doi: 10.1161/CIRCULATIONAHA.105.588020
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2006;113:2754-2774.)
© 2006 American Heart Association, Inc.
Controversies in Cardiovascular Medicine |
Management of hypertension: is it the pressure or the drug?
Blood Pressure Reduction Is Not the Only Determinant of Outcome
From Imperial College, London, UK (P.S.S., N.R.P.), and the Department of Preventive Medicine, RUSH Medical College of RUSH University at RUSH University Medical Center, Chicago, Ill (W.J.E., M.C.J., H.R.B.).
Correspondence to Professor Peter Sever, International Centre for Circulatory Health, Imperial College London, 59 N. Wharf Rd, London, W2 1PG, UK (e-mail p.sever{at}imperial.ac.uk), or William J. Elliott, MD, PhD, 1700 W. Van Buren, Suite 470, Chicago, IL 60612 (e-mail welliott@rush.edu).
 |
Introduction |
|---|
Top
Introduction
The Early Placebo-Controlled...Whether certain classes of antihypertensive drugs confer benefits beyond those associated with lowering blood pressure remains a highly controversial issue. Data from several meta-analyses have been used to support the notion that most, if not all, of the cardiovascular benefits reported with the use of different classes of antihypertensive drugs are simply a consequence of the extent to which they lower blood pressure. However, we submit evidence in this review that the diverse pharmacological actions of several antihypertensive medications may have benefits beyond their blood pressure–lowering effects and that, in the case of certain classes of drugs, notably ß-blockers, adverse metabolic effects of these drugs may actually mitigate the potential benefits of blood pressure lowering.
|
The Early Placebo-Controlled Hypertension Trials and the Shortfall in Coronary Heart Disease Prevention |
|---|
The early placebo-controlled trials of the treatment of hypertension, several of which were undertaken in high-risk patient populations, provided convincing evidence for substantial reductions in the risk of stroke but little or no evidence for benefits on coronary heart disease (CHD) events.1–6 However, the design, numbers of patients recruited, and event rates in individual trials provided inadequate power to evaluate the impact of treatment on CHD events. In only 1 trial, the Hypertension Detection and Follow-up Programme (HDFP),7 was a reduction in CHD events observed in those assigned "special care" compared with those assigned "usual care." The conduct of this particular trial, however, differed from the other early trials in that those in the special care group would have been likely to benefit from more comprehensive intervention on other cardiovascular risk factors.
In the first meta-analysis of the placebo-controlled trials of antihypertensive drug therapy,8 HDFP was excluded, presumably because the authors of the meta-analysis considered that outcome benefits could have occurred independently of blood pressure reduction. In subsequent meta-analyses, however, for reasons that are unclear, HDFP was included.9
From observational studies10,11 (Figure 1), it was possible to estimate the potential cardiovascular risk reduction associated with a blood pressure difference of 10– to 12–mm Hg systolic and 5– to 6–mm Hg diastolic pressure, the average reduction in blood pressure observed in the early trials. In the case of stroke, the relative risk reduction of 42% observed in the trials was compatible with the 35% to 40% difference associated in prospective observational studies, with a long-term difference of 5 to 6 mm Hg in diastolic pressures. In the case of CHD, however, the observed risk reduction of 14% to 16% (9% if HDFP is excluded) fell short of the 20% to 25% risk difference predicted from observational data for a similar difference in blood pressure. This apparent shortfall could have represented the play of chance because the upper 95% confidence limit for this significant CHD reduction in these trials was 22%. Alternatively, these observations could reflect a genuine shortfall in the protective effects against CHD events of what were exclusively older antihypertensive therapies (diuretics and ß-blockers).
|
Observations from 2 trials conducted by the Medical Research Council in the United Kingdom suggested that treatments based on diuretics or ß-blockers might confer different degrees of protection against stroke and CHD.3,12 Notably, in the Medical Research Council trial in older subjects with hypertension,12 protection against CHD events was observed with diuretic-based treatment but not with ß-blocker–based treatment. More recently, meta-analyses of ß-blocker–based trials in hypertension have suggested that this class of agent confers less reduction in cardiovascular risk than other classes of antihypertensive drugs,13 particularly in those without prior evidence of cardiovascular disease. In contrast, when ß-blockers have been assessed in long-term trials after myocardial infarction, allocation to active treatment or control resulted in differences of only a few millimeters of mercury (as little as 1 to 2 mm Hg in some individual trials).14 The average reduction in recurrent CHD events of 26% was too large to be attributed to this minor degree of blood pressure reduction and was perhaps indicative of pharmacological benefit independent of blood pressure lowering. These observations also highlight the potential for drug-induced benefits on cardiovascular events to be dependent on the patient subgroup.
