Author + information
- Received July 13, 2009
- Accepted August 25, 2009
- Published online December 29, 2009.
- Bryan Williams, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Bryan Williams, Professor of Medicine, Clinical Sciences Building, P.O. Box 65, Leicester Royal Infirmary, Leicester LE2 7LX, United Kingdom
The year 2008 was extraordinary because of the number of reports from major clinical outcome trials of blood pressure (BP)-lowering therapies. Consequently, this review of the past year in hypertension will primarily focus on these clinical trials and their likely impact on future treatment guidelines. However, there were also important reports on the global burden of hypertension, recommendations for home BP measurement, impact of nutritional factors on the incidence of hypertension, resistant hypertension, and treatment strategies for improving BP control in general.
The Global Burden of BP-Related Morbidity and Mortality
The burden of cardiovascular disease and mortality attributable to an elevated BP was estimated from the Global Burden of Disease 2000 study, recently updated to include more recent country-level data (1). The population burden of disease attributable to an elevated BP was estimated as the proportional reduction in average disease risk that would result from elimination of an elevated systolic BP (systolic BP ≥115 mm Hg), assuming the distributions of other risk factors remain unchanged over a defined time period.
In 2001, ≈7.6 million deaths worldwide were attributable to an elevated systolic BP, ≈14% of all deaths (1). Approximately 54% of all strokes and 47% of all ischemic heart disease was attributable to high BP. The majority (≈80%) of this BP-attributable disease burden occurred in low- and middle-income countries, with more than one-half occurring among people of working age (45 to 69 years). This finding suggests that since these data were generated in 2000, >60 million people have died worldwide as a consequence of high BP. It is interesting to reflect on this clear and present danger and the minimal worldwide attention focused on it, alongside the recent global panic regarding influenza. How many people wearing paper masks actually know what their BP is?
Another interesting aspect of the report is that approximately one-half of the BP-attributable deaths are due to systolic BP levels between the optimal systolic BP level (defined in the report as ≤115 mm Hg) and the current therapeutic intervention threshold in guidelines of ≥140/90 mm Hg (1). Thus, a major component of the population burden of BP-attributable disease occurs at BP levels that may be considered suboptimal (namely, systolic BP of 115 to 140 mm Hg) but for which we have no evidence of the benefit versus harm of further BP lowering. That presents both a clinical and a socioeconomic challenge: 1) can we generate the evidence for treatment benefit for people in this category; and 2) would populations accept and could they afford treatment of what would be the majority of the adult population?
Home Blood Pressure Measurement (HBPM)
A recent “call to action” statement from the American Heart Association has recommended wider use of HBPM to overcome “many of the limitations of traditional office BP measurement,” adding that it is “both cheaper and easier to perform than ambulatory BP monitoring” (2). This report recommended that HBPM should become a routine component of BP measurement for the majority of patients with known or suspected hypertension using validated oscillometric monitors that measure BP on the upper arm with an appropriate cuff size. It went on to recommend that 2 to 3 readings should be taken while the subject is resting in the seated position, both in the morning and the evening, over a period of 1 week. An average total of ≥12 readings is recommended for making clinical decisions. This method is widely used in Europe, and guidelines from Europe for HBPM were discussed in last year's review of the year in hypertension (3). Evidence is accruing to suggest that HBPM may be a better predictor of target organ damage and also helps reduce the “white coat effect” and determine the presence of “masked hypertension.” The levels of average HBPM considered normal by the majority of the guidelines are <135 mm Hg for average systolic BP and <85 mm Hg for average diastolic BP (see Pickering et al.  for references). The Japanese guidelines regard “definite normotension” as an average pressure of <125/75 mm Hg and “definite hypertension” as ≥135/85 mm Hg, and the British Hypertension Society stated that average home BP levels of <130/85 mm Hg can “probably be regarded as normal.” The World Health Organization–International Society of Hypertension Guidelines recommended an upper limit of normality as 125/80 mm Hg, whereas the 2007 European Society of Hypertension/European Society of Cardiology (ESH/ESC) guidelines suggested that the definition of hypertension for HBPM was ≥130 to 135 mm Hg for systolic BP and ≥85 mm Hg for diastolic BP. Thus, there are similarities in these thresholds but no international consensus as yet. Most patients are suitable for HBPM but do need instruction in the methods. Those unsuitable include patients with atrial fibrillation or other significant cardiac rhythm disturbances that make automated oscillometric devices unreliable.
