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

Abstract

Diabetes affects around 20 million persons in the United States and is the fifth leading cause of death worldwide, consumes a large percentage of our health care resources, and remains a leading cause of cardiovascular (CV) and kidney disease. With the increasing rates of obesity in the United States, there are 41 million people with prediabetes, 50% of whom will develop frank type 2 diabetes mellitus over the next several years. Several clinical trials of subjects with hypertension or underlying CV disease have suggested that agents that block the renin-angiotensin system (RAS) prevent the development of new-onset diabetes (NOD). Pooled analyses of these trials have found, on average, a 25% reduction in NOD with angiotensin-converting enzyme (ACE) inhibitor–or angiotensin II receptor blocker (ARB)–based treatment regimens (usually with a diuretic). It remains unclear, however, whether this finding is attributable to the specific use of these agents or to the effects of the comparator agents studied, usually a β-blocker or thiazide-type diuretic. In addition, the progression to diabetes in these trials has always been a secondary outcome, which is, by definition, only hypothesis-generating. NOD has never been studied as a primary outcome. The Diabetes Reduction Assessment With Ramipril and Rosiglitazone Medication (DREAM) trial was conducted by the Canadian Institutes of Health Research, with additional financial support from industry. Subjects from 21 countries in 191 centers were enrolled to evaluate whether the ACE inhibitor, ramipril, reduces the risk of NOD in persons at high risk for its development. They had to be at least 30 years of age and have impaired fasting glucose (IFG) (fasting plasma glucose levels at least 110 mg/dL but <126 mg/dL) or impaired glucose tolerance (IGT) (fasting plasma glucose levels of at least 140 mg/dL but <200 mg/dL 2 hours after an oral glucose load) but no history of diabetes, CV disease, or intolerance to an ACE inhibitor. This double-blind placebo-controlled trial was performed using a 2 × 2 factorial design whereby both the ACE inhibitor ramipril and the thiazolidinedione rosiglitazone were compared with placebo. The 2 pharmacologic agents were reported separately; this review will deal only with the ACE inhibitor ramipril compared with placebo. Between July 2001 and August 2003, 24,592 participants were screened and 5808 entered a 17-day run-in period. The most common reasons for exclusion were ineligibility (94%) and refusal to participate (3%). After the run-in period, 2623 participants were randomly assigned to receive ramipril and 2646 participants received placebo. Ramipril was given as 5 mg daily for the first 2 months and then increased to 10 mg at the 2-month visit and to 15 mg after 1 year. Visits were scheduled at 2 months and 6 months after randomization and then at regular 6-month intervals until the final visit between February and April 2006. At each visit, adherence to study drug was assessed and reinforced, as was a healthy diet and lifestyle. Electrocardiography was performed at baseline, at 2 years, and at the end of the study. At the 2-year and final visits, a glucose tolerance test was performed if the fasting plasma glucose was 126 mg/dL or higher, if the fasting plasma glucose level exceeded 95 mg/dL and the glycosylated hemoglobin level exceeded 93% of the upper limit of normal, and to confirm or eliminate the diagnosis of diabetes in people in whom diabetes had not developed. Diabetes was diagnosed if 2 consecutive plasma glucose levels performed on separate days exceeded the diagnostic thresholds within a 3-month period (ie, a fasting plasma glucose of ≥126 mg/dL or a 2-hour plasma glucose ≥200 mg/dL) or if the outside physician following the patient prescribed an antidiabetic agent after a single abnormal glucose value. Glucose levels were available for 92.6% of individuals who had not reached a primary outcome by the end of the study. Participants were followed for a median of 3.0 years. The mean systolic BP at baseline was 136.1 mm Hg in the ramipril group and 136.0 mm Hg in the placebo group. At 2 months, mean systolic BP decreased by 8.2 mm Hg among those receiving ramipril and 3.9 mm Hg among those receiving placebo, a difference of 4.3 mm Hg, which persisted throughout the trial. Changes in diastolic BP were 4.3 mm Hg in the ramipril group and 1.6 mm Hg in the placebo group at 2 months (difference of 2.7 mm Hg) and also remained significant throughout the trial. Mean serum creatinine level did not change significantly in either group throughout the trial. The prespecified primary outcome was the development of NOD or death. Key secondary outcomes were fasting glucose levels, regression to normal glucose levels (fasting glucose level <110 mg/dL or 2-hour post-load glucose level <140 mg/dL) and a composite of cardiorenal events, defined as either CV events (clinical or silent myocardial infarction, stroke, death from CV events, revascularization procedures, heart failure, newly diagnosed angina with objective evidence of ischemia, or a ventricular arrhythmia requiring resuscitation) or renal events (on the basis of measurements in urine or blood at a central laboratory). Renal event