Clinical Precision in the Management of Hypertension

Improving Upon the “One-Size-Fits-All” Approach

Corresponding Author

• Download figure
• Open in new tab
• Download powerpoint

• ACC/AHA 2017 hypertension guidelines
• cardiovascular risk
• hypertension management
• SPRINT

Guidelines for the evaluation and treatment of hypertension have become the subject of intense analysis and controversy in primary care and cardiovascular medicine practices in recent months. The newest guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in adults, published in November 2017 by the American College of Cardiology (ACC) and the American Heart Association (AHA) (1), has a different paradigm from prior guidelines that includes the incorporation of a cardiovascular (CV) risk score to advise clinicians in management decisions. Furthermore, the 2017 blood pressure (BP) guidelines reduced the threshold for the diagnosis of stage 1 hypertension from 140/90 mm Hg to 130 to 139 mm Hg systolic and/or 80 to 89 mm Hg diastolic, and recommended the introduction of pharmacological therapy in stage 1 hypertension when the 10-year CV disease risk score was 10% or greater (2). The new goal of therapy for just about every person with hypertension is an office-based measurement of <130/80 mm Hg.

Clearly, the results from SPRINT (Systolic Blood Pressure Intervention Trial) (3) played an important role in the development of the management paradigm in the 2017 ACC/AHA guidelines. In SPRINT, systolic BP reductions to <120 mm Hg resulted in highly significant reductions in cardiovascular mortality and acute decompensated heart failure admissions compared with the standard group (systolic BP <140 mm Hg). The 37% reduction in acute decompensated heart failure in SPRINT was perhaps of the greatest clinical importance as the patients with heart failure had a 27-fold increase in eventual cardiovascular death (4). Preventing heart failure decompensation in patients with high CV risk is very likely to lead to reduced CV death, including in patients with type 2 diabetes (5,6), which is a major clinical comorbidity excluded from SPRINT. One of the most important groups studied in SPRINT—those older than 75 years of age—also derived benefit from lower systolic BPs. The SPRINT-Senior subgroup analysis (7) demonstrated that the number needed to treat to prevent 1 primary outcome (nonfatal myocardial infarction, nonfatal stroke, death due to cardiovascular causes, acute decompensated heart failure, and non–myocardial infarction acute coronary syndromes) was 28 (vs. 61 in the study as a whole). To prevent 1 death, the number needed to treat was 41 (vs. 90 in the study as a whole).

The substantial benefits of intensive systolic BP lowering observed in SPRINT came at a cost of an increase in serious adverse events such as acute kidney injury (or acute renal failure), syncope, and hypotension and electrolyte disturbances (intensive vs. standard group: 4.7% vs. 2.5%) (3). Many have opined on the significance of the adverse event rates versus the benefits seen in SPRINT, and those considerations have had substantial influence on some, but not all, hypertension guidelines. For example, in addition to the ACC/AHA 2017 guidelines, the Canadian Hypertension Education Program recommended targeting intensive BP therapy only for those patients who are at high risk (8) and after having a benefit-risk discussion with patients. The American College of Physicians and the American Academy of Family Physicians have continued to recommend a systolic BP <150 mm Hg for most people >60 years of age and 140 mm Hg in certain high-risk individuals (9), citing the evidence for <140 mm Hg as “weak.” Undoubtedly, practicing physicians are concerned about intensively lowering BP in most patients with hypertension—both from a pragmatic standpoint as well as from a safety perspective.

It has been difficult for clinicians to extrapolate the findings in SPRINT to their patients with stage 1 or 2 hypertension (stage 2 is now defined as ≥140/90 mm Hg by the ACC/AHA 2017 guideline paper [1]). Patients entering into SPRINT had a long history of hypertension with comorbid chronic kidney or vascular disease or were >75 years of age, and were being treated with an average of nearly 2 antihypertensive drugs. SPRINT patients had a Framingham 10-year risk score of approximately 20% (3). Additionally, measurement of the BP in SPRINT has caused confusion among physicians, as automated BP devices were used that took most BP measurements in the absence of a clinician. Digital, unattended BP measurements may be several mm Hg lower than the more common BP measurement methods used in a medical care environment by a doctor or nurse (10,11). Taking all of these issues into consideration along with concerns about the effects of excessive BP reduction, particularly in older persons (9,12), will likely make implementation of the ACC/AHA 2017 guidelines difficult in clinical practice in the United States.