Nutritional Aspects: Dairy Products and Incident Hypertension
There has been much interest in the relationship between dairy product consumption and incident hypertension. A previous cross-sectional study from Framingham has suggested that children who consumed 2 or more servings of dairy products per day during their preschool years had a smaller rise in systolic BP throughout their childhood years and a lower systolic BP during adolescence (4). Other studies have also suggested an inverse relationship between dairy product consumption and metabolic syndrome, incident type 2 diabetes mellitus, and cardiovascular disease (5–7). A recent large prospective cohort study of middle-aged and older women also demonstrated an inverse association between low-fat dairy product intake and the risk of incident hypertension (8). The association was moderate, graded, and independent of known risk factors for hypertension. Of interest, there was no such relationship if the dairy intake was high fat, suggesting that the fat content of dairy intake influences the strength of the relationship.
Vitamin D levels influence calcium uptake, and there was also an inverse relationship between the calcium and vitamin D content of the diet and incident hypertension (8). Moreover, the association between low-fat dairy intake and incident hypertension was substantially attenuated after adjustment for the dietary calcium content, pointing to the importance of calcium intake in the relationship between low-fat dairy product consumption and incident hypertension. Of interest, there was no such relationship if the calcium and vitamin D were provided as supplements. It is unclear why high-fat dairy intake attenuates the aforementioned relationships, but it is conceivable that high saturated fat content offsets the cardiovascular benefits and/or impedes intestinal calcium absorption. These findings point to the potential importance of low-fat dairy product intake in the primary prevention of hypertension and cardiovascular disease and lend support to the 2005 dietary guideline from the U.S. Department of Agriculture, which recommended that the majority of Americans increase their intake of dairy products to 3 portions per day. These new data suggest that this recommendation should be qualified by emphasizing that the dairy products should be low fat. If implemented, such recommendations could have substantial public health implications.
Recent BP-Lowering Trials and Their Implications for Clinical Practice
In 2008, the results of 5 major clinical outcomes trials testing BP-lowering drugs were reported. These studies addressed some fundamental questions and succeeded in providing key information that will no doubt help frame future treatment guidelines—medical politics notwithstanding. These trials and their implications are reviewed below.
The ONTARGET (Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial) program
The ONTARGET program (9) addressed 2 important questions: 1) Is telmisartan, an angiotensin-receptor blocker (ARB) noninferior to ramipril, an angiotensin-converting enzyme (ACE) inhibitor, at preventing a composite of cardiovascular events in a population of patients at high risk for cardiovascular disease? 2) Is the combination of ramipril and telmisartan more effective than ramipril alone at preventing cardiovascular events? The demographics of the study population of 25,620 patients were similar to those of the HOPE (Heart Outcomes Prevention Evaluation) trial (10), which established that the ACE inhibitor ramipril reduced cardiovascular morbidity and mortality, namely, the patients had evidence of vascular disease or high-risk diabetes mellitus without heart failure at baseline. The mean age of ONTARGET participants was ≈66.5 years, 63% were male, 85% had cardiovascular disease, 69% were hypertensive, and 38% were diabetic. However, when compared with the HOPE trial, many more patients in the ONTARGET study were receiving proven cardiovascular risk-reducing therapies at baseline, including statins (62%), antiplatelet therapy (82%), beta-blockers (57%), and diuretics (28%). The primary outcome of the ONTARGET study was similar to that of the HOPE trial, namely, a composite of death from cardiovascular causes, myocardial infarction, or stroke, but also included hospitalization for heart failure. The median follow-up of the study was 56 months. An interesting feature of the design was a 3-week run-in period during which patients were exposed to low-dose ramipril and then ramipril plus telmisartan to establish tolerability of the treatment regimen before randomization. Of those who were randomly allocated to the run-in, 12% withdrew. The mean BP at randomization was 148/82 mm Hg, and the fall in BP during the trial was −6.0/4.6 mm Hg with ramipril, −6.9/5.2 mm Hg with telmisartan, and −8.4/6.0 mm Hg with the combination—the differences were small but statistically significant.
The comparison of telmisartan 80 mg daily versus ramipril 10 mg daily showed that telmisartan was noninferior to ramipril with regard to the primary end point, and this was consistent for all components of the primary outcome (9). This conclusion was unchanged after adjustment for the small differences in BP. Turning to the comparison of the combination of telmisartan with ramipril, versus ramipril alone, it was a surprise to many that the combination was not superior with regard to the primary outcome, mindful of the fact that BP was lower with the combination. Moreover, there were significantly more withdrawals from combination therapy and more adverse events, especially the onset of renal impairment.
The ONTARGET trial suggests that the ARB was likely to be equivalent to the ACE inhibitor at preventing cardiovascular events in this high-risk population and was better tolerated, with less withdrawal from therapy. The better tolerability of ARBs was most likely underestimated by the ONTARGET trial because drug exposure during the run-in excluded patients (12%) who failed to tolerate lower doses of the trial medications. The finding of noninferiority of the ARB challenges the controversial conclusions generated by some previous meta-analyses that ARBs may be less effective at preventing myocardial infarction than ACE inhibitors are. The findings of the ONTARGET study provide doctors with reassurance that when this controversial hypothesis was formally tested in a major randomized clinical trial, there was no signal of inferiority with the ARB in preventing myocardial infarction. Thus, in general, an ARB can be considered a better tolerated, equivalent, and suitable alternative to an ACE inhibitor.