data will be published at a future date. During the study, there was no difference in the occurrence of NOD or death (the primary end point) between individuals who received ramipril or placebo (18.1% of ramipril-treated patients compared with 19.5% of those who received placebo). There was also no difference in the development of NOD alone, which developed in 17.1% of the ramipril group and 18.5% of the placebo group. The rates of the primary outcomes were similar even after controlling for the use of other types of antihypertensive therapy, including thiazide-type diuretic (10% of participants), non-thiazide diuretic (6%), β-blocker (17%), calcium channel blocker (13%), or ARB (6%) therapy. The groups also did not differ in the incidence of CV events or hospitalization; however, significantly more ramipril than placebo recipients noted regression to normal glucose levels (43% vs 38%). There was no significant interaction with rosiglitazone; the hazard ratios for the primary outcome among those receiving ramipril were similar regardless of whether they received rosiglitazone. The authors conclude that the use of 15 mg of ramipril daily for 3 years in people with IFG or IGT and no evidence of established CV disease does not significantly prevent NOD or death.—Bosch J, Yusuf S, Gerstein HC, et al, for the DREAM Trial Investigators. Effect of ramipril on the incidence of diabetes. N Engl J Med. 2006;355:1551–1562. ACE inhibitors are among the most widely prescribed antihypertensive agents in the United States. They are currently recommended for the 6 compelling indications noted in the Seventh Report of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7). In addition to the benefits seen in patients with established CV disease, a number of widely publicized clinical trials have reported that the use of ACE inhibitors or ARBs decrease the incidence of NOD. These include the Heart Outcomes Prevention Evaluation (HOPE), where the use of the ACE inhibitor ramipril, given 10 mg daily (in addition to multiple other drugs), reduced the risk of NOD by 34% when compared with a group of patients on medications other than ramipril. This, however, was not a prespecified outcome, and the development of diabetes was ascertained only by self-report. By contrast, the use of ramipril was not associated with a reduction in NOD in the DREAM study. What explains the different results found in this recent trial? Unlike previous studies that evaluated the effect of antihypertensive drug therapy on the development of NOD, the DREAM investigation was specifically designed to evaluate the association of pharmacologic treatment and NOD as the primary end point. Previous studies, including HOPE, evaluated the development of diabetes either as a secondary outcome or in a post hoc analysis and, in many cases, the diagnosis of NOD was made primarily by self-report. Second, in contrast to other studies, patients in the DREAM study were only included in the trial if they had IFG or IGT and did not have established type 2 diabetes. Patients were screened at baseline with fasting glucose and an oral glucose tolerance test, and those who had overt diabetes or normal glucose levels were excluded. In previous studies, patients with undiagnosed diabetes may have been included, since baseline fasting plasma glucose and glucose tolerance were not routinely assessed. Finally, patients enrolled in the DREAM trial were younger (mean age, 55 years, compared with 65 years in other studies) and did not have evidence of CV disease on entry into the study. In previous studies, like HOPE, patients most often had known CV disease on entry. It is possible that ACE inhibitor therapy has a greater effect in individuals with established CV disease who may have increased activation of the renin-angiotensin-aldosterone system. It is also possible that the DREAM trial was too short in duration to detect a difference in incidence of NOD. In DREAM, the median follow-up was approximately 3 years compared with previous studies with ACE inhibitor and ARB therapy, which had a median follow-up of about 4.5 years. The Kaplan-Meier curves in DREAM suggest a trend toward benefit of ramipril in the prevention of NOD beginning at 3.5 years. Whether this is real or a chance phenomenon remains unclear, as does its clinical significance. Clinical trial evidence up to 6 years in duration continues to affirm the benefits and safety of thiazide-type diuretics despite their adverse effects on plasma glucose. For example, while there was a 3% increased risk of NOD noted with initial thiazide-type diuretic therapy compared with ACE inhibitor therapy in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT), CV outcomes with initial thiazide-type diuretics were unsurpassed. This result and other data led to the recommendation in JNC 7 that thiazide-type diuretics should be used as initial antihypertensive therapy for most patients with hypertension. Whether possible metabolic effects moderate the clinical benefits of thiazide-type diuretics over a longer follow-up period continues to be debated. The most recent estimates predict that the prevalence of type 2 diabetes mellitus will more than double, from 5.6% in 2005 to 12.0% in 2050, and that 48.3 million people will have a diagnoses of diabetes