If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Medicine and Surgery, University of Insubria, Varese and Department of Medicine and Cardiopulmonary Rehabilitation, Maugeri Care and Research Institute, IRCCS Tradate, Italy
The conclusiveness of major RCTs on prognostic impact of a more intensive versus a less intensive BP control remains uncertain.
•
We undertook an updated cumulative meta-analysis of aggregated data and a trial sequential analysis.
•
A more intensive BP control was conclusively superior to a less intensive control for prevention of stroke, HF, and MI.
•
It also showed a significant benefit, albeit not yet conclusive, for prevention of cardiovascular death.
Abstract
Outcome data from randomized trials which compared different blood pressure (BP) targets grew impressively after publication of recent trials. We conducted a cumulative updated trial sequential analysis of studies which compared a more versus less intensive BP control strategy, for a total of 60,870 randomized patients. The compared BP targets differed across the trials. Outcome measures were stroke, heart failure, myocardial infarction and cardiovascular death. The average duration of follow-up was 3.95 years and achieved systolic BP was 7.69 mmHg lower with the more intensive than the less intensive BP control strategy. The more intensive BP control strategy significantly reduced the risk of stroke (OR 0.79; 95% CI 0.67–0.93), heart failure (OR 0.73; 95% CI 0.55–0.96), myocardial infarction (OR 0.81; 95% CI 0.73–0.91) and cardiovascular death (OR 0.81; 95% CI 0.68–0.98) as compared to the less intensive strategy. In a trial sequential analysis, the more intensive BP control strategy provided conclusive benefits over the less intensive strategy on the risk of stroke, heart failure and myocardial infarction by definitely crossing the efficacy monitoring boundary. For cardiovascular death, the cumulative Z-curve of the sequential analysis touched the efficacy monitoring boundary, but did not cross it. In conclusion, data accrued from randomized trials conclusively demonstrate the superiority of a more intensive over a less intensive BP control strategy for the prevention of stroke, heart failure and myocardial infarction. Results also suggest a significant benefit, albeit not yet conclusive, of a more intensive over a less intensive strategy for prevention of cardiovascular death.
Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the global burden of disease study 2015.
]. The progressive decline in the incidence of cardiovascular disease, particularly stroke, which occurred over the past 50 years, has been largely attributed to the progressive reduction of elevated BP levels at population level [
Prospective Studies Collaboration Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies.
2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American college of cardiology/American heart association task force on clinical practice guidelines.
]. The debate is synthetized by the different position of the American College of Cardiology (ACC)/American Heart Association (AHA) 2017 Guidelines, which recommend a general BP goal of <130/80 mmHg, [
2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American college of cardiology/American heart association task force on clinical practice guidelines.
] and the European Society of Cardiology (ESC)/European Society of Hypertension (ESH) Guidelines which recommend a general BP goal of <140/90 mmHg, with lower goals in specific clinical scenarios [
According to the general principles of evidence-based medicine, a final answer to the question of the most appropriate BP target can stem from randomized controlled trials in which different BP lowering strategies are compared in their effects on BP reduction and outcome. An alternative solution may be the comparison of subjects with different achieved BP levels from post-hoc analyzes of observational and trial databases which, however, may expose to inherent methodological bias.
A few years ago, we published a trial sequential analysis of randomized trials with allocated patients to a more versus less intensive BP target [
]. The analysis definitely supported the superiority of a more intensive over a less intensive BP lowering strategy to reduce the risk of stroke and myocardial infarction (MI), whereas the benefits on the risk of heart failure (HF) and death were still inconclusive [
]. Subsequently, two important clinical trials have been published. The Final Report of the Systolic Blood Pressure Intervention Trial (SPRINT), published in year 2021 [
We addressed randomized controlled trials (RCTs) meeting all the following inclusion criteria: (a) comparison between different BP targets (more versus less intensive target), regardless of the specific BP goal; (b); stroke, MI, HF, or cardiovascular death as pre-specified outcome events, regardless of the definition of the primary end-point in each trial; (c) BP values measured both at baseline and at follow-up; (d) follow-up duration > 12 months with the exclusion of observational extensions phases; (e) publication before December 31th, 2021; (f) no age or language restriction, in order to avoid discriminating papers not written in English (‘tower of Babel bias’) [
Candidate studies were searched through MEDLINE (from Jan 1, 1950, to Sept 30, 2021), Embase (from 1966 to Sept 30, 2021) and the Cochrane Central Register of Controlled Trials (up to Sept 30, 2021), using research Methodology Filters [
]. We limited the search to the randomized phase of controlled clinical trials with more than 12 months of follow-up and hence excluded post-randomized extensions phases. We made a further screening of review articles, published proceedings of conferences, and regulatory agencies files [
] in order to identify other relevant studies. To focus our attention to cardiovascular outcome events, we excluded studies which reported all-cause mortality in the absence of data on different cardiovascular events or the possible cardiovascular origin of the death.
