ASCOT – More than a horse race!

The shortfall in coronary heart disease prevention observed in early hypertension trials has in part been attributed to disadvantages of the diuretics and beta‐blockers used. For a given blood pressure reduction, it has been suggested that newer agents would have advantages over diuretics and beta‐blockers. Thus, the objectives of the blood pressure lowering arm (BPLA) of the Anglo Scandinavian Cardiac Outcomes Trial (ASCOT) were to compare the effect on myocardial infarction and fatal coronary heart disease, mortality, other cardiovascular endpoints and new onset diabetes of the beta‐blocker atenolol+the diuretic bendroflumethiazide with the calcium‐antagonist amlodipine+the angiotensin‐converting enzyme (ACE) inhibitor perindopril.

As many as 19,257 hypertensive subjects (aged 40–79 years, with at least three other cardiovascular risk factors), were randomized to one of the two antihypertensive regimens. The study was planned to run until 1150 patients had experienced a primary endpoint. However, it was stopped prematurely, as there was a mortality and stroke benefit in the amlodipine+perindopril arm after having accumulated 106,153 patient‐years of observation [1]. Blood pressures had been lowered by 27/17 mmHg in the two groups combined but mean blood pressures averaged 2.7/1.9 mmHg higher on the atenolol+thiazide regimen. Compared with the atenolol+thiazide regimen, fewer deaths and cardiovascular events occurred among those allocated the amlodipine+perindopril regimen: primary endpoint (429 vs 474, p = 0.11), fatal and non‐fatal stroke (327 vs 422, p = 0.0003), total cardiovascular events and procedures (1362 vs 1602, p<0.0001), cardiovascular mortality (263 vs 342, p = 0.001) and all‐cause mortality (738 vs 820, p = 0.025). No significant heterogeneity was found among a number of pre‐specified subgroups. The incidence of developing diabetes was much lower on the amlodipine+perindopril regimen (567 vs 799 new cases), p<0.0001).

Thus, the amlodipine+perindopril antihypertensive treatment regimen prevented more major cardiovascular events and induced less diabetes than an atenolol+thiazide regimen. On the basis of previous trial evidence, these effects may not be entirely explained by better blood pressure control, which was addressed in an accompanying paper [2]. These results clearly have implications regarding optimal combinations of antihypertensive agents. However, three major issues of ASCOT must be discussed.

First, ASCOT‐BPLA did not reach its primary objective, which was to prevent myocardial infarction and fatal coronary heart disease significantly. However, in the daily clinical setting of aggressive coronary revascularization (PCI and CABG), leaving patients to get true myocardial infarction and die from it is a historical setting. In order to show this, ASCOT would have needed about 2 more years of follow up to reach the pre‐specified 1150 primary event and thus enough statistical power. However, along that road many lives would have been lost. This ethical dimension and safety reason put forward by the DSMB, that there already was a clear reduction of total mortality and strokes by the amlodipine+perindopril regimen, could not be argued by the study steering committee [1]. Omitting the few patients with silent myocardial infarction from the primary endpoint demonstrated that there was a significant benefit of the amlodipine+perindopril combination, also illustrating the marginal value of arguing the safety issue in ASCOT‐BPLA. Clearly, including silent myocardial infarction as an endpoint has never made any contribution in hypertension outcome research.

Second, ASCOT‐BPLA like most other comparative hypertension trials suffered from a difference in blood pressure favouring one of the treatment arms [2]. This is not a weakness per se, but rather a property inherited with the effectiveness of the drugs that were investigated, likely in this case amlodipine, and which also clouded the direct comparison of amlodipine and the angiotensin receptor blocker valsartan in the recent VALUE trial [3]. In VALUE [4], like in ASCOT [2], extensive explorative statistical work revealed that the difference in blood pressure matters but do not explain it all. By nature, post hoc statistical corrections for differences in blood pressure in intention‐to‐treat randomized clinical outcome trials are flown, but they do provide clinicians some insight into the mechanisms that may explain outcome benefits of some drugs up against others. In the case of ASCOT [2], not only a difference in blood pressure but also sustained differences in HDL‐cholesterol and triglycerides, as expected, had major impact in explaining the cardiovascular benefits of amlodipine+perindopril.

Third, was the benefit related to the treatment with amlodpine or perindopril? The design of ASCOT does not allow an answer to the question. However, in light of the remarkable similarities of the cardiovascular outcomes on amlodipine and valsartan in the VALUE Trial [3], in which some rather minor differential effects on preventing myocardial infarction and stroke were most likely fully explained by the differences in blood pressure [4], it is likely that the endpoint protection in ASCOT is caused by both amlodipine and perindopril, or in fact the combination. In a recent meta‐regression analysis, ACE inhibitors may have some benefit beyond blood pressure control in preventing coronary disease and calcium antagonists may have a similar benefit in preventing stroke [5].

ASCOT leaves us with a therapeutic regimen consisting of lipid lowering with atorvastatin [6] and blood pressure lowering with amlodipine+perindopril [1], or in more general terms, lipid lowering with statin and blood pressure lowering with calcium antagonist+ACE inhibitor effectively to prevent cardiovascular outcomes and diabetes in hypertensive patients with quite moderately increased risk of getting the diseases. With hundreds of millions of hypertensive patients worldwide, implementing these data has the potential to prevent millions of events beyond what can be expected of until now “standard” treatment, as well as to save lives and have major impact on public health.