in the United States by 2050. In addition, the tremendous increase in the rates of obesity and insulin resistance has led to an epidemic of individuals with IFG and/or IGT who are at high risk for progressing to type 2 diabetes. Whether small incremental changes in glucose over many years associated with some antihypertensive drug therapy carries the same risk as obesity-associated diabetes mellitus continues to be debated. For now, reduction of BP seems to be the most important target for improving clinical outcome in people both with and at risk for developing diabetes mellitus. Since most individuals require multiple antihypertensive agents to control their BP, combination regimens that target complementary pathophysiologic BP-lowering mechanisms are a rational approach to achieve BP control. While it is possible that a longer (ie, >3 years) and larger study would have shown a reduction in the development of NOD, the present 3-year trial with the use of the ACE inhibitor ramipril did not indicate that NOD was prevented among persons with IFG or IGT. The hypertension community awaits the outcomes of the Avoiding Cardiovascular Events Through Combination Therapy in Patients Living with Systolic Hypertension (ACCOMPLISH) study, evaluating initial fixed-dose combination therapy with an ACE/calcium channel blocker (benazepril/amlodipine) compared with an ACE/thiazide-type diuretic (benazepril/HCTZ) in individuals with and without diabetes. At present, continued use of thiazide-type diuretics with ACE inhibitors or ARBs as combination antihypertensive therapy is often required to control BP in most individuals with hypertension. With the publication of the DREAM results, enthusiasm for including RAS antagonists to decrease the risk of NOD has lessened. Traditional Chinese acupuncture (TCA) is advocated by some as an effective and well-tolerated treatment for essential hypertension. While a few case reports and several small clinical trials have suggested a benefit from acupuncture in patients with hypertension, the methodologic limitations of these trials have made it impossible to draw any definitive conclusions about the efficacy of TCA in lowering blood pressure (BP). The Stop Hypertension With Acupuncture Research Program (SHARP) pilot trial was designed to overcome the methodologic limitations of previous studies and evaluate the BP-lowering benefits of acupuncture. The SHARP trial was a prospective, double-blind, randomized, sham-controlled, parallel-group clinical pilot trial that was designed to test the safety and efficacy of 6 weeks of TCA in patients with moderate essential hypertension who were not receiving any antihypertensive medication. To be eligible, subjects had to be at least 18 years of age and have a systolic BP of 140–179 mm Hg or a diastolic BP of 90–109 mm Hg for 3 qualifying visits after cessation of all previous antihypertensive medications. In addition, they could not have had any TCA in the previous 6 months or have a medical contraindication to either TCA or the suspension of antihypertensive medications. Participants were randomly allocated into 1 of 3 treatment groups: (1) individualized TCA (IND); (2) standardized TCA (STD); or (3) sham acupuncture (CNTL). IND consisted of corporal acupuncture at 10–12 points selected and stimulated as prescribed for each individual by an experienced acupuncturist trained originally in China, plus auricular acupuncture at the 2 most active sites as identified by a point detector at the start of each session. STD consisted of corporal acupuncture with neutral stimulation at 5 bilateral points (10 total) plus auricular acupuncture at heart and jiang ya gou points as determined a priori by an expert panel convened for the purposes of this study. All diagnoses and treatment for IND and STD followed TCA principles previously established in the literature. CNTL consisted of sham acupuncture with no stimulation at 5 bilateral points (10 total) that do not correspond to any traditional Chinese meridians plus auricular acupuncture in Darwin's tubercle and the posterior ear lobe—all points considered inactive according to the tenets of TCA. Treatment consisted of no more than 12 twice-weekly 30-minute acupuncture sessions over 6 weeks. All participants received TCA from a team of licensed acupuncturists in a clinic established at the Massachusetts General Hospital. Treatment allocation was stratified by presence or absence of previous use of antihypertensive medications in the 6 months before study entry. All antihypertensive medications were tapered off according to a predetermined schedule before study initiation. After randomization, BP was measured at scheduled visits every 14 days until week 10 and subsequently at 4, 6, 9, and 12 months using a standardized technique by trained staff blinded to treatment allocation. In addition, blood chemistries, lipid profiles, complete blood cell counts, urinalysis, electrocardiography, and adverse events were analyzed at 10 weeks, 6 months, and 12 months. Quality-of-life data were obtained through use of the Medical Outcomes Study-Short Form self-administered questionnaires at baseline, 10 weeks, and 12 months. The predetermined primary end point was change in systolic BP from baseline to week 10, and the primary comparison