|
Trials With Calcium Channel Blockers and Drugs That Block the Renin-Angiotensin System in High-Risk Patients |
|---|
Placebo Controlled
Because of the cardiovascular benefits observed in these early placebo-controlled trials of blood pressure reduction, it became increasingly difficult to repeat placebo-controlled studies in hypertensive patients using newer drugs such as angiotensin-converting enzyme inhibitors (ACEIs) and calcium channel blockers (CCB), and with the exception of the Systolic Hypertension in Europe Trial (SYST-EUR),15 which compared nitrendipine with placebo in patients with isolated systolic hypertension, other trials comparing ACEI or CCB with placebo were carried out in very different patient populations that were not recruited on the basis of hypertension but included those with established cardiovascular or renal disease.16,17 Therefore, unlike many of the early trials with diuretics and ß-blockers, these trials did not assess reduction in CHD or stroke risk in primary prevention in hypertensive subjects. The risk reductions of 30% to 40% in stroke and
20% in CHD observed in placebo-controlled trials with ACEIs or CCBs in high-risk patients were associated with 5/2–mm Hg (ACEI) and 8/4–mm Hg (CCB) differences in blood pressure16,17 (Figure 2). These observations suggest that considerably greater risk reductions occur for a given difference in blood pressure than would have been predicted from the observational data (albeit in lower-risk populations).
|
In the Heart Outcomes Prevention Evaluation (HOPE),18 9000 patients, who were on average 66 years of age and had evidence of vascular disease or diabetes, were randomly assigned to ramipril 10 mg daily or placebo for a mean of 5 years. In this study, the primary outcome was a composite of myocardial infraction, stroke, or death from cardiovascular causes. This was reduced by 22% in favor of ramipril, together with a 20% risk reduction in myocardial infarction. The reported blood pressure reduction of 3/2 mm Hg was proposed by the authors to account for no more than one quarter of the reduction in the rates of myocardial infraction. However, in HOPE, owing to administration of the short-acting ACEI ramipril at night and the measurement of blood pressure the following day, the reported differences in blood pressure between the active and placebo treatment arms may have been underestimated.
In the European Trial on Reduction of Cardiac Events with Perindopril in Stable Coronary Artery Disease (EUROPA),19 >13 000 patients with previous coronary disease, were randomized to perindopril 8 mg daily or placebo. Follow-up was for 4.2 years, with a primary end point of cardiovascular death, nonfatal myocardial infarction, or cardiac arrest. The average age of patients was 60 years, and most were on concomitant ß-blockers and lipid-lowering therapy. In those assigned perindopril, there was a highly significant 20% reduction in the primary end point and a 22% reduction in nonfatal myocardial infraction. Although blood pressure was on average 5/2 mm Hg lower in those assigned perindopril, similar proportional risk reduction was seen in those who were not hypertensive at baseline compared with those who were, and in a post hoc analysis, similar risk reductions were observed in those in whom the ACEI had little or no effect on blood pressure, compatible with the hypothesis that in this particular patient group, some of the benefits observed would be independent of blood pressure.
Head-to-Head Comparisons of Active Treatments: Limitations of Trial Design
After the introduction of the ACEI and CCBs, several head-to-head studies were conducted comparing older treatments (diuretic or ß-blocker) with either ACEI- or CCB-based treatment.16,17 Most of the studies were underpowered to detect potential differences in CHD event rates and indeed failed to do so. Further prospective meta-analyses were conducted in an attempt to determine whether any particular drug class conferred advantages over the older drugs, which were a mixture of diuretic-based, ß-blocker–based, or diuretic- and ß-blocker–based strategies.16,17
Thus, any potential advantage or disadvantage of either of these drug classes could be masked. Additionally, in the case of diuretic-based treatment, there has been little attempt to assess whether long-term outcome is influenced by the dose of agent used, leading to widespread assumptions that "low-dose" diuretic would be equivalent to the older moderate- and even high-dose diuretics used in the earlier trials.
In the Losartan Intervention For Endpoint reduction in hypertension (LIFE) trial,20 in which the angiotensin receptor blocker losartan was compared with the ß-blocker atenolol, thiazide diuretics were added to each treatment arm in most patients. Pressures throughout the trial were apparently similar, although no detailed analysis of mean blood pressures for the 2 treatment limbs throughout the trial has been published. There were no significant differences in CHD outcome, but significant differences in stroke were seen in favor of losartan, which seem disproportionate to the negligible differences in blood pressure. In light of other studies with ß-blockers and subsequently the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) results, it seems likely that this differential outcome on stroke is attributed largely to the inefficacy of the ß-blocker–based treatment strategy and potentially explained by lower central aortic pressures21 with the losartan-based treatment than with the atenolol-based treatment for equivalent peripheral brachial artery blood pressures.
|
Recent Meta-Analyses |
|---|
In an effort to maintain high event rates and thereby to reduce trial costs, most of the more recent hypertension studies were carried out in much-higher-risk patient groups than earlier trials. The conclusion from pooled analyses of these trials comparing older and newer drugs was that there were marginal benefits of diuretics and/or ß-blockers versus ACEIs and of CCBs versus ACEIs on stroke outcome17 (Figure 3). In the case of CHD end points, hazard ratios approximated unity when ACEIs or CCBs were compared with diuretic- and/or ß-blocker–based regimens (Figure 3). Similarly, neither the composite end point of total cardiovascular events nor cardiovascular mortality differed significantly between newer and older treatment regimens. In these pooled analyses, overall mean systolic blood pressure differences between treatment groups during follow-up were between 1 and 2 mm Hg.
