Secondary Logo

Could adverse events offset the benefit of intensive blood pressure lowering treatment in the Systolic Blood Pressure Intervention Trial?

Kjeldsen, Sverre E.a,b; Os, Ingrida,c; Westheim, Arneb

doi: 10.1097/HJH.0000000000002018
EDITORIAL COMMENT
Free

aFaculty of Medicine, University of Oslo

bDepartment of Cardiology

cDepartment of Nephrology, Oslo University Hospital, Ullevaal, Oslo, Norway

Correspondence to Sverre E. Kjeldsen, MD, Department of Cardiology, Oslo University Hospital, Ullevaal, Kirkeveien 166, N-0407 Oslo, Norway. Tel: +47 22119100; e-mail: s.e.kjeldsen@medisin.uio.no

In recent years, there has been as increased focus on more intensive blood pressure (BP)-lowering treatment in routine follow-up of people with hypertension. In this scenario high baseline BP, in particular high SBP, visit-to-visit BP variability and treatment-induced adverse events pose challenges and may impact the benefit of treatment. Although lowering of BP prevents cardiovascular, renal and cerebral complications, the intensity of BP-lowering treatment may also be related to adverse events [1], reduced efficacy and possible outweigh the benefits of the intensive BP-lowering treatment.

In this issue of Journal of Hypertension Rueda-Ochoa et al. [2] assessed the impact of SBP over time or cumulative BP (defined by a statistical model integrating all SBP values in all patients) and serious adverse events on the intensive hypertension treatment efficacy in the Systolic Blood Pressure Intervention Trial (SPRINT). They used the original SPRINT database available by data request to National Heart, Lung and Blood Institute (BioLINCC repository) under the SPRINT data analysis challenge initiative, organized by The New England Journal of Medicine. Hypotension, bradycardia, falls, syncope, acute renal failure and electrolytes abnormalities were defined as serious adverse events included in the original SPRINT study. SPRINT was a randomized, controlled, open label trial performed at 102 clinical sites in the United States. Rueda-Ochoa et al. [2] included 9068 SPRINT participants with 128 139 repeated SBP measurements in their analyses. Study participants had been randomly assigned to intensive (SBP target <120 mmHg) vs. standard treatment (SBP target between 135 and 139 mmHg). Intensive treatment significantly reduced SBP by an average of 12.73 mmHg during follow-up. So-called cumulative joint models for longitudinal and survival data analyses were used in the statistical work [2].

The primary outcome was a composite of myocardial infarction (MI), other acute coronary syndromes, acute decompensated heart failure, stroke and cardiovascular mortality [2]. Similar to the original SPRINT report [3], the authors performed subgroup analyses among patients with chronic kidney disease (CKD)/non-CKD, female/male, black/nonblack race, age less than 75 years/at least 75 years, prevalent cardiovascular disease (CVD)/non-CVD and baseline SBP categories (≤132, 133–144 and ≥145 mmHg). Intensive BP treatment decreased the risk for the primary SPRINT outcome at the start of follow-up. However, when cumulative BPs, both for individuals and between treatment groups were taken into account, the initial beneficial effect of the intensive BP-lowering treatment was lost. A significant effect was lost after 1.1–3.4 years of follow-up in the total SPRINT population [2], dependent on characteristics of the patients. This was apparent for female sex after 1.1 years, patients with prevalent CKD or CVD after 1.3 years, black persons or persons with age below 75 years after 1.8 years, persons with baseline SBP more than 132 mmHg after 2.1 years and for patients with serious adverse events during follow-up after 3.4 years. Thus, it seems that the beneficial effect of the intensive hypertension treatment was offset by the cumulative SBP and by the development of serious adverse effects.

By scrutinizing the original SPRINT report [3], it seems that patients with serious side effects such as acute renal failure outnumbered the patients with reduced mortality caused by intensive treatment or the reduction in patients who had a primary endpoint (see definition above). Acute renal failure in the intensive treatment arm was observed in 84 more patients; and electrolyte disturbances that is low potassium below 3.0 mmol/l and serum sodium below 130 mmol/l were observed in 40 and 80 more patients, respectively. There were significant increases in the number of patients with either hypotension or syncope (n = 115) in the intensive treatment arm. It is well known that serious adverse events may cause patients to stop medication, or in a clinical trial – ‘drop out’. The SPRINT investigators expected a 2% drop out rate per year, but after follow-up of 3.3 years the drop-out rate was approximately 10.5%, or 489 vs. 497 in the two study treatment arms, respectively.

The analyses by Rueda-Ochoa et al. [2] suggest that the number of adverse events in SPRINT is substantial. Adverse events caused by intensive BP-lowering treatment are a major obstacle to attain BP target, and is problematic in high-risk hypertensive patients particularly in the early phase of the trial [2]. The adverse events offset the benefits in preventing the composite endpoint of MI, other acute coronary syndromes, acute decompensated heart failure, stroke and cardiovascular mortality. SPRINT has been criticized for a variety of reasons [4]. The analyses by Rueda-Ochoa et al. [2] add ammunition to SPRINT criticism.

However, the SPRINT investigators also defend their study [5]. They used unattended automated office BP measurements in SPRINT [6,7] but they have also questioned whether introducing this new method for measurement of BP in their study had any importance for the study outcomes [8]. Some investigators cannot detect lower BP when it is measured unattended [9]. Lower BP when measured unattended compared with taken attended may however depend on the level of BP in people with treated hypertension and on the amount of cardiovascular risk [10]. Regarding acute renal failure in SPRINT, recently some SPRINT investigators performed a case control study of 1 year duration of intensive BP reduction [11]. This study suggested that the incident acute renal failure during intensive BP-lowering was accompanied by decreases, rather than elevations, in level of renal damage biomarkers and thus may reflect benign changes in renal blood flow rather than intrinsic injury [11].

