Putting SPRINT into Practice

This month, Kidney NewsEditorial Board member Edgar V. Lerma, MD, FASN, interviewed George Bakris, MD, FASN, FASH, FAHA, about recent SPRINT (Systolic Blood Pressure Intervention Trial) results. Dr. Bakris is professor of medicine and director of the ASH Comprehensive Hypertension Center at the University of Chicago Medicine.

KN: Why is the SPRINT study significant?

Dr. Bakris: Because it was an appropriately powered study that addressed questions about the level of blood pressure (BP) control in high–cardiovascular (CV)-risk groups, including a very large group over age 75 and those with chronic kidney disease (CKD) stages 4 and 5. In addition to continuing follow-up of the CKD cohort, SPRINT is also looking at the effects of lower BP on dementia in the SPRINT Memory and cognition IN Decreased hypertension (SPRINT-MIND) trial.

KN: What message should primary care physicians and nephrologists get from the SPRINT study?

Dr. Bakris: UpToDate includes carefully crafted messages for general practitioners and specialists.

Regarding goal BP, major guidelines, published before the ACCORD BP Trial, suggested that the goal BP in patients with diabetes mellitus (DM) is <130/80 mm Hg. However, there are no convincing data supporting this approach, with the possible exception of patients who have diabetic nephropathy and proteinuria, for whom evidence suggests that such a goal may slow the rate of progression of the nephropathy. Thus, we agree with the eighth Joint National Committee (JNC 8) and the European Societies of Hypertension and Cardiology that goal BP in most patients with diabetes is <140/90 mm Hg.

We recommend a goal BP of <140/90 mm Hg compared with higher pressures in all patients (grade 1B).

We suggest (a weaker recommendation) an attempt to lower the systolic BP to <130 to 135 mm Hg if it can be achieved without producing significant side effects (grade 2B).

We recommend a goal BP of <130/80 mm Hg compared with higher pressures in patients with diabetic nephropathy and proteinuria (500 mg/d or more; grade 1B).

For patients who fulfill the entry criteria in the ACCORD BP Trial (type 2 diabetes plus either cardiovascular [CV] disease or at least two additional risk factors for CV disease), the author and reviewers of this topic suggest that the risks and burdens of aiming for a goal systolic BP of <120 mm Hg (more side effects, extra patient visits, and increased cost) plus the lack of experience of almost all clinicians in attaining such a goal may be too great a burden to achieve the small reduction in stroke that may be attained (absolute benefit: 1 in 89 patients at 5 years). However, such a goal may be considered in highly motivated patients who would accept more aggressive antihypertensive therapy to further reduce their risk of stroke.

On the basis of the entirety of the data, including the SPRINT study (although there were no diabetics in the SPRINT study, the meta-analysis performed by Perkovic is compelling [1]), the interaction with glycemic control found in the ACCORD Trial, the ABCD Trial, the post hoc analyses of normotensive subgroups in drug versus placebo trials, etc., do we not have enough evidence to suggest a goal BP similar to the goal BP now recommended for patients who meet criteria for the SPRINT Study?

Please note that, on the basis of BP measurement in the SPRINT study, what is routinely used in the office should give a systolic BP range about 5 mm Hg higher, and therefore, the goal should be 125–130 mm Hg. This is only one part of the UpToDate changes.

KN: What are the limitations of the study (and the implications in the results)? What do you think about the exclusion criteria: autosomal dominant polycystic kidney disease, diabetes, proteinuria >1 g/d, stroke patients, and nursing home patients? Do you think these exclusion criteria affected the results significantly, particularly the study’s generalizability?

Dr. Bakris: Limitations of the study are few and clearly stated in the paper. They did not want to look at diabetes because the ACCORD Trial did so, there was a need for CKD data in high-risk groups, and they did not make recruitment very difficult, because CV outcome was the primary end point (that is why high levels of proteinuria were excluded). The data are generalizable if you read the accompanying editorial, which makes the point clear, and we do so as well in UpToDate.

KN: Are the results applicable to CKD patients (with and without diabetes mellitus [DM])? ESRD patients? The elderly?

Dr. Bakris: Absolutely. The level of evidence in CKD is at least 2A if not 1B given factored power calculations. DM is less so, but we think so. ESRD is obviously no, and the elderly is absolutely yes (The results totally discredited JNC 8 recommendation and argue for more aggressiveness).

KN: Does the SPRINT study have any implications for resistant hypertension patients?

Dr. Bakris: Well, this is tough. They did not purposely recruit such patients, but many patients in the study were on three or more medications. The only thing you could say is that this group generally did not have resistant hypertension, because they were mostly well controlled with multiple agents and they took them.

KN: One cannot downplay the BP monitoring/documentation deployed by dedicated individuals involved in the study (e.g., three office BP readings with 5-min rest periods in between readings). Do you think this should be standard practice in all offices?

Dr. Bakris: Good point. Although I do not think that it is practical to do what they did in the study, I do think it is practical to do what I do with all my patients and that is make sure that they have a home BP monitor, know how to take BP, and report data through the Internet to the nurse or doctor, who will have given specific instructions as to when to take the BP (early morning preferred; not everyday but three times a week and after stable, once weekly). There are data that, in stable patients, seven BP readings a month over various times of day are as good as an ambulatory blood pressure monitoring. Patients need to take responsibility for their problem, including following a low-sodium diet.

KN: In this era of electronic medical records (EMRs), ICD-10 documentation, and staffing issues, is there any study looking into how rigorously BP monitoring is actually done in most primary care offices? Nephrologists’ offices?

Dr. Bakris: I am not aware of any such data; it will be challenging, especially in underserved populations and free clinics, but this is an opportunity for someone.

KN: The study did not seem to consider diastolic BP in the decision algorithm. Is this a concern?

Dr. Bakris: Well, yes. Interestingly, they did not have many people at all with a diastolic BP <60, even on treatment; however, it is well known that, in the cohort, studied risk is tied to systolic and not diastolic BP (so not inappropriate). I have mentioned in my interviews that physicians should avoid low diastolic BP in people younger than 60 and try to get systolic BP at least to 140–149 mm Hg in this small subgroup.

KN: Are the results going to warrant a revision of the JNC 8 2013 Guidelines? Kidney Disease Improving Global Outcomes (KDIGO) 2012 Guidelines?

Dr. Bakris: KDIGO is already in the process of making an update. Because there are no more JNCs, the American Heart Association/American College of Cardiology is crafting an updated and revised document to the JNC.

KN: How do you apply such results to your own clinical practice?

Dr. Bakris: In general, I was always applying them but not going as low as they did in the SPRINT study. I calculate the ASCVD (atherosclerotic cardiovascular disease) score with the application for all of my patients as I sit with them, and I discuss how much risk reduction they would get with a lower pressure. I then let them choose if they want more medications. Most opt for lower BPs.

KN: If you were to redo the SPRINT study, what would you have done differently?

Dr. Bakris: I was on the original planning committee and would have perhaps added more African Americans and a greater proportion with more advanced CKD, but there was no money to do this. It was a miracle that they funded what they did, and many investigators did this study as a loss. My administrator did not want me doing the study for that reason.



Emdin CA, et al. Blood pressure lowering in type 2 diabetes: a systematic review and meta-analysis. JAMA 2015; 6:603–15.