Content area
Background
Co-morbid hypertension is strong predictor of adverse cardiovascular (CV) outcomes in patients with atrial fibrillation (AF) but the optimal target for blood pressure (BP) control in this patient population has not been clearly defined.
MethodsThe Cardiovascular Risk reduction in patients with Atrial Fibrillation Trial (CRAFT) is an investigator-initiated and conducted, international, multicenter, open-label, parallel-group, blinded outcome assessed, randomized controlled trial of intensive BP control in patients with AF. The aim is to determine whether intensive BP control (target home systolic blood pressure [SBP] <120 mmHg) is superior to standard BP control (home SBP <135 mmHg) on the hierarchical composite outcome of time to CV death, number of stroke events, time to the first stroke, number of myocardial infarction (MI) events, time to the first MI, number of heart failure hospitalization (HFH) events, and time to the first HFH. A sample size of 1,675 patients is estimated to provide 80% power to detect a win-ratio of 1.50 for intensive versus standard BP control on the primary composite outcome. Study visits are conducted at 1, 2, 3, and 6 months postrandomization, and every 6 months thereafter during the study.
ConclusionsThis clinical trial aims to provide reliable evidence of the effects of intensive BP control in patients with AF.
Trial registrationThe trial is registered at ClinicalTrials.gov (NCT04347330)
Atrial fibrillation (AF) increases risks of cardiovascular (CV) death, stroke and heart failure (HF). AF affects approximately 60 million people worldwide, but this number is increasing as populations continue to age in developed and developing countries.1,2 Despite anticoagulation treatment being a proven strategy to prevent embolic strokes, outcomes remain poor including for HF which is a major cause of death in patients with AF.3
High blood pressure (BP) is associated with the development of AF,4,5 being present in approximately two-thirds of patients,6 but hypertension is also a risk factor for the main CV complications of AF.7,8 Although good BP control is recommended in this population, as in other high CV risk patients,9 the optimal BP level has not been clearly established. Observational studies have shown a J-shaped association between BP and risk of major CV events in patients with AF,10,11 whereas posthoc analyses of clinical trials indicate that patients with AF gain greater absolute CV benefits from intensive BP-lowering treatment.12-14 The Action in Diabetes and Vascular disease: preterAx and diamicroN-MR Controlled Evaluation (ADVANCE) study showed similar benefits with a fixed combination of perindopril and indapamide in patients with type 2 diabetes and AF.12 The Perindopril Protection Against Recurrent Stroke Study (PROGRESS) demonstrated significant reductions in major vascular events in AF patients.13 Additionally, the Systolic Blood Pressure Intervention Trial (SPRINT) trial revealed that intensive BP control provided substantial CV benefits for patients with AF.14
The Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events (ACTIVE-I) trial that included over 9,000 patients with AF did not find a clear reduction in CV events from use of irbesartan versus placebo,15 possibly because only a small (2.9 mmHg) between-group systolic BP difference was achieved. A recent meta-analysis of randomized controlled trials data showed that BP lowering treatment decreased the relative risk of major CV events in adults with AF to the same degree as those without AF,9 but the achieved BP levels in most of these studies were higher than is recommended in current guidelines. Direct evidence pertaining to the benefits and risks of intensive BP lowering treatment in the contemporary management of adults with AF is lacking.
Home BP monitoring (HBPM) has better reproducibility and prognostic significance than conventional office BP measurements16 and is recommended in current hypertension management guidelines,17 but few studies of antihypertensive therapy have been guided by HBPM. Herein, we outline the design of the Cardiovascular Risk reduction in patients with Atrial Fibrillation Trial (CRAFT) which aims to determine the effectiveness of intensive BP lowering supported by HBPM in adults with AF.
ObjectivesThe primary aim of this study is to determine the efficacy and safety of intensive HBPM guided therapy to a systolic BP target <120 mmHg, using a flexible stepped intensive BP lowering regime on top of standard care, compared to guideline-recommended usual standard of care (HBPM-SBP target <135 mmHg) in high-risk adults with AF. The primary outcome is the hierarchical composite endpoint of time to CV death, number of strokes, time to the first stroke event, number of myocardial infarction (MI) events, time to the first MI event, number of hospitalization for HF (HFH) events, and time to the first HFH.