was change in systolic BP for active (IND+STD) vs sham (CNTL) TCA. Between March 2001 and July 2002, 424 individuals underwent initial screening, of whom 192 met eligibility criteria and were randomized. Four subjects were lost to follow-up and 188 participants were available for the final evaluation. Patient characteristics in the 3 treatment groups were similar, with an average age of 55 years, 55% male, and an average baseline BP 149/93 mm Hg. Nearly all subjects (96%) received 12 treatments during an average of 42 days Mean systolic BP declined in all 3 groups between baseline and week 10 (the primary end point), with no statistically significant difference between those allocated to active (IND+STD, −3.56±1.92 mm Hg) compared with sham (CNTL, −3.84±1.93 mm Hg) acupuncture (P=.90). There was also no significant difference in diastolic BP decline between active (IND+STD) and sham (CNTL) acupuncture. In addition, there was no difference in systolic or diastolic BP between the 2 active treatment arms, IND and STD. When followed for 6 or 12 months, no significant treatment effect was noted on systolic or diastolic BP. Categorization of participants by age, race, sex, baseline BP, history of antihypertensive use, obesity, or primary TCA, diagnosis did not identify any subgroup that benefited from active TCA. Although BP increased modestly between measurements taken before and after an acupuncture session, no difference was seen between treatment groups. Ten-week change in physiologic and quality-of-life parameters did not differ between groups, except for fasting glucose, which was slightly lower in the CNTL group. Among the 141 participants who completed a masking assessment evaluation at 4 months, there was a nonsignificant trend toward subjects being able to determine their treatment assignment, but the number who guessed correctly was similar in all 3 groups (about 67%). Three study-related adverse events occurred—2 STD subjects experienced a hypertensive emergency and 1 CNTL subject experienced an episode of heart failure. In this well-designed prospective randomized clinical trial using an appropriate control group, 6 weeks of biweekly TCA has no demonstrable benefit on reducing BP in individuals with stage 1 hypertension who were not receiving antihypertensive drug therapy.—Macklin EA, Wayne PM, Kalish LA, et al. Stop Hypertension with the Acupuncture Research Program (SHARP): results of a randomized, controlled clinical trial. Hypertension. 2006;48:838–845. Previous evidence, mostly anecdotal observation, has suggested that TCA may have a role in the treatment of hypertension. The SHARP investigators appropriately considered these findings hypothesis-generating, leading them to design and execute a well-conceived prospective randomized clinical trial to test the hypothesis. Perhaps the most important design characteristics of this clinical trial were the use of appropriate blinding of acupuncturists and subjects, an appropriate control group (sham acupuncture in CNTL), and the use of 2 different previously validated types of TCA treatments (IND and STD). The study also benefited from careful measurement of an appropriate clinical outcome, the 10-week change in systolic BP. Like all randomized clinical trials, however, SHARP does have some shortcomings that may cause some to question the validity of the findings. As the investigators point out, the relatively wide confidence intervals in this small pilot study do not exclude the possibility of a small, but perhaps clinically significant, benefit (likely <4 mm Hg). The exclusion of subjects concomitantly treated with antihypertensive drug therapy does not rule out a possible beneficial effect when TCA is used in combination with BP-lowering medication, and the relatively short treatment period may not have allowed the full effects of TCA to be realized. Therefore, the results of this well-designed clinical trial make it unlikely that TCA provides substantial clinical benefit in patients with hypertension. Some of the perceived benefits of TCA, as pointed out by the SHARP investigators, may be the result of being treated by a clinician who “creates an expectation of medical improvement.” Practitioners of alternative medicine, including TCA, create confidence in the safety and efficacy of the treatments they offer, even though, in most cases and as in this study, they have not been shown to have a measurable clinical benefit. In contrast, much of the dialogue surrounding the use of drug therapy for the treatment of hypertension, both in the practitioner's office and in the media, focuses on the potential adverse effects associated with these agents. While it is important that clinicians not underestimate the potential tolerability issues of antihypertensive drug medications, it is also important to balance these issues with a clear explanation of the important clinical benefits and our confidence in these interventions. This should improve outcome and patient acceptance and adherence. As patients in the United States continue to use alternative medicines to treat hypertension, it is imperative that future clinical trials be designed and funded to determine the potential value of these treatments for improving the control of BP and other outcomes in hypertension.