ASH 2010: New insights in hypertension from ACCORD, TNT, VALUE studies

The American Society of Hypertension celebrated its 25th anniversary in New York on May 1, and the society president, Henry Black, MD, together with the program chairman, William White, MD, put together an outstanding program for this event. Clearly, space does not allow us to cover comprehensively what was presented in three and a half days. However, in the following column, I would like to briefly discuss a few outstanding presentations, and I will do so in no particular order.

William Cushman, MD, of the Memphis VA Medical Center in Tennessee, presented a re-analysis of the ACCORD trial. The principal findings of the ACCORD trial were presented in March at the American College of Cardiology Scientific Sessions and were published in the New England Journal of Medicine shortly thereafter. Briefly, patients with type 2 diabetes who were treated more aggressively to a goal systolic BP of less than 120 mm Hg did not have a significant reduction in the primary outcome (a composite endpoint of CV events) when compared with standard treatment, which lowered BP to less than 140 mm Hg. Not only was there no significant reduction in CV events, but serious adverse events were more common in the intensive treatment group (3.3% vs. 1.3%, P<.0001). The concluding words in the abstract were that targeting systolic BP to less than 120 mm Hg as compared with less than 140 mm Hg “did not reduce the rate of a composite outcome of fatal and nonfatal major CV events.”

This was certainly a disturbing finding, given that ever since the Syst Eur and HOT studies, we thought and taught that diabetes renders patients particularly susceptible to the effects of BP lowering and that indeed a given fall in BP conferred more benefits in patients with diabetes than in the patient without diabetes. Not surprisingly, most (if not all) of the recent hypertension guidelines adopted lower BP goals for those with diabetes and for those without diabetes.

Franz H. Messerli

Stroke is the most BP–dependent CV outcome, and Cushman’s presentation at ASH focused on stroke reduction, which indeed was significant (HR=0.59; 95% CI, 0.39-0.89). Cushman indicated that the number needed to treat to prevent new onset stroke during five years was 89. When asked why this highly meaningful finding was not included in the summary of the main article in the New England Journal of Medicine, Cushman indicated that the researchers were told by the editors/reviewers that it was most likely due to chance and therefore would not allow the researchers to put it in the summary. However, we should consider that the average BP difference between the two arms was more than 14 mm Hg/6 mm Hg and that stroke is the most BP–dependent complication of hypertensive CVD.

Cushman also discussed some interactions showing that patients with diabetes were easier to control, and perhaps those treated less intensively benefited more from BP reduction than those with more difficult to control diabetes.

It is perhaps not so surprising that no cardiac benefits were observed in ACCORD. We know that lowering diastolic BP excessively in patients with CAD can be counterproductive (J-curve). Thus, one reason for the absence of benefits may well be that in some patients, the diastolic BP was falling below critical levels and thereby, paradoxically, increased event rates. Such a J-curve has been much less evident with regard to cerebrovascular disease, and it is therefore unexpected that the same fall in BP significantly reduced the stroke rate in patients with diabetes.

Results from the TNT study

Prakash Deedwania, MD, a clinical professor of medicine at the University of California, San Francisco, presented a rather interesting re-analysis of the TNT (Treatment to Target) study. Much along the same line as ACCORD, he showed that systolic BP remained a powerful and significant risk factor for stroke but not for cardiac events. In contrast, LDL was a powerful risk factor for cardiac events but not for stroke. We have been suspecting these findings for quite some time. However, this is first time that in a 12,000 patient study such compelling data were documented. This again would indicate that the stroke-prone patient BP-lowering is more important than LDL control, whereas in the patient with CAD, the physician should primarily focus on LDL reduction.

Results and implications of the VALUE trial

Of great interest was the unique analysis of the VALUE study by one of its investigators, Stevo Julius, MD, of the University of Michigan Health System. We have known for decades that an elevated resting heart rate is associated with an increased risk for cardiac events. Although this concept is certainly somewhat speculative, it has been well documented that in certain disease states, heart rate is a rather powerful determinant of the risk.

