VALUE: Head-to-head comparison of amlodipine and valsartan a draw

Susan Jeffrey

June 14, 2004

Paris, France - Final results of the Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) trial show that valsartan-based and amlodipine-based therapy were essentially equal in reducing the primary end point of cardiac morbidity or mortality.

Secondary end points showed a trend to more congestive heart failure with amlodipine, but stroke trended higher with valsartan, and fatal and nonfatal MI were significantly higher in the valsartan group, although fatal MI was equal between groups. New-onset diabetes, a prespecified end point, was highly significantly reduced with valsartan-based therapy, the researchers report.

The findings were presented here at the 14th European Meeting on Hypertension, the annual meeting of the European Society of Hypertension, coinciding with their online publication at the Lancet[1].

Dr Stevo Julius

"This was a large study, well executed, with very good overall control of blood pressure in both arms," said VALUE trial chair Dr Stevo Julius (University of Michigan, Ann Arbor).

There was some early difference in BP between the groups favoring amlodipine that was minimized over time, but final systolic pressures were still about 1.8 mm Hg lower with amlodipine, Julius told heartwire . "In spite of that, the main cardiac end point was no different between groups, although there were some interesting and important time-related trends."

The VALUE of BP lowering

The VALUE trial was designed to compare two treatment strategies, one based on amlodipine (Norvasc®, Pfizer), the other on valsartan (Diovan®, Novartis). The aim was to see whether, if similar blood pressure control could be achieved with both drugs, there would be any beneficial effect on cardiac end points with valsartana drug that blocks the renin angiotensin systembeyond blood-pressure lowering alone.

The trial randomized 15245 patients from 31 countries. All were hypertensive with at least one high-risk feature, including coronary artery disease, previous stroke or TIA, peripheral vascular disease, or diabetes. The trial began with a six-month titration period, during which patients began either with valsartan 80 mg or amlodipine 5 mg. Patients on previous antihypertensive treatment92% of this populationwere "rolled over" onto the assigned treatment without a washout period, an important safety feature, Julius said. After a month, if the target BP of <140/90 mm Hg was not reached, patients were uptitrated to 160-mg valsartan and 10-mg amlodipine; if control was still not achieved, hydrochlorothiazide was added in increasing doses. Finally, other antihypertensive agents could be added.

At the end of the six-month titration, 60% of patients were at target BP, with both regimens successfully lowering blood pressure. "This is the best control achieved in any of the trials, including ALLHAT, that started with lower BP," Julius said. However, there were some time-related differences in the BP achieved between groups, with those on amlodipine getting pressures lower faster. At one month, the difference in systolic pressure between groups was 4.0 mm Hg. The disparity waned over time to 1.8 mm Hg but was still present at the end of the trial.

VALUE: Systolic BP over time by treatment group

Time point





154.5+19.0 154.8+19.0 <0.0001

One month

152.7+17.5 148.7+17.4 <0.0001

Three months

149.2+19.5 145.4+16.1 <0.0001


139.3+17.6 137.5+15.0 <0.0001

Despite the differences in BP-lowering seen, after a mean follow-up 4.2 years, there was no significant difference in the primary end point of cardiac morbidity or mortality.

VALUE: Primary composite end point

End point

Valsartan, n (%)

Amlodipine, n (%)

Hazard ratio (95% CI)


Primary composite 810 (10.6) 789 (10.4) 1.04 (0.94-1.15) 0.49
Cardiac mortality 304 (4.0) 304 (4.0) 1.01 (0.86-1.18) 0.90
Cardiac morbidity 586 (7.7) 578 (7.6) 1.02 (0.91-1.15) 0.71

However, there were some disparities seen in secondary end pointsspecifically, a trend to more stroke and significantly more fatal and nonfatal MI with valsartan-based treatment, although fatal MI was equal between groups. There was more fatal and nonfatal congestive heart failure and significantly more new-onset diabetes with amlodipine-based therapy.

VALUE: Secondary and prespecified end points

End point

Valsartan, n (%)

Amlodipine,n (%)

Hazard ratio (95% CI)


Fatal and nonfatal MI 369 (4.8) 313 (4.1) 1.19 (1.02-1.38) 0.02
Fatal MI 66 (0.86) 64 (0.84) 1.04 (0.74-1.47) 0.81
Fatal and nonfatal heart failure 354 (4.6) 400 (5.3) 0.89 (0.77-1.03) 0.12
Fatal and nonfatal stroke 322 (4.2) 281 (3.7) 1.15 (0.98-1.35) 0.08
All-cause death 841 (11.0) 818 (10.8) 1.04 (0.94-1.14) 0.45
New-onset diabetes 690 (13.1) 845 (16.4) 0.77 (0.69-0.86) <0.0001

Safety data showed no significant difference in treatment discontinuations between groups. Significantly more peripheral edema was seen with amlodipine, 32.9%, vs 14.9% with valsartan, but significantly more dizziness, headache, and diarrhea were seen in the valsartan group. More angina was also reported with valsartan-based therapy, they note, 9.3% vs 6.4% with amlodipine.