One of the surprises of the ONTARGET study was that the differential BP lowering between treatments and the longer duration of action of ARBs versus ACE inhibition, did not translate into a better outcome, especially for stroke. There are a number of considerations with regard to these findings. First, the reported data show only average BP changes, and it is conceivable that patients with higher baseline pressures achieved some benefit with greater BP lowering and patients with lower baseline BP achieved no benefit, or harm. Thus, the net effect was neutral. This hypothesis has been subsequently evaluated by examination of outcomes stratified according to baseline BP and the in-trial BP change (11). Second, another consideration is that the baseline BP was not especially high and many patients were receiving concomitant BP-lowering medications. Previous studies have shown that telmisartan is significantly more effective than ramipril at lowering BP over 24 h (12), and it is conceivable that had the study been conducted among patients who were overtly hypertensive, this BP-lowering advantage of telmisartan may have translated into a greater differences in BP throughout the trial, and a potentially a different outcome. In this regard, analysis of the 24-h BP data from an ONTARGET substudy will be of interest. Third, this study also generates uncertainty about the benefits of aggressive BP lowering in this population of patients and underscores the need for further trials exploring the safety and efficacy of BP lowering in patients at high cardiovascular disease risk in whom BP is not overtly elevated at baseline.
The data from the ONTARGET study with regard to the lack of benefit and potential harm of an ACE inhibitor/ARB combination has had a more telling impact. It has effectively shelved the launch of such combinations by pharmaceutical companies and suggests that doctors should not be using this combination in routine clinical practice. There are some circumstances where the use of an ACE inhibitor/ARB combination will still be used, such as for patients with severe heart failure or patients with advanced chronic renal disease and heavy proteinuria, but this requires careful consideration of the balance between benefit and harm, as is discussed in the following text in relation to renal disease. This finding from the ONTARGET study also raises intriguing questions about whether other strategies to deliver dual blockade of the renin-angiotensin system, namely, ACE inhibition or ARB plus a direct renin inhibitor, will be more effective. Such important questions are being addressed by the ongoing clinical trial program with aliskiren, a direct renin inhibitor that is being added to existing therapy for patients already being treated with an ACE inhibitor or ARB who have heart failure or diabetic nephropathy.
Dual renin-angiotensin system (RAS) blockade and renal disease: on or off target?
A pre-specified secondary analysis of the ONTARGET trial examined the impact of ramipril and telmisartan and the combination on regression of proteinuria and a primary composite outcome of dialysis, doubling of serum creatinine, and death (13). The number of events for the composite primary outcome was similar for telmisartan (n = 1,147 [13.4%]) and ramipril (n = 1,150 [13.5%]) but was significantly increased with combination therapy (n = 1,233 [14.5%]). The urinary albumin/creatinine ratio increased in all groups during the trial, but least with the combination. Moreover, the risk of developing new microalbuminuria was similar with ramipril or telmisartan but was significantly less with the combination. Thus, there appeared to be dissociation between the effects of the treatments on albuminuria and risk of developing the primary composite outcome. This was a red rag to the nephrological bull, challenging the holy grail of nephrology, which has long associated more effective reduction of albuminuria with nephroprotection. There are, however, a number of important caveats to consider with regard to this renal outcome data before jumping to simplistic conclusions. 1) The primary composite end point examined in the ONTARGET study renal outcomes report was dominated, and primarily driven, by death. The number of patients progressing to chronic dialysis, or experiencing a doubling of serum creatinine, was not significantly different between groups. This is an important message in itself, highlighting that patients at high cardiovascular disease risk and/or with diabetes, whether or not they had significant albuminuria or renal impairment, were much more likely to die than reach a conventional renal end point. 2) The percentage of patients in the ONTARGET study with significant renal impairment at baseline was low, and the scrutiny of renal disease progression was not as rigorous as would normally be employed in a dedicated renal outcomes study—moreover, only 98 of >25,000 patients progressed to chronic dialysis. Thus, the finding from the ONTARGET renal substudy should not undermine previous abundant clinical trial data demonstrating the importance of RAS blockade in delaying the progression of more advanced chronic renal disease (namely, in patients with an estimated glomerular filtration rate <60 ml/min and overt proteinuria >300 mg/day) due to diabetic and nondiabetic nephropathy. 3) It is also possible that patients who experienced renal end points such as dialysis with the combination therapy did so because more aggressive RAS blockade predisposed them to more acute renal failure due to acute disturbances to volume homeostasis, rather than accelerated progression of chronic renal impairment. Review of the data supports this conclusion—the significant difference in requirement for dialysis between treatment groups reflected more acute dialysis, rather than chronic dialysis, of patients treated with the combination.