Methods Trial designThe CRAFT study is an international, multicenter, investigator-initiated and conducted, open-label with blinded outcome evaluation, parallel-group randomized controlled trial. A total of 1,675 patients with persistent or paroxysmal AF, a baseline HBPM-SBP 125-154 mmHg and at least 1 additional CV risk factor will be recruited from 160 to 180 clinical centers in China and Japan. Patients are centrally randomized (1:1) to intensive BP control (HBPM SBP <120mmHg) or standard BP control (HBPM SBP <135 mmHg). Study outcomes, BP levels, use of antihypertensive drugs, assessment of health-related quality of life (HRQoL), and all serious adverse events (SAEs), are collected at 1, 2, 3, 6 months postrandomization, and every 6 months thereafter. Patients are recruited over a 4-year period and the last patient is to be followed for for ≥1 year.
Participants, inclusion/exclusion criteria, and proceduresPatients with documented AF and a standard resting office SBP of 140-179 mmHg (if not on any BP lowering medication) or 125-164 mmHg (if on any BP-lowering medication), are screened for inclusion in an open run-in assessment phase lasting 2 weeks. During the run-in phase, patients have their antihypertensive medications titrated to a appropriate range for SBP and are required to measure their BP 3 times in the morning and evening for ≥1 week, and upload these readings for central monitoring. Patients with average HBPM-SBP 125-154 mmHg during the run-in phase are eligible for inclusion. If the HBPM-SBP is ≥155 mmHg or <125 mmHg during the initial run-in phase, the run-in period may be prolonged by up to 2 weeks (Figure 1).
Inclusion criteriaPatients who meet all of the following criteria are eligible for inclusion in the study: age ≥18 years; documented AF, either persistent or paroxysmal (≥2 episodes in the previous 6 months); average HBPM-SBP 125-154 mmHg; have at least 1 of the following CV risk factors: history of thromboembolism, diabetes mellitus (DM), coronary artery disease (CAD), peripheral artery disease (PAD), chronic kidney disease (CKD), or age ≥65 years. Further details of the inclusion criteria can be found in the supplementary material.
Exclusion criteriaPatients are excluded if they meet any of the following criteria: successful AF ablation; moderate-to-severe mitral stenosis or mechanical valve replacement; HBPM-SBP ≥145 mmHg after taking ≥4 antihypertensive medications at full-dose; inability to upload home BP recordings over ≥5 days in the run-in phase; secondary hypertension; SBP <110 mmHg after 1 minute standing; diagnosis of polycystic kidney disease; glomerulonephritis treated with immunosuppressant therapy; an estimated glomerular filtration rate (eGFR) <30 mL/min/1.73m2); any CV event, CV-related procedure or hospitalization for unstable angina in the past 3 months; HF with reduced left ventricular ejection fraction (<40%) or clinical severity III-IV in the New York Heart Association(NYHA)classification; diagnosis of dementia; or pregnancy. Further details of the exclusion criteria are outlined in the supplementary material.
Setting, location and recruitmentParticipants will be recruited from the outpatient clinics of participating hospital sites. Methods for identifying potentially eligible participants include: a targeted review of medical records or databases; prospective screening of inpatient and outpatient settings and referrals from medical institutions/healthcare professionals; and advertisements or other written material. All participants who are preliminarily screened as being eligible and have provided consent for the study will proceed to the run-in phase; those who finish the run-in phase and remain eligible will be randomized.
Study treatmentsEligible participants are randomized to intensive BP control group (HBPM-SBP <120 mmHg) or standard BP control group (HBPM-SBP <135 mmHg). As the aim is to test a treatment strategy of SBP targets rather than a specific medication regime, the treatment protocol is flexible in terms of choice and doses of antihypertensive medications. However, there are preferences for the drug classes to be used, based upon evidence and current guidelines, with the early use of fixed-dose combinations being encouraged.
Based on previous trials,18-20 it is anticipated that most participants in the intensive group will achieve the target SBP within 8-12 months. Most patients in the intensive group are recommended to commence with 2 or 3 antihypertensive agents. If the HBPM-SBP remains elevated (ie ≥120 mmHg), another type of antihypertensive drug should be added or the dose of the initial antihypertensive drugs increased. Dosage adjustments or drug additions are made every month until a stable level of HBPM-SBP <120mmHg is achieved. If a participant's HBPM-SBP is >120 mmHg during any 6-month follow-up period, a different antihypertensive medication should be added rather than increasing the dose of a current drug, unless there are contraindications to this approach. If severe symptoms of hypotension should occur, a gradual reduction in medication is recommended. Figure 2 describes the treatment algorithm for the intensive BP control group.