|
In the most recent meta-analysis by Verdecchia and colleagues22 (Figures 4 and 5
), when outcome was plotted against blood pressure differences, benefits on stroke events were greater than expected in trials with CCBs and benefits on CHD outcome were beyond those expected from the differences in blood pressure in the case of ACEI trials. These observations would, in general, support the favorable evidence on stroke outcome with CCBs reported in earlier meta-analyses and the benefits of ACEIs demonstrated in post–myocardial infarction trials in which the relative risk reduction of 25% is more than that expected from the observed differences in blood pressure.
|
|
|
Most Recent Data From Head-to-Head Trials: ALLHAT, VALUE, and ASCOT |
|---|
Data from large individual trials may provide more insight into the benefits or harm of particular drug treatment strategies than information provided by meta-analyses in which the inappropriate inclusion of individual trials may distort the overall conclusions, eg, the inclusion of HDFP in the Collins and Peto9 meta-analysis and in more recent meta-analyses the inclusion of the CAPtopril Prevention Project (CAPPP),23 in which randomization was biased and results likely were confounded by 24-hour blood pressure differences.
Three large trials, Antihypertensive and Lipid Lowering treatment to prevent Heart Attack Trial (ALLHAT),24 Valsartan Antihypertensive Long term Use Evaluation trial (VALUE),25 and ASCOT,26 focused on CHD as the primary end point comparing various treatments. In both ALLHAT and VALUE, second-line drugs added to the "newer" first-line treatments and taken by most subjects were invariably "older" drugs such as ß-blockers, clonidine and reserpine (ALLHAT), and thiazide diuretics (VALUE). In ASCOT, the newer treatment regimen of a CCB with or without an ACEI was not contaminated by either ß-blockers or thiazide diuretics and therefore represents the only true comparison of new versus old combinations of treatment.
ALLHAT24 randomized >40 000 patients in a double-blind study comparing 3 first-line agents: the CCB amlodipine, the ACEI lisinopril, and the 
-blocker doxazosin with the diuretic chlorthalidone as the reference drug. The -blocker limb of the study was terminated early because of an excess of certain cardiovascular end points in the doxazosin arm compared with chlorthalidone. First-line drugs were titrated monthly to achieve a target blood pressure of <140/90 mm Hg (amlodipine 2.5 to 10 mg daily, lisinopril 10 to 40 mg daily, chlorthalidone 12.5 to 25 mg daily). If goal blood pressures were not achieved, step 2 medications, including reserpine (0.05 to 0.2 mg daily), clonidine (0.1 to 0.3 mg BID), or atenolol (25 to 100 mg daily), were added. Step 3 medications included hydralazine (25 to 100 mg BID). Hypertensive subjects recruited into ALLHAT were an average of 67 years of age, 47% were female, and 32% were black. Patients were at moderately high risk (2% CHD risk per year). The average length of follow-up was 4.9 years. Blood pressure levels at baseline were evenly matched falling from 146/84 mm Hg in all 4 groups to 133.9/75.4 mm Hg (chlorthalidone), 134.7/74.6 mm Hg (amlodipine), 135.9/75.4 mm Hg (lisinopril), and 137.4/76.6 mm Hg (doxazosin). Hence, compared with the chlorthalidone group, the mean follow-up systolic blood pressure was 2 mm Hg higher in the lisinopril group, 1 mm Hg higher in the amlodipine group, and 3 mm Hg higher in the doxazosin group. Differences in blood pressure between the treatment limbs were greatest during the first 2 years of follow-up, after which dose titration and the addition of second- and third-line therapy reduced these differences.
Despite these mean blood pressure differences, CHD outcomes (the primary end point) were not different among those in the 4 comparator drug groups. There was a 15% excess of stroke in the lisinopril arm compared with chlorthalidone and a 26% excess of stroke when the doxazosin and chlorthalidone groups were compared. Notably, in blacks, there was a 40% excess of stroke in those assigned lisinopril compared with chlorthalidone. The ALLHAT authors concluded that these differences in stroke outcome could not be explained by differences in blood pressure. However, this depends on which reference group is chosen for the derivation of "expected outcome" to compare with that observed. The ALLHAT authors based their expected outcome on the original observational studies referred to previously,10,11 not the more recent intervention trials involving high-risk patients in whom smaller differences in blood pressure have been associated with larger differences in outcome. Does ALLHAT provide evidence for benefits beyond blood pressure? Indirectly, it does in that almost identical rates of coronary events occurred among the 4 blood pressure drug classes evaluated despite different degrees of blood pressure reduction, raising the intriguing possibility that had blood pressure levels been equivalent throughout the trial in the ACEI, CCB, and doxazosin and chlorthalidone arms, would the newer treatments have conferred greater protection against CHD events?