In summary, the renal data from the ONTARGET study suggest that an ACE inhibitor/ARB combination has no advantages and should not be routinely used for patients with chronic renal disease and low-grade albuminuria and a low risk of developing end-stage renal disease. For such patients, treatment with optimal doses of an ACE inhibitor or ARB as part of their BP-lowering treatment regimen would be more appropriate. As for patients with advanced chronic renal disease and more significant proteinuria, further studies are required to better define the place in treatment for an ACE inhibitor/ARB combination or other combinations designed to further enhance the inhibition of the RAS—the data from the ONTARGET study does not preclude such studies; indeed, it mandates them.
Turning to other combinations to deliver dual blockade of the RAS with regard to nephroprotection, Parving et al. (14) recently reported the results of the AVOID (Aliskiren in the eValuation of prOteinuria In Diabetes) study, which studied 599 patients with diabetic nephropathy and proteinuria. The patients had a baseline estimated glomerular filtration rate of ≈68 ml/min/1.73 m2and a baseline albumin/creatinine ratio of ≈500, indicative of proteinuria. They were all treated with a maximum recommended dose of losartan 100 mg daily in addition to other BP-lowering medication, which achieved an average baseline BP of ≈135/78 mm Hg. The patients were then randomly assigned to the addition of aliskiren 300 mg daily, a direct renin inhibitor, or placebo, in addition to their existing medication. That resulted in a comparison of dual RAS blockade, namely, direct renin inhibitor plus ARB versus an ARB plus placebo. The primary outcome was the reduction in the albumin/creatinine ratio in an early morning urine sample, 6 months after randomization. The addition of aliskiren-based treatment significantly reduced the mean urinary albumin/creatinine ratio by 20%. Moreover, when comparing treatment groups, there were minimal differences in BP (−2/1 mm Hg in favor of aliskiren) and no significant differences in adverse event rates. Careful monitoring of potassium levels will be required in patients with renal impairment who receive dual RAS blockade, but the incidence of significant hyperkalemia (≥6 mmol/l) was surprisingly low (aliskiren plus ARB, 4.7%; placebo plus ARB, 1.7%) and was often transient, leading to study withdrawal of very few patients (14). These results are promising, and it will be of interest to see whether an ongoing clinical outcomes study of patients with diabetic nephropathy confirms the superiority of dual RAS blockade with an ARB and aliskiren versus ARB-based therapy alone (15).
The TRANSCEND (Telmisartan Randomised Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease) study
The TRANSCEND study was complementary to ONTARGET and randomized 6,666 patients with a profile similar to that of patients in the ONTARGET study but who were intolerant of ACE inhibition. The patients were randomly allocated to telmisartan 80 mg/day versus placebo (16). The composite primary end point was identical to that in the ONTARGET study (see preceding text). The median follow-up was 56 months, mean BP at baseline was 141/82 mm Hg, and 76% of the patients were hypertensive, the majority receiving concomitant treatment with other BP-lowering medications. The BP difference throughout the trial was 4.0/2.2 mm Hg lower with telmisartan versus placebo. The event rate for the primary outcome was lower with telmisartan (15.7%) than with placebo (17%), but this was not significant. Review of the components of the primary composite outcome showed that there were fewer myocardial infarctions and strokes in the telmisartan group than in the placebo group, but this was not significant. There were also nonsignificant reductions in new diabetes and new left ventricular hypertrophy in the telmisartan group. Total mortality, the number of cardiovascular deaths, and hospitalizations for heart failure were similar between the 2 groups. This result will come as a surprise to many because active BP-lowering treatment would be expected to be more effective than placebo at reducing cardiovascular events. Moreover, based on usual data with BP-lowering monotherapy, one might have expected a greater difference in BP between active treatment and placebo. There are a number of potential explanations. First, it is often the case that when active BP-lowering treatment is compared with placebo, despite the blinded nature of the trial, add-on BP-lowering therapy is greater in the placebo group, reducing the anticipated difference in BP between groups. This clearly happened in the TRANSCEND trial because there was significantly more additional use of ARBs, diuretics, calcium-channel blockers (CCBs), and ARBs in the placebo versus telmisartan groups. That would serve to minimize the difference in BP between groups and hence the power of the study to detect a difference in the primary outcome. This confounding, namely, greater add-on therapy, in the placebo arm has happened before, in the SCOPE (Study on Cognition and Prognosis in the Elderly) trial (17), when a comparison between candesartan versus placebo in elderly hypertensive patients also generated a null result because there was a less than expected difference in BP between groups. Second, BP lowering would also be expected to reduce the rate of hospitalization for heart failure, but this was not different in the TRANSCEND study between groups. This observation is most likely explained by the use of more add-on diuretic therapy among the placebo-treated patients. Moreover, as heart failure was the second most commonly occurring component of the primary outcome, that would have further diminished the power of the study. Third, overall, the TRANSCEND study seems to lack statistical power for reasons highlighted in preceding text but also because of the number of patients included in the trial and its duration. It is difficult to reconcile how a study modeled on a patient population and design similar to the HOPE trial could have adequate statistical power with only 5,926 patients compared with 9,000 patients in the HOPE trial (10). Moreover, the TRANSCEND study patients received better background treatment, for example, statins and antiplatelet therapy and thus experienced a lower event rate when compared with the HOPE trial patients. Thus, one has to conclude that the TRANSCEND study result looks like a problem of statistical power. With this in mind, the result of the ONTARGET study becomes especially important and reassuring in showing noninferiority of telmisartan in a head-to-head comparison with ramipril (9).