One cannot however get around one major criticism of SPRINT [12] which is the fact that the SPRINT investigators included incident heart failure into the primary endpoint in a study with open design. In addition, incident heart failure was driving the difference in the primary endpoint [3]. The changes in the specific first-line treatment with diuretic or heart failure medication may open up or conceal the typical symptoms of heart failure in high-risk hypertensive patients with latent or mild degree of this endpoint. In other words an apparent difference in heart failure as endpoint is explained by a systematic error like in a previous large outcome trial [13].

As many as 3136 patients were randomized into SPRINT with baseline SBP 130 to 132 mmHg or less and the treatments of these patients visualize the problem. For the 1553 patients who then needed elevation of SBP towards 135–139 mmHg, some medication was down-titrated or discontinued. Heart failure protective or symptoms hiding drugs in patients with high-risk hypertension were discontinued in this arm, and they were intensified in the 1583 patients with baseline SBP 130 to 132 mmHg or less in the other arm for them to lower SBP to less than 120 mmHg. This is in our opinion the explanation for the 38 patient-difference in incident heart failure between the two treatment arms. Thus, the difference in patients with incident heart failure appeared as a consequence of the SPRINT design and the finding is therefore artificial [14]. In addition, it does not matter whether the changes in drug treatment caused some new cases of heart failure or just masked or demasked symptoms of heart failure; the difference appeared because of the study design itself [3] which led to an average of two heart protective drugs in the less intensive arm and an average of three heart protective drugs in the intensive arm.

In a well designed open study the cases of incident heart failure had counted as secondary endpoints, and SPRINT had run its full course with a chance of showing a difference in cerebral stroke which would have been expected with an average difference of 12.73 mmHg in SBP between the two arms [2,3]. Discussing the impact of adverse events in SPRINT would have been much less interesting if a reduction in a robust endpoint like cerebral stroke had been driving a difference in the primary endpoint.

Back to Top | Article Outline

ACKNOWLEDGEMENTS

Conflicts of interest

S.E.K. has received lecture and consulting honoraria from Bayer, Merck & Co., Servier and Takeda. I.O. reports no conflicts of interest. A.W. reports lecture honoraria from Novartis.

Back to Top | Article Outline

REFERENCES

1. Thomopoulos C, Parati G, Zanchetti A. Effect of blood pressure lowering treatment in hypertension: 8. Outcome reductions vs discontinuations because of adverse drug events – meta-analyses of randomized trials. J Hypertens 2016; 34:1451–1463.
2. Rueda-Ochoa OL, Rojas LZ, Ahmad S, van Duijn CM, Ikram MA, Deckers JW, et al. Impact of cumulative SBP and serious adverse events on efficacy of intensive blood pressure treatment. a randomized clinical trial. J Hypertens 2019; 37:1058–1069.
3. The SPRINT Research Group. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med 2015; 373:2103–2116.
4. Kjeldsen SE, Narkiewicz K, Hedner T, Mancia G. The SPRINT study: outcome may be driven by difference in diuretic treatment demasking heart failure and study design may support systolic blood pressure target below 140 mmHg rather than below 120 mmHg. Blood Press 2016; 25:63–66.
5. Oparil S, Cushman WC, Johnson KC, Kitzman DW, Whelton PK, Wright JT Jr. Sprinting towards the optimal blood pressure target for hypertensive targets. Circ Res 2018; 123:531–534.
6. Kjeldsen SE, Lund-Johansen P, Nilsson PM, Mancia G. Unattended blood pressure measurements in the Systolic Blood Pressure Intervention Trial. Implication for entry and achieved blood pressure values compared to other trials. Hypertension 2016; 67:808–812.
7. Williamson JD, Supiano MA, Applegate WB, Berlowitz DR, Campell RC, Chertow GM, et al. SPRINT Research Group. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years a randomized clinical trial. JAMA 2016; 315:2673–2682.
8. Johnson KC, Whelton PK, Cushman WC, Cutler JA, Evans GW, Snyder JK, et al. Blood pressure measurements in SPRINT (Systolic Blood Pressure Intervention Trial). Hypertension 2018; 71:848–857.
9. Papademetriou V, Tsiofis C, Chung A, Geladari C, Andreadis EA. Unobserved automated office BP is similar to other clinic BP measurements: a prospective randomized study. J Clin Hypertens 2018; 20:1411–1416.
10. Seo J, Lee CJ, Oh J, Lee S-H, Kang S-M, Park S. Large discrepancy between unobserved automated office blood pressure and ambulatory blood pressure in high cardiovascular risk cohort. J Hypertens 2018; (In press).
11. Zhang WR, Craven TE, Malhotra R, Cheung AK, Chonchol M, Drawz P, et al. SPRINT Research Group. Kidney damage biomarkers and incident chronic kidney disease during blood pressure reduction: a case–control study. Ann Intern Med 2018; 169:610–618.
12. Kjeldsen SE, Mancia G. A critical review of the Systolic Blood Pessure Intervention Trial (SPRINT). Eur Heart J 2017; 38:3260–3265.
13. The ALLHAT Officers Coordinators for the ALLHAT Collabortative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic. The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA 2002; 288:2981–2997.
14. Kjeldsen SE, Mariampillai JE, Nilsson PM. Optimal blood pressure target in diabetic and nondiabetic patients. Circ Res 2018; 123:528–530.
Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.