The VALUE investigators25 designed a study also focused on CHD end points, but in this case, on a composite end point that included revascularization procedures and hospital-based heart failure, in addition to nonfatal myocardial infarction and fatal CHD. More than 15 000 patients were randomized in a double-blind fashion to either the angiotensin receptor blocker valsartan (80 to 160 mg daily) or the CCB amlodipine (5 to 10 mg daily). Both arms had the diuretic hydrochlorthiazide (12.5 to 25 mg daily) added in an attempt to achieve target blood pressures of <140/90 mm Hg. Further add-on drugs were used as needed to achieve target blood pressure at the discretion of the investigator. VALUE recruited patients of an average age of 67 years, 42% of whom were female. The patient population was at high cardiovascular risk, most with established coronary, cerebral, or other arterial disease. Blood pressure fell from 154.5/87.4 to 139.3/79.2 mm Hg with valsartan-based regimens and from 154.8/87.6 to 137.5/77.7 mm Hg with amlodipine-based regimens. Blood pressure reductions from baseline until the study end were 15.2/8.2 and 17.3/9.9 mm Hg in the valsartan and amlodipine arms, respectively. Differences between the treatment arms were again greatest in the first year of treatment, maximally 4/2 mm Hg in favor of the amlodipine-based regimen, and on average throughout the trial differed by 3/2 mm Hg.
There was a 3% nonsignificant difference between the 2 treatment arms in the primary composite cardiac end point favoring the amlodipine-based limb and a 15% nonsignificant excess of strokes in the valsartan-based regimen and a significant 19% excess of fatal and nonfatal myocardial infarction in those receiving valsartan. The authors of VALUE pursued their belief that blood pressure differences in this trial accounted for the differences in outcome by a time-dependent analysis (Figures 6 and 7) of CHD and stroke events throughout the trial that showed that the excess of stroke and CHD events in the valsartan group was maximal during the first year when blood pressure differences were greatest and that the differences between the arms diminished with time as the blood pressure curves came together. There are problems, however, with this time-dependent analysis because separating out the first year’s data and reanalyzing subsequent time intervals lead to a loss of randomization. In this trial, it is difficult to conclude that any benefits were clearly independent of blood pressure reduction with respect to CHD events, although in terms of new-onset diabetes, blood pressure–independent benefits of valsartan were apparent.
|
|
SBP) between treatment groups with odds ratios for stroke and 95% CIs during consecutive time periods in the study.
ASCOT26 recruited 20 000 patients 40 to 79 years of age who had either untreated or previously treated hypertension who were then randomized to receive either amlodipine (5 to 10 mg daily) or atenolol (50 to 100 mg daily). After dose titration, the second-line drugs perindopril (4 to 8 mg daily) and bendroflumethiazide-K (1.25 to 2.5 mg daily), respectively, were added as required to achieve goal blood pressures of <140/90 or <130/80 mm Hg in those with diabetes. Thereafter, the third-line drug doxazosin-GITS (4 to 8 mg daily) and other drugs were added to either drug regimen as required to achieve target blood pressures. By way of a factorial design, >10 000 patients with total cholesterol levels of 
250 mg/dL were randomized to atorvastatin 10 mg or placebo.27
The patient population recruited into ASCOT differed substantially from those recruited into several other recently reported hypertension trials in that a history of prior myocardial infarction or current CHD excluded patients from participation. Although 3 additional cardiovascular risk factors were required for entry, the overall risk of the ASCOT patient population was low (<1% per annum CHD event rate) and much less than among those recruited into ALLHAT and VALUE. The average age of the patients was 63 years, 23% were female, and 95% were white.
Blood pressure levels fell from 163.9/94.5 to 137.7/79.2 mm Hg in the atenolol-based treatment arm and from 164.1/94.8 to 136.1/77.4 mm Hg in the amlodipine-based treatment arm. Again, better blood pressure lowering in favor of the CCB treatment regimen was seen particularly early in the trial. Overall, blood pressure (integrated mean) was 2.7/1.9 mm Hg lower on the amlodipine-based regimen than the atenolol-based regimen, with maximal differences of 5 mm Hg systolic in the first year but only 1.6 mm Hg systolic by the end of the trial.
Anglo-Scandinavian Cardiac Outcomes Trial–Blood Pressure Lowering Arm (ASCOT-BPLA) was stopped prematurely because of significant all-cause mortality differences between the 2 treatment arms and concerns that those patients assigned the ß-blocker/thiazide regimen would continue to be disadvantaged if the trial went to its planned completion.
All-cause mortality and cardiovascular mortality were reduced significantly (11% and 24%, respectively) among those allocated to the amlodipine/perindopril regimen. The primary end point (nonfatal myocardial infarction and fatal CHD) was reduced by 10%, but this did not achieve statistical significance. Other prespecified coronary end points, including the primary end point and excluding silent myocardial infarction, and a composite total coronary end point, however, were significantly reduced (13% and 13%, respectively) among those allocated to the amlodipine/perindopril regimen, as were stroke events (23%).
These observations raised the question as to what extent the blood pressure differences, which occurred predominantly early in the trial, explained the differences in cardiovascular events seen in the 2 blood pressure arms of the study.
