SPRINT Will Change But Not Revolutionize Blood Pressure Treatment

After all the hype and hoopla it turns out that the SPRINT trial  will in all likelihood really have a significant impact on clinical practice and future guidelines, but it also also seems clear that it will not bring about a revolution, as some have recently speculated, in the treatment of high blood pressure. That’s the broad  consensus emerging from hypertension experts and others who have started to dig into the details of the large NIH-funded trial. Although SPRINT offers solid support for a more intensive approach to lowering blood pressure for many people similar to those studied in the trial, the experts say it does not support the establishment of a universal systolic blood pressure target of 120 mm Hg.

In SPRINT 9,361 people with systolic blood pressure of 130 mm Hg or higher and who were at elevated risk for cardiovascular disease but did not have diabetes were randomized to standard treatment with a blood pressure target of 140 mm Hg or intensive treatment with a blood pressure target of 120 mm Hg.

Results of the trial were presented on Monday at the American Heart Association meeting in Orlando and published simultaneously in the New England Journal of Medicine. When the trial was stopped back in September the announcement of the preliminary results created a media sensation.

The SPRINT Details

The investigators reported a “rapid and sustained between-group difference in systolic blood pressure.” At 1 year blood pressure levels were 121.4 in the intensive treatment group and 136.2 mm Hg in the standard treatment group. The mean number of BP medications was 2.8 in the intensive treatment group and 1.8 in the standard treatment group.

After 3.26 years the trial was stopped due to a significantly lower rate of the composite endpoint (MI, ACS, stroke, heart failure, or CV death): 5.2% in the intensive treatment group versus 6.8% in the standard treatment group (HR 0.75, CI 0.64-0.89, p<0.001). In absolute numbers there were 243 events in the intensive treatment group and 319 in the standard treatment group.

SPRINT logoImportantly, the investigators also recorded significant reduction in both all cause and cardiovascular deaths. There were 155 deaths in the intensive treatment group versus 210 in the standard treatment group ( 3.3% versus 4.5%, HR 0.73, CI 0.60-0.90, p=0.003). There were 37 cardiovascular deaths in the intensive treatment group versus 65 in the standard treatment group (0.8% versus 1.4%, HR 0.57, CI 0.38-0.85, p=0.005).

For the individual components of the primary endpoint there were no significant differences in MI, ACS, or stroke. Most of the difference in the composite endpoint was driven by the significant reductions in cardiovascular death and a 38% reduction in heart failure.

The authors calculated the numbers need to treat (NNT)

  • to prevent a primary endpoint event: 61
  • to prevent death: 90
  • to prevent cardiovascular death: 172

There were more serious adverse events in the intensive treatment group, including syncope, electrolyte abnormalities, and acute kidney injury or failure, but there was no difference in injurious falls.

SPRINT in Perspective

The authors sought to explain the apparent discrepancy with an earlier trial, ACCORD, which failed to demonstrate that more intense blood pressure treatment was beneficial in a diabetic patient population. The discrepancy “could be due to differences in study design, treatment interactions, or the play of chance,” they wrote. They thought it unlikely but could not rule out the idea that there is “an inherent difference in the cardiovascular benefits of systolic blood-pressure lowering between the population with diabetes and the population without diabetes.”

Striking a theme repeated by many other observers, the authors said that the results should not automatically be generalized to groups not included in the trial, including younger patients under 50, diabetics, and patients with a prior stroke. They also noted that it was important to wait for results of forthcoming substudies to learn the effect of intensive treatment on the brain and the kidney.

The New England Journal of Medicine and other journals published a whole slew of accompanying commentaries and articles seeking to interpret the trial.

In a NEJM accompanying editorial Vlado Perkovic and Anthony Rodgers (University of Sydney, Australia) said that SPRINT will “have far-reaching implications.” Responding to concerns that benefits may be overestimated when trials are stopped early, they wrote that “this is unlikely for SPRINT, which had more than 500 primary outcome events.” They write that the results of SPRINT and ACCORD are “generally consistent,” with SPRINT providing the statistical power lacking in ACCORD. “More broadly, labeling trials as ‘positive’ or ‘negative’ is seductive but ultimately counterproductive; it is more helpful to look at the totality of available data.”

The editorialists write that SPRINT “strongly supports pharmacotherapy decisions based on absolute risk levels, in a similar way to current recommendations for lipid lowering.” a goal of 120 mm Hg may be appropriate for people at high risk of a cardiovascular event.

