BLOOD PRESSURE VARIABILITY EMERGING AS A BETTER PREDICTOR THAN RANDOM BP OF CARDIOVASCULAR RISK

BP variability is the next big target in hypertension
June 29, 2010 | Lisa Nainggolan


Oslo, Norway - New research that is only just starting to be digested by the hypertension community indicates that visit-to-visit variability in blood-pressure readings will likely become another way of looking for "at-risk" hypertensive patients and in fact is likely to be more reliable as an indicator of cardiovascular risk than the currently used mean BP.

Discussing the importance of this issue for guidelines and clinical practice, Dr Tony Heagerty (University of Manchester, UK) told the recent European Society of Hypertension (ESH) European Meeting on Hypertension 2010: "We are poking around in the dark, offering treatment blankly across a large community, and probably treating a lot of people who don't need to be treated, while not necessarily treating the highest-risk patients. We should stop being reassured by 'occasional' normal BPs. The whole game now is, can we improve the identification of our 'at-risk' individuals?"

Heagerty was speaking at a special plenary session on late-breaking research discussing BP variability as a risk factor. This issue has emerged following new analyses reported at the ACC meeting and published in a number of papers in the Lancet and Lancet Neurology earlier this year, which showed that variability in blood pressure is a much stronger determinant of both stroke and coronary disease outcome than average blood pressure.

The reports also showed, for the first time, that calcium antagonists and diuretics seem to bring about the greatest reduction in visit-to-visit BP variability and are associated with the best stroke prevention, independently of mean systolic blood pressure, whereas beta blockers—despite reducing mean BP—actually increase BP variability. Neurologist Dr Peter M Rothwell (John Radcliffe Hospital, Oxford, UK), who was the lead author on a number of these papers, recapped all of the findings for ESH attendees at the special plenary session.

Heagerty said: "BP variability is here to stay; it has reached a profile of public consumption so that clinicians must acknowledge it. More research is needed to define its clinical usefulness, and of course we will need to identify the criteria for applying it to everyday practice. But we should certainly look harder at choice of antihypertensive agents, doses, and combinations that we administer. Treatment algorithms should confirm the downgrade of beta blockers, and perhaps we should be looking at the introduction of calcium antagonists into treatment regimens for everyone. And we should be urging the pharmaceutical industry to develop new drugs that actually stabilize BP."



Variability can be factored into clinical trials


The doctors discussing blood-pressure variability all agreed that more research is needed on this parameter and how it brings about stroke, including more prognostic studies. It should not be difficult to incorporate variability as a predictor of outcomes into clinical trials, they said.

"In a clinical trial lasting six months or three years, with, let's say, five to six visits, you take the average BP of each visit, you calculate the standard deviation on this number of visits, and then you get variability," session chairperson Dr Stephane Laurent (Hôpital Europeén Georges Pompidou, Paris, France) explained to heartwire. "Then you introduce variability into the multivariate analysis or logistic regression to determine which parameters contribute to the cardiovascular events."

Visit-to-visit variability is, Laurent says, quite simple to differentiate from white-coat hypertension. "Imagine I have one patient. Every time that I see her she has BP of 180/80 mm Hg, but then I look at her home BP measurement and it's absolutely perfect, 120 mm Hg [systolic] every visit for the past five years, so this is typically 'white-coat' and there is no visit-to-visit variability. I think that now we are at the point where we should incorporate BP variability, 24-hour BP variability, visit-to-visit variability, and various types of variability [into trials]."

Another session chair, Dr Suzanne Oparil (University of Alabama at Birmingham), asked for advice on "behalf of us backward Americans who are in the process of designing the Systolic Blood Pressure Intervention Trial (SPRINT)." This trial is comparing a strategy of lowering systolic BP to <120 mm Hg vs <140 mm Hg in patients 55 years or older with hypertension (>140 mm Hg) or prehypertension and stage 3 chronic kidney disease with one or more additional CVD risk factors. The primary outcome will be a composite of CV outcomes and progression of kidney disease. "We haven't taken into account any of the things you have been discussing. What can we do to improve our trial?" she inquired.

Rothwell replied: "Collecting data on BP variability would be fascinating, particularly at the lower level, because in all the data sets we've looked at we've had this strange finding that the lower the mean BP, the more predictive variability becomes."



Changing the mindset of GPs; treat the episodically hypertensive


Rothwell says he views "the most important point about BP variability, from a stroke perspective, as being for diagnosis rather than for treatment. The problem we find with general practitioners treating hypertension is that they don't treat people with variable BP. The guidelines ask them to bring patients back and measure their BP again, and one of those measurements will be low, and they believe the low reading. And, as we would all argue, anyone who has a high BP once or twice probably needs treatment even if in-between their BP is low, because that is a very high-risk situation."

Laurent agrees, and says that, conversely, doctors also often don't treat an occasional high BP, because they make the mistake of thinking that the patient was a little bit stressed or running late, "which translates into physician inertia."

But many at the session said it would take time to develop an everyday test for BP variability that can be easily utilized by family doctors. "We could accept all the findings of BP variability 'as is,' but we will need to have some kind of consensus on how we assess BP variability that can be applied in everyday practice," said Heagerty. Twenty-four-hour ambulatory blood-pressure monitoring (ABPM) is not necessarily a good tool, he said, pointing out that it is often poorly reproducible in studies, "because you might do your ABPM at a time when the BP is stable or low or alternatively apply the investigation when there has been an episodic rise in pressure."

But Rothwell said there is "one way we can do this straightaway," by following the ESH guidelines, taking one week of at-home measures of BP three times a day, "which measures the medium-term variability pretty well," and then "if you add in an ABPM at the beginning or end of the week, you will get most of the information you need in a week.

"What we can get family doctors to do is to change their method of diagnosis of hypertension, in terms of trying to measure variability in clinical practice and start treating patients who are episodically hypertensive, because their risk is high," Rothwell stressed.



Not yet ready for primetime


But despite his enthusiasm for the concept of variability, Laurent says it is not yet ready for clinical daily practice. "It's a very good concept for research" and for integrating into new trials, he notes, but points out how long it took to incorporate left ventricular hypertrophy (LVH)—detected 25 years ago—into clinical trials and then into guidelines.

His own pet project, arterial stiffness—which can be assessed by pulse-wave velocity (PWV)—is a case in point, he says. Although it entered the guidelines three years ago, it is not yet ready for widespread clinical practice, because it depends on the availability of equipment and training of doctors, he observes.

In fact, many believe that BP variability could be a measure of arterial stiffness, and Laurent says further work to investigate this is needed.

A hotline presentation on the predictive value of PWV for cardiovascular events in more than 15 000 subjects was also reported at ESH and will be featured on heartwire in an upcoming story.