The observed blood pressure difference of <3/2 mm Hg seen in ASCOT-BPLA might explain an 4% to 8% reduction in coronary outcome and an 8% to 14% reduction in strokes based on prospective observational studies10,11 and the most recent pooled analysis of clinical trials reported by the Blood Pressure Lowering Treatment Trialists Collaboration.17
Correcting for blood pressure differences in randomized trials, however, is problematic, and there is no ideal way to do so. Nevertheless, further analyses using in-trial data were undertaken28 in an attempt to ascertain to what extent the beneficial effect of the amlodipine/perindopril regimen could have been explained by the differences in blood pressure and the other variables that occurred after randomization.
First, analyses were performed to evaluate any temporal association between blood pressure differences and coronary and stroke end points using differing censoring points throughout the trial (Figure 8). These analyses were then extended using a technique similar to but more rigorous than the serial median matching carried out in the VALUE trial analyses. It was clear in the ASCOT analyses that for coronary and stroke end points, there was no apparent temporal link between the size of blood pressure differences and the difference in end points between the amlodipine/perindopril regimen and the atenolol/thiazide regimen.
|
In addition to blood pressure differences, assignment to the atenolol/thiazide regimen was associated with significant metabolic differences compared with the amlodipine/perindopril regimen. Although there were no differences in low-density lipoprotein cholesterol between the 2 limbs (which would not have been expected), serum high-density lipoprotein (HDL) cholesterol, pulse rate, and potassium were lower, and body weight, serum triglycerides, fasting glucose, and creatinine were higher among those on the ß-blocker/thiazide regimen. All of these parameters have previously been reported in association with ß-blockers and diuretics, and the possibility arose that these adverse metabolic changes could have contributed to differences between the 2 arms of the trial, given that all have been implicated as independent cardiovascular risk factors.
Further analyses were undertaken using updated Cox regression techniques to provide additional information on the role that differences in various measures of blood pressure and serum HDL cholesterol, triglycerides, potassium and creatinine, body weight, pulse rate, and blood glucose could have played in explaining the differential risk reductions observed in ASCOT-BPLA. In these analyses, which should be interpreted cautiously, it appears that blood pressure differences contributed in a minor way to the risk reduction in coronary events but contributed more to stroke differences.28 For stroke, blood pressure appeared to contribute 30% of the benefits of the amlodipine/perindopril regimen, and for coronary events, HDL cholesterol accounted for 30% of event rate differences. Overall, in multivariate analyses, differences in blood pressure and the other variables considered accounted for about half the differences in coronary events and 40% of stroke events (Figures 9 and 10).
|
|
The possibility that other factors could have contributed to the outcome differences seen in ASCOT-BPLA has arisen from 2 other sources of information within the trial.
First, an investigation of the differential impact of the amlodipine/perindopril and atenolol/thiazide regimens on central aortic pressure and clinical outcomes (the Conduit Artery Function Evaluation [CAFE] Study) was a substudy of ASCOT, and the results may shed some light on the mechanisms contributing to the differences in cardiovascular outcomes observed between the 2 treatment limbs in ASCOT-BPLA.21
In CAFE, >2000 ASCOT patients had observations on radial artery applanation tonometry and pulse-wave analysis using the Sphygmocor device, and central aortic pressures and hemodynamic indexes were derived on several occasions during the 5 years of the trial. The key results of this substudy were that despite almost identical brachial artery pressures in the 2 blood pressure limbs of the trial, there were substantial reductions in central aortic pressures and other hemodynamic indexes in favor of the amlodipine/perindopril regimen. Central aortic systolic pressure and central aortic pulse pressure were 4 and 3 mm Hg lower, respectively, on the amlodipine/perindopril regimen. In addition, the augmentation index was 6% lower in the amlodipine/perindopril regimen. This study confirms other reports29–31 that ß-blockers lower central aortic pressure to a lesser extent than other drugs for an equivalent reduction in peripheral arterial pressure. In this substudy, there was a significant relationship between central pulse pressure and a composite end point of cardiovascular and renal events. The authors of CAFE hypothesized that some of the differences in outcome in ASCOT-BPLA could be explained by differences in central aortic pressures. Another important conclusion from this study was that the reduced beneficial effect of atenolol-based treatment on central arterial hemodynamics was dependent on heart rate slowing; thus, the observations could be extrapolated to other ß-blockers and may not be specific to atenolol-based treatment.
Second, a possible further mechanism underlying the observed differences in cardiovascular end points between the 2 treatment arms in ASCOT-BPLA has come about as a result of the prespecified evaluation of any synergy between the use of lipid-lowering therapy and antihypertensive treatments.22 Experimental studies and at least 1 clinical study have indicated the potential for synergy between amlodipine and statins on cardiovascular outcomes.32,33 The physicochemical properties of atorvastatin and the dihydropyridine CCB amlodipine, by virtue of their lipophilicity and oppositely charged molecules, lend themselves to tight bonding in the lipid bilayer of cell membranes and hence the potential for prolonged actions on a number of molecular and cellular processes involved in the biology of atherosclerosis in the vessel wall (R.P. Mason and R. Kay, personal communication).