Most participants in the standard BP control group are likely to achieve the HBPM-SBP target (<135mmHg) within 3-6 months. The treatment approach aims to achieve a HBPM-SBP of 130-134 mmHg in as many participants as possible. On the basis of a careful review of prior antihypertensive medications at the randomization visit, investigators are required to determine the best medication strategy in order to maintain a HBPM-SBP of 130-134mmHg, using combination or single-drug therapy. Either a dose adjustment or the addition of another drug may be needed if HBPM-SBP is ≥135 mmHg during follow-up. Down titration (a reduction of the dose or number of antihypertensive drugs) should be carried out if HBPM-SBP is <125 mmHg. Figure 3 illustrates the treatment algorithm for the standard BP control group.
Home BP is calculated from BP readings taken over 7 consecutive days before each clinic visit using a validated BP monitor (Omron HEM-7134). Measurements are undertaken twice daily, in the morning before taking any antihypertensive medication and in the evening before bedtime. During each session, 3 BP readings are to be recorded with a 1-2-minute interval between each measurement.
Follow-upThe follow-up of participants in the CRAFT study includes BP measurements scheduled at 1, 2, 3, and 6 months postrandomization, followed by every 3 months. BP follow-up can be conducted in outpatient clinics or via phone calls, solely for adjusting antihypertensive treatment plans. Study follow-ups are scheduled at 1, 2, 3, and 6 months postrandomization, followed by every 6 months. Study follow-ups require an outpatient visit to complete the study case report forms, including home and clinic BP data, laboratory results, physical examination results, endpoint events, and adverse events. Table 1 illustrates the timeline for study follow-up.
OutcomesThe original protocol included a conventional time-to-first-event composite endpoint of CV death, nonfatal stroke, nonfatal MI, or HFH as the primary outcome in the study. However, due to recruitment challenges exacerbated by the impact of COVID-19 and an inability to secure major funding to broaden the study internationally, the primary outcome of the trial was modified to a hierarchical composite of time to CV death, number of strokes, time to the first stroke, number of MI events, time to the first MI, number of HFH events, and time to the first HFH, compared by win-ratio approach, after 1,197 patients randomized with a mean of 14 months follow-up completed, in November 2023.
The main secondary outcomes are each component of the primary outcomes; all-cause death; time to first occurrence of progression of CKD or incident albuminuria; HRQoL; major bleeding according to the International Society on Thrombosis and Haemostasis (ISTH) criteria; PAD; and coronary artery revascularization. Adjudication of the primary outcomes is performed by a blinded and independent Clinical Event Committee (CEC). All SAEs and adverse events according to standard definitions will be collected until study end. Definitions of CV endpoints follow the 2017 Cardiovascular and Stroke Endpoint Definitions for Clinical Trials consensus developed by the Standardized Data Collection for Cardiovascular Trials Initiative (SCTI).21 Detailed information of the outcomes and safety events are available in the supplementary material.
Sample sizeThe primary objective of the study is to determine whether more-intensive long-term BP control is superior to standard guideline-recommended BP control. This objective will be assessed using a hierarchical composite endpoint and analyzed using the win-ratio method.22 We assume annual rates of CV death, nonfatal stroke, nonfatal MI and HFH of 2%, 2%, 0.5%, and 4%, respectively, in the standard arm, according to the blinded event rate in our study. We assume that intensive treatment will result in relative reductions of 43%, 20% 17%, and 38% for CV death, nonfatal stroke, nonfatal MI, and HFH, respectively, as in the SPRINT study (Table 2).19 Based on a power of 80% and a 2-sided α of 5%, a win-ratio effect size of 1.50 between the intensive group and standard group will be achieved with a sample size of 1,624 participants followed up for a mean of 3 years. The goal is to enroll 1,675 participants, assuming 3% will be lost to follow-up.
