Meta-analysis challenges drug use in mild hypertension

Marlene Busko

August 15, 2012

Vancouver, BC - A review of close to 9000 patients in four randomized controlled trials comparing pharmacotherapy and placebo for primary prevention of cardiovascular events in mild hypertension found no benefit with treatment[1]. But the study, published online August 15, 2012 in the Cochrane Database of Systematic Reviews by Dr Diana Diao (University of British Columbia, Vancouver) and colleagues, had too few events to be able to draw any concrete conclusions, outside experts caution.

Diao et al say the outcomes of pharmacotherapy in this patient subgroup have not been previously rigorously studied. With this analysis, "[using] the best available evidence . . . we have a clear answer that we do not know that the benefits of treatment outweigh the harms for this population," senior author Dr James M Wright (University of British Columbia) told heartwire in an email.

However, two outside experts speaking to heart wire voiced major concerns about this review.

"The most important problem in this analysis is that it's just simply too small—it has too few events and it's grossly underpowered to make any kind of statement," said American Society of Hypertension president Dr William B White (University of Connecticut, Farmington). "I think that the entire conclusion of this analysis is flawed. It really has no clinical importance, [and] it should not be adopted by any practicing physician."

Dr Suzanne Oparil (University of Alabama, Birmingham) echoed these statements, saying, "I don't think the paper is particularly important, because it's looking at a very small number of participants . . . followed for only four to five years . . . [and] because of the very small number of events."

Both White and Oparil stressed that treating mild (stage 1) hypertension—defined according to the Seventh Report of the Joint National C ommittee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC7) criteria as systolic pressure of 140 to 159 mm Hg and diastolic pressure of 90 to 99 mm Hg—prevents progression to more severe hypertension.

Focusing on stage 1 hypertension

In a search of the literature for randomized controlled trials specifically looking at primary prevention in patients with mild hypertension, the authors found four trials that met their criteria:

  • Australian Therapeutic Trial in Mild Hypertension (ATTMH)—chlorothiazide vs placebo.

  • UK Medical Research Council (MRC) trial of drug therapy for mild hypertension—Bendrofluazide and propranolol vs placebo.

  • Systolic Hypertension in the Elderly (SHEP)—chlorthalidone and atenolol vs placebo.

  • Veterans Administration - National Heart, Lung, and Blood Institute (VA-NHLBI) Evaluation of Drug Treatment in Mild Hypertension—Chlorthalidone vs placebo.

The SHEP trial was published in the early 1990s, and the others were published in the late 1970s.

The authors analyzed individual patient data from all those in the VA-NHLBI trial but only from some patients from the other studies, including only seven patients from over 4000 patients in SHEP.

Treatments did not result in significantly reduced mortality, CHD, stroke, or total CVD events. However, event numbers were few: just 77 deaths in the treatment arm vs 90 deaths in the placebo arm, for example.

In the treatment group there were 10 strokes vs 20 strokes in those on placebo, a 50% reduction. "If they simply had larger studies or more data, this would have been obviously a very striking finding," White said.

Oparil and White have no qualms about the authors' thorough literature review. But they stress that studies need to have sufficient events and the resultant power to be able to exclude harm or rule in benefit. "You can't do that with 30 events—and that's what they had for stroke," White said."It's just too small."

The 9% of patients who discontinued treatment due to adverse events is to be expected with the old studies, which used "old-fashioned drugs . . . higher-dose diuretics and higher-dose beta blockers," said Oparil.

Take findings with a grain of salt

But the authors stand by their findings. "Clinicians treating these patients (who probably represent at least half of the hypertensive patients they treat) most likely thought that treatment was supported by gold-standard evidence," third author of the new analysis, Dr David K Cundiff (formerly University of Southern California Medical Center, Long Beach, CA, now retired), told heart wire . "Now they know that is not the case. In my opinion, physicians should inform their mild-hypertension patients that this systematic review shows no benefit with drug treatment and that the DASH diet (per JNC 7 recommendations) and exercise will now be prescribed instead of pills," he told heart wire .

While White and Oparil agree there is little in the way of randomized controlled trial data to support the treatment of mild hypertension, they vehemently disagree with anyone stopping antihypertensive therapy on the basis of this new study.

"While it is true that the highest level of evidence—randomized controlled trials with hard outcomes—is lacking," Oparil noted, "we do know that over your lifetime your blood pressure goes up, [and] if you have mild hypertension in middle age, likely you will go on to more severe disease when you are older." Guidelines recommend starting first with lifestyle changes, she noted, but the ability of physicians to influence behavior is limited. "People with [mild hypertension] can't ignore it. Their situation as they age will not get better by itself," she added. Oparil is cochair of JNC 8, which is expected to release its guidelines by the end of this year.

White says that, based on this study, it "would be bad . . . for patients to believe that they no longer need to take blood-pressure medicine that their doctors have started them on, and then [stop their treatment] . . . and accelerate to a much more severe form of hypertension that would lead to major problems."

The authors and commenters have no known conflicts of interest.

 

 

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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.

processing....