Overall, when both blood pressure treatments in ASCOT-LLA are combined, CHD events (nonfatal myocardial infarction and fatal CHD) were significantly lower in the atorvastatin group than in the placebo group (36%; P=0.005). However, in those assigned the amlodipine/perindopril regimen, CHD events were reduced by 53% (P=0.0001) in association with allocation to atorvastatin compared with placebo, whereas in those assigned the atenolol/thiazide regimen, the risk reduction associated with allocation to atorvastatin compared with placebo was 16% (P=NS). Formal testing for interaction between lipid-lowering and blood pressure–lowering therapy was of borderline statistical significance for a tertiary end point (P=0.025).34
In our evaluation of the extent to which the observed differences in the 2 blood pressure–lowering strategies in ASCOT-BPLA could be explained by differences in blood pressure and other risk factors that were differentially affected after randomization,28 we concluded that it remained possible that additional mechanisms could have contributed to the event rate differences. The most recent analyses indicate that there may be synergy between the amlodipine/perindopril regimen and lipid lowering with atorvastatin and that this could be one such additional mechanism.
|
New-Onset Diabetes |
|---|
In ALLHAT24 and ASCOT,26 important differences were observed in the extent to which different antihypertensive treatment strategies were associated with new-onset diabetes. In seems clear from these 2 studies and earlier observations20,35 that ß-blocker and thiazide drugs, particularly when used in combination, increase the incidence of new-onset diabetes compared with strategies based on CCBs or drugs that block the renin-angiotensin system. In VALUE,25 in which the thiazide diuretic was added to either a CCB or an angiotensin receptor blocker, new-onset diabetes also occurred less frequently in those originally assigned the angiotensin receptor blocker. Although the duration of new-onset diabetes observed in these trials was short and thus unlikely to be associated with excess cardiovascular risk, it seems probable in the longer term that cardiovascular event rates would be higher in those with new-onset diabetes, a phenomenon clearly independent of drug effects on blood pressure.
|
Conclusions |
|---|
Placebo-controlled trials in hypertensive patients of diuretics and ß-blockers conferred less-than-expected protection against CHD events than would have been predicted from observational studies. When similar placebo-controlled trials were carried out with newer agents such as the CCBs in hypertensive patients, benefits on stroke and CHD were compatible with the reductions in blood pressure seen in the trials. However, when placebo-controlled trials have been carried out, notably in patients with ACEIs, in patients recruited on the basis of high cardiovascular risk rather than high blood pressure, the benefits observed appeared greater than would be expected from the differences in blood pressure.
When ß-blockers are compared with other classes of drugs, they are less effective in preventing strokes in hypertensive subjects and are probably less effective than diuretics and other classes of drugs, including ACEIs and CCBs, in the primary prevention of CHD for equivalent blood pressure reduction. The disadvantages of ß-blocker–based treatment may be partly explained by the hemodynamic effects of ß-blockers on central aortic pressure. In addition, adverse metabolic effects of ß-blockers, notably on HDL cholesterol, also may contribute to differential effects on CHD outcome as seen in ASCOT. By way of contrast, in the secondary prevention of CHD, ß-blockers appear to prevent recurrent CHD to a greater extent than blood pressure reduction would predict.
For ACEIs, their poorer outcome in some trials such as CAPPP and ALLHAT on stroke events is compatible with and hence best explained by blood pressure differences. Recent meta-analyses of trials with ACEIs and CCBs involving high-risk patients seemingly provide evidence of greater cardiovascular risk reductions than expected on the basis of observational studies and the earlier placebo-controlled trials of antihypertensive therapy. These results suggest that at least some of the benefits are independent of blood pressure reduction. Although this article has focused on CHD and stroke outcomes, for other cardiovascular end points such as heart failure and renal end points, benefits of particularly ACEIs are largely independent of their effects on blood pressure.
We therefore believe that there is strong evidence to support the view that some of the cardiovascular benefits of antihypertensive agents arise from properties beyond blood pressure lowering as measured conventionally in the clinic and may be differential among particular subgroups of patients.
We accept that evidence to the contrary is extensive36 and that this issue will remain controversial. However, standing back from either the meta-analyses of potentially heterogeneous data or specific findings cherry-picked from various trials, we should consider whether it is likely that all antihypertensive agents would have an equal impact on all cardiovascular outcomes for the same degree of lowering clinic blood pressures.
Given the established multifactorial origin of cardiovascular outcomes and that different antihypertensive agents have differential impacts on many of the established cardiovascular risk factors, it would be extraordinary if all these agents were "equal" once brachial artery pressures were standardized. Different antihypertensive agents for the same degree of blood pressure will have significantly and clinically important differential effects on multiple variables, including lipid profiles, glucose, insulin, potassium, creatinine, angiotensin, catecholamines, pulse rate, body weight, and central pressure, not to mention 24-hour blood pressure control. Why would all, or any, of these effects be considered trivial (compared with, say, 2–mm Hg systolic pressure) or together exert a neutral effect such that the only property of all antihypertensive agents that is of value is the lowering of brachial blood pressure?
It is our view that, on the basis of differential mechanisms of action and diverse effects on major cardiovascular risk factors, different classes of antihypertensive agents are likely to provide different cardiovascular protection and that, given sufficiently sensitive tests, these differences could be shown more clearly.
