analysis is the possible effects of crossovers between study groups, removal of study drug, or additions of other agents designed to affect outcomes, in this case, blood pressure reduction. In the losartan group, 23% were off study drug at end point or end of follow-up, as were 27% of the atenolol group. Further, another 66% of the losartan group and 62% of the atenolol group received additional antihypertensive agents, mostly hydrochlorothiazide. This left approximately 11% in each group on study drug alone during the course of the entire study. In the diabetic group, this raises an important question since the combination of an angiotensin II blocker with a diuretic has been shown to produce more beneficial effects on renal function than a β blocker and a diuretic. Another question is raised by the determination of drug efficacy. For example, in both groups, efforts were made to lower blood pressure to a normal target level. However, the beneficial effects of heart rate reduction by a β blocker are optimal at 60 beats/min. The β blocker group averaged 66 beats/min, indicating suboptimal dosing. As for blood pressure reduction, the target of <140/90 mm Hg was achieved in only 49% of the losartan patients and 46% of the atenolol patients. Finally, the study group was diverse in inclusion of participants with and without previous cardiovascular events, which may mitigate comparison of results with other studies confined to more homogeneous groups. The results of this study indicate that the angiotensin II blocker losartan is at least as beneficial as a β blocker in modulation of cardiovascularevents in a high-risk, hypertensive, upper middle age–older age, white population of diverse geographic area and nationalities. Summary. ARBITER9 was a single-center, randomized clinical trial of 161 mostly male participants, approximately one half with known cardiovascular disease, to determine the efficacy of two statin agents of different potency (atorvastatin and pravastatin) in affecting carotid atherosclerosis over 1 year. Results. The results indicated that marked lowdensity lipoprotein cholesterol (LDL-C) reduction was associated with atherosclerosis regression measured by carotid intimal-medial thickness (IMT), but moderate reduction of LDL showed no change in carotid IMT. Comment. The purpose of this study was to evaluate the effects of two statins of different potencies on LDL-C lowering but also different properties possibly influencing pleiotropic effects. Although there are many clinical end points for atherosclerosis, there is only one validated surrogate noninvasive end point—change in carotid artery IMT. It is possible that lowering LDL-C to moderate levels, i.e., between 100–130 mg/dL, may be as efficacious as more aggressive LDL-C lowering because of pleiotropic effects that outweigh lipid-lowering effects per se. Furthermore, there is still considerable controversy as to whether lowering LDL-C to <100 mg/dL would have more significant clinical outcome effects. The study methodology was straightforward. In a single center, the 161 participants were recruited because they met National Cholesterol Education Program II guidelines for lipid-lowering therapy intervention. The study participants were mostly men (71.4%) and late middle-aged (mean, 60 years). A total of 79 received atorvastatin, 80 mg/d, and 82 received pravastatin, 40 mg/d. The study lasted for 1 year. Lipid levels and IMT of the distal common carotid artery far wall were evaluated at baseline, 6 months, and 1 year. The baseline level of the lipid profile for the entire group was: total cholesterol 232 mg/dL, LDL-C 152 mg/dL, high-density lipoprotein cholesterol (HDL-C) 49 mg/dL, and triglycerides 207 mg/dL. There were no significant differences in baseline values between groups. The reduction in lipid values, as expected, was greater in the atorvastatin group (40 mg pravastatin has the potency of 10 mg atorvastatin). By 12 months, total cholesterol had decreased from 229 to 146 mg/dL in the atorvastatin group and from 234 to 181 mg/dL in the pravastatin group. This was produced primarily by a decrease of LDL-C (atorvastatin 148 to 76 mg/dL; pravastatin 155 to 110 mg/dL). These changes were reached by 3 months. HDL-C values were virtually unchanged and triglycerides decreased by 40% in the atorvastatin group and by 10% in the pravastatin group. Of other serologic results, C-reactive protein decreased only in the atorvastatin group. Fibrinogen values did not change significantly in either group. The major effects were a significant decrease in mean carotid IMT in the atorvastatin group with a slight increase in the pravastatin group at 1 year (virtually equal to baseline values). However, interestingly, there was a similar decrease in carotid IMT in both groups at 6 months, which continued to decrease by a similar increment at 1 year in the atorvastatin group while the carotid IMT returned to baseline in the pravastatin group. The net regression of mean carotid IMT with atorvastatin was –0.038 mm compared with no significant change in the pravastatin group (0.026 mm; p=0.03). In terms of individual participants, adecrease in mean carotid IMT was seen in 54% of atorvastatin patients and 39% of pravastatin patients at 1 year. The beneficial effect of this result in the atorvastatin group was of borderline significance (p=0.06). There was a 4% incidence of coronary events, divided equally between groups, and no significant liver enzyme changes or myositis. This study is the latest in a series evaluating the effects of moderate vs. aggressive lipid-lowering therapy on outcome events (Table). These studies indicated a benefit of aggressive lipid-lowering therapy over moderate intervention. The Post Coronary Artery Bypass Graft (Post-CABG) trial10,11 demonstrated that aggressive lipid lowering conferred incremental benefit in quantitative coronary atherosclerosis progression in both the grafts and native coronary arteries. The Atorvastatin vs. Revascularization Treatment (AVERT) trial12 showed a borderline benefit in reduced ischemic events. The Agressive vs. Conventional Lipid Lowering on Atherosclerosis Progression in Familial Hypercholesterolemia (ASAP) trial13 reported regression of carotid IMT in patients with familial hypercholesterolemia. The Heart Protection Study (HPS)14 revealed that lowering LDL-C from less than 116 mg/dL to below 77 mg/dL reduced vascular disease risk by about 25%. These findings suggest that there is as yet no clear threshold below which reduction of LDL-C is ineffective. For example, in the HPS,14 reduction of LDL-C from 97 to 65 mg/dL produced a similar reduction in risk as in those with initially higher LDL-C levels. The results of the ARBITER trial cannot be confidently dissected into pleiotropic vs. lipid-lowering effects. Regression analysis between LDL-C lowering and change in carotid IMT was not indicated but if the correlation were high, it would suggest that the lipid-lowering effects were paramount. As noted, despite the marked disparity in drug potencies with regard to lipid lowering, IMT regression was seen in 39% of the patients in the pravastatin group. This could suggest that pleiotropic changes might have contributed to the results. Since atorvastatin is a synthetic statin as opposed to pravastatin, there are possible differences in activity that could influence pleiotropic actions. An additional experimental method, which was not assessed, would have been determination of a doseresponse effect by utilizing graded statin doses. The results of this study indicate that there are beneficial effects of more potent statins but that, on an individual basis, regression of IMT is still seen in a substantial proportion of participants despite lower statin potency. This study adds to the growing body of evidence for the benefit of statins in atherosclerotic processes. It also serves to stimulate further investigation to dissect the pleiotropic from the lipid-lowering effects of statins.