Data synthesis and quality assessment
Table 1 shows the trials identified on the basis of the above criteria. Overall, trials accrued 60,870 patients randomized to a more intensive or less intensive BP lowering strategy. The average duration of follow-up was 4.2 years. Fig. 1 shows the flow diagram with the criteria used for selection of trials. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) statement [
]. Data were independently extracted by two authors (FA and PV) on the basis of an intention-to-treat approach. Disagreements were discussed in conference. In order to evaluate the methodological quality of trials (sequence generation of allocation, allocation concealment, masking of participants, personnel, and outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias), we adopted the methods recommended by the Cochrane Collaboration [
Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the hypertension optimal treatment (HOT) randomised trial. HOT study group.
UK Prospective Diabetes Study Group Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38.
Cardiac and renal effects of standard versus rigorous blood pressure control in autosomal-dominant polycystic kidney disease: results of a seven-year prospective randomized study.
We used the raw counts for each outcome event and calculated the cumulative odds ratios (ORs) and 95% confidence intervals (CIs) through a random-effects model. In case of multi-arm trials we analyzed the pre-specified comparisons as defined in the original design [
]. We tested the null hypothesis of homogeneity across individual trials by using the Q-test, and the I2 statistics to assess heterogeneity across pooled estimates. We tested the publication bias by a weighted regression test [
]. CMA evaluate data from studies which are sequentially included according to the year in which they first became available. The earliest available study is entered into the analysis first. At each step of the CMA, one more study is added to the analysis, and the effect size and 95% CI are re-calculated. Hence, CMA of successive RCTs addresses the potential impact of more recent studies on prior pooled results. It can be used to settle whether enough evidence has been accrued comparing an intervention versus a control treatment, or whether a new RCT deserves to be initiated. Notably, no adjustment is made in CMA for repeatedly testing the null hypothesis.
A cumulative TSA can be helpful to decide whether and when a firm and conclusive evidence favoring a specific intervention has been reached from data accrued so far [
]. Cumulative TSA requires a pre-specified and clinically relevant estimate of the effect of intervention, as well as an overall estimate of the risk of type I error. In the present analysis we used the relative risk reductions (RRR) derived from CMA for each outcome event, with alpha set at 5% (two-sided) and a power of 80% considering early and repetitive testing. Based on this information we calculated a diversity adjusted required information size (i.e., the number of participants in the meta-analysis required to accept or reject the specific intervention effect) [
]. Diversity (D2) is estimated as the percentage of the variability between trials (i.e., sum of the variability between trials). D2 differs from the adjusting factor based on the quantification of heterogeneity, the inconsistency (I2), which might underestimate the required information size [
]. On the above basis, we built monitoring boundaries for benefit, harm, or futility with regard to the observed RRR to control the overall type I error as statistical tests were repeated throughout the accumulation of studies [
]. We generated cumulative Z-curves (i.e., Z-statistics after each trial) for each cumulative random-effects meta-analysis, and assessed its crossing of the conventional significance level and the monitoring boundaries. Of particular note, the efficacy monitoring boundaries should be crossed by the cumulative Z-curve to obtain firm evidence for a beneficial intervention effect. Crossing of Z = 1.96 provides a “conventionally” significant result, whereas crossing of the monitoring boundary denotes firm evidence adjusted for random error risk.
Analyzes were performed using R version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria) and TSA software version 0.95 (Copenhagen Trial Unit, DK, http://www.ctu.dk/tsa/).
Role of the funding source
The funder of the study did not have any role in the design of the study, data collection, statistical analysis, interpretation of results and writing the manuscript. All the authors of this study had full access to data and share responsibility for their submission and dissemination.