Meanwhile, we believe that the apparent anomalies in the results of several trials in hypertension described above may well reflect not just chance variation but real differences beyond the magnitude of lowering blood pressure between the agents used in the trials.
|
Acknowledgments |
|---|
Disclosures
Dr Sever has received research grants from, served on speaker’s bureaus for or received honoraria from, and has served as consultant to Pfizer Inc, Servier Laboratories, AstraZeneca, Takeda Pharmaceuticals and Novartis AG. Dr Poulter has received research grants from Pfizer Inc, Servier Laboratories, Mars Inc, and Menarini. He has served on speaker’s bureaus for, received honoraria from, and served as consultant for Pfizer Inc, Servier Laboratories, AstraZeneca, Takeda Pharmaceuticals, Novartis AG, Squibb/BMS, Sanofi-Aventis, and other pharmaceutical companies.
|
References |
|---|
- Veterans Administration Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension: results in patients with diastolic blood pressure averaging 115 through 129 mm Hg. JAMA. 1967; 202: 1028–1034.[CrossRef][Medline]
[Order article via Infotrieve]
- Australian National Blood Pressure Management Committee. The Australian therapeutic trial in mild hypertension. Lancet. 1980; 1: 1261–1267.[Medline]
[Order article via Infotrieve]
- Medical Research Council Working Party. MRC trial of treatment of mild hypertension: principal results. BMJ. 1985; 291: 97–104.[Medline]
[Order article via Infotrieve]
- Veterans Administrative Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension, II: results in patients with diastolic blood pressure averaging 90 through 114 mm Hg. JAMA. 1970; 213: 1143–1152.[CrossRef][Medline]
[Order article via Infotrieve]
- Amery A, Birkenhager W, Brixko P, Bulpitt C, Clement D, Deruyttere M, De Schaepdryver A, Dollery C, Fagard R, Forette F, et al. Mortality and morbidity results from the European Working Party on High Blood Pressure in the Elderly trial. Lancet. 1985; 1: 1349–1354.[Medline]
[Order article via Infotrieve]
- Coope J, Warrender TS. Randomised trial of treatment of hypertension in the elderly in primary care. BMJ. 1986; 293: 1145–1151.[Medline]
[Order article via Infotrieve]
- HDFP (Hypertension Detection and Follow-up Program) Cooperative Group. Five-year findings of the Hypertension Detection and Follow-Up Program: reduction in mortality in persons with high blood pressure, including mild hypertension. JAMA. 1979; 242: 2562–2571.[Abstract]
- MacMahon SW, Cutler JA, Neaton JD, Furberg CD, Cohen JD, Kuller LH, Stamler J. Relationship of blood pressure to coronary and stroke morbidity and mortality in clinical trials and epidemiological studies. J Hypertens Suppl. 1986; 6: S14–S17.
- Collins R, Peto R. Antihypertensive drug therapy: effects on stroke and coronary heart disease. In: Textbook of Hypertension. London, UK: Blackwell Scientific Publications; 1994: P1156–P1164.
- MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke and coronary heart disease, part 1: prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990; 335: 765–774.[CrossRef][Medline]
[Order article via Infotrieve]
- Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, for the Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002; 360: 1903–1913.[CrossRef][Medline]
[Order article via Infotrieve]
- Medical Research Council Working Party. MRC trial on treatment of hypertension in older adults: principal results. BMJ. 1992; 304: 405–412.[Medline]
[Order article via Infotrieve]
- Lindholm LH, Carlberg B, Samuelsson O. Should beta blockers remain first choice in the treatment of primary hypertension? A meta-analysis. Lancet. 2005; 366: 1545–1553.[CrossRef][Medline]
[Order article via Infotrieve]
- Beta-Blocker Heart Attack Trial Research Group. A randomised trial of propranolol in patients with acute myocardial infarction. JAMA. 1982; 247: 1717–1724.