The results of SPRINT are not applicable to people with diabetes, stroke patients, or institutionalized elderly people, since they were excluded from the study, writes Aram Chobanian in an accompanying perspective in the New England Journal of Medicine. Chobanian supports more aggressive treatment– certainly a target less than 150 mm Hg–but takes “a conservative view” about an overall target of 120 mm Hg “for most people with hypertension,” especially since many people in the intensive-treatment group didn’t reach 120 in any case. “In my opinion, the results from SPRINT warrant reducing the treatment goal for systolic blood pressure to less than 130 mm Hg in most people with hypertension who are over 50 years of age and do not have diabetes or a history of stroke.”

But Chobanian notes that the more strict levels of 120 or 130 mm Hg “create major challenges for clinicians.” Suddenly a majority of US hypertensives would be defined as having uncontrolled hypertension. Further, a strict target would necessarily require that many people take  at least 3 drugs, though “many clinicians and their patients with hypertension are reluctant to go beyond two different antihypertensive drugs.”

A paper by Paul Munther (University of Alabama at Birmingham) and colleagues published in the Journal of the American College of Cardiology used NHANES data to estimate the size of the US population that would meet the SPRINT criteria. They estimated that there are more than 25 million US adults 50 years of age or older with a systolic blood pressure of 120 mm Hg or greater, at high risk for cardiovascular disease, and who do not have any of the trial’s exclusion criteria.

In an article in Hypertension, Daniel Jones, Lyssa Weatherly, and John Hall also raise the question: “how generalizable are the results?” They argue that recommendations for diabetic patients will necessarily rely on expert opinion rather than a firm evidence base, given the ambiguity and uncertainty around the ACCORD results. More generally, they write, “Given that SPRINT only included patients with, or at high risk for, cardiovascular disease, the committee might consider whether treatment recommendations should be based on global cardiovascular risk rather than BP alone.”

Some of the SPRINT investigators, in another commentary in Hypertension, warn that hypertension could be overestimated and overrated by clinicians who seek to incorporate the SPRINT results but who don’t follow the strict standards used in the trial for diagnosing high blood pressure.

In yet another commentary in Hypertension, Murray Esler expresses doubt that SPRINT has established “a new ‘universal’ target blood pressure. Although the results should certainly guide treatment strategies for patients like the SPRINT population, he writes that patients with more severe grades of hypertension will likely not derive the same benefit from a lower target.

And in yet another Hypertension paper, Rhian Touyz writes that “the adverse outcomes of SPRINT should not be overlooked lightly, especially since hypotension and acute kidney disease are associated with their own mobility and mortality.”

Jim Stein (University of Wisconsin) delivered a thoughtful response to the trial:

The results of SPRINT, along with last week’s meta-analysis and persistent controversy surrounding the last HTN guidelines now solidify the need for a HTN guideline re-do.  The previous focus on higher targets was misguided and based on an overly narrow interpretation of the data.  But the key question from SPRINT is “how can we reproduce these highly favorable results in our practices in real patients who differ from trial participants and the setting of a clinical trial in meaningful ways?” Key aspects will be patient selection (age over 50, higher CVD risk, BP >130) , initial drug choice (usually combo therapy), and careful titration (monthly follow-up) as in SPRINT.  Extrapolation beyond the types of subjects in this study should be done with caution,  For example, the results likely can be extrapolated safely  in  younger and otherwise healthy adults facing a lifetime of potential complications but should be done much more cautiously in diabetic patients, those with previous strokes, the very old, those with severe kidney disease (stage 5), or significant comorbidities – including those already on too many drugs thus excluded from SPRINT.  In these latter types of patients, the risk/benefit ratio of more aggressive treatment is likely narrow than seen in SPRINT.
I love target-based guidelines but they need to be implemented with common sense. IF we say a target of 120 mmHg is correct, we need to recognize that even in SPRINT, the mean on treatment BP was 121 mmHg and more drugs for a few additional points might not be necessary.
 Previous  Coverage of SPRINT:

Comments

  1. The study should not have been stopped early. But even a 4-year trial might not have been adequate to fully assess the effects of intense treatment. Given the increased adverse events in the intense treatment group, it seems possible that after, say, 5 or more years of treatment benefits might disappear.

Leave a Reply

%d bloggers like this: