Analysis of Recent Papers in Hypertension
Jan Basile, MD, Senior Editor

Authors

  • Michael J. Bloch MD,

    1. From the Division of General Internal Medicine/Division of Cardiology, University of Nevada School of Medicine, Reno, NV;1the Risk Reduction Center, Saint Mary's Regional Medical Center, Reno, NV;2the Primary Care Service Line, Ralph H. Johnson VA Medical Center, Charleston, SC;3 and the Division of General Internal Medicine/Geriatrics, Medical University of South Carolina, Charleston, SC4
    Search for more papers by this author
  • and 1,2 Jan N. Basile MD 3,4

    1. From the Division of General Internal Medicine/Division of Cardiology, University of Nevada School of Medicine, Reno, NV;1the Risk Reduction Center, Saint Mary's Regional Medical Center, Reno, NV;2the Primary Care Service Line, Ralph H. Johnson VA Medical Center, Charleston, SC;3 and the Division of General Internal Medicine/Geriatrics, Medical University of South Carolina, Charleston, SC4
    Search for more papers by this author

Michael J. Bloch, MD, Risk Reduction Center, Saint Mary's Regional Medical Center, 645 North Arlington Street, Suite 460, Reno, NV 89503
E-mail: mbloch@aol.com

THRESHOLDS FOR DIASTOLIC BLOOD PRESSURE REDUCTION IN OLDER PATIENTS WITH ISOLATED SYSTOLIC HYPERTENSION APPEAR TO BE DIFFERENT DEPENDING ON THE PRESENCE OR ABSENCE OF UNDERLYING CORONARY HEART DISEASE

Whether a J-curve relationship exists between achieved diastolic blood pressure (BP) and cardiovascular (CV) events or total mortality in persons with hypertension is a subject of ongoing debate. It is clear that coronary heart disease (CHD) events are high in patients with diastolic BP >90 mm Hg and lowest in patients with diastolic BPs <80 mm Hg, but do they increase again when diastolic BPs get below 60–70 mm Hg? A recent meta-analysis of 10 prospective randomized trials from the Hypertension Detection and Follow-Up Program Cooperative Group concluded that lowering diastolic BP to <70 mm Hg was not associated with a poor outcome. This analysis, however, did not differentiate between patients with systolic-diastolic hypertension and those with isolated systolic hypertension (ISH). A post hoc analysis of the Systolic Hypertension in the Elderly Program (SHEP) did observe a J curve for CV events in older patients with ISH in whom on-treatment diastolic BP <60 mm Hg was achieved, but it did not stratify for coronary heart disease (CHD) at baseline. The numbers of patients in the lower diastolic BP groups were small and, for example, one subset of patients in the diastolic BP group of 45 to 50 mm Hg did not experience more CHD events than those in the 70 to 80 mm Hg group. The Systolic Hypertension in Europe (Syst-Eur) trial provided an opportunity to examine this question. The current post hoc analysis was designed to determine the relationship between on-treatment diastolic BP and mortality or CV events in older patients with ISH with or without established CHD at baseline.

The Syst-Eur trial was a prospective randomized, placebo-controlled double-blind intervention trial performed in 198 centers in Europe. Eligible patients had to be at least 60 years of age, with an average systolic BP of 160 to 219 mm Hg and diastolic BP <95 mm Hg during the 3-month run-in period. Patients were then randomized to either placebo or active treatment consisting of the calcium channel blocker (CCB) nitrendipine (10–40 mg/d), which could be combined with or replaced by enalapril (5–20 mg/d) and/or hydrochlorothiazide (12.5–25 mg/d), as needed, to reduce sitting systolic BP by at least 20 mm Hg and to <150 mm Hg. After the main phase of the trial (phase 1) was completed in 1997 (median follow-up, 2 years [range, 1 month–8.1 years]), patients in the placebo group were changed to active therapy while those in the original active therapy group continued to receive active treatment for an additional 4 years (phase 2) ending on December 31, 2001.

In the present post hoc analysis of the Syst-Eur trial, follow-up data were analyzed separately for placebo and active treatment; data for placebo treatment are from phase 1 only, and data on active treatment are from both phase 1 and phase 2. Outcomes measured included fatal and nonfatal cerebrovascular events (stroke and transient ischemic attack); fatal and nonfatal CHD events (sudden death, myocardial infarction, and coronary revascularization); an aggregate of all CV events (CV death, myocardial infarction, stroke, heart failure, coronary revascularization, aortic aneurysm, and transient ischemic attack); CV mortality; and non-CV death. To determine the effect of achieved diastolic BP on the risk of outcomes, multivariate Cox regression models were performed after correcting for age, sex, smoking status, previous antihypertensive treatment at baseline, weight, and previous history of stroke, myocardial infarction, and diabetes. Hazard ratios were calculated for each 5-mm Hg decrement of achieved diastolic BP. These analyses were repeated with and without systolic BP as a continuous time-dependent or dichotomous variable.

A total of 4695 patients were randomized in phase 1 of the Syst-Eur trial with no significant differences in baseline characteristics between those randomized to active treatment or placebo. Mean age was 70.2±6.7 years, 66.9% were women, and 14.5% had a previous history of established CHD. Mean systolic BP was 173.8±9.9 mm Hg, and mean diastolic BP was 85.5±5.8 mm Hg. After the completion of phase 1, as originally reported, there was a statistically significant (42%) reduction in the incidence of fatal and nonfatal cerebrovascular events. Significant reductions were also demonstrated for fatal and nonfatal CHD (−30%), heart failure (−29%), and all CV events (−31%).

When data from phase 1 and phase 2 were combined for the present report, there were 14,511 patient-years of follow-up on active treatment and 5219 patient-years of follow-up with placebo. On active treatment, the mean achieved systolic BP was 149.0±10.3 mm Hg and mean achieved diastolic BP was 78.8±6.0 mm Hg. During placebo treatment, the mean achieved systolic BP was 161.8±13.6 mm Hg and mean achieved diastolic BP was 83.6±6.4 mm Hg. Of interest is that only 1.7% of patients with active treatment and 0.44% of patients on placebo treatment had a mean on-treatment diastolic BP level <65 mm Hg.

The results of the regression models presented for diastolic BP as a continuous variable revealed that on-treatment diastolic BP was not associated with an increase in CV mortality with either placebo or active treatment. Non-CV mortality, however, was significantly increased with decreasing diastolic BP. For cerebrovascular events, decreasing on-treatment diastolic BP was associated with increased risk in the placebo group, but not in the active treatment group. For CHD events, there was a significant increase in risk in patients in the active treatment only group among patients with a history of established CHD at baseline; conversely, with placebo there was an increase in the risk of CHD with decreasing diastolic BP only in patients without CHD at baseline. These results did not change appreciably when on-treatment systolic BP was included as a continuous variable in the model or when on-treatment diastolic BP was examined using 5-mm Hg cutoffs as opposed to a continuous variable.

The authors of this post hoc analysis of the Syst-Eur trial conclude that these findings “support the hypothesis that antihypertensive treatment can be intensified for the prevention of CV events when systolic BP is not under control in older patients with ISH, at least until diastolic BP reaches about 55 mm Hg.” They did, however, suggest a more “prudent approach” in patients with ISH and concomitant CHD, in which diastolic BP “should probably not be lowered to <70 mm Hg.”—Fagard RH, Staessen JA, Thijs L, et al. On-treatment diastolic blood pressure and prognosis in systolic hypertension. Arch Intern Med. 2007;167(17):1884–1891.

COMMENT

The major finding of the present post hoc analysis suggests that in older patients with ISH who are on active antihypertensive therapy and have no history of established CHD, there is no significant J-shaped curve or increase in CV mortality or CV events when diastolic BP is reduced to as low as 55 mm Hg. Furthermore, since with both active antihypertensive therapy and placebo non-CV mortality was increased with decreasing achieved diastolic BP, these findings support the hypothesis that the increased mortality seen with low achieved diastolic BP in some previous retrospective studies may have been the result of confounding factors such as overall ill health or reverse causation. Of interest, none of these findings were not confounded by the achieved systolic BP.

