Bob Harrington, MD: Hi, this is Bob Harrington from Stanford University on Medscape Cardiology and theheart.org. Over the years we have talked a lot about various clinical-trial results that may be influencing clinical practice and that may result in some change in healthcare policy, reimbursement issues. We try to address trials that have some controversial issues that are worth talking about in the community.
Today I'd like to talk about the SPRINT trial.[1] You might ask yourself, "Didn't we hear those results about a year ago?" And that is certainly true. This was a large trial looking at two different levels of blood-pressure [BP] lowering: less than about 140 mm Hg systolic in the control arm and less than about 120 mm Hg systolic in the active-intervention arm. It was stopped early. Results were then presented at the heart association meetings about a year ago with subsequent publication in the New England Journal of Medicine. You might think, "If the trial showed a positive result to the magnitude reported, what is still being discussed?"
A lot is being discussed. My guest today is a terrific person to help lead us through issues on stopping trials early, early announcement of overall results without details, and open data and open science. We will discuss controversy arising from this summer's European Society of Cardiology [ESC] meeting regarding the method of BP measurement in the SPRINT trial and whether this is relevant to clinical practice today.
I'm very pleased to be joined today by a colleague located just a few miles north of me, Ethan Weiss. Ethan is an associate professor of medicine at the University of California, San Francisco [UCSF] School of Medicine. Ethan is a cardiologist interested in preventive cardiology and metabolic disorders. I am an active follower of Ethan on Twitter (@ethanjweiss) and I would advise my listeners to do the same. Ethan, thanks for joining us here today on Medscape Cardiology.
Ethan Weiss, MD: Thank you, Bob. It's absolutely my pleasure. I would only add the warning that anybody who does choose to follow me on Twitter should be aware that I do not limit my discussions to medicine.
The Intent of SPRINT
Dr Harrington: Let's dive right into it. Could you give some context to the question that was being asked in SPRINT? What was the issue regarding systolic blood pressure [SBP]140 mm Hg vs less than 120 mm Hg? And particularly for the readers who may not follow hypertension literature, why was this such an important question for National Heart, Lung, and Blood Institute [NHLBI] to tackle?
Dr Weiss: My perspective, Bob, comes from the fact that, until this year, I have been teaching the hypertension lecture to the first-year medical students since I first came to UCSF back in the late 1990s. Every so often the NHLBI would put out a new set of guidelines—what used to be called the Report of the Joint National Committee (JNC) on the Management of Hypertension. JNC 7[2] was published in the early 2000s, and when I would give this lecture in the fall, I would tell the students, "We expect JNC 8 out anytime." It would never show up.
Then about 3 years ago, JNC 8[3] was published with a relaxed definition of hypertension. I do not need to get into the details, but they changed the targets and they did so based not on data, but on the absence of data. That caught the attention of a lot of us in the prevention community. We all wondered aloud whether this was a good or bad decision. Fortunately, NHLBI sponsored a big bear of a trial. It was designed to ask the question, in a very data-driven way, what is the optimal target for treatment of BP in people with high risk? Is the SBP target of 140 mm Hg adequate, safe, and optimal, or should we be more aggressive?
Dr Harrington: That is very much how I interpreted it. Whenever I talked about hypertension, I was always referring back to a fairly old set of documents. It was interesting to see how delayed the new hypertension guidelines were.
At the same time the hypertension guidelines were being updated, another set of controversial guidelines[4] was being constructed around risk assessment and cholesterol lowering. It was a time when there was a lot of interest in areas that are pivotal to understanding prevention. We got the hypertension guidelines, and not that much later, the SPRINT trial stopped early. It's unusual for us to have a large cardiovascular trial stopped early, particularly one for efficacy, because that means that the signal for efficacy had to be pretty profound to meet the criteria for stopping. What did you think when you heard that the trial was being stopped early?
Stopping SPRINT and the Data Wait
Dr Weiss: My reaction was colored mostly by what had happened with the publication of JNC 8. In case the listeners do not remember exactly, the JNC 7 guidelines published in 2003 defined hypertension as BP greater than 140/90 mm Hg and normal BP as less than 120/80 mm Hg. Those guidelines recommended to treat people to the target of 120/80 mm Hg.
In the new set of guidelines[3] published in 2013, the recommendation was to treat people to a target of less than 140/90 mm Hg if they were less than 60 years old, but to 150/100 mm Hg if they were over 60 years old. That meant that a normal SBP in patients 60 years old and older would be considered anything less than 160 mm Hg, which really was a dramatic change.
I thought the early release from SPRINT made a lot of sense. We'd had big fanfare about this new set of guidelines that was really a significant departure from what we had all thought about and how we had all practiced medicine for at least as long as I can remember. In large part, I thought that the announcement that came when it did was important, even though it did not come with much more than top-line results. It was something that the NHLBI felt they had to report. Here was a big trial that not only found a really remarkable reduction in the primary end point but also found a significant reduction in all-cause mortality. I was happy to see that they had done this.