Results
As shown in Fig. 1, literature search initially yielded 13,590 reports. After removal of duplicates and studies not focused on BP regulation, we reviewed 367 randomized studies in full text. Among these, we excluded 344 studies because of lack of data on the precise number of outcome events, no identification of BP lowering strategies or follow-up extension after the randomized phase. Among the remaining 23 studies, we removed 7 studies because of unclear reporting of outcome events, BP strategy based on ambulatory BP, or follow up time < 12 months. Thus, 16 trials entered the final analysis. Of these, 16 trials reported data on cardiovascular death, 14 on MI, 10 on congestive HF and 14 on stroke.
In the included trials, a total of 61,916 patients were randomized and the average duration of follow-up was 3.95 years (244 568 patient-years of exposure). Baseline BP averaged 149/86 mmHg and 150/87 mmHg in the less intensive and more intensive arm, respectively. On follow-up, average BP was 138/81 mmHg and 129/76 mmHg in the less intensive and more intensive arm, respectively. The weighted BP difference between randomized strategies from baseline to follow-up (assessed by weighting the difference observed in each contributing trial by the number of randomized subjects) was 7.69 mmHg (95% CIs 7.64–7.71) for systolic BP and 4.29 mmHg (95% CIs 4.27–4.31) for diastolic BP. Overall, there were 1323 stroke events, 1403 MI events, 543 HF events and 959 cardiovascular deaths.
The methodological quality varied across the trials, albeit most were at low risk of bias, and a formal test of plot asymmetry showed no evidence of publication bias (p-value = 0.208).
Fig. 2 shows the cumulative MA of the examined studies. Overall, the more intensive BP lowering strategy was associated with a reduction in the cumulative risk of stroke (OR 0.79; 95% CI 0.67–0.93; p = 0.004), HF (OR 0.73; 95% CI 0.55–0.96; p = 0.025), MI (OR 0.81; 95% CI 0.73–0.91; p < 0.0001) and cardiovascular death (OR 0.81; 95% CI 0.66–0.98; p < 0.026) compared to the less intensive strategy. Notably, the heterogeneity across studies was moderate to null (I2 = 38% [p = 0.07], 44% [p = 0.07], 0% [p = 0.95] and 27% [p = 0.15] for stroke, HF, MI and cardiovascular death, respectively).
Fig. 2. Benefits of more intensive versus less intensive blood pressure control on the risk of stroke, heart failure, myocardial infarction and cardiovascular death. Diamonds represent the 95% confidence interval for pooled estimates.
Results of TSA are shown in Figs. 3 and 4. For the outcome stroke (Fig. 3), the cumulative Z-curve reached the sequential monitoring boundary with the study by Wei et al. [
] thus providing firm evidence of the beneficial effect of the intervention. For HF (Fig. 3), the cumulative Z-curve definitely crossed the conventional significance boundary after the SPRINT study [
] and definitely crossed it after all the subsequent six studies. For cardiovascular death, the cumulative Z-curve reached the sequential monitoring boundary with the STEP study, without fully crossing it. Subsequently to the three latest trials (i.e., the study by Wei et al. [
]) the cumulative Z-curve for cardiovascular death moved far away from the futility area.
Fig. 3. Trial sequential analysis of more intensive versus less intensive blood pressure control on the risk of stroke and heart failure. The dashed red lines denote the sequential monitoring boundaries and the gray area the futility inner wedge.
Fig. 4. Trial sequential analysis of the effect of more intensive versus less intensive blood pressure control on the risk of myocardial infarction and cardiovascular death. The dashed red lines denote the sequential monitoring boundaries and the gray area the futility inner wedge.
Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies.
Systematic review for the 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American college of cardiology/American heart association task force on clinical practice guidelines.
Effects of blood pressure lowering on outcome incidence in hypertension: 7. Effects of more vs. less intensive blood pressure lowering and different achieved blood pressure levels - updated overview and meta-analyses of randomized trials.
] with regard to the potential of a more intensive versus a less intensive BP lowering strategy for reducing the risk of stroke, HF, MI and cardiovascular death in hypertensive subjects. Second, our cumulative TSA showed that data accrued so far conclusively support the superiority of a more intensive BP control strategy for the prevention of stroke, HF and MI, whereas the benefits are not yet conclusive, although numerically significant from a statistic viewpoint, for prevention of cardiovascular death.