The present findings from the Syst-Eur trial are in contrast with those from SHEP, the other major randomized prospective trial of antihypertensive therapy in older patients with ISH. In SHEP, very low achieved diastolic BP was associated with an increased risk of CHD, stroke, and CV events in actively treated but not in placebo-treated patients. The increased relative risk for combined CV events in SHEP became significant for achieved diastolic BP at <70 mm Hg; with a greater risk at diastolic BP of <55 mg Hg. As noted, at even lower levels, however, (ie, 45–50 mm Hg) the risk was not as great, although numbers of events were small.

There are a number of possible explanations for the differing findings between SHEP and the present Syst-Eur report. Most important, there was a difference in baseline and achieved diastolic BP between the 2 studies. In SHEP, the inclusion criteria was a diastolic BP of <90 mm Hg, with a baseline diastolic BP level of 77 mm Hg; the mean diastolic BP level achieved with active treatment was 68 mm Hg. In the Syst-Eur trial, the inclusion criterion was a diastolic BP level of <95 mm Hg, with a baseline diastolic BP of 86 mm Hg; the mean diastolic BP achieved with active treatment was 79 mm Hg. In the Syst-Eur trial, only 1.7% of actively treated patients had a mean achieved diastolic BP of <65 mm Hg and only 0.2% of actively treated patients had a mean achieved diastolic BP of <55 mm Hg; in the placebo group, the percentages were even lower (0.44% for <65 mm Hg and 0.04% for <55 mm Hg). With lower baseline and achieved diastolic BP level, SHEP may have been more likely to demonstrate evidence of a J-shaped curve with active treatment. Another possible explanation for the divergent results might have been the different active treatment regimens for each study; initial therapy with a CCB in the Syst-Eur trial and a thiazide-type diuretic in SHEP.

Neither SHEP nor the Syst-Eur trial results conclusively settle this issue. Given what we understand about pulse pressure in patients with similar systolic BP levels, further lowering of diastolic BP might be associated with an increased risk of CV events. Observational epidemiologic data from the Framingham study suggest that in untreated older patients, decreasing diastolic BP and increasing pulse pressure are both associated with an increase in the risk of CV events. There is some reason to believe that this relationship would also hold true for older individuals treated for ISH. The question therefore becomes whether the known benefits of lowering elevated systolic BP in people with ISH are mitigated by further decreasing diastolic BP. For example, in a 70-year-old patient without established CHD who has a BP of 150/60 mm Hg, will the addition of further antihypertensive therapy decrease the risk of subsequent CV events? All of our available clinical trials addressing ISH have baseline and on-treatment systolic BP values that are too high to answer this question. Both SHEP and Syst-Eur used the stage 2 baseline systolic BP of >160 mm Hg as an entry criteria and each had a mean baseline diastolic BP >75 mm Hg. In fact, we have no randomized clinical trial data to support the recommendation of using antihypertensive therapy in older patients with ISH who have stage 1 ISH (systolic BP, 140–160 mm Hg). Our current guidelines for intervening in these patients are extrapolated from the results of randomized trials that enrolled patients with higher baseline systolic BP, like SHEP and the Syst-Eur trial.

The question of a possible J-shaped curve in older patients with ISH remains important and fundamental enough to warrant a future randomized clinical trial comparing active treatment to placebo in older patients with baseline systolic BP between 140 and 160 mm Hg and a diastolic BP <70 mm Hg. Established CHD should probably be an exclusion criterion in such a clinical trial since evidence from multiple sources, including the Syst-Eur trial and data from the International Verapamil-Trandolapril Study (INVEST), suggest that very low diastolic BP achieved with antihypertensive therapy may be associated with an increase in events in this patient population. Pending the results of such a clinical investigation, the following recommendations are suggested by existing evidence. In the majority of older patients with ISH, antihypertensive therapy should be intensified until systolic BP reaches <140 mm Hg or <130 mm Hg in patients with diabetes or chronic kidney disease, at least until the diastolic BP reaches 55 mm Hg. For individuals with underlying CHD, antihypertensive therapy should be intensified with caution in patients whose diastolic BP is lowered to <70 mm Hg.

Ancillary