Dr Harrington: Whenever I see a clinical trial of such public-health importance stopped early for efficacy, particularly when it includes a reduction in mortality, there is a sense of, "Wow, this is a step forward for our patients." It gave us more solid evidence upon which to base the treatment of patients with hypertension, and it certainly helped with what seemed to be the controversy over the last hypertension guidelines—not everyone on the committee agreed with the way that the final document was written and the way the final set of recommendations was put forward.[5]
This seemed to be an awfully good thing to bring clarity to the community. The first thing being discussed was, why did the NHLBI feel the imperative to release the results at a qualitative level without giving us data until the American Heart Association (AHA) meeting a couple of months later? Certainly, I understood the issue of getting everything ready for presentation, but many in the community were a bit chafed by this. What did you think?
Dr Weiss: I took the more sympathetic approach like you did. I recognized that there is a lot that goes into getting these data together and organized, ready for presentation and publication. I thought that the time that they actually were able to do that in was really remarkable, given the size of the trial and the complicated nature of the trial.
I do understand people's strong desire to see data. I wanted to see data but felt I could be patient for a couple of months. I was not going to change my practice over 2 months. I thought, "It's an important finding. Here are some 'bells and whistles' reminding all of us that something big is coming. Pay attention and see what this looks like." It really was not that long.
Dr Harrington: It was a couple of months. I said the same thing as you, which is, "This is complicated." It gave us a heads up—[results are] coming. There was a lot of work to do. The investigators and the sponsor, in this case the National Institutes of Health (NIH), worked hard to get results ready for presentation and publication. Seeing the full disclosure both in print and at the AHA meeting was really pretty impressive. It generated a lot of discussion right out of the box, didn't it?
SPRINT in the Spotlight
Dr Weiss: Yes, I do not recall anything like this. It brings up something we may not have time to get into today, which is there seems to be—I'm sure people would characterize my views one way as opposed to the other way—a pretty strong "religious" disagreement over the principles underlying this trial.
I was really surprised how much energy there was pointed against this trial. Here was a trial that was sponsored by the NIH—there was no pharmaceutical involvement. It was addressing an important question, one that had been missing and that had everyday clinical practice and relevance.
If you go back to the 2002 World Health Organization report[6] on attributable risk, their conclusion was that suboptimal treatment of BP was the number-one attributable risk for preventable death in the world. They estimated that 62% of cerebral vascular disease and 49% of ischemic heart disease could be preventable by treating suboptimal BP.
This is a hugely important area, and here we have a trial that was really, really well done, at least in my opinion, that answers the same question.
Dr Harrington: You've rightly framed it. This is a major public-health problem for the United States and the world. SPRINT was a large trial, with hard clinical end points, a well-defined hypothesis, and potential global public-health implications. They achieved what they set out to achieve, which was a big separation in blood pressure, roughly 136 mm Hg or so vs 120 to 121 mm Hg, and they demonstrated an effect on the hardest of clinical outcomes, including cardiovascular death.
You used an interesting phrase when you said there was almost a "religious disagreement." I'm not a hypertension investigator, but I'm certainly interested in it from a clinical perspective. What would you describe as the schools of belief on this?
Why the SPRINT Divide?
Dr Weiss: I'm not a hypertension investigator, either. In fact, I don't even study BP in the lab. The area is fascinating, and maybe my naïveté is a good thing or maybe it's a bad thing. This may extend to medicine in general or to the concept of prevention, overall.
There are two general camps here. One believes that we can trigger a combination of global lifestyle measures and medicine, identify and treat people who are at risk for disease, and prevent or at least delay the onset of that disease. The other camp says that it's not possible and they believe that less is more. That people should just go out and live their life and if the impact is there it's small and probably not worth the cost (side effects) that come with the increased intensity of intervention. It comes down to classic interventionalists vs minimalists. We see this a lot in medicine.
Dr Harrington: You have summarized it nicely. Maybe it's good that two investigators outside the area are having this conversation, since neither one of us necessarily has a preconceived camp to which we belong. I was delighted to see the data; I took them as a positive step for the treatment of patients and potentially a positive step for public-health issues of better BP control.
I interpreted a lot of the [results] as better BP control. It took more drugs to get you to better BP control, and you did accept some risk of hypotension with that better control. But on balance, when you weighed the risks vs the benefit—and here the benefit is on pretty big events—it seemed worthwhile.
A lot of the discussion around SPRINT was still simmering in the background until the ESC meeting this summer in Rome, when there was a series of debates on what the results meant.[7]
SPRINT BP Measurement Controversy
Dr Harrington: During that debate, someone speaking on the side against putting SPRINT into practice made a comment like, "Wait a minute. The way BP was measured in SPRINT was really different from how it is typically done in clinical practice. It would be very difficult to duplicate the way they measured BP. Therefore, these results do not have direct meaningfulness to clinical practice." In other words, an SBP of 120 mm Hg may be way too aggressive in the context of clinical practice given the measurement issues. What do you think?