The PROFESS (Prevention Regimen for Effectively Avoiding Second Strokes) trial and BP lowering after stroke
The PROFESS trial compared telmisartan and placebo in 20,332 patients who had recently suffered an ischemic stroke (18). The primary outcome was recurrent stroke. The median time from stroke to randomization was only 15 days, and the mean follow-up was 2.5 years. The BP difference between groups was once again less than anticipated at only 3.2/2.0 mm Hg. There was no difference in the primary outcome, although recurrent stroke, major cardiovascular events, and new onset of diabetes were modestly but nonsignificantly lower among the patients assigned to telmisartan. This result was also surprising when one considers the prior evidence, for example, the PROGRESS (Perindopril Protection Against Recurrent Stroke Study) (19), showing that BP lowering is an effective strategy to reduce the risk of recurrent stroke. So, how do we reconcile this result? First, the baseline BP was slightly higher and the BP lowering substantially greater (−12.3/5.0 mm Hg) in the PROGRESS study. This result is explained in that the PROGRESS study used a combination of an ACE inhibitor/diuretic (perindopril/indapamide) for most patients. Indeed, among the patients in the PROGRESS study who only received perindopril and thus experienced a lesser fall in BP (−4.9/2.8 mm Hg), there was no significant reduction in stroke. The smaller than anticipated fall in BP with telmisartan in the PROFESS trial is most likely explained by the fact that add-on BP-lowering therapy in the placebo group was greater, reflecting the desire of physicians to improve BP control in patients after stroke. Second, the median time to randomization to BP-lowering therapy was substantially different when comparing the PROFESS study (15 days) with the PROGRESS study (8 months), and the durations of the studies were also different: PROFESS 2.5 years versus PROGRESS 4.5 years. This latter point may be crucial because the outcomes between placebo versus telmisartan in the PROFESS trial did appear to be diverging after ≈2 years in favor of telmisartan. Indeed, in previous studies such as HOPE (10) and LIFE (Losartan Intervention for Endpoint Reduction in Hypertension) (20), which ultimately showed significant benefits of an ACE inhibitor or ARB, respectively, on stroke rates, there was little difference in outcomes between treatment groups in the first 12 months, suggesting the importance of longer follow-up. Maybe the 2.5 years of follow-up in the PROFESS study was simply too short to reveal a significant difference in outcomes, especially in light of the smaller than expected difference in BP between groups. Finally, the PROFESS study adopted a factorial design that also compared different antiplatelet strategies—there was no evidence of an interaction between these treatment strategies with regard to the outcomes, so this is unlikely to have influenced the outcome (18).
The somewhat surprising results of the TRANSCEND and PROFESS studies may thus be explained by inadequate statistical power and study duration, respectively. A pre-specified combined analysis of data from the TRANSCEND and PROFESS studies provided more robust power to examine the effects of telmisartan versus placebo and showed a significant benefit of telmisartan therapy on a primary composite outcome of cardiovascular death, myocardial infarction, and stroke with or without heart failure (16). This finding lends support to the aforementioned interpretation of the TRANSCEND study and PROFESS study results.