Randomization and allocationRandomization will be conducted through an internet-based web browser using a computer algorithm, which is accessed via password with encryption protection. Once security requirements are satisfied, a series of questions are asked to verify the eligibility of the potential participant for the study. Eligible participants are randomly assigned to either intensive or usual-care BP treatment in a 1:1 allocation with stratification according to study centre, age (<75 vs ≥75 years), and use of anticoagulation.
Statistical methodsThe analysis will follow the intention-to-treat (ITT) principle. Baseline characteristics will be summarized by treatment group. Continuous variables will be summarized by means (SD), and medians (IQR). The hierarchical primary outcome will be analysed by win-ratio analysis and each participant in the intensive group will be compared with each participant in the usual-care group to determine the win/loss/tie in each pair across each of the hierarchical outcomes. All secondary time-to-event outcomes will be analysed using Cox proportional hazard models between the 2 treatment arms. One formal interim analysis will be performed for this study. The Haybittle-Peto rule of an α <0.001 for the interim analysis in favour of intensive BP control will be applied to be considered significant to justify halting or modifying the study before the planned completion of recruitment. The α-level for the final analysis will be the conventional significance level (α = 0.05) given the infrequent interim analysis.
Descriptive statistics will be provided for safety data, where SAEs and treatment discontinuation will be tabulated using standard terminology. Treatment effect on the primary endpoint will be assessed in predefined subgroups that include the stratification variables and baseline BP (above and below median), sex, age, and histories of CKD and CV disease. A detailed statistical analysis plan (SAP) will be prepared prior to data lock.
Data monitoringAn independent data safety monitoring board (DSMB) is monitoring the accumulating data, integrity of the study and patient safety. The DSMB comprises experts in CV disease and hypertension, kidney disease, clinical trials, geriatrics, biostatistics, and health outcomes. The DSMB meets at least twice yearly to monitor safety, advise on study progress and performance, and makes any recommendations regarding study continuation and any changes to the protocol. The DSMB is responsible for advising the Steering Committee as to whether to continue the study, make modifications to the protocol, or terminate the study. Any recommendations will be approved by the Steering Committee before implementation.
Trial managementThe CRAFT study is overseen by a Steering Committee. Central coordination center is from the Heart Health Research Center in Beijing, with responsibility for communications with participating hospital sites and ensuring the study is undertaken according to the protocol, timelines, and recruitment objectives.
FundingThis study was supported financially by the Beijing Municipal Science and Technology Commission (D171100006817001) and the Chinese Society of Cardiology Foundation (HFCSC2019A02).
Trial registration, human research ethics, and dissemination planThe study has received approval from ethics committees of Beijing Anzhen hospital (2020-005) and other study centers, and is registered at ClinicalTrials.gov (NCT04347330). The primary results will be published in the name of the CRAFT investigators. The authors will be responsible for the study design, implementation, analysis, and drafting, editing, and final content of the manuscript. Publications will adhere to the International Committee of Medical Journal Editors' guidelines. Data access will be granted through a protocol approved by the Steering Committee. The data of the trial will not be made publicly available or placed in an open-access repository. The central coordination center will retain the data for at least 15 years.
Current statusThe CRAFT study was lauznched at Beijing Anzhen Hospital on August 31, 2020. By April 1, 2024, the CRAFT study has officially started at 133 centers in China and 17 centers in Japan, with a total of 1,850 patients screened and 1,438 patients randomized.
DiscussionOur study is the first randomized controlled trial to specifically evaluate the effectiveness of more-intensive long-term BP control compared to contemporaneous usual-care BP management in high-risk patients with AF. The optimal BP for patients with AF remains uncertain. Observational studies suggest a systolic BP level of 120-129 mmHg is the optimal range for preventing major CV events.10 A posthoc analysis of the pivotal SPRINT study showed that patients with AF gain greater absolute CV benefit (12.3 vs 5.6 events per 1,000 person-years) from more-intensive BP control.14 However, the ACTIVE-I trial, which enrolled over 9,000 patients with AF, did not demonstrate a significant reduction in cardiovascular events with the use of irbesartan compared to placebo. This lack of benefit might be attributed to the modest reduction in BP achieved, which was only 2.9 mmHg.15 If the results of the CRAFT study indicate that reducing SBP to <120mmHg in patients with AF can reduce the risk of major CV events without increasing adverse events, it would support a more aggressive approach to BP control in this patient group.