The treatment of BP in this group of patients needs to be prompt, and blood pressure brought down vigorously, most likely starting with combination treatment.

Julius speculated that amlodipine's antianginal effects might account for the "peculiar" finding of more fatal and nonfatal MI with valsartan, but no difference between groups in fatal MI. "It is conceivable that these patients did not realize [they were having an event], did not go and get it diagnosed, but that's speculation," he said.

He also noted that the excess of MI and stroke with valsartan-based treatment was largely accounted for by events occurring during the early phase of treatment, when BPs were significantly different between groups. Similarly, the trend to a cardioprotective effect against CHF with valsartan-based treatment did not appear until later in the trial, when BPs were more similar.

These findings lead them to conclude that patients with stage 1 hypertension and cardiovascular risk factors should be treated aggressively, he said. "The treatment of BP in this group of patients needs to be prompt, and blood pressure brought down vigorously, most likely starting with combination treatment," he said.

Finally, he pointed to the difference in new-onset diabetes as a key finding. "This is important to know because this was the first time that we compared a drug that is not supposed to be bad for glucose metabolism, amlodipine, with valsartan, and we still see a major difference in favor of valsartan," he said.

Serial median matching

In a separate report, Dr Michael A Weber (State University of New York, Brooklyn) presented some post hoc analyses looking specifically at the effects that the BP differential between groups may have had on the outcomes seen[2].

Weber pointed out that because similar BPs had not been reached between groups, it was impossible for them to actually test the hypothesis they set out to explorethat at similar levels of BP lowering, blocking the renin angiotensin system might provide additional benefit. They did an analysis using a computerized technique called serial median matching, in which 5006 pairs of patients from each treatment group were matched exactly for systolic BP, age, sex, and the presence or absence of previous coronary disease, stroke, or diabetes at the six-month point. With these identical characteristics, he said, most end points, including MI and stroke, did not favor either drug, except heart failure, which significantly favored valsartan

VALUE: Results of serial median matching analysis

End point

Hazard ratio (95% CI)


Composite cardiac events

0.90 (0.79-1.03) 0.111


1.02 (0.81-1.28) 0.899


0,96 (0.84-1.10) 0.566


0.97 (0.80-1.19) 0.791

Heart failure

0.81 (0.66-0.99) 0.040

They also found that reaching BP control levels (<140 mm Hg systolic) by six months was associated with significant CV benefits regardless of drug type and that the BP response seen after only one month of treatment predicted events and survival.

"Blood-pressure control and rapidity of response are critical for reducing events in high-risk hypertensives," Weber concluded.

Did VALUE add value?

Asked for comment by heartwire , session cochair Prof Peter Sleight (Oxford University, UK) said the VALUE trial shows the importance of lowering blood pressure quickly, "and I think that's the take-home message from these results," he said.

"It's possibly disappointing for the sponsors that their drug didn't turn out to be fantastically better than what is a very effective drug, but it's actually, I think, heartening to them that it's about the same in the long run," he said. "But I think the design of the trialit was designed many years agoreally harmed them in a way because they titrated, which was the standard practice at the time, but now we know we need to go more quickly, and we need to go with more effective BP-lowering combinations of agents."

Cochair Prof Guiseppe Mancia (Universit degli Studi Milano-Bicocca, Monza, Italy) said the VALUE findings provide confirmation that valsartan is cardioprotective, just as amlodipine is known to be from such trials as ALLHAT. "This increases to a significant degree the evidence that we can count on this class of angiotensin 2 antagonists," Mancia told heartwire .

In addition, the data show again, "very forcefully," that control of blood pressure is of fundamental importance, he said. "The new finding is that prompt control of blood pressure is of immediate importance in high-risk patients, and once again, confirms the wisdom of the European guidelines that really make this the major point. It doesn't matter which drug is used, differences between drugs have to be seen on top of good blood pressure control, not instead of good control."

The time-trend analyses shown in VALUE represent a new method of analyzing the data that has not been done before, Mancia added, "and I found this very, very interesting."

Finally, in a commentary article published online with the VALUE findings, Prof Lars H Lindholm (Umeå University, Sweden) speculates on whether VALUE "adds value" to data gained from other "'sartan" trials, including the LIFE trial with losartan and the SCOPE trial with candesartan[3].

He points out that the failure in VALUE to achieve comparable BP lowering makes the comparison of CV outcomes "difficult" and called the serial median matching analysis "interesting, but its reliability remains to be proven."

However, results of the three trials do suggest that this group of drugs appears to be at least as effective as other antihypertensives in reducing CV events when BP is well lowered, that BP lowering is important to cardiovascular outcomes, and that they appear to reduce new-onset diabetes, he writes.

What's not clear is whether there are differences between the agents, Lindholm concludes, information that could be gained only from a randomized controlled trial comparing them. "I suspect, however, that we will have to wait a very long time for the results of such a trial, if it is ever carried out."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.