Are you never too old to start treatment for hypertension? The results of HYVET (Hypertension in the Very Elderly Trial)
International hypertension treatment guidelines have remained appropriately cautious with regard to treatment recommendations for people over the age of 80 years because of a lack of data confirming the safety and efficacy of treatment. There were legitimate concerns that the potential benefits of BP lowering could be offset by harm due to complications such as hypotension-related syncope and mortality. Indeed, a pilot study for a trial of BP lowering in the very elderly reported a reduction in stroke risk that was offset by an increased risk of mortality with BP-lowering therapy (21). Understanding the balance of benefit versus harm with BP lowering has been compounded by the very elderly invariably having been excluded from recruitment into prior major clinical outcomes trials of BP-lowering therapy, and as a consequence, there was inadequate evidence upon which to base clinical guidance. These important issues were addressed by the HYVET study of BP-lowering treatment in the very elderly (22). The HYVET study randomized 3,845 patients from Europe, China, Australasia, and Tunisia who were 80 years of age or older (mean age 83.6 years) and had a sustained systolic BP of 160 mm Hg or more (mean seated BP at baseline was 173.0/90.8 mm Hg). The patients received active treatment with the diuretic indapamide (sustained release, 1.5 mg daily) or matching placebo. The ACE inhibitor perindopril (2 or 4 mg), or matching placebo, was added when required to achieve the target BP of 150/80 mm Hg. The primary end point was fatal or nonfatal stroke. The baseline history of cardiovascular disease was only 11.8%, indicating that this was a relatively healthy elderly population. Blood pressure was reduced to ≈143.5/75.4 mm Hg with active treatment and to 158.5/84 mm Hg with placebo, a difference in BP lowering of ≈15/8.6 mm Hg for a median follow-up of 1.8 years. The study was discontinued on the recommendation of the Data Safety Monitoring Board after a median follow-up of 1.8 years because of a significant reduction in all-cause mortality of 21% in favor of active treatment. This finding was unexpected, mindful of the data from the HYVET pilot study, which had suggested the possibility of an increase in mortality with BP-lowering treatment. At the time of study cessation there had also been a 30% reduction in the rate of fatal or nonfatal stroke, which just failed to reach significance, and a 39% significant reduction in the risk of fatal stroke. There was also an impressive 64% reduction in the rate of heart failure, which was highly significant. Reassuringly, fewer serious adverse events were reported in the active-treatment group. These findings are important and are likely to influence recommendations for the treatment of the very elderly in future guidelines. An important caveat is that the HYVET study population appeared healthier than is often the case for people at this age, emphasizing that doctors must always assess the potential of treatment with regard to the expectations of benefit in the individual patient. That said, the HYVET study has provided unequivocal evidence that the benefits of BP lowering in the very elderly can be very substantial, removing justification for equivocation. As a consequence of this landmark finding, many more very elderly patients will be actively treated with BP-lowering therapy.
BP-lowering therapy and cognitive function
The HYVET study also included an assessment of the impact of BP-lowering therapy on incident dementia in the very elderly (23). There were fewer cases of dementia in the actively treated group but this result did not reach significance, probably because of the very short duration of the study. When the data from the HYVET study were combined in a meta-analysis with data from other placebo-controlled studies that have assessed the impact of BP-lowering therapy on incident dementia, there was a reduction in incident dementia of marginal significance with active treatment (23). Nevertheless, it should be recognized that clinical trials are of relatively short duration and by necessity use unsophisticated methods to assess cognitive function. If the impact of BP-lowering therapy over a lifetime of treatment serves to better preserve cognitive function later in life, which I think it will, then this could turn out to be one of the most compelling reasons for treating hypertension in younger people.
Incident heart failure in hypertension: new data from ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial)
The HYVET study demonstrated a very large reduction in incident heart failure among hypertensive very elderly patients treated with diuretic-based therapy. A recent analysis of the ALLHAT study underscores the strong influence of diuretic therapy on incident heart failure among high-risk hypertensive patients (24). The report examined incident hospitalized heart failure in patients with preserved ejection fraction (≥50%) or reduced ejection fraction (<50%). In the ALLHAT study, 42,418 high-risk hypertensive patients were randomly assigned to chlorthalidone, amlodipine, lisinopril, or doxazosin. In a pre-specified substudy, ejection fraction was estimated by contrast ventriculography, echocardiography, or radionuclide study in 910 of 1,367 patients (66.6%) with hospitalized heart failure events meeting ALLHAT study criteria. Chlorthalidone significantly reduced the risk of heart failure with preserved ejection fraction as compared with amlodipine (−31%), lisinopril (−26%), or doxazosin (−47%). Chlorthalidone also significantly reduced the risk of heart failure with reduced ejection fraction when compared with amlodipine (−26%) or doxazosin (−39%). The risk of heart failure with reduced ejection fraction was similar with chlorthalidone and lisinopril. The report also provided a stark reminder that the development of heart failure in hypertensive patients has serious consequences. Once heart failure had occurred in the ALLHAT study, death was more common among patients with reduced ejection fraction (42%) when compared with patients who had preserved ejection fraction (29%). These data underscore the importance of early and optimal intervention to control BP to prevent the onset of heart failure and the effectiveness of diuretic-based therapy in achieving this objective.