The win-ratio method was first proposed by Pocock et al.22 in 2012 as an alternative indicator to the hazard ratio (HR), and it has been increasingly applied in CV trials.23,24 The win-ratio method can take into account the clinical importance of the components of composite outcomes as well as the relative timing of these components. Its advantages include the objective assessment of endpoints, saving sample size, and enhancing statistical power. Considering the slow enrollment rate in the CRAFT study and the different importances of various composite endpoints, we believe that the win-ratio method may be more suitable for the CRAFT study. However, this method, while statistically useful, does not clearly convey the treatment's benefit, making it challenging for patients to grasp the effectiveness of the treatment.
In summary, intensive BP control, as compared to usual standard of BP management, may have additional CV benefits in patients with AF but this remains unproven. The CRAFT study will provide direct evidence on the efficacy and safety, or otherwise, of more-intensive long-term BP control through the conduct of large randomized controlled trial. The results are expected to offer new perspectives on the management of BP in patients with AF.
CRediT authorship contribution statementChao Jiang: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. Zhiyan Wang: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. Xin Du: Writing – review & editing, Supervision, Methodology, Investigation, Data curation, Conceptualization. Yufeng Wang: Investigation. Mingyang Gao: Investigation. Zhaoxu Jia: Investigation. Zhongyi Chai: Investigation. Zhiyun Yang: Investigation. Chi Wang: Data curation, Formal analysis, Project administration. Liu He: Data curation, Formal analysis, Project administration. Rong Hu: Investigation, Supervision. Qiang Lv: Investigation, Supervision. Jiahui Wu: Investigation, Supervision. Xu Li: Investigation, Supervision. Changqi Jia: Investigation, Supervision. Rong Han: Investigation, Supervision. Hisatomi Arima: Investigation, Supervision. Xia Wang: Writing – review & editing, Methodology, Supervision, Project administration. Bruce Neal: Writing – review & editing, Methodology, Supervision, Project administration. Anthony Rodgers: Writing – review & editing, Methodology, Supervision, Project administration. Graham S. Hillis: Supervision, Methodology. Anushka Patel: Supervision, Methodology. Qiang Li: Writing – review & editing, Methodology, Supervision, Project administration. Jianzeng Dong: Conceptualization, Supervision, Writing – review & editing, Methodology, Project administration, Resources. Craig S. Anderson: Conceptualization, Supervision, Writing – review & editing, Methodology, Project administration, Resources. Changsheng Ma: Conceptualization, Supervision, Writing – review & editing, Methodology, Project administration, Resources.
Conflict of interestThe principal Investigators of CRAFT study have no disclosures of financial and other competing interests. CSA reports receiving funding from the Medical Research Council (MRC) and Medical Research Foundation (MRF) of the UK, and the National Health and Medical Research (NHMRC) of Australia, and Penumbra and Takeda. AP is the nonremunerated Chair of the Board of George Medicines, a commercial spin-out of The George Institute that is aiming to bring a novel fixed-dose antihypertensive combination to market, and reports receiving funding from NHMRC.
AcknowledgmentsWe would like to thank all the investigators and participants from the CRAFT study centers for their support of the study.
| Time relative to randomization | -2 to 0 w | 0 w | 1m | 2m | 3m | 6m | 12m/24m/36m/48m | 18m/30m/42m/54m | Close out |
| N/A | N/A | ±10d | ±10d | ±10d | ±30d | ±30d | ±30d | ||
| X | X | X | |||||||
| X | |||||||||
| X | X | ||||||||
| X | X | ||||||||
| X | X | ||||||||
| X | X | X | |||||||
| X | X | X | X | X | X | X | X | X | |
| X | X | X | X | X | X | ||||
| X | X | X | |||||||
| X | X | ||||||||
| X | X | ||||||||
| X | |||||||||
| X | |||||||||
| X | X | X | |||||||
| X | X | X | X | X | X | X | X | ||
| X | X | X | X | X | X | X | |||
| X | X | X | X | X | X | X | |||
| X | X | X | |||||||
| X | X | X | |||||||
| X | X | X | |||||||
| X | X | X | |||||||
| X | X | X |
| CV death | Nonfatal stroke | Nonfatal MI | HFH | |
| Annal events in the control group | 2% | 2% | 0.5% | 4% |
| Relative risk reduction | 43% | 20% | 17% | 38% |
Copyright Elsevier Limited Dec 2024