- Staessen JA, Fagard R, Thijs L, Celis H, Arabidze GG, Birkenhager WH, Bulpitt CJ, de Leeuw PW, Dollery CT, Fletcher AE, Forette F, Leonetti G, Nachev C, O’Brien ET, Rosenfeld J, Rodicio JL, Tuomilehto J, Zanchetti A, for the Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. Lancet. 1997; 350: 757–764.[CrossRef][Medline]
[Order article via Infotrieve]
- Blood Pressure Treatment Trialists’ Collaboration. Effects of ACE inhibitors, calcium-antagonists and other blood pressure lowering drugs: results of prospectively designed overviews of randomised trials. Lancet. 2000; 355: 1955–1964.[CrossRef][Medline]
[Order article via Infotrieve]
- Blood Pressure Lowering Treatment Trialists’ Collaboration (BPLTTC). Effects of different blood-pressure-lowering regimens on major cardiovascular events: results of prospectively-designed overviews of randomised trials. Lancet. 2003; 362: 1527–1535.[CrossRef][Medline]
[Order article via Infotrieve]
- HOPE (Heart Outcomes Prevention Evaluation) Study Investigators. Effect of ramipril on cardiovascular and microvascular outcome in people with diabetes mellitus: results of the HOPE study and MICRO-HOPE substudy. Lancet. 2000; 355: 253–259.[CrossRef][Medline]
[Order article via Infotrieve]
- Fox KM, for the European Trial on Reduction of Cardiac Events With Perindopril in Stable Coronary Artery Disease Investigators. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease: randomised, double-blind, placebo-controlled, multicentre trial (the EUROPA study). Lancet. 2003; 362: 782–788.[CrossRef][Medline]
[Order article via Infotrieve]
- Dahlöf B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U, Fyhrquist F, Ibsen H, Kristiansson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H, for the LIFE Study Group. Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint Reduction in Hypertension Study (LIFE): a randomised trial against atenolol. Lancet. 2002; 359: 995–1003.[CrossRef][Medline]
[Order article via Infotrieve]
- Williams B, Lacy PS, Thom SM, Cruickshank K, Stanton A, Collier D, Hughes AD, Thurston H, O’Rourke M; CAFE Investigators; Anglo-Scandinavian Cardiac Outcomes Trial Investigators; CAFE Steering Committee and Writing Committee. Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation. 2006; 113: 1213–1225.
[Abstract/Free Full Text] - Verdecchia P, Reboldi G, Angeli F, Gattobigio R, Bentivoglio M, Thijs L, Staessen JA, Porcellati C. Angiotensin-converting enzyme inhibitors and calcium channel blockers for coronary heart disease and stroke prevention. Hypertension. 2005; 46: 386–392.
[Abstract/Free Full Text] - Hansson L, Lindholm LH, Niskanen L, Lanke J, Hedner T, Niklason A, Luomanmaki K, Dahlof B, de Faire U, Morlin C, Karlberg BE, Wester PO, Bjorck JE. Effect of angiotensin-converting-enzyme inhibition compared with conventional therapy on cardiovascular morbidity and mortality in hypertension: the Captopril Prevention Project (CAPPP) randomised trial. Lancet. 1999; 353: 611–616.[CrossRef][Medline]
[Order article via Infotrieve]
- ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial: major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA. 2002; 288: 2981–2997.
[Abstract/Free Full Text] - Julius S, Kjeldsen SE, Weber M, Brunner HR, Ekman S, Hansson L, Hua T, Laragh J, McInnes GT, Mitchell L, Plat F, Schork A, Smith B, Zanchetti A, for the VALUE Trial Group. Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on valsartan or amlodipine: the VALUE randomised trial. Lancet. 2004; 363: 2022–2031.[CrossRef][Medline]
[Order article via Infotrieve]
- Dahlof B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, Collins R, Kjeldsen SE, Kristinsson A, McInnes GT, Mehlsen J, Nieminen M, O’Brien E, Ostergren J, for the ASCOT Investigators. Prevention of cardiovascular events with an antihypertensive regimen of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial–Blood Pressure Lowering Arm (ASCOT-BPLA): a multicentre randomised controlled trial. Lancet. 2005; 366: 895–906.[CrossRef][Medline]
[Order article via Infotrieve]
- Sever PS, Dahlof B, Poulter NR, Wedel H, Beevers G, Caulfield M, Collins R, Kjeldsen SE, Kristinsson A, McInnes GT, Mehlsen J, Nieminen M, O’Brien E, Ostergren J, for the ASCOT investigators. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (ASCOT-LLA): a multicentre randomised controlled trial. Lancet. 2003; 361: 1149–1158.[CrossRef][Medline]
[Order article via Infotrieve]
- Poulter NR, Wedel H, Dahlof B, Sever PS, Beevers DG, Caulfield M, Kjeldsen SE, Kristinsson A, McInnes GT, Mehlsen J, Nieminen M, O’Brien E, Ostergren J, Pocock S, for the ASCOT Investigators. Role of blood pressure and other variables in the differential cardiovascular event rates noted in the Anglo-Scandinavian Cardiac Outcomes Trial–Blood Pressure Lowering Arm (ASCOT-BPLA). Lancet. 2005; 366: 907–913.[CrossRef][Medline]
[Order article via Infotrieve]
- Morgan T, Lauri J, Bertram D, Anderson A. Effect of different antihypertensive drug classes on central aortic pressure. Am J Hypertens. 2004; 17: 118–123.[CrossRef][Medline]
[Order article via Infotrieve]
- Hirata K, Vlachopoulos C, Adji A, O’Rourke MF. Benefits from angiotensin-converting enzyme inhibitor "beyond blood pressure lowering," beyond blood pressure or beyond the brachial artery? J Hypertens. 2005; 23: 551–556.[Medline]
[Order article via Infotrieve]
- London GM, Asmar RG, O’Rourke MF, Safar ME. Mechanism(s) of selective systolic blood pressure reduction after a low-dose combination of perindopril/indapamide in hypertensive subjects: comparison with atenolol. J Am Coll Cardiol. 2004; 43: 92–99.
[Abstract/Free Full Text] - Mason RP, Walter MF, Day CA, Jacob RF.