Combination therapy for hypertension: mission accomplished?
Continuing the theme of trials that are likely to influence clinical practice, the ACCOMPLISH (Avoiding Cardiovascular Events Through Combination Therapy in Patients Living With Systolic Hypertension) trial reported its principal findings in 2008 (25). We know that combinations of BP-lowering drugs are required to achieve optimal BP control for most patients. Despite this, the majority of trials of BP-lowering therapy have focused on initial treatment with monotherapy, and there has been too little emphasis on the potential importance of the composition of subsequent combinations of therapies. The ACCOMPLISH trial specifically evaluated whether the type of combination therapy is important in influencing clinical outcomes by comparing 2 different single-pill combinations, namely, an ACE inhibitor (benazepril) plus a diuretic (hydrochlorthiazide), versus the same ACE inhibitor combined with a CCB (amlodipine). This randomized, double-blind trial included 11,506 high-risk hypertensive patients with a mean age of 68 years; 60% had diabetes, and many had clinical coronary heart disease (23% with previous myocardial infarction and 36% with previous coronary revascularization), 13% had prior stroke, and a similar percentage had electrocardiographically documented left ventricular hypertrophy. Furthermore, before randomization, most patients (97%) were already being treated for hypertension, with 75% receiving 2 or more classes of antihypertensive medications. Despite this, only 37.3% had a BP <140/90 mm Hg at baseline, indicating that as well as being high risk, they were also most likely a group with quite severe hypertension. The primary outcome of the ACCOMPLISH trial was a composite of death from cardiovascular causes, nonfatal myocardial infarction, nonfatal stroke, hospitalization for angina, resuscitation after sudden cardiac arrest, and coronary revascularization. During the trial, achieved BP values were impressive (≈132/74 mm Hg), differed little (≈1/1 mm Hg at trough) between treatment groups, and ≈75% achieved the BP goal of <140/90 mm Hg. The trial was terminated early after a mean follow-up of 36 months, when the boundary of the pre-specified stopping rule was crossed. There was a significant ≈20% reduction in the primary end point in favor of the benazepril-amlodipine treated group, and all components showed this trend. This difference does not appear to be due to differences in BP control between groups, a conclusion likely to be supported by data from an ambulatory BP monitoring substudy. Whatever the explanation for the difference in primary outcomes between treatment groups, the ACCOMPLISH study has established that combining a CCB with ACE inhibition, or presumably other forms of RAS blockade, is a very effective treatment option for high-risk patients with hypertension.
So, how do we interpret the data from the ACCOMPLISH trial alongside that from other trials? Well, no doubt some will suggest the era of the diuretic is waning and that a CCB should now be the preferred component of any combination, and perhaps for this specific population it should. But I would urge caution about generalization of data from a single trial. As impressive as the data are, 1 trial should not dictate a treatment guideline. What we need is some balance and pragmatism. There is no single optimal treatment for everybody with hypertension, and once we reach a requirement for 2-drug combinations, physicians need choices. Most physicians would choose some form of RAS blockade as part of their preferred treatment strategy for hypertension. The data from the ACCOMPLISH trial, alongside a wealth of data with diuretics, now provides 2 excellent options to combine with RAS blockade, namely, a diuretic or a CCB, or perhaps both for patients with more resistant hypertension.
Single-pill combination therapy as initial therapy for hypertension?
The data from the ACCOMPLISH study also showed that when patients were transferred to either of the single-pill combinations tested in that study, there was a dramatic improvement in BP control, with >70% achieving their BP target (25). The reasons for this are not immediately obvious but may in part relate to improved concordance with therapy in a clinical trial setting. It does lend some support to current guideline recommendations of initial therapy with a low-dose, single-pill combination for patients whose BP is >20/10 mm Hg above their target BP (26,27). In my view, the preferred combination based on current evidence would include RAS blockade and either a CCB or a diuretic. The likelihood of this approach producing greater BP lowering than monotherapy alone seems obvious. The main concern has been safety and tolerability with regard to potentially large initial BP declines in treatment-naïve patients. The ACCOMPLISH trial cannot address the latter because almost all patients were already treated for their hypertension immediately before randomization.
Resistant hypertension is defined as BP that remains above goal despite the concurrent use of 3 antihypertensive agents of different classes. Ideally, 1 of the 3 agents should be a diuretic, and all agents should be prescribed at optimal dose amounts (28). This practice is more common for older patients and for obese patients. It is also more common for patients with diabetes, chronic kidney disease, and target organ damage, namely, high-risk populations for whom the benefits of optimizing BP control would be greatest. A recent scientific statement from the American Heart Association discusses the evaluation and treatment of resistant hypertension (28). Of interest, a recent paper in the Journalreported that a common pathogenetic factor contributing to treatment resistance in hypertensive patients is large artery stiffening (29). In that study, the cohort with the highest carotid-femoral pulse wave velocity at baseline (indicative of arterial stiffening) was the least likely to have their BP controlled. This observation suggests that the many factors associated with the development of resistant hypertension might exert their effects through accelerated aging and stiffening of large conduit arteries, putting the aorta and its functional characteristics at the heart of resistant hypertension (30). That would also explain why most resistant hypertension is due to poorly controlled systolic rather than diastolic BP (31).