] in the meta-analysis allowed to conclude that a more intensive BP lowering strategy specifically reduces the risk of stroke, HF, MI and cardiovascular death by 21% (p = 0.004), 27% (p = 0.025), 19% (p < 0.0001) and 19% (p = 0.026), respectively. Of note, the risk of stroke showed a non-significant 19% reduction in the intensive strategy arm in SPRINT and a 33% reduction in STEP (Hazard Ratio [HR] 0.67; 95% CI 0.47–0.97). Conversely, the risk of cardiovascular death showed a significant 42% reduction in the intensive strategy arm in SPRINT, but a non-significant 28% reduction in STEP (HR 0.72; 95% CI 0.39–1.32).
In a comprehensive meta-analysis, which did not include SPRINT and STEP, a more intensive BP target was associated with a significant 22% reduction in the risk of stroke and a significant 13% reduction in the risk of MI, along with a non-significant reduction in the risk of HF (15%) and cardiovascular death (9%) [
], inclusion of the two latter trials consolidated the overall benefit on HF in the more intensive arm. Similarly, SPRINT contributed 112 new cases, and STEP 43 new cases, of cardiovascular death which was less frequent by 42% and 28%, respectively, in the more intensive arm of the SPRINT and STEP [
We used a cumulative TSA of randomized trials to estimate whether the evidence progressively accrued on pivotal specific outcomes may be considered strong and conclusive [
]. The logic of ‘early stopping rules’ used in the management of randomized intervention trials to evaluate whether it is still ethical to continue a trial on the basis of data accrued so far, can be applied to a TSA to understand whether the accrued data are definitive and no further studies are needed [
]. In our analysis, the cumulative Z-curves for the hypotheses of a 20%, 25% and 18% relative risk reduction for stroke, HF and MI, respectively, in the intensive strategy arm crossed not only the conventional Z = 1.96 (P < 0.05) boundary, but also the monitoring boundaries, thereby providing firm and conclusive evidence of benefit. The monitoring boundaries were definitely crossed after the SPRINT trial for stroke, and after the STEP trial for HF. For MI, the monitoring boundaries were definitely crossed with the ACCORD study and the trend consolidated after the subsequent trials.
For cardiovascular death, the benefit associated to the intensive strategy crossed the conventional Z = 1.96 (P < 0.05) boundary after SPRINT [
]. However, the efficacy monitoring boundary was touched, but not crossed, after STEP. Hence, albeit significant from a statistic viewpoint, the benefit of an intensive BP lowering strategy for prevention of cardiovascular death may not yet been considered conclusive.
Several methodological aspects suggest that meta-analyzes should be interpreted with some more caution in comparison with well-designed and rigorously conducted controlled randomized clinical trials [
]. Unfortunately, however, several randomized intervention trials may have been underpowered to robustly test the superiority of intensive BP control strategies on organ-specific outcomes, thus justifying meta-analyzes of aggregated and individual data, as well as further trials [
Unattended blood pressure measurements in the systolic blood pressure intervention trial. implications for entry and achieved blood pressure values compared with other trials.
In the ongoing Stroke in Hypertension Optimal Treatment (SHOT) trial, promoted by the European Society of Hypertension and the Chinese Hypertension League, hypertensive patients with a history of stroke or transient ischemic are being randomized to different systolic BP targets [
Continuation of the ESH-CHL-SHOT trial after publication of the SPRINT: rationale for further study on blood pressure targets of antihypertensive treatment after stroke.
]. The present analysis suggests that the primary outcome of the Stroke in Hypertension Optimal Treatment (SHOT) study, a composite of fatal and non-fatal stroke, will be likely to be observed, while further important evidence will be accrued in relation to cardiovascular death and other secondary outcomes of the study [
Continuation of the ESH-CHL-SHOT trial after publication of the SPRINT: rationale for further study on blood pressure targets of antihypertensive treatment after stroke.
In the present meta-analysis of aggregated data, we found that achieved BP was around 8/4 mmHg lower in the more intensive than in the less intensive BP lowering arm, with at least eight studies targeting a systolic BP level < 130 mmHg and a diastolic target < 80 mmHg in the more intensive arm. These data support the view that a universal BP target <130/80 mmHg is strong and that, in general, hypertension guidelines should reconsider the BP targets moving towards lower values [
Blood Pressure Lowering Treatment Trialists Collaboration Pharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis.