Similarly, Julius has now shown in the VALUE study that resting heart rate as a risk factor should not be neglected. Little difference was seen as long as it remained less than 79 beats per minute. However, patients who had resting heart rate more than 79 beats permitted a highly significant increase in CV morbidity and mortality. Perhaps the most interesting finding was that heart rate was not only a strong risk factor at baseline but also remained so even in patients whose BP was otherwise well controlled. An increase in heart rate of 10 beats per minute from baseline increased HF by 25%, cardiac death by 20%, MI by 9% and stroke by 12%. Patients in the highest quintile had an OR of 1.46 for the primary endpoint, an OR of 1.88 for HF and an OR of 1.52 for sudden death. Whether we can conclude from this subanalysis of VALUE in more than 15,000 patients that heart rate-lowering by pharmacological means will turn out to be beneficial remains an open question. In a recent meta-analysis of 11 studies and 34,000 hypertensive patients, the effects of heart rate lowering by beta-blockade have been disappointing.

Similarly, in ASCOT, patients with the highest heart rate had a better outcome when treated with amlodipine rather than with atenolol (Tenormin, AstraZeneca). As for ivabradine (Procoralan, Servier), the only outcome study (BEAUTIFUL) had disappointing results in patients with CAD who usually benefit from beta-blockade. Thus, before we subject our patients with a resting heart rate above 79 beats per minute to heart rate cosmetics, we need more information as to the specific reasons of the heart rate elevation and the connection with morbidity and mortality.

A study of BP control in VA hospitals

Very convincing are the data presented by Vasilios Papademetriou, MD, of the Washington Hospital Center in Washington D.C., and colleagues, regarding BP control in VA hospitals. The researchers reported on an eight-year study of 15 VA medical centers across the United States. A decade ago, control rates were fluctuating between 40% and 50% in all patient groups. More recently, this number has increased so that now a stunning 79% of hypertensive patients in the VA system have had their BP at goal.

This success has been reached at practically no increased cost. Because all drugs are free in the VA system, cost was obviously not a barrier to the patient. In addition, electronic records as well as the use of automatic reminders were deemed to be helpful in improving compliance. The study was based on almost 500,000 hypertensive patients and was initiated more than a decade ago. Some centers actually reached a BP control that exceeded 80% of all patients. Of note, the VA study was even successful in patient groups who are known to be difficult to control, such as in blacks and Hispanics. Papademetriou also showed in a subsegment of 43,000 patients that BP control was related to outcome. They subdivided patients into three groups, including those whose BP was always controlled; those whose BP was elevated at between 1% and 25% of all visits; and those whose BP was elevated at between 76% and 100% of all visits. There was an astounding 47% reduction in mortality between the optimum BP control group and the poorly controlled BP group.

BP control with a novel angiotensin receptor antagonist

Another story that deserves our attention includes data with the new angiotensin receptor antagonist azilsartan medoxomil (Takeda). Even in monotherapy, this new angiotensin receptor antagonist seems to be an excellent antihypertensive drug, lowering BP somewhat more than most of the currently available angiotensin receptor antagonists. What makes this compound particularly interesting is that the company will provide it in a fixed combination with chlorthalidone instead of with hydrochlorothiazide. In the study presented by George Bakris, MD, director of the hypertension center at the University of Chicago Medical Center, azilsartan plus hydrochlorothiazide was compared with azilsartan plus chlorthalidone. After one month of treatment, patients assigned to the combination with chlorthalidone had a 5.6-mm greater reduction in systolic BP in the clinic and also had a highly significant better reduction in 24-hour ambulatory systolic pressure. We have known for years that chlorthalidone is a better antihypertensive — it has a smoother antihypertensive effect and is more powerful.

Furthermore, there are well-documented outcomes data for chlorthalidone, whereas for 12.5 mg to 25 mg of hydrochlorothiazide, there are no data showing that it reduces MI or stroke. Clearly, Takeda is to be congratulated for thinking outside the box and for combining their angiotensin receptor antagonist azilsartan with an efficacious thiazide diuretic.

Franz H. Messerli, MD, is the Director of the Hypertension Program at St. Luke’s Roosevelt Hospital in New York. Messerli is also a member of the Vascular Medicine section of the Cardiology Today Editorial Board.