With regard to treatment of resistant hypertension, the evidence base is thin, and this could be a fruitful avenue for therapeutic development, especially as the numbers of patients affected will continue to rise as populations age. This interest is reflected by the various classes of BP-lowering drugs in various phases of development that may emerge as potential therapies for patients with resistant hypertension. These drugs include selective endothelin receptor antagonists, aldosterone synthase inhibitors, longer-acting formulations of clonidine, and novel diuretic agents.
A novel approach to treating patients with resistant hypertension using radiofrequency ablation (RFA) of renal sympathetic nerves was recently reported (32). The rationale for this approach is that renal sympathetic tone plays a role in the pathogenesis of resistant hypertension. The RFA was used to destroy renal sympathetic innervation to the kidney through a percutaneous catheter in the renal artery. This technique was used for 50 patients with resistant hypertension who were then followed up for as long as 12 months. There was 1 intraprocedural renal artery dissection, but no other renovascular complications. The RFA was associated with impressive systolic and diastolic BP reductions of >20 mm Hg and 10 mm Hg, respectively, which were sustained for as long as 12 months of observation. These BP reductions are similar to those observed in the therapeutic studies of resistant hypertension reviewed last year (3). These initial studies of RFA of the renal nerves seem to justify this novel approach to treatment, but further large-scale studies are required to determine the place of this approach in more routine clinical practice.
Treatment Strategies and Lessons for Guidelines: Simplicity Wins!
A concerning aspect of modern guideline development is their increasing complexity. This contrasts with my own guiding principle, which is, if clinical guidance cannot be summarized on a single page, it probably will not work. Of course, the analysis and explanation of how the single page was derived should be detailed, but the treatment algorithm should be simple. The BHS/NICE (British Hypertension Society/National Institute for Clinical Excellence) guidance for hypertension treatment in the United Kingdom exemplified this with its simple A/CD algorithm, which not only adopted a traditional step-wise approach but also specified the classes of treatment and preferred combinations to be used (33). In this regard, the recent report of a similar simple treatment algorithm tested in Canada was of interest (34). This study cluster randomly designated 45 general practices to use a simplified hypertension treatment algorithm or the current Canadian hypertension education program guidelines. The primary outcome was the proportion of patients achieving their target BP after 6 months. The simplified treatment algorithm was similar to the BHS/NICE guidelines in that it focused on the use of RAS blockade, namely, ACE inhibition or ARB in combination with a diuretic or a CCB. There was, however, an important difference in the Canadian study approach in that all patients began treatment with low-dose combination therapy, using an ACE inhibitor or ARB with a diuretic, up-titrating the combination therapy to the highest dose at step 2. Addition of a CCB was recommended at the next step. The proportion of patients achieving the target was significantly higher with the simplified guideline (64.7% vs. 52.7%). This finding sends out an important message to guideline developers: simplicity and pragmatism win in the real world of hypertension treatment. With that, here ends a remarkable year in hypertension research.
- Received July 13, 2009.
- Accepted August 25, 2009.
- American College of Cardiology Foundation
- Pickering T.G.,
- Houston Miller N.,
- Ogedegbe G.,
- Krakoff L.R.,
- Artinian N.T.,
- Goff D.
- Williams B.
- Bostick R.M.,
- Kushi L.H.,
- Wu Y.,
- Meyer K.A.,
- Sellers T.A.,
- Folsom A.R.
- Parving H.H.,
- Brenner B.M.,
- McMurray J.J.,
- et al.
- The Telmisartan Randomised Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease (TRANSCEND) Investigators
- Peters R.,
- Beckett N.,
- Forette F.,
- et al.
- Davis B.R.,
- Kostis J.B.,
- Simpson L.M.,
- et al.
- Protogerou A.,
- Blacher J.,
- Stergiou G.S.,
- Achimastos A.,
- Safar M.E.
- Williams B.
- Higgens B.,
- Williams B.,
- Guideline Development Group
- The Global Burden of BP-Related Morbidity and Mortality
- Home Blood Pressure Measurement (HBPM)
- Nutritional Aspects: Dairy Products and Incident Hypertension
- Recent BP-Lowering Trials and Their Implications for Clinical Practice
- Treatment Strategies and Lessons for Guidelines: Simplicity Wins!