SPRINTing towards lower BP goals: A re-analysis of the ACCORD-BP trial

There is an old adage in the fashion industry that if you wait long enough, everything comes back into style. While that may not be true (yet) for Saturday Night Fever polyester suits, it certainly seems to apply to blood pressure (BP) goals in patients with type 2 diabetes mellitus (T2DM). The seventh report of the Joint National Committee (JNC-7)¹ and American Diabetes Association (ADA) guidelines for many years recommended a BP goal of <130/80 mmHg for patients with T2DM. This all changed with the publication of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial², which failed to show a reduction in cardiovascular (CV) events when an intensive systolic BP (SBP) goal <120 mmHg was targeted when compared to the “standard” <140 mmHg SBP goal in patients with T2DM at high risk for CV events (see the iForumRx Commentary – Blood Pressure Control And Diabetes: What Should Our Goal Be ACCORDing To The Evidence?). Despite some limitations in the ACCORD trial, a BP goal of <140/90 mmHg subsequently became the standard of care in both the ADA and other guidelines.^3,4

Fast forward to 2018 and the new 2017 ACC/AHA guidelines⁵ recommend a BP goal of <130/80 mmHg for everyone – including patients with diabetes (see the iForumRx Commentary – Ten Things Every Clinician Should Know About the 2017 Hypertension Guidelines). The 2018 ADA guidelines also recommend a goal of <130/80 mmHg, but only in patients at high risk of cardiovascular disease and only when it can be achieved without undue treatment burden.⁶ This change in recommendations is largely driven by results of the Systolic Blood Pressure Intervention Trial (SPRINT), which demonstrated a 25% reduction in the primary composite outcome of CV events with intensive BP control (SBP target <120 mmHg)⁷ (see the iForumRx Commentary – Pulling Ahead After a SPRINT – Evidence for Lower Blood Pressure Goals). However, extrapolating these findings to patients with T2DM has been challenging as patients with diabetes were excluded from SPRINT.

In an attempt to address these concerns, a recent post-hoc, multivariate subgroup analysis was conducted with data from the ACCORD-BP study.  The investigators selected subjects who were randomized to the standard glucose control arm but also had additional cardiovascular disease (CVD) risk factors which met the inclusion criteria for the SPRINT trial.⁸ Of the nearly 5000 patients enrolled in the ACCORD-BP study, a total of 2,592 would have met SPRINT inclusion criteria (termed SPRINT-eligible ACCORD-BP participants).⁸ After excluding 1,308 patients who were randomized to the intensive glucose control arm of ACCORD, the investigators compared the outcomes in 652 participants assigned to intensive BP control (goal SBP<120 mmHg) and 632 participants assigned to standard BP control (goal SBP<140 mmHg).⁸ Similar to the findings in SPRINT, this post-hoc analysis found that intensive BP control significantly reduced the composite of CVD death, non-fatal myocardial infarction, nonfatal stroke, any revascularization, or heart failure by 21% among patients with T2DM and additional CVD risk factors (Table 1).⁸

Table 1: Effect of Intensive BP control on clinical outcomes among SPRINT-eligible ACCORD-BP patients

^{Legend: CI = confidence interval; CV = cardiovascular; MI = myocardial infarction}

A strength of this post-hoc analysis was that the data was generated from a well-designed randomized control trial (ACCORD-BP). The original ACCORD-BP study was likely underpowered due to lower than anticipated event rates and thus failed to detect a difference in the primary composite outcome.² However, event rates in the post-hoc analysis were significantly higher than those in the original ACCORD-BP study allowing for demonstration of a clinically important and statistically significant reduction in CVD event rates despite the smaller sample size.⁸ The authors also analyzed outcomes for heterogeneity between those with and without T2DM by pooling original SPRINT participants with SPRINT-eligible ACCORD participants demonstrating no difference in effect on CVD outcomes between these groups.⁸

A common concern among clinicians while aiming for strict BP targets is the risk of treatment-related adverse effects. Treatment-related serious adverse events (SAEs) occurred more frequently in intensive BP control SPRINT-eligible ACCORD patients compared with standard control (4.1% vs. 2.1%, respectively; P = 0.003).⁸ This is similar to the findings in SPRINT (4.7% intensive vs. 2.5% standard, HR = 1.88, P < 0.001) where intensely-treated patients experienced significantly higher rates of hypotension, syncope, electrolyte abnormalities, acute kidney injury, or acute renal failure.⁷ However, comparing the adverse event rates between SPRINT and ACCORD-BP is difficult due to differences in definition and classification. While higher rates of treatment-related adverse effects may occur with intensive treatment, it is important to note that absolute event rates were low and with the reduced rate of cardiovascular outcomes, the benefits of intensive BP control in patients with T2DM likely outweigh the risks.

With any post hoc subgroup analysis, caution should be applied when interpreting results. Unplanned subgroup analyses can increase the chance of false positives, especially as the number of analyses increase. It is unknown based on this analysis if certain adverse events may occur more or less frequently in patients with T2DM who are treated to intensive BP goals.

While the findings of this study appear to support the recommendations from the new ACC/AHA Hypertension guidelines, should we be more judicious in which patients with T2DM we select for a SBP goal of less than 130 mmHg compared to those with a SBP goal of less than 140 mmHg? The authors of the SPRINT-eligible ACCORD-BP analysis encourage clinicians to use clinical judgment and consider patient preference before recommending an intensive BP goal in patients with T2DM.⁸ Based on the body of evidence evaluating intensive BP control compared with less intensive control in T2DM, reduction of CV events and death is most pronounced in patients with a baseline SBP > 140 mmHg and who are at high risk of atherosclerotic cardiovascular disease (ASCVD) or have a history of CV events.⁶

Determining ASCVD risk should help guide antihypertensive treatment.⁵ Patients with T2DM who are at high risk of CV events and willing to risk potential adverse effects with additional BP treatment should be treated to a BP goal of <130/80 mmHg. A BP goal of <140/90 mmHg is reasonable for patients at lower risk of CV events, frail elderly, those unable to tolerate the adverse effects associated with antihypertensive medications, and when the pill burden or cost are unacceptable. Ambulatory care pharmacists should engage in a shared-decision making process when selecting BP treatment goals, with an exploration of benefits and potential adverse effects.

So which BP goal would you recommend for your patients with T2DM? Should we SPRINT towards more intensive BP lowering for all or should we reserve more strict BP goals (<130/80 mmHg) for certain high-risk patients only? What do you think?