Full results of the Systolic Blood Pressure Intervention Trial (SPRINT) have been released. Much acclaim has followed along with numerous calls to lower blood pressure targets. The main conclusion of SPRINT is that in patients 50 years or older with an increased risk of cardiovascular events, those assigned to intensive blood pressure (BP) control (<120 mmHg) compared to standard control (<140 mmHg) had a significantly lower rate of cardiovascular events including death. Despite these results there are several reasons to pause in the race to lower BP targets.
First, the beneficial effect size seen in the trial was very small. On a yearly basis, 185 patients would need to be treated to prevent 1 from experiencing a major cardiovascular event. Furthermore, at this point, results still need to be assessed by what medications were used (3 in intensive group and 1.9 in standard group). The results may not be due to lower BP but rather the medications used to achieve it.
Next, the SPRINT results must be taken in context with results from the Action to Control Cardiovascular Risk in Diabetes Blood Pressure Trial (ACCORD BP). ACCORD BP applied the same BP lowering intervention, with similar BP lowering results, in patients with diabetes and an elevated cardiovascular risk and found discordant outcomes (refer to table).
Table. Event rates per year in patients assigned to intensive (I) versus standard (S) control in the SPRINT and ACCORD BP trials. An (*) indicates a statistically significant difference between groups in the respective trial.
|CHF||MI||Stroke||Death (CV)||Death (all-cause)||Serious adverse events|
Some investigators have suggested the results from SPRINT and ACCORD BP are actually in agreement but ACCORD BP was underpowered to detect a difference in outcomes. Notably, the SPRINT trial included nearly twice as many patients; however, the actual number of events in the two trials was quite similar. Based on this similarity the treatment effects should not have differed if the two trials were in agreement (refer to table). Therefore, we are left to wonder, why is SPRINT different than ACCORD? Is there a fundamental difference between the effects of intensive BP lowering in diabetics versus non-diabetics? The diabetic patients in ACCORD had significantly higher baseline cardiovascular disease burden than did those in SPRINT. In ACCORD 34% of patients had a previous cardiovascular event compared to only 17% of patients in SPRINT with clinical cardiovascular disease. The average age of patients in ACCORD and SPRINT was 63 and 68 years respectively. There is no way to easily reconcile these discrepant results.
While the beneficial effects of intensive BP lowering differed between ACCORD and SPRINT, harms associated with more aggressive treatment did not. In SPRINT, a significantly higher number of intensively treated patients experienced serious adverse events including hypotension, syncope, electrolyte abnormality, and acute kidney injury. Interestingly, SPRINT found no difference in injurious falls (which is counterintuitive given the significant differences in hypotension and syncope). Perhaps fall risk was mitigated by the close monitoring of patients in the trial – clinical and laboratory data were obtained at baseline and every 3 months thereafter – which may not be feasible in real world settings.
Multiple studies including several in JAMA Internal Medicine over the preceding years have found associations between antihypertensive medication use and fall injuries. Tinetti et al. reported that antihypertensive medication use was associated with an increased risk of serious fall injuries, particularly among those with previous fall injuries. Butt et al. reported that exposure to antihypertensive drugs was associated with an immediate increased risk of hip fracture in hypertensive community-dwelling elderly patients. Based on the findings of these observational studies, it is possible that intensive BP lowering could result in adverse event rates that exceed those from the trial itself.
One additional consideration is that the population of patients enrolled in SPRINT had a high cardiovascular risk (with 61% having a 10 year Framingham risk score ≥15%). This leaves us wondering how intensive blood pressure lowering would affect a population with average cardiovascular risk.
So what is the practicing clinician to do in the wake of SPRINT? In my opinion, physicians should continue to exercise caution in the pharmacologic management of blood pressure and stick to JNC 8 guidelines. Decisions to intensify therapy in patients who would be represented by SPRINT should be based on a careful measurement of BP as well as a shared-decision making approach that provides information on the expected benefits and risks in clinically relevant terms such as absolute risk or NNT.