In this issue of the Journal, Phillips et al. (13) have performed an independent analysis of the impact of baseline 10-year CV risk in the SPRINT population to ascertain benefit (reduction in the primary endpoint) and harm (increases in the rates of all-cause serious adverse events) in those patients randomized to the intensive treatment group (goal systolic BP <120 mm Hg). In performing this patient-level analysis, the investigative team provides useful information to clinicians for determination of who derived benefit in excess of harm, and conversely, which patients experienced more harm than benefit. Using the ACC/AHA CV risk calculator (2) to stratify the SPRINT cohort into quartiles of baseline 10-year CV disease risk, Cox proportional hazard models were used to evaluate the associations of treatment with both the primary endpoint and serious adverse events. Further Poisson regression modeling of the data determined benefit-to-harm ratios within the quartiles of CV disease risk.

Importantly, within each quartile of 10-year CV disease risk (the ranges were <11.5% in the first quartile to >28.9% in the fourth quartile), there was still a lower rate for the primary outcome in the intensive-treatment group compared with the standard-treatment group. Hence, in absolute terms of benefit, lowering the systolic BP to <120 mm Hg (i.e., using the SPRINT BP methodology that has been projected to translate to about <130 mm Hg in real-world practice [1,10,11]) appears to be appropriate regardless of baseline CV risk in “SPRINT-like” patients. Serious adverse event rates in SPRINT, particularly hypotension, electrolyte disturbances, and acute kidney injury or acute renal failure, were observed more frequently in the intensive group in the highest quartiles of CV disease risk. The authors’ predictive model of benefit-to-harm of intensive systolic BP treatment as a function of the 10-year CV disease risk quartile showed ratios of decreased benefit-to-harm in quartiles 1 and 2 (hazard ratio: 0.5 ± 0.15 and 0.78 ± 0.26, respectively) and ratios of increased benefit-to-harm in quartiles 3 and 4 (hazard ratio: 2.13 ± 0.73 and 4.80 ± 1.86, respectively). Hence, if one believes that fairly large increases in serious adverse events such as acute kidney injury, hypotension, syncope, and falls are as clinically important as relatively smaller reductions in heart failure or CV mortality, then one might not treat patients with higher-risk hypertension in quartiles 1 and 2 to intensively reduced levels of systolic BP.

Another recent analysis from SPRINT evaluated beneficial and harmful outcomes in the intensive versus standard group in those patients who entered the trial with low diastolic BPs (14). In that analysis, patients were divided into quintiles of baseline diastolic BPs—those with the lowest quintile (<68 mm Hg) had the highest CV event rate, creating a J-shaped curve for the population’s CV event rates versus baseline diastolic BP. However, regardless of the baseline diastolic BP quintile, the intensive reduction in systolic BP still appeared to be beneficial (lowest quintile of diastolic BP hazard ratio: 0.78; 95% confidence interval: 0.57 to 1.07). Of note, the risk for incident chronic kidney disease in the nonkidney disease subgroup in the lowest diastolic BP quintile (hazard ratio: 3.16; 95% confidence interval: 1.42 to 7.00) was similar to the incident chronic kidney disease hazard in the higher 4 diastolic BP quintiles (hazard ratio: 3.51; 95% confidence interval: 2.37 to 5.41).

Despite some of these controversies, the study results by Phillips et al. (13) are important and should be helpful for physicians who manage high-risk patients with hypertension, particularly older individuals. It was noted in their study that in patients above the age of 75 years, all but 2.9% were in the higher 2 CV disease-risk categories (defined as a 10-year CV disease score of >18.2%). Hence, despite concerns of harm of intensive antihypertensive therapy in patients older than 75 years (12,15), benefits are likely to substantially outweigh serious adverse events in this important high-risk group. In contrast, the evidence for younger patients with CV disease risk estimations <18.2% seems to support a therapeutic target systolic BP of <140 mm Hg.

• 2018 American College of Cardiology Foundation