]. In particular, some authors believe that an excessive BP reduction may exert an adverse impact on coronary flow and myocardial perfusion in patients with coronary artery disease [
]. In a post-hoc analysis of the Ongoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial (ONTARGET) study, we investigated 19,102 patients with coronary artery disease at baseline and did not find any excess risk of myocardial infarction after adjustment for potential determinants of reverse causality including cancer and heart failure [
]. Notably, an achieved BP of 118/68 mmHg was associated with a markedly reduced risk of stroke, without any significant increase in the risk of myocardial infarction [
]. These results have been confirmed by a post-hoc analysis of the Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) trial, in which there was no evidence of an excess risk of major events with lower achieved BP values [
]. Overall, these data suggest that the so called ‘J-curve phenomenon’ might mostly result from reverse causality induced by some coexistent diseases associated with low BP and poor outcome (i.e., cancer, heart failure, etc.), rather than any direct effect of BP lowering per se [
The present analysis was not designed to investigate safety issues. However, despite the evidence of episodes of hypotension, electrolyte abnormalities and renal failure associated with a more intensive control of BP [
Effect of intensive versus usual blood pressure control on kidney function among individuals with prior lacunar stroke: a post hoc analysis of the secondary prevention of small subcortical strokes (SPS3) randomized trial.
], the bulk of evidence suggests that the benefits of a more intensive BP control in terms of prevention of major cardiovascular disease and death outweigh the risk of adverse events. We recognize that the magnitude of treatment benefit and the risk of adverse events may depend on the age of initiation, especially if therapies are intended for life-long use. Specifically, younger patients have a lower risk of adverse events and longer projected survival times and longer anticipated exposure to BP lowering treatments over their lifetime.
In other words, it remains to be fully determined whether intensive BP lowering is well-tolerated and if its effects are uniform across the age spectrum.
Limitations of the study
Our aggregate meta-analysis lacks individual patient data (IPD), which would have probably allowed a better precision of estimates. However, the superiority of IPD analyzes on aggregate meta-analyzes has been questioned [
]. Our analysis shares the major limitations of systematic reviews including publication bias, different inclusion and exclusion criteria in the individual trials, different adjudication of major outcome events, different follow-up duration, and different methodological quality of the original studies. By using the Cochrane Collaboration's tool for assessing risk of bias, we found no evidence of heterogeneity between higher and lower quality trials. Finally, our aggregate analysis could not investigate the prognostic impact of more versus less intensive strategies across different strata of cardiovascular risk. Similarly, we could not assess the potential interactions between the intensity of BP control and long-term BP variability [
Our meta-analysis and TSA of aggregate data from randomized clinical trials conclusively demonstrate that a more intensive BP control strategy is superior to a less intensive control strategy for prevention of stroke, HF and MI. We also found a significant benefit, albeit not conclusive, of a more intensive over a less intensive strategy for prevention of cardiovascular death. These data support the view that hypertension guidelines should reconsider lower BP targets in the management of hypertension.
Sources of funding
This study has been funded in part by the no-profit Fondazione Umbra Cuore e Ipertensione – ONLUS, Perugia, Italy.
Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the global burden of disease study 2015.
Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies.
2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American college of cardiology/American heart association task force on clinical practice guidelines.
Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies.
Systematic review for the 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American college of cardiology/American heart association task force on clinical practice guidelines.
Effects of blood pressure lowering on outcome incidence in hypertension: 7. Effects of more vs. less intensive blood pressure lowering and different achieved blood pressure levels - updated overview and meta-analyses of randomized trials.
Unattended blood pressure measurements in the systolic blood pressure intervention trial. implications for entry and achieved blood pressure values compared with other trials.
Continuation of the ESH-CHL-SHOT trial after publication of the SPRINT: rationale for further study on blood pressure targets of antihypertensive treatment after stroke.
Pharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis.
Effect of intensive versus usual blood pressure control on kidney function among individuals with prior lacunar stroke: a post hoc analysis of the secondary prevention of small subcortical strokes (SPS3) randomized trial.
Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the hypertension optimal treatment (HOT) randomised trial. HOT study group.
Cardiac and renal effects of standard versus rigorous blood pressure control in autosomal-dominant polycystic kidney disease: results of a seven-year prospective randomized study.
00132-7&id=ga1.jpg){kind=link}
00132-7&id=gr1.jpg){kind=link}
00132-7&id=gr2.jpg){kind=link}
00132-7&id=gr3.jpg){kind=link}
00132-7&id=gr4.jpg){kind=link}