Dr Weiss: That is fascinating and brings up the bigger question of how do you apply clinical-trial results to everyday clinical practice. I had my own opinion about how I was going to apply this to my practice when I saw the results. Starting with the basic idea that in a mostly primary, albeit high-risk, prevention population—I think the average Framingham risk score was 20%, and 80% of those were true primary-prevention patients—we had a dramatic reduction in the primary end point and a big reduction in all-cause mortality. Not cardiovascular mortality, but all-cause mortality. So, I thought, "Wow, there is something really interesting here, and it's our job to figure out what it is." My initial reaction was, "Obviously I'm not going to treat all patients to an SBP target of 120 mm Hg, irrespective of whether they are falling down, having syncope, or having renal failure. I'm going to aspire to treat them as low as I can get them without them having these side effects." That was how I reacted.
Then the controversy over the BP measurement came out. I've thought a lot about this and talked to a bunch of different people who are much more schooled in this area. My conclusion—again from talking to the people I respect the most in the field—is what they did in this trial was the gold standard. This was the best and most accurate way to measure BP. By using these ambulatory BP monitors, unobserved, and over time, they got people's true BP. The reaction was, this is not what happens in practice, this is not real world. That may be the case. Do you say, "These trial results are meaningless"? Or do you say, "There is a really strong signal here, and maybe we should be doing a better job of trying to measure BP in practice"?
My reaction was to try to adopt the methods they used in the trial to get a real and accurate BP reading on my patients, which I think we all aspire to do. My reaction was polar opposite from what was presented at ESC and some of the things that were said afterward.
Dr Harrington: It's interesting. You and I have not prepped for this conversation, but you have articulated my interpretation of this almost to the letter. After looking at the results of SPRINT, I'm aspiring to lower BP almost as much as I can without causing symptoms or renal dysfunction in my patients. That is my goal.
I look at SPRINT as an efficacy trial, where they used gold-standard measurement, as you have called it, and demonstrated that the 15-mm Hg difference in SBP carries with it an important reduction in hard clinical outcomes. Yes, I may not be able to measure BP like that in my busy clinic, but shouldn't we try? At the very least, shouldn't we strive in whatever environment we are in to lower BP further, without causing side effects? Our interpretations of the data are consistent.
Are SPRINT Methods Achievable?
Dr Harrington: A question that comes up a lot is, how do you take clinical-trial results, which are by definition a slice of a population and done in a certain way, and make them more generalizable? Most of our clinical trials are really focused on a certain question and a certain population. Yet as clinicians who see a broader population of patients in a broader amount of settings, we are trying to extrapolate those data to fit that group of people.
Dr Weiss: That is absolutely right. There is ardent science to how you do that. It will be really interesting to see how the guidelines tackle this next time—if we ever see a new set of BP guidelines emerge from this controversy. These are the questions everyday doctors and practitioners want to know. How do I apply the results, and how do I apply the methods?
In this case, I do not think the method of how they measured BP is completely unrealistic in terms of something that we could do in an ambulatory setting or even in an office setting. It obviously takes time, but maybe the idea of doing an unobserved BP for our patients is the right way. I was always taught as a medical student that the most accurate BP measurement is one you did with a manual cuff right there at the bedside. I may start to rethink that. Maybe me standing there with a white coat on in an office setting does not make the manual cuff even close to what is accurate.
Dr Harrington: You are exactly right. In my own clinic, the medical assistant typically takes the BP when somebody walks in. Then either myself or one of the fellows or the nurse retakes it. We may even take it a third time, and try to put it into context of everything else that is going on with that person.
When we start to see more and more data collected in the ambulatory setting by people with their own cuffs, or new devices striving to measure BP on a more regular basis, I wonder what BP we will pay attention to? Is it going to be the average over some period of time? Is it going to be the median? Is it going to be the upper-quartile range? Is it going to be some summated amount of BP? SPRINT opens my eyes—there are a lot of questions about how to do this. I do hope the guideline writers help us with this.
Dr Weiss: Absolutely. That is such a great point, too, about the natural variations in BP. I often joke that my BP may be 120/80 mm Hg when I'm sitting here in my office, but when I'm driving across town in San Francisco traffic, it sure isn't anything close to that. All these questions are so important. We are going to learn so much from new ambulatory trials, longitudinal studies, and new devices. SPRINT just tells us that we have something to learn about how we collect BP. Maybe the lesson is that we should be aspiring to do it as accurately as possible and not accept all the natural variations as noise.
Dr Harrington: That is a perfect way to bring our conversation to an end. SPRINT has definitely taught us something and, I suspect, will continue to teach us. The New England Journal of Medicine announced an open competition where you can analyze the SPRINT data set. As the community dives into SPRINT, maybe more interesting work will come out of it from people who approach it with a very, very different set of views than people who are usually working in the hypertension space.
Dr Weiss: I agree. It's fantastic, and I commend them for doing that. We will learn a lot.
Dr Harrington: Thank you for joining me here on Medscape Cardiology. My guest today has been Dr Ethan Weiss, an associate professor of medicine at UCSF School of Medicine. Thanks, Ethan, for having the conversation with me today.
Dr Weiss: Thank you, Bob. I really enjoyed it and I look forward to hearing from everybody about what they think.
Dr Harrington: Thank you.
Cite this: SPRINT: Why Hasn't The Low BP Goal Been Embraced in Practice? - Medscape - Oct 21, 2016.
Comments